Rt Hon Lord Lilley


    Paying for Success

    How to make contracting out work in employment services

    Edited by Peter Lilley MP and Oliver Marc Hartwich

    Policy Exchange is an independent think tank whose mission is to develop and promote new policy ideas which will

    foster a free society based on strong communities, personal freedom, limited government, national self-confidence and

    an enterprise culture. Registered charity no: 1096300.

    Policy Exchange is committed to an evidence-based approach to policy development. We work in partnership with academics

    and other experts and commission major studies involving thorough empirical research of alternative policy outcomes.

    We believe that the policy experience of other countries offers important lessons for government in the UK. We

    also believe that government has much to learn from business and the voluntary sector.


    Charles Moore (Chairman of the Board), Theodore Agnew, Richard Briance, Camilla Cavendish, Richard Ehrman,

    Robin Edwards, Virginia Fraser, George Robinson, Andrew Sells, Tim Steel, Alice Thomson, Rachel Whetstone.

    About the


    and editors

    Thomas Bredgaard

    is an Associate

    Professor at the Centre for Labour Market

    Research at Aalborg University, Denmark.

    Dr Oliver Marc Hartwich

    is the Chief

    Economist at Policy Exchange.

    The Rt. Hon. Peter Lilley MP

    is the

    Member of Parliament for Hitchin and

    Harpenden and a former Secretary of State

    for Social Security.

    Peter Saunders

    is a Professor Emeritus at the

    Centre for Independent Studies, Australia,

    and Adjunct Professor at the Australian

    Graduate School of Management.

    Dr Hilmar Schneider

    is the Director of

    Labor Policy at the Institute for the Study

    of Labor, Germany.

    Els Sol

    is an Associate Professor at the

    Hugo Sinzheimer Institute, University of

    Amsterdam, The Netherlands.

    Jason Turner

    was one of the designers of

    the welfare reforms in Wisconsin. He was

    also a commissioner for the New York

    Human Resources Administration under

    Mayor Rudy Giuliani.


    © Policy Exchange 2008

    Published by

    Policy Exchange, Clutha House, 10 Storey’s Gate, London SW1P 3AY


    ISBN: 978-1-906097-27-6

    Printed by Heron, Dawson and Sawyer

    Designed by SoapBox, www.soapboxcommunications.co.uk


    Executive summary 4

    1 Paying for success – the international experience of contracting out 7

    by Peter Lilley MP

    2 The experience of contracting out employment services in Australia 15

    by Peter Saunders

    3 The experience of privatization of welfare services in Wisconsin 34

    by Jason Turner

    4 The labour market reform in Germany and its impact 49

    on employment services

    by Hilmar Schneider

    5 The creation and regulation of markets in employment services 59

    – Danish experiences

    by Thomas Bredgaard

    6 It’s the client, stupid! An active role for the client in 71

    Dutch employment services

    by Els Sol


    • 3

    Executive Summary

    In December 2006 the Department for

    Work and Pensions commissioned David

    Freud to investigate welfare reform. The

    Freud Report, which was published in

    March 2007, had one core recommendation:

    to use the private and voluntary sectors

    in the provision of employment services.

    But while Freud briefly mentioned

    experiences made abroad, there was no

    detailed analysis of the reforms in other

    countries. Among UK policy makers,

    knowledge of international experience is


    This gap spurred Policy Exchange to

    commission research about five countries

    that have reformed the way in which they

    provide employment services to jobseekers:

    Australia, the United States (Wisconsin),

    Germany, Denmark and the Netherlands.

    These countries are most frequently mentioned

    in welfare reform debates. Their

    experiences are assessed with regard to the

    lessons they hold for the UK by former

    Secretary of State for Social Security, Peter

    Lilley MP.

    Welfare systems are notoriously complex.

    Understanding them requires considerable

    knowledge of the political systems,

    culture and labour markets in which they

    operate. Policy Exchange thus commissioned

    local experts who had in-depth

    knowledge of their (welfare) states. Their

    task was twofold: on the one hand they

    were asked to judge the effects of contracting

    out employment services on employment

    and, consequently, welfare spending.

    On the other hand, they were tasked with

    identifying any difficulties that had been

    experienced in the process.

    The overall results are encouraging. The

    essays in this compilation show how use of

    the private and voluntary sectors has

    brought improvements in the employment

    services sector. While it is sometimes hard

    to isolate the effects of reforms in this area

    from the more general changes to the welfare

    system, it is clear that contracting out

    employment services can improve the jobseekers’

    chances to find work quickly.

    However, potential difficulties arise from

    the design of the contracting out regimes.

    Some of the authors have reported cases of

    ‘creaming’ and ‘parking’, where service

    providers concentrated on jobseekers that

    were the easiest to deal with or delayed and

    sometimes even ignored the most challenging

    cases. In Australia, success fees were

    sometimes fraudulently paid to employers

    taking on jobseekers for a limited period.


    Paying for Success


    insights into the design, implementation

    and pitfalls of contracting out regimes in

    employment services. While welfare reform

    experiences are hardly ever directly transferable

    from one country to another because of

    national peculiarities, the reforms documented

    in this collection of essays will be

    valuable to UK policy makers.

    There appears to be growing cross-party

    recognition that welfare reform along the

    lines suggested in the

    Freud Report

    is the

    way forward, and indeed

    Paying for Success

    confirms its benefits. But in doing so, mistakes

    made abroad should and can be


    Key lessons:

    Peter Lilley: Paying for success


    The welfare reforms documented in

    this volume show the potential of contracting

    out employment services. In

    Wisconsin welfare rolls fell by 80%

    over three years. If similar changes in

    the UK achieved only a quarter of this

    change, the annual budget for

    Incapacity Benefit claimants would be

    cut by £1 billion, and funding for lone

    parents with children over seven by

    around £300 million.



    If the UK matched the 50% drop in

    job placement costs achieved in

    Australia, the cost of operating the welfare

    system would be cut by £250 million.


    In mainland Europe, too, welfare

    reforms have had positive effects.

    Germany’s unemployment count fell

    by 1 million in the two years after it

    started to reform its welfare state. Both

    Denmark and the Netherlands have

    been more successful in getting lone

    parents and the disabled back to work

    than other EU countries.

    The main lessons that should be learnt

    from the international employment services

    reforms are as follows:


    Markets for welfare-to-work schemes

    evolve rapidly, so policy makers must

    be prepared to change regulations, particularly

    incentives, when they do.


    Each country has unique circumstances,

    and must pay attention to

    them when designing its own system.


    Deadweight losses are significant in systems

    that pay by results. ‘Creaming’,

    where companies concentrate on the

    people who are easiest to get back to

    work, and ‘parking’, where private

    providers ignore the hardest to get into

    employment, can be significant problems.

    The state may therefore have to

    identify the hardest to place individuals

    if private contractors do not have a positive

    incentive to do so. However, for all

    parties it is hard to identify such people.


    Negotiating contracts that give private

    providers appropriate incentives is difficult.

    To get the right outcomes, the

    state needs to be shrewd and may need

    to bring in negotiating skills from outside

    when it draws up such agreements.


    Choice is important, but experience in

    some countries suggests that benefit

    recipients do not pick and choose

    between welfare to work providers as

    much as scheme designers would like.


    Rating systems can reinforce payment

    by results to ensure desired outcomes.


    The conditions and sanctions faced by

    benefit recipients are central to their

    participation in back-to-work schemes.

    Australia: The experience of contracting out

    employment services in Australia

    (Peter Saunders)


    Contracting out cut the cost of getting

    claimants back in to work significantly.

    The Australian Productivity Commission

    found that in the first four years

    after the changes, the cost of the country’s

    labour market programmes fell

    from AUD 3.7bn to AUD 1.3bn.


    The pursuit of cost savings has damaged

    the service quality received by the

    most needy welfare recipients.


    The ‘choice’ aspect of the market-based

    system has had little impact. Claimants

    have not acted like customers in a real



    The contracting out model worked

    well when its goal (getting claimants

    back into work) was simple. But when

    its aims became less clear (i.e. once it

    was given more social targets), it

    became less effective.

    The United States (Wisconsin):

    The Experience of Privatization ofWelfare

    Services inWisconsin (Jason Turner)


    The state saved around 15% of its welfare

    budget during the first contract

    period after the start of the reforms.


    Claimants showed themselves to be far

    more resilient to changes in their benefits

    and demands for them to work

    than previously thought.


    Flexibility among system designers and

    managers was important when they

    had to steer the system past unexpected



    Big changes in what the state politicians

    were saying about welfare reform

    significantly affected the behaviour of

    the private contractors.


    • 5

    Executive Summary

    The vitality of private providers, plus

    the incentives they faced, was vital to

    the success of the scheme.

    Germany: The Labour Market Reform in

    Germany and its Impact on Employment

    Services (Hilmar Schneider)

    The reforms of employment services

    were part of a bigger welfare reform

    package that made significant contributions

    to the falls in unemployment

    after their introduction between 2003

    and 2005.

    Because claimants have not had strong

    incentives to demand high quality services

    from providers, the system has not

    worked as well as it could have done.

    The lack of a price mechanism that

    could reveal the costs of working in certain

    areas has made placing the hardest

    to reach claimants unprofitable.

    The introduction of private providers

    has induced a significant amount of

    deadweight loss.

    Denmark: The creation and regulation of

    markets in employment services – Danish

    experiences (Thomas Bredgaard)

    The introduction of the contracting

    out approach gave policy makers an

    opportunity to shift welfare policy

    towards a ‘work first’ approach.

    The Danish reforms moved through

    two phases: the regime was created

    first, but it was tweaked when contractors

    were asked to ‘make a difference’

    rather than just relieve the state of its

    administrative burdens.

    The creation of the market exposed

    politicians to the problems of dealing

    with behaviour that was profit-seeking

    but not supportive of the needs of

    many claimants. They thus had to deal

    with a situation that was working well,

    but not as efficient and simple as they

    had expected. Ultimately, the designers

    of the changes were made aware of the

    need to strike a balance between the

    possibilities of a market system and the

    social objectives of welfare policies.

    The Netherlands: It’s the client stupid!

    An active role for the client in Dutch

    employment services (Els Sol)

    The emphasis on short-term job outcomes

    meant there were no incentives

    for providers to help the most disadvantaged.

    Outcomes were improved once

    Individual Reintegration Agreements

    were introduced in 2004 and the system

    was thus moved away from a ‘one

    size fits all’ structure.

    Comparing the quality of providers has

    been difficult, but the introduction of a

    ratings system will help.

    Paying for Success



    Paying for success –

    the international

    experience of

    contracting out

    By Peter Lilley MP

    Payment by results is not a difficult concept

    to grasp. Nor is it novel.

    Reward success and people will be more

    likely to deliver success. Invite organisations

    to compete to deliver the success

    which brings that reward and success is

    even more likely. But it has taken a long

    time even to consider applying these simple

    concepts to the public services.

    There are few areas where success is

    more valuable than in helping people off

    welfare and into work; it is in the interest,

    above all, of the workless person; it is in the

    interest of their families and communities;

    it is in the interest of taxpayers through

    reducing the welfare burden; and it is in

    the interest of the economy through

    increasing the workforce.

    The only people at present who do not

    stand to benefit directly from success in

    enabling people tomove fromwelfare towork

    are those whose job it is to help them do that.

    Maybe that is one reason the UK has

    been less successful at boosting employment

    than might be expected after a long

    period of steady growth with a deregulated

    labour market.

    Even among men of prime working age


    – 25 to 49 – more than one in nine is not

    in gainful employment after 15 years of

    steady growth. Their employment level has

    stagnated between 88 and 89% throughout

    this decade and is still below the recent

    peak employment rate of 90.3% reached in

    1990. Even that is far short of the employment

    rates of prime working age men in the

    1950s and 60s which were over 96%, albeit

    not on an entirely comparable basis.

    Among older workers and lone parents the

    proportion currently not in work is far

    higher even than one in nine. And the proportion

    of young people not in work or

    education has actually risen since 1997.

    The New Deal has not achieved the success

    that its supporters expected and even

    its critics hoped for. So it is not surprising

    that attention is returning to the idea of

    harnessing reward and competition to

    make our efforts to get people back into

    work more effective.

    The last Conservative government

    announced plans to pilot a scheme called

    Britain Works loosely modelled on

    AmericaWorks which features in the studies

    below. It sought to harness the profit

    motive to the task of getting the long term

    unemployed into work on the basis of payment

    by results. As Secretary of State for

    Social Security I was disappointed that we

    had not started to do this earlier. The division

    of responsibilities between the

    Department of Social Security (DSS) and

    the Department of Employment (DEmp)

    was a major factor slowing things down.


    • 7

    1 For comparisons over time it is

    best to look at the employment

    rate of males of prime working

    age since the proportion of

    women who wish to work has

    changed over time for cultural

    reasons as has the proportion of

    young people staying in education

    and of old people approaching

    retirement age who choose

    early retirement.

    Anyone with experience of Whitehall will

    know that it was easier to get the USA and

    USSR to cooperate at the height of the

    cold war than to get two Whitehall departments

    to work together.

    Whether by accident or design, the

    Labour government removed that obstacle

    by merging the DSS and DEmp into

    the Department for Work and Pensions

    (DWP) in 2001. Had the plans to bring

    in private initiative and payment by

    results not been sidelined before then,

    that merger would have speeded up their

    implementation. It will make it easier to

    implement them now that they are being



    In the meantime other countries have

    taken up the idea of contracting out

    employment services to the private and

    voluntary sectors and of paying them by

    results for helping people get off benefit

    and into work. If we are to make up for our

    lost decade we need to learn from experience

    built up abroad.

    That is why Policy Exchange commissioned

    five studies carried out by experts in

    each of the main countries which have

    been experimenting on these lines.

    These studies reveal that the potential of

    this approach is, indeed, considerable.

    Wisconsin saw welfare rolls fall by 80%

    over three years. If the UK were a quarter

    as successful over the long run we could

    hope to see savings in the annual budget

    for Incapacity Benefit of over £1 billion

    and of the order of £300 million on lone

    parents with children over seven.

    Australia reportedly reduced the cost of

    job placement by half. If DWP’s placement

    costs are proportionate to those in

    Australia and savings were a quarter of

    those in Australia the British taxpayer

    could save around £1/4billion annually on

    operating costs alone.

    Above all the process of contracting out

    focuses attention on the central objective

    of helping people back into work. In

    Germany, before this was done, only 10%

    of the Federal Labour Agency’s 90,000

    employees were involved in job placement.

    It is hard to quantify the impact of contracting

    out in most of the countries studied

    since it formed one part of more comprehensive

    reforms which often came in

    over several stages. Nonetheless, it is significant

    that, for example, in the two years

    following the contracting out of some

    placement, Germany has seen unemployment

    fall by 1 million – the largest and

    most rapid decline in its history. And both

    Denmark and the Netherlands, who make

    extensive use of the voluntary and private

    sectors, have been more successful than

    other EU countries in increasing employment

    among previously excluded groups

    like the disabled and lone parents.

    We have made considerable progress in

    the UK in tackling unemployment, i.e. in

    reducing the number of people on Job

    Seekers’ Allowance. There is more to be

    done particularly for the long term unemployed,

    the repeat claimants and the growing

    number of NEETS (youngsters not in

    employment, education or training).

    However, the largest category of people

    who are excluded from employment is

    those on Incapacity Benefit. This is an

    international phenomenon. It reflects two

    factors. First, changes in the labour market

    in developed countries have reduced the

    number of jobs available to the least skilled

    requiring purely physical effort or repetitive

    actions. They have been replaced by

    jobs mostly in the service sector which may

    not require formal skills but involve a

    degree of mental or personal involvement

    and adaptability. The second factor is that

    people displaced from the old physical and

    repetitive jobs have discovered how to

    access incapacity and similar benefits

    which are usually higher than unemployment

    benefits and do not require them

    actively to seek work. Once on those benefits

    they tend to lose motivation and

    employability so that they become permanently

    detached from the labour market.

    Paying for Success


    2 Although, when Chancellor,

    Gordon Brown was hostile to the

    idea for contracting out put forward

    in the Freud report, he has

    recently given them his backing.

    In the UK access to the old Invalidity

    Benefit simply required a note from the

    claimant’s own doctor which was usually

    forthcoming. So the number on it trebled

    over 15 years. After I replaced this with

    Incapacity Benefit (IB), which required an

    objective medical test, the number of

    claimants fell slightly, but then stabilised at

    the still appallingly high level of 2.6million.

    Welfare reform is not a static process. I

    always envisaged further reforms and will

    wholeheartedly support measures by this

    government or the next to actively help

    people move from IB into work. Clearly

    some people with particularly severe disabilities

    will be unable to do so. However,

    the majority probably could do some form

    of work and would benefit from it. The

    more successful we are at helping them

    into work the more generous we will be

    able to be towards those who are genuinely

    too handicapped to do so.

    The problem is that most people on IB

    face a chicken and egg situation. Out of

    work they lose motivation, their health

    deteriorates, and they become less attractive

    to employers. If they could get into

    work it would restore their motivation,

    often be good for their health, and enable

    them to acquire on the job skills. Even

    unskilled jobs require some knowledge and

    familiarity with the task – most of which

    can only be acquired on the job. Although

    most IB claimants could do valuable work,

    far fewer have the ability to identify what

    they can do beyond their previous occupation,

    fewer still have the ability to market

    themselves. And even if they could, a

    potential employer may require compensation

    for the period during which a new

    recruit is learning on the job.

    So if we are to make serious progress in

    helping people off IB they will need support

    from an organisation which is motivated

    – to boost their confidence; to identify

    their aptitudes; to ‘market’ them to

    potential employers; and, if need be, subsidise

    them for a period. We take it for

    granted in most other spheres that those

    involved in marketing and sales themselves

    require incentivising with rewards for success.

    It is high time we introduced that into

    the business of helping to sell the substantial

    but currently unrealised abilities of

    those on Incapacity Benefit.

    Key lessons

    The five studies commissioned by Policy

    Exchange reveal that, although the concepts

    of payment by results and contracting

    out are simple, their application is

    complex and fraught with difficulties.

    Every country that has adopted this

    approach has largely had to learn from trial

    and error how to implement it effectively.

    Hence the value of these studies. We can

    at least learn from their experience, emulate

    their successes and avoid some of their


    An adaptable system is essential


    Markets for welfare to work schemes

    evolve rapidly, so policy makers must

    be prepared to change regulations, particularly

    incentives, when it does.


    Each country has unique circumstances,

    and must pay attention to

    them when designing its own system.

    However much we benefit from foreign

    experience and however well planned the

    system may be at the start, it will still need

    further adaptation in the light of experience.

    That has been true of every other

    country. One reason is that the market

    evolves as new suppliers gain expertise or

    prove inadequate and fall by the wayside.

    In Wisconsin the system was so successful

    that it had to be scaled down to deal with

    a far smaller case-load. In Germany illjudged

    contract terms bankrupted the

    major provider of placement services.

    Most countries found that incentives

    needed to be modified to avoid inducing

    perverse responses or because they were

    exploited in unintended ways. And systems


    • 9

    Paying for success – the international experience of contracting out

    need to adapt to each country’s unique circumstances

    – its labour law, economic structure,

    employment culture etc. So it is important

    that the initial system is not too rigid,

    top down, over specified or inflexible to be

    adapted in the light of experience.

    Getting the incentives right


    Deadweight losses are significant in systems

    that pay by results.


    ‘Creaming’, where companies concentrate

    on the people who are easiest to

    get back to work, and ‘parking’, where

    private providers ignore the hardest to

    get into employment, can be significant


    The experience of all five countries suggests

    that one of the most difficult problems is

    designing the right structure of incentives

    for the providers. The most intractable issue

    inherent in any system of payment by

    results is how to minimise ‘deadweight losses’,

    discourage ‘creaming’ and prevent ‘parking’.

    A ‘deadweight loss’ is payment for

    helping people to get jobs who would have

    got them by their own efforts. ‘Creaming’ is

    where suppliers select (if they have any say

    in who they take on), or focus their efforts

    on, those clients who are easiest to get into

    work – many of whom would have

    obtained a job with little or no help. This is

    particularly likely if providers are paid the

    same amount for everyone they help into

    work rather than being paid more for the

    harder to place or for increasing the proportion

    of their case load who find work above

    a base level. ‘Parking’ is ignoring, or devoting

    least attention to, the hardest to place

    clients who actually need the most intensive

    support. ‘Parking’ is likely to be a particular

    problem if the remuneration system provides

    additional resources or rewards for

    helping the hard to place only after they

    have been unemployed beyond a threshold


    There is no simple solution to these

    problems in practice.

    But in theory they can be minimised by

    designing incentives to align providers’

    interests with those of the taxpayer – which

    largely coincide with those of the jobless.

    The taxpayers’ interest is to maximise savings

    on benefits (by helping people get

    work and stay in work) net of the cost of

    providing such help.

    Bizarre as it may seem, not even the

    DWP’s interests have been aligned with

    those of the taxpayer. The Department has

    not been allowed to use any of the savings

    in the benefit bill that it makes from welfare-

    to-work programmes to fund those

    programmes. The Conservative Green

    Paper on Welfare Reform promised to

    remedy this problem. This seems to have

    prompted the Chancellor to announce in

    his 2008 Budget statement that he will do

    likewise in future years.

    Ideally, providers’ interests would be

    aligned with those of the taxpayer if, in

    respect of each cohort of benefit claimants

    for which they were given responsibility,

    the providers were rewarded for reducing

    the benefits this cohort claimed in each

    future year by being paid a share of those

    benefit savings. The providers would use

    their share of benefit savings to meet the

    costs they incur in getting and keeping

    people off benefits – any surplus representing

    their profit for doing so effectively.

    One way to establish a fair and competitive

    level for the share of benefit savings needing

    to accrue to the provider would be to

    invite them to bid in terms of the amount

    they would require to undertake the task.

    To assess how much the cost of benefits

    had been reduced by the provider’s efforts

    would require a reliable forecast of what

    the benefits incurred by that cohort would

    have been in each future year without the

    provision of a welfare-to-work programme.

    Such forecasts can be made. Experience

    shows what proportion of a given cohort

    typically leave benefits each year and how

    many subsequently return and for how


    Paying for Success


    For example, for a given cohort of new

    Incapacity Benefit claimants about 25%

    leave within three months; by six months

    some 42% have left; by nine months 52%

    have left and by the end of the first year it

    is about 58%. The pace then slows and by

    the end of the second year the original

    number of claimants has fallen by some

    67%. So a provider would only be remunerated

    in any period to the extent he had

    reduced the welfare rolls below the projected

    base level for that period. This would

    eliminate the ‘deadweight loss’ problem.

    Logically, the provider should be



    for each cohort of

    individual claimants assigned to him. He

    would then have an incentive to identify

    potential long term claimants and help

    them into work early, even if that requires

    disproportionate resources, since he will

    then be remunerated in respect of every

    future year they are not claiming benefit.

    Thus there would be no incentive for

    ‘parking’ the hard to place. On the contrary,

    providers would have an incentive to

    identify those who, without help, would

    probably spend the longest time on benefit

    and focus his efforts on them because if he

    can reduce the number of long term

    claimants he will be rewarded annually.

    Moreover, the provider would only be

    remunerated in respect of additional

    claimants helped into work for as long as

    they remain in work. The provider would

    not have any incentive to focus on the easy

    to place unless he could accelerate their

    return to work relative to what they would

    have achieved without extra help. And he

    would only do so as long as the cost did

    not exceed his promised share of benefit

    savings. So there would be no incentive for


    I emphasise that this is the theoretically

    optimal system of remunerating providers.

    It has not been adopted in full in any of

    the countries studied. Wisconsin Works

    has come nearest to it, with remuneration

    for welfare-to-work providers based on

    savings relative to welfare spending plus

    administration costs in the pre-contract


    One reason this model is likely to

    remain a purely theoretical ideal is that it

    implies indefinite responsibility for each

    cohort of claimants allocated to a provider.

    That would require indefinitely long contracts

    for a given cohort – which is unlikely

    to be acceptable and would limit competition

    and contestability. It would also

    require the ability to forecast with confidence

    the base level of benefit the long

    stayers would be receiving were it not for

    the extra welfare-to-work services provided

    under the contract. That information

    exists nationally on the basis of past experience

    but may not be available for local

    markets and specialist subgroups of

    claimants who may be assigned to individual

    contractors. The longer the contract

    period the more it is likely to be affected by

    changes in the economic environment

    which can neither be predicted by the

    Department nor influenced by the

    provider. Also, once contracting out

    becomes the norm there will no longer be

    any information about what would have

    happened in the contemporary economic

    climate in the absence of welfare to work

    programmes. So that too would make the

    theoretical model unacceptable in practice.

    Nonetheless, the model provides a useful

    theoretical template against which actual

    and proposed contract terms can be evaluated.

    Interestingly, David Freud suggested

    a three year contract period in respect of

    each cohort of claimants. This is considerably

    longer than applied elsewhere and

    apparently twice as long as the Treasury is

    prepared to consider. It would, however,


    • 11

    Paying for success – the international experience of contracting out

    The aim must be not just to get people into work but to

    help them stay in work and preferably in rewarding jobs

    give scope – subject to getting other contract

    features right – greatly to reduce the

    incentive to ‘cream’ and ‘park’ benefit

    claimants. In addition, the problem of

    ‘deadweight losses’ would be reduced by

    only allocating to a provider a cohort of

    claimants who had already been on benefit

    for, say, six or twelve months.

    One reason the Treasury is reportedly

    reluctant to envisage three year contracts is

    that they tacitly assume that providers will

    not be paid until the end of the contract

    period and will therefore demand substantially

    higher rewards. In fact, payment

    related to the saving in benefits relative to

    that expected in each period for each

    cohort could result in payment to

    providers as and when savings in benefits

    actually accrue to the Treasury.

    One point this ideal template highlights

    is that remunerating providers simply on

    the basis of the absolute number of

    claimants helped into work in each period

    can lead to distortions. It gives no incentive

    to help them remain in work. Indeed,

    Germany found that making the first payment

    to the provider immediately after one

    of their clients was placed in a job led to

    fraud as well as ‘deadweight loss’. Still less

    does it give providers any incentive even to

    identify claimants who are likely to be

    hardest to place – let alone to devote proportionately

    more effort to prevent them

    becoming long term unemployed and

    eventually unemployable.

    Identifying hard to place claimants early


    The state may have to identify the

    hardest to place individuals if private

    contractors do not have a positive

    incentive to do so. However, for all parties

    it is hard to identify such people.

    If the provider of welfare-to-work services

    does not have an incentive to identify

    those likely to be hardest to place, the

    DWP or its prime contractor will have to

    undertake this. It will need to establish a

    triage system to categorise claimants from

    the beginning by the degree of help they

    are likely to need and assign rewards and

    resources for helping them into work correspondingly.

    Unfortunately, it is not easy to identify

    those likely to be hard to place. Attempts

    to find criteria which correlate with the

    effort needed to help any individual into

    work have not been very successful. Even a

    past benefit record is only a partial guide –

    partly because a majority of all claimants of

    JSA, for example, have been previous

    recipients. The fact that any individual is

    particularly difficult to place often only

    becomes clear with the passage of time

    after the basic interventions have failed.

    Yet we know that the longer people spend

    out of work the harder it becomes for them

    to get a job. So the earlier the intervention,

    the better.

    Developing a triage system with criteria

    which accurately identify hard to place

    individuals early on should be a high priority.

    The methods used in Australia to classify

    claimants and in the Netherlands to

    relate support packages to individuals may

    be valuable in developing such a system.

    Some categories of jobless people are

    self evidently more likely to be hard to

    integrate into employment than others –

    particularly drug and alcohol abusers, exprisoners,

    those with serious mental

    problems or learning difficulties. They

    will also probably need specialist help

    and may well need to be assigned to specialist

    organisations, appropriately remunerated

    for the more intensive effort

    needed to help them into work. It may

    well be that other categories are also

    identifiable and could usefully be

    assigned to a specialist provider – those

    unable to read or write or lacking numeracy,

    for example.

    Acquiring skills in negotiating contracts


    Negotiating contracts that give private

    providers appropriate incentives can be

    Paying for Success


    difficult. To get the right outcomes,

    the state needs to be shrewd and may

    need to bring in negotiating skills from

    outside when it draws up such agreements.

    A key feature of the contracting out of welfare

    to work – as Denmark established as

    early as 1992 – is that it requires a split

    between purchaser and provider. The purchaser/

    provider split has been found to be

    beneficial in other areas of public services.

    It focuses attention on defining the objective

    of the service. And it prevents decisions

    being dominated by the vested interests

    of the provider.

    However, it raises a central issue which

    every country has faced in trying to contract

    out their welfare-to-work system

    based on payment by results: how to negotiate

    contracts which provide appropriate

    incentives with the right choice of organisations

    to take on the task of helping people

    get into work.

    As the UK goes down this route we will

    need to be clear about our objectives. The

    aim must be not just to get people into

    work but to help them stay in work and

    preferably in rewarding jobs. Remuneration

    should reflect and incentivise that. At the

    same time the contract must encourage

    suppliers to focus on helping the hardest

    to place back into work while avoiding

    paying for those who would have got jobs

    anyway. Ideally contracts should allow

    suppliers time to build up expertise (particularly

    at the start of the process) in the

    art of helping people into work. Yet it is

    also desirable that suppliers face challenges

    from new entrants and from other suppliers

    who prove more successful. The ultimate

    aim is to ensure the services provided

    are best suited to the needs of those

    who are being helped into work – which

    may differ between different categories

    like the long-term unemployed, older

    workers, those with different incapacities –

    as well as between individuals. That may

    require specialist organisations contracted

    to focus on specific groups. Voluntary and

    other organisations with such skills may

    exist in some locations but need to be

    built up from scratch in others. In theory,

    at least, claimants are the ‘clients’ and

    should be able to choose between

    providers thereby exercising consumer

    pressure on them.

    Devising and negotiating contracts

    which appropriately incentivise a suitable

    array of suppliers through remuneration

    terms, duration, degree of monopoly, geographical

    extent and scope of specialisation

    is a highly complex task. It requires

    skills that government Departments rarely

    possess and will need either to master or

    delegate to others – most likely through a

    appointing a prime contractor. It will be

    necessary to harness a large number of

    providers working in collaboration. But

    as the Wisconsin study suggests it is

    important that there is a lead contractor

    who can decide who does what. Relying

    on a partnership of equals can result in


    Giving claimants a choice of provider, though

    desirable, may not be worth pursuing


    Choice is important, but experience in

    some countries suggests that benefit

    recipients do not pick and choose

    between welfare to work providers as

    much as scheme designers would like.

    The idea of allowing claimants to choose

    between alternative providers is attractive

    in theory. It should put pressure on

    providers to offer jobless people the services

    they want in ways they find most accessible.

    And it should make it more likely

    that jobless people are matched to

    providers offering the approaches best suited

    to their needs. But Australian experience

    suggests that these theoretical attractions

    do not emerge in practice. Most

    claimants failed to choose and had to be

    assigned to a provider. The Danes and


    • 13

    Paying for success – the international experience of contracting out

    Germans also found that in practice few

    claimants exercised choice. The Australian

    authorities attribute this to the fact that

    benefit claimants are often people who are

    least used to exercising choice. It may also

    be that choice is most effective where people

    have to repeat the decision process – as

    when we buy necessities. But choice is least

    effective where, as in this case, the user is

    making a one off choice between providers

    with no previous experience of any of them

    to draw upon.

    So it would probably be unwise to rely

    too heavily on claimant choice to force

    providers to compete to offer ever improving

    services to jobless people. If choice can

    only be provided by artificially creating

    alternative suppliers, who may be suboptimal

    in size, it is probably not worth the


    On the other hand the Dutch experience

    of providing each claimant with a personally

    tailored and transferable work plan

    may well be worthwhile studying.

    Successful outcomes, rather than a prescribing

    process, should drive up quality


    Rating systems can reinforce payment

    by results to ensure desired outcomes.

    The Australians, finding claimant choice

    to be ineffectual, have developed a star rating

    system to drive up quality. This may

    have a degree of merit. But the overriding

    incentive to succeed should be provided by

    the system of payment by results.

    Conditionality is key to success


    The conditions and sanctions faced by

    benefit recipients are central to their

    participation in back-to-work schemes.

    The success of contracting out and payment

    by results ultimately rests on making

    benefits conditional on engagement

    with work. The conditions and sanctions

    regimes are crucial – though the subject

    of future Policy Exchange study rather

    than these papers. The rigour with which

    conditions and sanctions are enforced will

    markedly affect providers’ success in getting

    people back to work. If those powers

    were given to the provider there could be

    a severe temptation to exercise them too

    vigorously. So it is important that

    enforcement powers are exercised by a

    separate body – in the UK the DWP



    There can be little doubt in the light of

    foreign experience that it is right to seek

    to harness the profit motive to help people

    from welfare to work. But it will meet

    with resistance from those with vested

    interests in the status quo. That includes

    not only the public sector unions but

    some of the welfare lobbies who are more

    interested in seeking compensation for

    their client group than helping people

    leave it.

    So implementing a successful programme

    of contracting out will require

    both courage and skill. But the intellectual

    tide worldwide is moving in this direction.

    The practical evidence suggests it can

    hardly fail to be an improvement on the

    status quo. And, above all, the needs of several

    million of our most disadvantaged fellow

    citizens demand that we harness the

    powers of competition and incentive to

    give them the opportunities from which

    they have been excluded.

    Paying for Success



    The experience

    of contracting

    out employment

    services in Australia

    by Peter Saunders


    Welfare reform in Australia

    Over the last twenty years, the Australian

    welfare system has been radically reshaped.

    The reforms started under the Hawke and

    Keating Labor governments, and gathered

    pace with the election of John Howard’s

    Liberal/National Coalition government in

    1996. With the return of a Labor government

    in November 2007, further reform is

    likely, although the direction that this will

    take is not yet clear.

    The reforms of the last two decades have

    affected both the payment of benefits,

    where claimants have increasingly been

    required to undertake some prescribed

    activity in return for income support, and

    the organisation of employment services for

    jobless people, where government has come

    to rely on private and community sector

    organisations to deliver key functions.

    The changes to the benefits system are

    not dramatic when compared with, say, the

    US experience, but they are significant.

    Activity requirements were initially applied

    to young, unemployed claimants, but they

    have gradually been extended both to older

    unemployed people and to sole parents

    with children of school age. There have

    also been changes to the definition of

    ‘incapacity’ in an attempt to limit the

    growth of Disability Support Pension

    numbers. Nevertheless, most working-age

    Australians drawing income support are

    still not expected to perform any activity in

    return for their benefit, and total spending

    on income support has continued to rise.

    About one in seven working-age

    Australians is wholly or mainly dependent

    on welfare, even though the economy has

    been booming and jobs appear plentiful.


    The privatisation of employment services

    was a more dramatic change, for when the

    old Commonwealth Employment Service

    (CES) was shut down and replaced by a new

    purchaser-provider model in 1998, Australia

    went down a path that no other country had

    explored. This system has now been in operation

    for almost ten years and has attracted

    growing attention around the world.

    The origins of contracting out

    The welfare reform process started in 1988

    when a voluntary training, job placement

    and child care service (JET) was introduced

    for single parents seeking to return

    to the labour market. At the same time, a

    new unemployment benefit, called

    Newstart Allowance, was introduced and

    the long-term unemployed were for the

    first time required to participate in

    ‘Intensive Assistance’ in return for their



    1 Saunders P, “What are Low

    Ability Workers to do When

    Unskilled Jobs Disappear?’

    Issue Analysis (Centre for

    Independent Studies)

    , no 91,

    December 2007

    2 Intensive Assistance offered

    personalised support to help

    claimants get back into work.

    Following assessment of skills

    and job capability, it aimed to

    improve employability through

    tailored education and training.


    • 15

    These innovations were followed in 1994

    by the

    Working Nation

    program, the core of

    which was a ‘Job Compact’ which guaranteed

    claimants who had been out of work for

    18 months or more a job placement for 6 to

    12 months, backed up by training as appropriate.

    To deliver this, the government

    enlisted the help of a variety of non-government

    agencies. The CES remained responsible

    for finding unemployed people jobs (the

    ‘job matching’ function), but management

    of labour market programs was shared

    between the CES agency, ‘Employment

    Assistance Australia’, and a variety of specialist

    voluntary and commercial agencies

    judged to have expertise in handling people

    with special difficulties. Contracts were

    awarded by fixed-price tender, and by 1996-

    97, 20%of cases had been contracted to private

    firms and 25% were being handled by

    community sector agencies, with the

    remaining 55% still managed by CES or

    other public sector agencies such as further

    education colleges.


    The new Liberal/National Coalition government

    that came to power in 1996

    scrapped Labor’s

    Working Nation


    on the grounds that it was expensive and

    largely ineffective in getting people permanent

    jobs. In its place, it introduced a policy

    of ‘mutual obligation’ which required that

    young unemployed people should undertake

    a recognised activity in return for their payments.

    Over time, this mutual obligation

    policy has been extended. Although nearly

    two million people (almost one-fifth of the

    working age population) are still drawing

    welfare benefits of one kind or another, and

    the welfare dependency rate among working-

    age adults is as high as it has ever been.

    What was not scrapped, however, was

    Labor’s policy of contracting out employment

    services. Indeed, within a year of

    coming to office, the new government shut

    down the Commonwealth Employment

    Service altogether and replaced it with a

    completely new system based on the purchaser-

    provider model.

    The new employment services system:

    Centrelink and the Job Network

    From 1946 to 1998, the federal government’s

    employment service, the CES, was

    responsible for assessment of job seekers,

    employment assistance, management of

    labour market programs, and monitoring

    compliance with activity requirements.


    in the 1996-97 Budget, the government

    announced that CES was to be closed and

    replaced by a single statutory authority,

    called Centrelink, working with several

    hundred non-governmental employment

    agencies, organized as a ‘Job Network’ (JN).

    Centrelink was conceived as the ‘gateway’

    for unemployed people seeking

    labour market assistance. Operating under

    a Business Partnership Agreement with the

    Department of Employment and Workplace

    Relations (DEWR), its role included:

    determining eligibility of job seekers

    for Job Network services;

    providing information to job seekers

    about Job Network services;

    registering job seekers;

    assessing the job seeker’s relative labour

    market disadvantage;

    referring job seekers to Job Network

    providers; and

    administering job seeker participation

    and compliance requirements.


    Under this system, Centrelink is the first

    point of entry for anyone seeking unemployment

    assistance. At its 321 Customer

    Service Centres, job seekers are assessed,

    registered and then referred on to a JN

    service provider. The original idea was

    that, once assessed, clients would choose

    their own JN provider, but most do not in

    fact do this and are instead allocated

    through an automated referral system.


    All other employment services were contracted

    out to JN agencies. These represented

    a mixture of for-profit and not-forprofit

    (charitable, religious and community)

    service providers, many of which

    3 Eardley T, “Outsourcing

    Employment Services: What

    have we learned from the Job

    Network?” Paper to Centre for

    Applied Economic Research

    conference on the Economic

    and Social Impacts of

    Outsourcing, University of New

    South Wales, pp 6, 4-5

    December 2003

    4 Wong P, “Reward for effort:

    Meeting the participation challenge”,

    Australian Labor Party

    Discussion Paper,



    5 Australian National Audit

    Office, “DEWR’s Oversight of

    Job Network Services to Job


    Audit Report

    , no 51,


    6 Productivity Commission,

    “Independent Review of the Job

    Network”, report no 21, June



    Paying for success

    7 Bruttel O,


    Competition in a Public Service

    Market: The Job Network in an

    International Perspective

    , Centre

    for Labour Market Research

    Discussion paper, no 3, pp 6,


    8 Bruttel O,


    Competition in a Public Service

    Market: The Job Network in an

    International Perspective

    , Centre

    for Labour Market Research

    Discussion paper, no 3, 2003

    9 Australian National Audit

    Office, “DEWR’s Oversight of

    Job Network Services to Job


    Audit Report

    , no 51,


    The experience of contracting out employment services in Australia


    • 17

    already had some experience of running

    labour market programs under previous

    Working Nation


    The employment assistance agency formerly

    run by CES (Employment Assistance

    Australia) was turned into a commercial

    company in which government owned

    100% of the shares, and it joined the Job

    Network under its new name of

    Employment National. At the first round of

    tenders (Employment Service Contract 1, or

    ESC1), Employment National won a 37%

    market share,making it the biggest single JN

    service provider, but its share declined

    sharply to just 8% in ESC2, and in 2003, at

    the end of the ESC2 contract period, the

    company was wound up when the federal

    government failed to find a buyer.


    JN service providers are answerable to

    DEWR, although their customers are

    referred to them by Centrelink (see Figure

    1). They were given responsibility for three

    core functions which had previously been

    the responsibility of the CES: Job

    Matching (i.e. finding vacancies for unemployed

    people to fill), Job Search Training

    (a total of 15 days training in writing

    application letters, compiling CVs, and

    interview techniques), and Intensive

    Assistance, which is aimed at those who

    have been out of work for an extended

    period or are deemed to be at risk of longterm

    unemployment. JN providers also

    manage some minor programs, like the

    New Enterprise Incentive Scheme (to help

    job seekers set up businesses) and Harvest

    labour services (organising vacancies for

    seasonal farm work), and they are required

    to monitor job seeker compliance with

    activity requirements and to inform

    Centrelink of any breaches.

    JN members enjoy flexibility in the services

    they provide and how they do it.

    Some specialize (e.g. by providing services

    for those with disabilities or for Indigenous

    job seekers), although unemployed people

    with special needs, such as those with mental

    health or substance abuse problems, are

    referred to the Personal Support Program

    which operates outside the Job Network.

    The relationship between DEWR (the

    purchaser), job seekers (the clients) and

    Centrelink combined with the JN agencies

    (the providers) is set out in Figure 1.

    From the outset, the idea was that JN

    members should compete for referrals and

    be paid by results (‘outcomes’). Fees-forservice

    were paid by DEWR when a job

    seeker was taken on by a JN provider, and

    outcome payments were added when


    Job seekers services

    Business Partnership Arrangement

    Job seeker services









    Figure 1: The new purchaser-provider model for delivery of

    employment services


    10 Strictly speaking it is not a

    ‘Work First’ strategy, for clients

    are registered for unemployment

    allowances by Centrelink and

    are then referred to a JN

    provider to find them work. This

    contrasts with the US ‘Work

    First’ approach where the

    emphasis is on finding clients a

    job before they register for benefits

    (see Mead L,



    , Princeton University

    Press, 2004). Nevertheless,

    compared with the previous system,

    the new arrangements

    undoubtedly emphasise job

    placement as the primary outcome.

    11 ‘New job network to replace

    the CES’, media release, 26

    February 1998.

    12 Bruttel O,


    Competition in a Public Service

    Market: The Job Network in an

    International Perspective


    Centre for Labour Market

    Research Discussion paper, no

    3, table 1 pp 8, 2003

    13 Productivity Commission,

    “Independent Review of the Job

    Network”, report no 21, June



    Paying for success

    clients were successfully placed in a job (a

    lower outcome fee is payable if they complete

    an education or training course). This

    structure of fees and outcome payments has

    been revised over time as DEWR’s policy

    priorities have shifted between rewarding

    rapid placements of job-ready clients and

    encouraging long-lasting placements of disadvantaged

    job seekers. But the key measure

    of success has always been placement of

    unemployed people in work (a so-called

    ‘Work First’ strategy).

    The more successful

    a JN provider is in getting unemployed

    people back into work, the more DEWR

    pays them, and the more likely they are to

    have their contracts renewed at the next

    round of bidding.

    The aim of the reform was simple. It

    was hoped that an incentivised system

    would result in more unemployed people

    getting jobs more quickly, and in cost savings

    (efficiency gains) accruing to the government.

    As the relevant Minister, David

    Kemp, explained at the launch of the JN in

    1998: ‘An overhaul of employment services

    was needed because the old system

    under the CES didn’t get enough people

    into jobs. It tended to ‘manage’ unemployed

    people rather than place them in

    jobs. Job Network focuses on results. Job

    Network members will be paid when they

    place a job seeker in a job for a sustained

    period of time.’


    Job Network contracts

    The first round of competitive tendering

    for JN contracts took place in mid-1997.

    Over 1,000 organisations submitted a total

    of 5,300 tenders for $1.7 billion worth of

    contracts. DEWR vetted all tenders on

    quality and those that met the required

    standard were then ranked by price.

    Agencies submitting tenders could decide

    which of the three core service functions

    they wished to provide (it was not a

    requirement that all agencies offer

    Intensive Assistance as well as Job

    Matching and Job Search Training), and

    some bidders offered specialized services

    aimed at certain kinds of clients.

    Contracts were awarded to the agencies

    promising to deliver the required services

    at the lowest price.

    Contracts were awarded to 223 successful

    bidders (a slight drop compared with

    the 243 agencies involved in contracting

    Intensive Assistance services under

    Working Nation

    Half of these were private

    (for-profit) contractors, 44% were

    community sector organizations, and the

    remaining 6% were government agencies

    (including Employment National, which

    ended up with a 37% market share). 79%

    of successful bidders at ESC1 had previously

    held government contracts for managing

    and administering labour market

    programs under the old

    Working Nation



    Following an early DEWR evaluation of

    how the JN was operating, a second round

    of contract tenders (ESC2) took place in

    mid-1999 when all first-round contracts

    expired. The Department was concerned

    that some service providers were neglecting

    the most difficult of their three core functions

    – Intensive Assistance to prepare the

    most disadvantaged job seekers to return to

    employment – and were instead making

    easy money finding rapid placements for

    job seekers who would probably have

    found jobs anyway if left to themselves. To

    counter this, ESC2 contracts required

    providers to agree an individual ‘Preparing

    for Work’ plan for anyone who had been

    unemployed for 13 weeks. This would

    identify any barriers to employment and

    put in place a strategy for overcoming

    them. Some observers thought this sig-

    It was hoped that an incentivised system would result

    in more unemployed people getting jobs more quickly

    nalled a move back towards an emphasis

    on training and labour market subsidies

    which had been scrapped when



    was abolished.

    At ESC2, the original 29 Job Network

    regions were reduced to 19 in an attempt

    to make contracts more financially viable.


    These regions were then divided into a

    total of 137 Employment Service Areas

    (ESAs) so bidders could target their services

    at particular populations (e.g. Indigenous

    people in remote areas). ESAs covered

    between 100 and 18,000 job seekers,

    but most catered for between 4,000 and

    6,000. DEWR wanted to recruit two or

    three competing providers in each ESA,

    but in some remote areas, no bids were

    received, and agency arrangements had to

    be negotiated outside the JN bidding


    In the end, 168 bidders (including the

    government-owned Employment National)

    were given contracts worth $3 billion at


    The decline in the number of JN

    providers as compared with ESC1 was

    mainly due to the withdrawal of agencies

    offering only Job Matching and Job Search

    Training, for 80% of JN expenditure went

    on Intensive Assistance, and without this

    money, most providers found participation

    in the Job Network was not financially

    worthwhile (by the time of ESC3, held

    in 2003, every JN provider was required

    to offer Intensive Assistance).


    although the number of different providers

    fell, the number of sites at which they

    operated rose to 2,010 (a 54% increase

    over ESC1). The Job Network was therefore

    expanding at the same time as it was


    Incumbency proved a strong advantage

    when it came to getting a contract at this

    second round of bidding, for 87% of

    ESC1 contractors succeeded at ESC2.

    The market share achieved by not-forprofit

    community groups increased from

    30% to 45%, and for-profit companies

    increased their share from 33% to 47%.

    These gains were made at the expense of

    Employment National whose share fell

    from 37% to 8%.

    Incumbency proved even more important

    at the third round of contracts (ESC3)

    in 2003, for sixty per cent of business was

    now reserved for existing high performers.

    These were identified by ‘star ratings’ based

    on DEWR’s estimation of the value added

    by each JN member over the previous two

    years. To measure ‘value added’, DEWR

    applied a complex formula that took

    account of the characteristics of clientele

    and of the local labour market as well as

    raw job outcomes data. Star ratings had

    been designed to help job seekers select the

    best provider, but as we saw earlier, few of

    them make an active choice. But the ratings

    system also fulfils a second key function

    by allowing DEWR to assess the quality

    of different JN providers when their

    contracts come up for review. With the

    abandonment of price competition from

    2003 onwards (see below), these quality

    ratings have become vital in securing

    future contracts.

    At ESC3, the 86 best performing JN

    members were ‘invited to treat’, and 74 of

    them took up the offer. Only 40% of business

    was put out to open tender, and in the

    event, most of this was mopped up by

    existing JN members. Only seven new

    providers joined the Job Network in 2003,

    and they won just 1.5% of the market

    (four existing providers who had not been

    invited to treat also won contracts in the

    open competition).

    In all, 2,100 bids were received in

    2003, an average of 15 in each ESA, and

    109 organisations eventually won contracts.


    As in the previous contract

    rounds, therefore, the number of agencies

    shrank, and so this time did the number

    of JN offices (down to 986 sites, with 110

    offering specialist services such as those

    for Indigenous job seekers). This reduction

    in the number of JN outlets was,

    however, balanced by the licensing of 266

    14 Eardley T, “Outsourcing

    Employment Services: What

    have we learned from the Job

    Network?” Paper to Centre for

    Applied Economic Research

    conference on the Economic

    and Social Impacts of

    Outsourcing, University of New

    South Wales, pp 6, 4-5

    December 2003

    15 The larger number of labour

    market regions and providers in

    ESC1 was intended to promote

    competition, but was reduced to

    improve the economic viability of

    providers (Productivity

    Commission, “Independent

    Review of the Job Network”,

    report no 21, pp 4.8–4.9, June


    16 Bruttel O,


    Competition in a Public Service

    Market: The Job Network in an

    International Perspective


    Centre for Labour Market

    Research Discussion paper, no

    3, 2003

    17 Productivity Commission,

    “Independent Review of the Job

    Network”, report no 21, June


    18 Bruttel O,


    Competition in a Public Service

    Market: The Job Network in an

    International Perspective


    Centre for Labour Market

    Research Discussion paper, no

    3, 2003

    The experience of contracting out employment services in Australia


    • 19

    19 Australian National Audit

    Office, “DEWR’s Oversight of

    Job Network Services to Job


    Audit Report

    , no 51,


    20 Bruttel O,


    Competition in a Public Service

    Market: The Job Network in an

    International Perspective


    Centre for Labour Market

    Research Discussion paper, no

    3, 2003

    21 Department of Employment

    and Workplace Relations,


    Network Employment Services

    Contract 2003-2006

    , March


    22 Full details of 2006-09 Job

    Network contracts can be found







    23 Productivity Commission,

    “Independent Review of the Job

    Network”, report no 21, pp xxii,

    June 2002

    24 Productivity Commission,

    “Independent Review of the Job

    Network”, report no 21, table

    6.1, June 2002

    25 Productivity Commission,

    “Independent Review of the Job

    Network”, report no 21, pp 6.23-

    6.25, June 2002


    Paying for success

    organisations offering only

    Job Matching

    services (see below for details). These new

    licensed agencies share information on

    vacancies and job seekers with the 109

    full JN members, so the total capacity of

    the job matching system was expanded

    substantially. According to a National

    Audit Office report in 2005, 250 localities

    now have an employment office

    where there was not one previous to



    ESC3 continued the trend towards

    market concentration. Existing providers

    consolidated their position and got bigger

    while outsiders found it increasingly difficult

    to break in. After 2003, the top five

    JN providers had 39% of market share,

    and the top ten had 55%. The average

    market share for those outside the top ten

    was 0.5%.

    The two largest JN members

    were both community organizations: the

    Salvation Army (with a 15% market

    share) and Mission Australia (8%).


    profit providers between them

    accounted for 54% of Job Network members

    (with a 50% market share), while

    commercial providers made up 43% (with

    a 47% market share). Following the closure

    of Employment National, the public

    sector shrank to just 3% (with 3% market


    A fourth round of contract tendering

    was scheduled for 2006, but it never took

    place. When ESC3 expired in 2006, all

    existing contracts were extended for three

    more years, except where performance

    (measured by star ratings) was deemed


    This trimmed the number of JN members

    from 109 to 103. Rolling local area

    tenders have now been introduced so

    poor performers can be replaced

    throughout the extended contract period

    at six monthly reviews. It is unclear

    whether these arrangements will now

    become permanent, or if another round

    of open tendering will take place in 2009

    when the current extension period ends.


    Learning by experience

    By the time the third round (ESC3) of

    contracts took place in 2003, substantial

    changes had been introduced into the

    organization and management of the Job

    Network. These followed the publication

    in 2002 of an independent review by the

    Productivity Commission (PC), a statutory

    authority charged with analyzing and

    evaluating the delivery of government services.

    The PC review found that the ‘new

    framework has many advantages and

    should be retained.’

    It thought the purchaser-

    provider model made the objectives

    of employment services providers clearer

    (by setting out contractual requirements in

    tender documents), and that incentives for

    achieving these objectives had been

    strengthened. There had also been cost

    efficiency gains. The new system was

    cheaper than earlier programs (like

    Working Nation

    ), and it gets more

    favourable feedback from job seekers (e.g.

    a 1999 survey recorded a ‘strongly positive

    view overall by job seekers about the JN’).


    Employer satisfaction too was running

    quite high (on balance, employers preferred

    the new system to the old CES),


    although only 20% of employers in 1999

    had actually used the Job Network to

    recruit labour. The PC also forecast that

    outcomes would improve over time as poor

    performers lost their contracts. It quoted

    evidence that the best JN performers were

    achieving outcome rates 12 percentage

    points higher than the average. The star

    ratings system meant they would increase

    their market share over time, thereby driving

    up the average level of performance of

    the network as a whole.

    Not everything was positive, however.

    There was still a problem of ‘perverse

    incentives’ with hard cases being ‘parked’

    on benefits while agencies focused their

    efforts on easier-to-place clients in order to

    maximize outcome payments. There were

    also substantial ‘deadweight costs’ since

    26 Productivity Commission,

    “Independent Review of the Job

    Network”, report no 21, pp xxviii,

    June 2002

    27 Productivity Commission,

    “Independent Review of the Job

    Network”, report no 21, pp 5.20,

    June 2002

    28 Star ratings are issued every 6

    months (at each ‘contract milestone’)

    on a rolling 2 year basis.

    70% of providers score 3 stars or

    better (5% get 5 stars, 4% get 1

    star, indicating ‘room for improvement’

    (Department of Employment

    and Workplace Relations,


    Network Star Ratings

    , Australian

    Government July 2007). Stars are

    awarded based on job placement

    and outcome data. Where there

    are significant divergences in

    stars awarded to different providers

    within a single Employment

    Services Area, business is redirected

    from weak to stronger performers.

    29 Productivity Commission,

    “Independent Review of the Job

    Network”, report no 21, pp xxxiv,

    June 2002

    30 In 2007, initial job placements

    attracted a fee of $165 – $385

    (depending on duration of previous

    unemployment). Intensive

    Assistance recipients placed in

    jobs lasting 13 weeks pay $550 –

    $4,400, and in jobs lasting 26

    weeks (a ‘final outcome’) they

    attract an additional fee of $825 –

    $2,200 (again depending on their

    prior length of unemployment).

    Where IA people are placed in

    jobs which reduce but don’t terminate

    their welfare benefits, fee

    paid = $550 – $1,100. See

    Department of Employment and

    Workplace Relations, Job

    Network Star Ratings, Australian

    Government, pp 5, July 2007.

    31 Senate Community Affairs


    A Hand Up, Not a

    Hand Out

    , Commonwealth of

    Australia, pp 67, 2004

    The experience of contracting out employment services in Australia


    • 21

    many of the successful ‘outcomes’ achieved

    by JN members would, in the view of the

    PC, have occurred even without the intervention

    of JN providers.

    Nor was the PC impressed by the JN’s

    record in getting unemployed people back

    into work through Intensive Assistance

    programs. The report accepted that active

    labour market programs around the world

    generally achieve little success, and that

    Working Nation

    had achieved little in this

    regard either. Nevertheless, it concluded

    that Australia’s new employment services

    system was not having the impact that had

    been hoped: ‘Job Network programs have

    so far probably only had modest effects on

    job seekers’ chances of gaining employment.’


    It also took issue with DEWR’s

    claim in its own second stage evaluation of

    the JN that Intensive Assistance programs

    were achieving a 10% net impact on

    unemployment of participants. According

    to the Productivity Commission, this estimate

    was ‘significantly overstated’.


    In the light of these problems, several

    key changes were implemented in the

    organization of the Job Network in time

    for the third contract round (ESC3).

    First, price competition was abandoned

    as DEWR returned to the fixed price tenders

    which had operated under



    . This change implicitly recognized

    that some providers had been winning

    contracts with low tenders, but had then

    been maximizing their outcome payments

    by focusing their efforts on job-ready customers

    (who actually needed little assistance)

    while neglecting the harder cases.

    In 2003, the Department specified in

    detail the services it wanted to buy, and it

    fixed the price for each service, choosing

    among tenders according to its judgement

    of the quality of the service each agency

    could provide. This, of course, favoured

    existing providers who had already established

    reputations and enjoyed strong star

    ratings. Indeed, we have seen that the best

    performers were issued with invitations to

    treat and did not even have to subject

    themselves to the competitive tender



    Secondly, fees-for-service and outcome

    payments were re-weighted to offer higher

    returns for sustained outcomes achieved by

    long-term unemployed job seekers.

    Commencement fees (paid when job seekers

    were referred to a JN member) were

    scrapped. Instead, JN providers are now

    paid a fee when they accept someone onto

    Intensive (and Customised) Assistance. If

    the client is subsequently placed in a job

    for at least 13 weeks, they get an ‘interim

    primary outcome’ payment (a ‘full primary

    outcome’ payment is made after 26 weeks,

    and smaller ‘secondary outcome’ payments

    are made when they place clients in education

    courses or in part-time jobs where

    they still draw some welfare benefits).

    In ESC2, only 15% of commencements

    resulted in a job lasting 13 weeks, and

    another 8% resulted in an ‘interim secondary

    outcome.’ Most (about 70%) of the

    income of JN members therefore came

    from non-outcome-based commencement


    Since 2003, however, this balance

    between fees and outcome payments has

    shifted, and the rewards for work with

    long-term unemployed clients have been


    Under ESC3, someone unemployed

    for 3 years who gets a job lasting 26

    weeks would earn his or her JN provider a

    $6,600 outcome payment, which compares

    with an average of $4,500 earned

    over the previous 3 years in fee for service

    and other (Job Seeker Account) payments.


    As a result of this restructuring of fees and

    payments, the balance of income between

    JN providers are now paid a fee when they accept

    someone onto Intensive Assistance… If the client is

    subsequently placed in a job for at least 13 weeks,

    they get an ‘interim primary outcome’ payment

    fees and outcome payments seems to have

    shifted (one of the leading non-profit

    providers, CentaCare, operated by Catholic

    Social Services, reports, for example, that

    47% of its revenue in 2006 came from

    placements and outcomes).

    Fees paid in

    respect of ‘Customised Assistance’ (offered

    when claimants have been unemployed for

    12 months) were also increased (in 2003-

    04, 48% of all service fees were paid in

    respect of customized assistance).


    A third change made in 2003 was the

    introduction of Job Seeker Accounts.

    These consist of money set aside by

    DEWR to cover expenditure incurred by

    JN providers on behalf of their clients in

    helping them get work. These accounts are

    intended to cover things like purchase of

    work clothing or equipment, payment of

    travel costs, course fees or payment of subsidies

    to employers (which again looks very

    much like a return to the days of



    ). This initiative was designed to

    encourage JN members to put extra effort

    into finding work for their toughest cases,

    but we shall see later that it has generated

    new problems, with JN members complaining

    about the detailed level of scrutiny

    their claims receive from DEWR, and

    some suggesting that their rivals are ‘buying

    jobs’ by blowing large sums on shortterm

    employer subsidies.

    Fourthly, to meet continuing concerns

    about poor service quality, ESC3 introduced

    an Employment Services Code of

    Practice and built a Job Network Service

    Guarantee into the Employment Services

    Contract: ‘The Code and Service Guarantee

    require that JNMs deliver a guaranteed set

    of services in accordance with specified

    principles and processes in a manner that is

    sensitive to the job seeker’s culture, circumstances

    and background.’

    For example,

    the third Employment Services Contract

    stipulated that Job Network members

    should contact and meet face-to-face with

    each job seeker once every fortnight during

    their first period of

    Customised Assistance


    This meant a total of twelve service contacts

    were required in the course of a sixmonth

    period of Customised Assistance.

    Breach of this contractual requirement

    could result in a provider losing star ratings,

    and therefore being disadvantaged at

    any later period of contract renewal.

    However, stipulations like this can only set

    a quality baseline (there is no incentive to

    exceed the minimum quality standards),

    and they depend on a high level of surveillance

    to see if they are being followed.

    A fifth change saw the introduction of

    licensing for job matching agencies wishing

    to link to the Job Network. ESC3

    contracts for the first time required all JN

    members to offer Customised Assistance.

    The exclusion of agencies offering only

    job matching functions was compensated

    by these new licensing agreements under


    Job Placement Only


    were given access to the government’s new

    JobSearch database listing around 90,000

    vacancies, and were paid a fee of up to

    $550 if they succeeded in placing a job

    seeker in work. In return, these organizations

    were required to list all their nonexecutive

    vacancies on the database, for

    other providers to share. At the inception

    of this scheme in 2003, 375 licenses were

    awarded (including 109 to JN members).


    The Active Participation Model

    All these changes in 2003 were important,

    but the most significant of all was the

    adoption of an integrated ‘Active

    Participation Model’ for managing job

    seekers. This was key to the new rule that

    32 Excluding payments for Job

    Seeker Accounts – see Murray P,

    “A Job Network for Job


    Catholic Social

    Services Discussion Paper

    , pp

    25, November 2006

    33 DEWR estimates that service

    fees of $250 million were paid in

    2003–04 for provision of

    Intensive Assistance (customized

    assistance), and that

    this represented 48 per cent of

    total service fees, quarterly payments

    and one-off payments

    (excluding outcome payments)

    to JN members in that financial

    year. Outcome payments for

    intensive support job seekers

    commencing and remaining in

    employment or education for

    periods of at least 13 or 26

    weeks duration totaled a further

    $171 million (Australian National

    Audit Office, “DEWR’s Oversight

    of Job Network Services to Job


    Audit Report

    , no 51,

    pp 127, 2005)

    34 Australian National Audit

    Office, “DEWR’s Oversight of

    Job Network Services to Job


    Audit Report

    , no 51,

    pp 54, 2005

    35 Bruttel O,


    Competition in a Public Service

    Market: The Job Network in an

    International Perspective


    Centre for Labour Market

    Research Discussion paper, no

    3, 2003; Wong P, “Reward for

    effort: Meeting the participation

    challenge”, Australian Labor

    Party Discussion Paper,

    November 2006


    Paying for success

    If they are still jobless after six months, FJNE job

    seekers under the age 50 are required to participate in

    a six-month

    Work for the Dole


    job seekers must remain with the same JN

    member throughout any single unemployment

    episode. This not only made it easier

    to monitor their progress over time, but it

    also meant a single plan for case management

    could be devised which all JN members

    would have to follow. In principle,

    this meant it should no longer be possible

    to ‘cream off ’ the easy cases while ‘parking’

    the difficult ones.

    It also meant



    Intensive Assistance

    could run together

    in a continuum of increasing intensity as

    time passes. This new ‘service delivery

    continuum’ is outlined in Figure 2.

    There are two main classes of job seekers

    who can access JN services:

    the ‘Fully

    Job Network Eligible’ (FJNE), who are

    registered as looking for work and who are

    receiving income support (all job seekers

    aged 15 to 20 and not in full-time education

    or training are also included in this


    ; and ‘Job Search Support Only’

    (JSSO) job seekers, who are looking for

    work but are not FJNE (this second category

    might include, for example, single

    parents with pre-school age children or

    Disability Support Pensioners who are

    seeking employment even though they are

    not required to work).

    Individuals in

    both categories are eligible to receive

    JobSearch Support services, which means

    they get access to the Australian JobSearch

    national vacancy database, where their CV

    or ‘vocational profile’ is automatically

    matched with available vacancies on a daily

    basis. But only FJNE job seekers qualify

    for the additional services offered in the

    Active Participation Model


    If an FJNE job seeker remains unemployed

    after three months, they move into

    Intensive Support.

    This begins with a period of Job Search

    Training that includes help with writing

    job applications, interview skills, and confidence

    building. Job search support continues

    during this period, so clients continue

    to search actively for employment.

    If they are still jobless after six months,

    FJNE job seekers under the age 50 are

    required to participate in a six-month

    Work for the Dole

    project (or other Mutual

    Obligation activity). Participation (which

    normally takes up two days per week) is a

    condition of continuing to receive income

    support payments.

    If they are still unemployed after 12

    months, job seekers move into Intensive

    Support customised assistance (ISca). ISca

    is a six-month period when job seekers

    receive substantial, intensive and person-

    36 The CEO of Job Futures

    noted at a recent conference

    that keeping job seekers with

    the same provider throughout

    means providers have to pay

    more attention to difficult cases

    rather than just parking them.

    Under ESC3 ‘providers who do

    not successfully achieve this

    goal not only find it difficult to

    remain financially viable but also

    find it increasingly difficult to

    keep their service provision contract

    with DEWTR’ (Dudley S,

    Not Just Any Job

    , Paper to

    Australian Social Policy

    Conference, University of New

    South Wales, pp 3, July 2005).

    37 Australian National Audit

    Office, “DEWR’s Oversight of

    Job Network Services to Job


    Audit Report

    , no 51,


    38 FJNE Job seekers do not

    receive income support until

    they have attended their first

    meeting with their Job Network


    39 Since 2005, all DSP and PP

    claimants have been allowed to

    volunteer for JN services and

    register directly with a JN

    provider if they want to.

    40 Australian National Audit

    Office, “DEWR’s Oversight of

    Job Network Services to Job


    Audit Report

    , no 51,


    The experience of contracting out employment services in Australia


    • 23

    Job Search


    Mutual Obligation Customised


    Job Search Support Intensive Support

    Referral from


    Month 1 Month 3 Month 6 Month 12

    Service delivery continuum

    Figure 2: Stages in the Active Participation Model


    Source: ANO

    Note: Job seekers who are the most disadvantaged in the labour market receive immediate access to ISca

    alised assistance. It is tailored to individual

    needs and includes training, work experience

    in a subsidized placement, or referral

    to a language or literacy and numeracy

    training program. At this stage, JN

    providers may also draw on the Jobseekers

    Account for funds to support these various


    Job seekers who are identified by

    Centrelink as most disadvantaged skip the

    first 12 months of this continuum and

    move straight into ISca.

    In the year ending June 2005, 144,300

    job seekers participated in JobSearch

    Training, 298,900 had Customised

    Assistance, and 148,000 did a Mutual

    Obligation activity, of whom 81,900 did a

    Work for the Dole placement.


    More employment outcomes at

    less cost?

    Not even the critics of the JN reform deny

    that it has succeeded in one of its key

    objectives – to reduce costs. DEWR’s 2001

    Net Impact Study claimed that the JN

    Intensive Assistance program had achieved

    better net job outcomes (i.e. after factoring

    out cases where people would have

    achieved employment even without assistance)

    compared with

    Working Nation

    , but

    this claim was later challenged by the

    Productivity Commission, among others.

    What was not challenged, however, was

    the claim that the system was delivering

    outcomes more cheaply than the



    program that preceded it.

    According to DEWR, each person

    placed in a job was costing between

    $5,000 and $6,000, compared with

    $10,000 to $16,000 under the previous


    Each net employment

    outcome from Intensive Assistance programs

    was on average costing $22,000

    using the JN, compared with $35,100

    under the old system.

    The methodology

    used to calculate net outcomes may have

    been suspect, but given that the same

    methodology was applied to both the

    ‘before’ and ‘after’ data, the cost savings

    were clearly real.

    In its 2002 report, the Productivity

    Commission noted that the aggregate cost

    of labour market programs fell by half in

    the first four years of the Job Network

    (from $3.7bn to $1.3bn), and that this saving

    was achieved with no apparent change

    to unemployment levels.

    Similarly, Tony

    Eardley of the Social Policy Research

    Centre accepts that, ‘The Network does

    seem to have made efficiency gains compared

    to earlier employment services in

    terms of public expenditure,’ although he

    immediately adds that the changes have led

    to ‘only modest improvements, if at all, in

    macro employment outcomes.’


    DEWR continues to insist that the JN is

    not only cheaper than its predecessor – it is

    also delivering better employment outcomes.

    For its 2005 Net Impact Study,

    DEWR used a new methodology for estimating

    net job impacts based on a complex

    logistic regression modelling procedure

    which it claimed was the one approved by

    both the Productivity Commission and

    OECD. This factored out compliance

    effects (where jobseekers leave the welfare

    system as soon as they are required to begin

    some prescribed activity)

    as well as deadweight

    losses (people who would have

    found work anyway) and substitution

    effects (people who find a job but push

    someone else into unemployment as a

    result). And the results were impressive.

    DEWR found that three months after

    completing the relevant program, 55% of

    those in JobSearch Training were in

    employment, as were 46% of those who

    did Customised Assistance and 32% of

    those who did Work for the Dole.


    outcomes were calculated by comparing

    these actual outcomes with those predicted

    for people not having done these programs.


    DEWR reported a positive 11%

    net impact on employment outcomes for

    those doing Job Search Training, a 10% net

    41 Department of Employment

    and Workplace Relations,

    Customised Assistance, Job

    Search Training, Work for the

    Dole and Mutual Obligation – A

    Net Impact Study, Department of

    Employment and Workplace

    Relations, report 1/2006, April


    42 Department of Employment

    and Workplace Relations, Job

    Network Evaluation Stage 3:

    Effectiveness Report,

    Department of Employment and

    Workplace Relations, Report

    1/2002, pp 4, 2002

    43 Australian National Audit

    Office, “DEWR’s Oversight of

    Job Network Services to Job

    Seekers”, Audit Report , no 51,


    44 Productivity Commission,

    “Independent Review of the Job

    Network”, report no 21, pp 5.24,

    June 2002

    45 Eardley T, “Outsourcing

    Employment Services: What

    have we learned from the Job

    Network?” Paper to Centre for

    Applied Economic Research

    conference on the Economic

    and Social Impacts of

    Outsourcing, University of New

    South Wales, pp 10-11, 4-5

    December 2003,

    46 Eardley T, “Outsourcing

    Employment Services: What

    have we learned from the Job

    Network?” Paper to Centre for

    Applied Economic Research

    conference on the Economic

    and Social Impacts of

    Outsourcing, University of New

    South Wales, pp 9, 4-5

    December 2003. Suggests the

    biggest portion of net impact

    was compliance effect.

    47 Department of Employment

    and Workplace Relations,

    Customised Assistance, Job

    Search Training, Work for the

    Dole and Mutual Obligation – A

    Net Impact Study

    , Department

    of Employment and Workplace

    Relations, report 1/2006, pp 7

    table 1, April 2006

    48 Department of Employment

    and Workplace Relations,

    Customised Assistance, Job

    Search Training, Work for the

    Dole and Mutual Obligation – A

    Net Impact Study

    , Department

    of Employment and Workplace

    Relations, report 1/2006, pp 8

    table 2, April 2006


    Paying for success

    impact for those undergoing Customised

    Assistance, and a 9% net impact for those

    undergoing a Mutual Obligation activity

    (7% in the case of Work for the Dole).

    Customised Assistance participants were

    also more likely than a comparable control

    group to have left income support, and to

    have done so more quickly.

    Applying its new methodology retrospectively

    to its own, earlier, JN net impact

    studies, DEWR found that Intensive/

    Customised Assistance had achieved a net

    impact of just 0.6% in 2001, rising to

    6.2% in 2002 and to 10.1% in 2005. It

    claimed this dramatic improvement could

    be explained by the adoption of the Active

    Participation Model, including the new fee

    structures and the Job Seeker Account.

    The Department also compared these

    outcomes with those achieved by roughly

    comparable interventions overseas. It

    found, for example, that the 10 percentage

    point net impact of Customised Assistance

    compared very favourably with the 5 percentage

    point net impact achieved for similar

    long term unemployed clients in the

    UK who went through the New Deal.


    concluded: ‘These impacts are…equal to

    or better than those of high performing

    programs internationally.’


    Looking at the raw figures, in 2006-07

    the Job Network placed 186,400 longterm

    unemployed job seekers in jobs lasting

    at least 13 weeks (all these were either

    disadvantaged job seekers or people unemployed

    for more than 3 months). Another

    6,600 on Intensive Support got education

    outcomes. The 2006-07 placements figure

    is a new record compared to 2000, when

    only 47,200 long-term/disadvantaged job

    seekers were placed in jobs.


    Problems and criticisms

    Notwithstanding these impressive results

    on cost efficiency and (it seems) on net job

    impacts, the contracted-out employment

    services system continues to attract criticism,

    and there clearly are a number of

    unresolved problems which will need to be

    addressed in the future.

    Consumer choice

    The original intention that Job Seekers

    should choose their service provider never

    really worked – most simply get allocated a

    JN provider by Centrelink after their needs

    have been assessed. Part of the reason for

    this may be inadequate information – the

    National Audit Office found that

    Centrelink offices were failing to give job

    seekers enough information about different

    JN members to allow them to make an

    informed choice.

    But it also undoubtedly

    reflects the characteristics of the clients

    themselves. DEWR’s core client group –

    ‘disadvantaged job seekers’ – are not people

    who are used to making active choices and

    decisions in their lives, and it always was a

    faint hope that they would put time and

    effort into selecting a JN provider, rather

    than simply leaving it up to Centrelink to

    give them one.

    Not only do they fail to choose their

    own provider, but since the introduction of

    the Active Participation Model, job seekers

    are not allowed to switch between

    providers, except under very limited circumstances,

    so dissatisfied consumers can

    no longer ‘vote with their feet.’ All of this

    undermines one of the core advantages of a

    market model, which is that producers

    have to compete to attract consumers. In

    reality, even though unemployed clients

    are still called ‘customers,’ the real customer

    for JN members is and always was


    The absence of real consumer sovereignty

    in this quasi-market means there has

    been a recurring problem of rewarding or

    penalising the quality of the service being

    provided to job seekers by different JN

    members. When consumers do not choose

    between providers, other, generally inferior,

    ways have to be found to judge the

    quality of service they are being given.

    49 Department of Employment

    and Workplace Relations,

    Customised Assistance, Job

    Search Training, Work for the

    Dole and Mutual Obligation – A

    Net Impact Study

    , Department

    of Employment and Workplace

    Relations, report 1/2006, pp 9,

    April 2006

    50 Department of Employment

    and Workplace Relations,

    Customised Assistance, Job

    Search Training, Work for the

    Dole and Mutual Obligation – A

    Net Impact Study,

    Department of

    Employment and Workplace

    Relations, report 1/2006, pp 4,

    April 2006

    51 ‘Job network continues to

    break records’ Media Release,

    Minister for Workforce

    Participation, 27 July 2007;

    Karvelas P, “Agencies put More

    Jobless into Work”,



    , 16 January 2006.

    52 The audit assessed the quality

    of the information provided to

    job seekers by Centrelink at four

    centres, to see whether it was

    good enough for job seekers to

    make an informed choice of JN

    member. It found: ‘The provision

    of information and information

    products at information seminars

    and in information display areas

    was variable, often poor, and did

    not meet minimum requirements

    specified in the Business

    Partnership Arrangement. Many

    information seminars were not

    conducted prior to the job seeker

    making a choice of their JNM,

    and some job seekers did not

    attend a seminar at all’

    (Australian National Audit Office,

    “DEWR’s Oversight of Job

    Network Services to Job


    Audit Report

    , no 51,


    The experience of contracting out employment services in Australia


    • 25

    Sometimes this has been done by surveys

    (DEWR maintains that job seeker satisfaction

    levels are high, although this has been

    disputed by critics); the number of complaints

    can also be measured, although

    there are doubts about how effective the

    complaints procedures are in allowing

    clients to express dissatisfaction.

    Codes of

    practice establish minimum standards, but

    their effectiveness depends on how well

    they are policed, and they fail to distinguish

    varying quality standards above the

    minimum level.

    Increasing complexity and administrative costs

    Given that the immediate consumers of

    the service do not discriminate between

    different providers, DEWR has to take on

    this role itself. This has created a recurring

    tension between the desire to give JN

    providers autonomy to run their own businesses

    and to innovate in response to varying

    local circumstances, and the need to

    monitor them from the centre to ensure

    they are fulfilling their contractual obligations

    and that public money is not being


    The CEO of one provider says there was

    very little prescription from the centre in

    the first 3 or 4 years of the Job Network,

    but that criticisms from the Productivity

    Commission and elsewhere led to more

    DEWR control, so that today there is, ‘a

    high degree of prescription of process combined

    with an awful lot of detailed administration’.


    This is borne out by Catholic

    Social Services, which runs JN member

    Centacare. It complains of micro-management

    from DEWR, citing cases where tiny

    claims for reimbursement from the Job

    Seekers Account have been subject to

    intense scrutiny. According to Catholic

    Social Services, the old ‘bureaucratic mindset’

    of the CES has re-emerged in recent



    Part of the administration burden carried

    by JN members has been the cost and

    time involved in preparing tenders,


    this issue has to some extent been

    addressed by scrapping the 2006 contract

    round and renewing all existing contracts.

    Nevertheless, a recent Labor Party review

    of the Job Network claims feedback from

    JN members is still raising concerns about

    the time and money spent on administration,

    and it also draws attention to the cost

    of DEWR’s own administration ($411.5m

    spent on administering labour market programs

    in 2006 – 18% of its total budget

    for these programs).

    Similarly, Catholic

    Social Services points to the apparent inefficiency

    in having 1,200 DEWR staff in

    Canberra overseeing local management in

    just 1,100 offices.


    Attenuation of competition

    We have seen that by ESC3, it had become

    very difficult for new entrants to win JN

    contracts, for 60% of business was reserved

    for existing providers and most of the

    remaining 40% ended up with them too.

    The advantages of incumbency were reinforced

    by the decision in 2003 to merge

    Intensive Assistance, job search training

    and job matching in the new Active

    Participation Model, for this knocked out

    providers who could not offer IA services

    (although the new licensing system did

    bring in new suppliers of job matching

    services). The result was a dwindling in

    the eligible pool of service providers and a

    concentration of business in a smaller

    number of JN members. This can clearly

    be seen by comparing the number of bids

    received at ESC1 (5,300) with the number

    received at ESC3 (2,100).

    Coupled with the increasing direction

    from the centre and the closing down of

    effective competition in the tendering

    process, this concentration of JN membership

    suggests that ten years after the old

    CES model was scrapped, the new system

    has begun to look increasingly like the old

    one. As Tony Eardley suggests: ‘The JN has

    swung back to being highly regulated and

    government-controlled and only open to

    Paying for success


    53 The 2002 PC review recorded

    a ‘strongly positive view overall

    by job seekers about the JN’

    (Productivity Commission,

    “Independent Review of the Job

    Network”, report no 21, pp 6.6

    table 6.1, June 2002), and found

    complaints running at less than

    0.1% (pp 6.15). However, the

    National Audit Office report

    found job seeker awareness of

    the Code, Service Guarantee

    and associated complaints

    mechanisms remains ‘very low,’

    and thought that this might prevent

    them from complaining

    about poor service (Australian

    National Audit Office, “DEWR’s

    Oversight of Job Network

    Services to Job Seekers”,



    , no 51, pp 66, 2005).

    54 David Thompson, CEO of

    Jobs Australia, interviewed on

    Counterpoint (ABC Radio

    National) Transcript: “


    Network Burn Out?

    ”, 31 July


    55 Murray P, “A Job Network for

    Job Seekers”,

    Catholic Social

    Services Discussion Paper

    , pp

    20, November 2006

    56 Productivity Commission,

    “Independent Review of the Job

    Network”, report no 21, pp xxxiii,

    June 2002

    57 Wong P, “Reward for effort:

    Meeting the participation challenge”,

    Australian Labor Party

    Discussion Paper

    , pp 88,

    November 2006

    58 Murray P, “A Job Network for

    Job Seekers”,

    Catholic Social

    Services Discussion Paper

    , pp

    20, November 2006

    new entrants to a limited degree. It now

    makes even less sense than before to think

    of it as a genuine ‘contestable market.’


    Quantity, not quality

    Part of the problem that DEWR has

    encountered in monitoring and regulating

    the performance of JN members is that

    objectives such as speed, sustainability and

    equity of service cannot easily be measured.

    The result is that easily-measured outcomes

    (the quantity of successful job placements)

    are rewarded, while other objectives which

    DEWR says it wants to encourage (particularly

    those to do with service quality) are not.

    In a recent, well-publicised and scathing

    report, Catholic Social Services claimed that

    although service quality is emphasized in JN

    contracts, it cannot be measured and is

    therefore not enforced. It claims that:

    ‘Ultimately the only real reward is for quantity

    of outcomes – Government expectations

    with respect to outcome quality and service

    quality are platitudes… providers who pursue

    [quality objectives] receive the same unit

    outcome fees and star ratings as those who

    give little attention to them… Job seekers

    are frequently met by a one size fits all service

    from providers focusing on quick fix and

    process oriented solutions (such as outcome

    buying) which often result in a mismatch

    between a job seeker and a job.’


    This complaint that quality considerations

    are being neglected received support

    from the 2005 National Audit Office

    report which found that levels of contact

    between JN providers and job seekers

    rarely met contractual specifications. Job

    seekers receiving Customised Assistance

    are meant to meet with their service

    provider once a fortnight, but the Audit

    revealed an average of only 6.7 appointments

    per job seeker over a 22 week period,

    when eleven appointments were scheduled.

    The report concluded: ‘The nature

    and level of problems identified raises concerns

    about whether assistance is actually

    intensive and personalised’.


    The Audit report went on to note that,

    while the introduction of the Code of

    Practice and Service Guarantee was a positive

    step, these documents contain service

    commitments that are ‘largely subjective.’


    DEWR’s ‘Quality Key Performance

    Indicator,’ introduced at ESC3, is based on

    the Department’s satisfaction that services

    have been delivered in compliance with the

    Code and the Service Guarantee, but

    according to the National Audit Office,

    this assessment has to be based on ‘subjective

    judgements by DEWR contract managers,

    because most of the service commitments

    in the Code and Service Guarantee

    are not clear, measurable statements of

    service requirements.’

    A related problem is that DEWR’s

    Quality Indicator takes the form of a simple

    pass/fail hurdle. The default is that all JN

    providers pass unless a clear reason has been

    identified for applying a fail. The National

    Audit Office found that this pass/fail system

    did not allow DEWR to assess changes in

    performance over time, nor to compare

    variations among those who pass: ‘DEWR’s

    ability to gain assurance that job seekers

    receive high quality services from JNMs is

    limited by the lack of objective and measurable

    performance indicators relating to

    DEWR’s specified service standards.’


    Perverse incentives and unethical practices

    From the outset, DEWR has struggled to

    get the incentives right. The 2002

    Productivity Commission report found

    problems of ‘creaming’ and ‘parking’, and

    although the Active Participation Model

    attempted to resolve both of these problems,

    these criticisms continue to surface.


    • 27

    The experience of contracting out employment services in Australia

    59 Eardley T, “Outsourcing

    Employment Services: What

    have we learned from the Job

    Network?”, Paper to Centre for

    Applied Economic Research

    conference on the Economic

    and Social Impacts of

    Outsourcing, University of New

    South Wales, pp 2, 4-5

    December 2003

    60 Murray P, “A Job Network for

    Job Seekers”,

    Catholic Social

    Services Discussion Paper

    , pp

    54, November 2006

    61 Australian National Audit

    Office, “DEWR’s Oversight of

    Job Network Services to Job


    Audit Report

    , no 51,

    pp 131, 2005

    62 Australian National Audit

    Office, “DEWR’s Oversight of

    Job Network Services to Job


    Audit Report

    , no 51,

    pp 55, 2005

    63 Australian National Audit

    Office, “DEWR’s Oversight of

    Job Network Services to Job


    Audit Report

    , no 51,


    Catholic Social Services claimed that although service

    quality is emphasized in JN contracts, it cannot be measured

    and is therefore not enforced

    It is regularly claimed by critics that

    some JN members are short-circuiting the

    system to maximize their payments, and

    that more scrupulous providers are being

    disadvantaged and are having to cut the

    quality of their service just to stay in business.

    These concerns were forcibly

    expressed in 2006 by the Social Affairs correspondent

    of the

    Sydney Morning Herald

    who claimed: ‘The experiment has gone

    wildly wrong… rorts, unethical practices,

    waste of taxpayers’ money and shortchanging

    of the unemployed are endemic…

    Competition is the wrong model if it

    rewards the cynical and punishes the ethical.’


    A report issued by Catholic Social

    Services detailed some of the powerful perverse

    incentives that still remain in the system

    of payments. For example, if a JN

    member fails to place a job seeker in work

    in the first three months, it pays to minimize

    further assistance until twelve months

    have elapsed so as to qualify for the higher

    outcome payments attaching to placement

    of long-term unemployed clients.


    Similarly, it pays to place a client in a

    short-term position, knowing they will

    soon return to unemployment, rather than

    putting in extra effort to find a more sustainable


    This report also identified a variety of

    practices which are arguably fraudulent as

    well as unethical. It claimed, for example,

    that the drive to maximize outcome payments

    was resulting in manipulation of

    recorded outcomes (a claim that was also

    made back in 2002 by the Productivity

    Commission). It suggested that some

    providers get an extra payment by starting

    clients on a low level job and then having

    them upgraded. Some JN members pay

    employers when placements do not turn

    up for work so as to maintain the continuous

    record of employment that is needed

    to prove a successful outcome. And some

    are said to engage in ‘outcome buying,’

    using money from the Job Seekers Account

    to subsidise employers so they will create

    short-term jobs that would otherwise not

    exist and that will collapse once the subsidy

    is withdrawn ($102m was spent in

    2005-6 on subsidizing 40,000 placements).


    It is impossible to tell how widespread

    these practices are, and other JN agencies

    have rejected these claims. The main industry

    body (National Employment Services

    Association) maintains that DEWR scrutiny

    is very tight and there is ‘no evidence of

    people working to perverse incentives.’

    Similarly, the CEO of one of the for-profit

    JN members says the participation

    requirements of the Active Participation

    Model mean it is now ‘impossible’ to delay

    placing people.

    But malpractice does still

    occur, and it does not only involve the forprofit


    In 2006, even the

    Salvation Army was ordered to repay $9m

    after a DEWR investigation found they

    had wrongly classified some of their clients

    as needing Customised Assistance.

    Undermining the autonomy and ethos of the

    community sector

    Some of Australia’s biggest religious charities

    have become heavily reliant on government

    financing as a result of their participation

    in the Job Network. Public funds

    make up half the turnover of the non-profit

    JN providers, while commercial recruitment

    firms rely on government for only

    3% of their profits.

    The Salvation Army,

    Centacare (run by the Catholic Church),

    Mission Australia and Wesley Uniting

    Employment together now rely on JN contracts

    for one-third of all their income. As


    Australian Financial Review

    noted, JN

    contracting ‘has been a key to the transformation

    of the charitable sector into big

    business enterprises under the Howard


    This high level of financial dependency

    has undoubtedly compromised the integrity

    of the community sector, and this is

    nowhere more apparent than in the repeat-

    Paying for success


    64 Horin A, “Jobless Just Pawns

    in a Nasty Numbers Game”,

    Sydney Morning Herald

    , 16

    November 2007 (‘rort’ is

    Australian slang for fraudulent

    practice). This article was

    prompted by the Catholic Social

    Services report which claimed:

    ‘Some providers are sacrificing

    service quality for outcome volume

    just to survive’, Murray P,

    “A Job Network for Job


    Catholic Social

    Services Discussion Paper


    November 2006

    65 Murray P, “A Job Network for

    Job Seekers”,

    Catholic Social

    Services Discussion Paper

    , pp

    36 and 38, November 2006

    66 Murray P, “A Job Network for

    Job Seekers”,

    Catholic Social

    Services Discussion Paper

    , pp

    31, November 2006

    67 Morris S, “Job Network

    Agencies Split on System


    Australian Financial


    , 24 November 2006

    68 Maiden S, “Millions Owed on


    The Australian

    , 13

    February 2006. As we saw earlier,

    disadvantaged job seekers

    are referred immediately to customized

    assistance under the

    Active Participation Model. In

    2005, responsibility for classifying

    job seekers passed from

    Centrelink to JN members themselves,

    and this led to a spike in

    special payments, triggering the


    69 Bruttel O,


    Competition in a Public Service

    Market: The Job Network in an

    International Perspective


    Centre for Labour Market

    Research Discussion paper, no

    3, pp 6, 2003

    ed strains between the government and the

    not-for-profit JN members over so-called

    ‘breaching penalties.’


    If a jobseeker breaks the terms of his or

    her Preparing for Work Agreement, their

    Job Network provider is required to submit

    a ‘participation report’ to Centrelink

    which may then impose a financial

    ‘breaching penalty’ on the individual concerned.


    Breaches can result in a reduction

    of welfare payments, and repeated breaches

    can lead to full suspension of payments

    for up to eight weeks, or until the job seeker

    complies. In the early years of the Job

    Network, many thousands of people were


    but following inquiries by

    the Ombudsman and the Productivity

    Commission, procedures were changed

    and the numbers have fallen.


    the most recent figures (for the period

    from October to December 2006) still

    show nearly 4,000 people were suspended

    for eight weeks for breaching the conditions

    of their income support payment for

    the third time.


    Welfare organisations campaigned against

    the breaching penalties system almost from

    the outset, and in 2002 they even set up

    their own ‘independent inquiry’ into

    breaches and penalties.

    It concluded that

    the breaching system had operated in an

    ‘arbitrary, unfair or excessively harsh’ manner;

    that ‘breaches are imposed too frequently’;

    and that penalties are ‘often too

    severe’ and cause ‘unjustifiable hardship’


    It recommended that Activity Agreements

    should be watered down, that Centrelink

    should consider waiving penalties altogether

    if they are likely to cause hardship, and

    that no penalty should involve withholding

    more than 25 percent of benefits or

    should run for more than 8 weeks in total.

    Although the federal government agreed

    in 2003 to reduce the penalty for a first

    breach, and set up a taskforce to ensure

    that Centrelink enforces the rules ‘fairly,’

    the welfare groups have persisted with their

    demand that no penalty should last more

    than eight weeks or exceed a 25% reduction

    in payments.

    If this were ever implemented,

    it would bury the principle of

    mutual obligation, for claimants who did

    not wish to meet the requirements

    imposed on them could simply settle for a

    three-quarters payment, knowing that

    nothing more could be done to force them

    to comply.

    The truth is that many welfare bodies

    have never felt comfortable managing the

    punitive aspects of their JN role. These are

    organisations whose traditional rationale

    has been helping people in need, and they

    have continually resisted punishment of

    non-compliance. The result is that they

    weaken the compliance regime at the same

    time as they compromise their own principles.

    The 2002 Productivity Commission

    report found that non-profit service

    providers under-report breaches by 12 per

    cent. This reflects their willingness to

    indulge transgressors, rather than any overzealousness

    on the part of other JN members.


    As one sympathiser put it, ‘They

    often push their contractual obligations to

    the legal limits in order to avoid reporting

    a client to Centrelink for breaching.’


    In 2006, these tensions came to a head

    when the government extended mutual

    obligation requirements to single parents

    whose children had reached school age.

    The welfare groups fiercely opposed this


    and some (like the St Vincent de

    Paul Society) refused to have anything to

    do with it. Yet despite their opposition,

    eighty organisations agreed to participate

    in a new ‘financial case management

    scheme’ designed to monitor families

    where breaching penalties had been

    imposed and to dispense special payments

    where there was evidence of hardship

    affecting children. Within a year, however,

    twelve welfare organizations had pulled

    out of the scheme, led by Catholic Social

    Services, arguing that the policy of breaching

    single parents was immoral and that

    they wanted nothing more to do with it.



    • 29

    The experience of contracting out employment services in Australia

    70 Tingle L, “Charity Faces Notso-

    Tender Test”,

    Australian Financial


    , 24 September 2002. See

    also Ramia G and Carney T, “New

    Public Management, the Job

    Network and Non-Profit


    Australian Journal of

    Labour Economics

    , 6:2, 2003

    71 There are two ways in which

    claimants may find themselves in

    breach of the conditions governing

    their payment: administrative

    non-compliance (e.g. failing to

    attend at the office of Centrelink

    or a Job Network service provider

    when required to do so, or failing

    to provide information when

    requested to do so); and activity

    test breaches (e.g. a failure to

    undertake an activity, such as

    training or job search, laid down in

    an Activity Agreement, or obstructiveness

    in the process of finding

    a job). Administrative breaches are

    seen as less serious than Activity

    Test breaches, and they attract

    lesser penalties. I have discussed

    breaching in more detail in

    Saunders P,

    Australia’s Welfare

    Habit and How to Kick it


    CIS/Duffy & Snellgrove, 2004.

    72 Even though the total number

    of unemployed claimants was

    falling between 1997 and 2001,

    the total number of breaches

    rose by 220 per cent. Most of

    these were first offences but

    over 30,000 people were

    breached for the third time in

    two years in 2001-02 and were

    therefore cut off from payments

    altogether for eight weeks. See

    Saunders P,

    Australia’s Welfare

    Habit and How to Kick it


    CIS/Duffy & Snellgrove, 2004.

    73 The 2002 investigation of

    Centrelink and the Job Network

    by the Commonwealth Ombudsman

    identified ‘some deficiencies

    in Centrelink procedures and

    practice’ which resulted in mistakes

    being made – McLeod RN,

    Social Security Breach Penalties

    – I

    ssues of Administration


    Canberra, pp ii, 2002

    74 Peatling S, “Little Centrelink

    Cheer for the Poorest of the


    Sydney Morning Herald


    18 December 2007.

    75 Pearce D, Disney J, and

    Ridout H,

    Making it Work

    , Report

    of the Independent Review of

    Breaches and Penalties in the

    Social Security System, 2002.

    The inquiry was sponsored by

    The National Welfare Rights

    Network, ACOSS, the

    Brotherhood of St Laurence,

    Jobs Australia, Job Futures,

    Mission Australia, the Salvation

    Army, the Smith Family and the


    Clearly, the welfare organisations feel

    increasingly compromised by the longterm

    relationship they have established

    with the government. They have become

    dependent on the government’s money,

    but they do not want to dance to its tune.

    Management of the hardest cases

    The Job Network was set up when unemployment

    in Australia was running at over

    8 per cent. It has virtually halved since

    then. This means there are fewer cases for

    JN members to manage, and that most

    people looking for a job can find one without

    much assistance. The cases that remain

    on JN members’ books tend to be a lot

    more difficult and labour-intensive. This

    has caused problems for all JN members

    (reduced unemployment has forced some

    agencies to lay off staff), and it has prompted

    suggestions that the JN framework may

    be ill-suited to delivering the sorts of services

    and outcomes that are now required.

    Caseloads have certainly got tougher.

    Catholic Social Services reports that the

    proportion of its clients with less than a Year

    Ten education has increased since ESC3

    from 19% to 25%.

    The President of

    ACOSS (the peak welfare body) reports that

    64% of those on the Newstart unemployment

    payment for more than two years have

    a YearTen education or less. So too do 62%

    of those with disabilities and 72% of single

    parents claiming Parenting Payment. Nor is

    lack of education the only problem, for

    45% of sole parents on Parenting Payment,

    30% of Disability Support Pension

    claimants and 35% of long term Newstart

    Allowance recipients have mental health


    These are the people JN

    providers are increasingly trying to help. As

    the CEO of Jobs Australia suggests: ‘A significant

    number of the people left in the

    queue have very complex needs, typically

    mental health issues, housing issues, family

    relationship issues, all sorts of things that

    may make it difficult for them to comply.’


    Some JN members now openly doubt

    whether the JN system in general, or the

    Active Participation Model in particular,

    is appropriate for managing cases like

    these. The Brotherhood of St Laurence,

    for example, criticizes the emphasis on

    ‘rapid movement into any job without

    ongoing support for career advancement

    or skill development,’ and it claims the

    system is not well suited to handling people

    who are not job ready. It is a system

    geared to fast throughput of job-ready

    unemployed people, but these aren’t the

    clients any more.


    Sheriden Dudley (CEO of Job Futures, a

    combined community groups JN member)

    echoes these sentiments. He says the continuum

    model that underlies Active Participation

    is premised on the assumption that

    people can follow a pathway into work, but

    some cannot: ‘We know that many job seekers

    are not able to follow this linear pattern

    and that the reality for many people is they

    move in and out of the system. This is particularly

    true for those with a mental illness…

    One of the many challenges facing JN

    is how do we support people through periods

    when they simply can’t work, when

    employment outcomes seem at times to be

    the sole driver of the system.’

    He goes on to

    argue for a system of rewards that acknowledges

    non-economic outcomes – ‘many of

    those who have been referred to these programs

    in the past require a great deal of support

    in dealing with personal issues before

    they are ready to join an employment program.’


    Under present arrangements, for

    example, a JN provider can put a huge

    amount of effort and resources into someone,

    but if they then drop out of work before

    13 weeks have elapsed, their JN provider

    receives no payment.

    The National Employment Services

    Association (NESA) is another voice arguing

    that the ‘work first’ ethos now needs

    changing. It says there should be more

    emphasis on ‘proper’ skills training rather

    than on rapid placement, and that this

    could make a positive contribution to

    Paying for success


    76 Pearce D, Disney J, and

    Ridout H,

    Making it Work

    , Report

    of the Independent Review of

    Breaches and Penalties in the

    Social Security System, paragraphs

    18-21, 2002

    77 Senate Community Affairs


    A Hand Up, Not a

    Hand Out

    , Commonwealth of Australia,

    recommendation 14, 2004

    78 The Productivity Commission

    found (Productivity Commission,

    “Independent Review of the Job

    Network”, report no 21, pp 6.18,

    June 2002) that for-profit JN providers

    had higher breaching rates

    (12 percentage points higher) than

    not-for-profits. This is because they

    aremore loath to report clients (6.20)

    79 Kinnear P, “Putting Obligation

    in its Place”,


    , pp 9,

    February 2002. Kinnear is a former

    branch manager in the

    Australian public service and

    Research Fellow at the left-wing

    Australia Institute.

    80 The Salvation Army complained

    that ‘vulnerable groups

    within the community will almost

    certainly be adversely affected by

    this policy’ (

    The Australian


    pull out of welfare to work”, The

    Australian, 31 October 2006.

    81 The Australian, “Twelve groups

    pull out of welfare to work”,



    , 16 February 2007.

    82 Murray P, “A Job Network for

    Job Seekers”,

    Catholic Social

    Services Discussion Paper

    , pp

    15, November 2006

    83 Davidson P,

    Strengths and

    Weaknesses in Australian Welfare

    to Work Policy

    , Paper to

    Australian Social Policy conference,

    University of New South

    Wales, July 2007. These cannot,

    presumably, be serious mental

    health problems, otherwise these

    claimants would all be on DSP.

    84 ABC Radio National


    Job Network Burn


    , 31 July 2006.

    85 Brotherhood of St Laurence,

    “Brotherhood Welcomes Labor

    Call to Overhaul Job Network”,

    Brotherhood of St Laurence

    Media Release

    , 25 October

    2006. See also Nicholson T,

    “Social Inclusion the Path to


    The Australian

    , 23

    November 2007.

    86 Dudley S,

    Not Just Any Job


    Paper to Australian Social Policy

    Conference, University of New

    South Wales, pp 5, July 2005

    87 Dudley S,

    Not Just Any Job


    Paper to Australian Social Policy

    Conference, University of New

    South Wales, pp 5, July 2005

    meeting growing skills shortages in the


    Whether people with mental

    health and other problems, who have failed

    to complete a Year 10 education, are actually

    capable of being trained to fill skilled

    vacancies is, however, an issue that these

    commentators rarely address.


    Before it lost office, the Howard government

    made clear it was aware of some of

    these problems and that it intended to

    address them by further reforming the JN

    system at the 2009 contract round.


    new Rudd government has also swiftly signalled

    its intention to change fee structures

    in order to reward those dealing with the

    hardest cases, probably by strengthening

    interim payments.

    Whether this will be

    enough to answer the concerns of the JN

    providers is, however, doubtful, for there

    appears to be a fundamental problem in

    measuring non-specific outcomes like getting

    someone off drugs or building their

    self-confidence to a point where they can

    face going to a job interview. After ten years

    of wriggling within the JN system, the notfor-

    profits seem now to be suggesting that

    the whole system needs fundamentally

    rethinking, and at the time of writing, the

    new Rudd government has announced a

    review which will look into these concerns

    prior to the scheduled 2009 contract round.


    Nearly ten years after employment services

    were contracted out to a new Job Network,

    there have clearly been significant economies

    and cost savings for the Australian

    taxpayer. There is also some evidence that

    placement of unemployed people into jobs

    has improved as compared with the outcomes

    achieved before 1998, and that the

    purchaser-provider model is now generating

    positive net impacts for most categories

    of job seekers that are as good as any in the


    Yet having recognized this, there have

    been recurring concerns that savings have

    been achieved at the expense of the quality

    of service offered to the most disadvantaged

    job seekers. To some extent, these

    claims reflect the anti-market and anticompetitive

    prejudices of many of the

    journalists, academics and not-for-profit

    JN providers who comment on these matters,

    for they instinctively prefer a cooperative,

    public sector model to a competitive

    private sector one. However, there is also

    some real substance to these claims, for

    some JN service providers have cut corners,

    and some (including religiouslybased

    not-for-profit contractors) have been

    guilty of unethical or fraudulent practices.

    An incentivised system like the Job

    Network will always run the risk that service

    producers will sacrifice quality and look

    for the easiest returns. In some cases,

    unethical and fraudulent practices are

    motivated by greed (a desire to maximize

    outcome payments); in others they may

    have been motivated by shrinking margins

    and the intensified struggle to keep their

    heads above water as a result of a declining

    client base and the increasing difficulty of

    placing the marginal clients who remain

    on their books. But whatever the immediate

    reasons, the root of the problem is

    structural, for this is a ‘market’ where consumers

    enjoy little or no sovereignty.

    Job Network members cater to people

    who are not paying for what they receive

    and who have limited power to switch

    between providers if they are unhappy

    with what they are getting. The Australian

    experience suggests this is probably

    unavoidable, for much of the client base

    lacks the experience, capacity or desire to

    act as rational and discriminating consumers.

    This failure of the ‘customer

    choice’ ideal has inevitably led to a tightening

    of Departmental control, monitoring

    and regulation of JN providers as the only

    effective way of ensuring service quality.

    This has in turn resulted in increased

    administrative costs, both for DEWR and

    for JN members. It has also subverted one


    • 31

    The experience of contracting out employment services in Australia

    88 National Employment

    Services Association,


    of the Future

    , NESA (South

    Melbourne), pp 4, 2007,

    89 Saunders P, “What are low

    Ability Workers to do When

    Unskilled Jobs Disappear?”,

    Issue Analysis (The Centre for

    Independent Studies

    ), no 91, 6

    December 2007

    90 Karvalas P, “Job Seekers to

    Face Radical Shakeup”,



    , 21 July 2007

    91 Karvalas P, “ALP Eyes New

    Job Network system”,



    , 18 December 2007.

    See also Thomas M, “A Review

    of Developments in the Job


    Parliamentary Library

    Research Paper

    , no 15, 24

    December 2007

    of the original aims of the reform, which

    was to give service providers the flexibility

    to innovate and to free them from the

    stranglehold of the central bureaucracy. As

    things have turned out, increasing central

    control has curtailed flexibility and has to

    some extent reproduced the old top-down

    system of bureaucratic management.

    Increasing central control has also exacerbated

    the tensions between the

    Department (the final and arguably ‘real’

    customer for their services) and the notfor-

    profit service providers. This tension

    has been present right from the start, as is

    apparent from the long-running war of

    attrition over breaching penalties. As the

    mutual obligation policy has been extended

    to increasing numbers of welfare

    claimants (and to new categories of

    claimants, such as single parents), so these

    strains have become worse, culminating in

    a number of providers reneging on their

    agreements to participate in the new

    ‘financial case management scheme’ for

    single parents.

    Many of the not-for-profit JN members

    are lukewarm (at best) about mutual obligation

    and they dislike the ‘Work First’

    principle around which the JN system has

    been constructed. They feel their mission

    has increasingly put them at odds with

    what their government masters are telling

    them to do. By 2006/07, some of them

    were openly campaigning to get the whole

    system scrapped, or at least radically overhauled.

    This tension at the heart of the Job

    Network has arguably been bad for the

    not-for-profit sector (whose autonomy and

    defining purpose have been compromised),

    and bad for the government and

    the Department (which has struggled to

    have its policies implemented in the way

    that it intended).

    The basic problem, therefore, is that this

    is not a ‘market’ in any recognizable sense

    of the word. Unemployed clients are called

    ‘customers,’ but most of them do not even

    choose their service provider, let alone pay


    As for the supply side, the JN

    providers are paid by results, but they can

    hardly be said to be competing for business

    any more, given that 97% of business allocated

    at the 2003 contract round went to

    existing JN members, and the 2006 round

    was scrapped as existing contracts were

    rolled over for another three years.

    Provided they meet minimum quality

    standards and maintain an adequate level

    of job placement outputs, JN members are

    now in little danger of being usurped by

    new competitors.

    The real danger to their business is not

    competition from other suppliers, but government

    financial stringency.

    The not-for-profits, in particular, have

    been active in campaigning for a change to

    the funding formula, arguing that the existing

    system of fees and outcome payments is

    unfair and outdated now that most of their

    cases consist of ‘disadvantaged’ clients.

    They want ‘Work First’ scrapped and more

    generous (and less conditional) payments

    introduced so that they can work long-term

    with disadvantaged clients in highly intensive,

    one-to-one, supportive relationships.

    Rather than concentrating on getting these

    people into jobs, the not-for-profits want to

    be paid for helping them overcome their

    ‘barriers to work’ and for getting them fully


    The push is therefore on to transform an

    employment services system with an economic

    rationale into a social services system

    with a social work rationale, and with

    a change of government in Canberra, it is

    clear that the Job Network model has

    reached a critical fork in the road.

    Paying for success


    92 They might become genuine

    customers if they relied on personal

    Unemployment Insurance

    or ‘Temporary Earnings

    Replacement Accounts’ rather

    than government welfare benefits

    – see Feldstein M and

    Altman D,


    Insurance Savings Accounts


    National Bureau of Economic

    Research, December 1998;

    Brunner L and Colarelli S,

    “Individual Unemployment


    The Independent


    , no 8, pp 569–585, 2004;

    Saunders P, “A Welfare State for

    Those Who Want One, Opt Outs

    for Those Who Don’t”,


    Analysis (Centre for Independent


    , no 79, January 2007

    Rather than concentrating on getting these people into

    jobs, the not-for-profits want to be paid for helping them

    overcome their ‘barriers to work’ and for getting them

    fully ‘job-ready

    As for whether a country like the UK

    would be well advised to follow in

    Australia’s footsteps, the lesson seems to

    be that the purchaser-provider model

    works well if the desired outcomes are

    kept simple. If the clear aim is to get welfare

    claimants into work, then a system

    like the Job Network (coupled with

    appropriate reforms to the conditions

    attaching to receipt of welfare payments,

    and a tough sanctions regime for when

    these conditions are breached) can perform

    a lot better than a public sector

    bureaucracy. But if other objectives get

    tacked onto this, or the aim gets watered

    down or subverted (e.g. by moving away

    from a Work First policy), then the purchaser-

    provider model is likely to become

    less effective and manageable. The more

    difficult it is to measure outcomes, the

    less attractive a Job Network model is

    likely to be.

    It is a case of horses for courses. A quasimarket

    in employment services will deliver

    lower costs and somewhat better job outcomes

    (although no labour market program,

    public or private, will generate

    strong net outcomes). But the Job

    Network is not a system designed or suited

    to delivering social work support to people

    with mental health problems, those who

    never finished school, drug and alcohol

    abusers, or those who have never become

    habituated to the routines of working. It is

    for the government to decide how it wants

    to manage these ‘hard cases,’ and if it wants

    to adopt a ‘therapeutic’ or ‘treatment’

    approach, a profit-driven, contracting-out

    model is probably not the best option for

    managing it.


    • 33

    The experience of contracting out employment services in Australia

    93 An estimated figure; Central

    Intelligence Agency (2008)


    2008 World Factbook


    Central Intelligence Agency).

    94 Australian Bureau of

    Statistics (latest relevant year)

    Labour Force Survey


    Australian Bureau of Statistics).

    95 Australian Bureau of

    Statistics (latest relevant year)

    Labour Force Survey


    Australian Bureau of Statistics).

    Plus author’s calculations.

    96 Australian Bureau of

    Statistics (latest relevant year)

    Year Book Australia


    Australian Bureau of Statistics).










    100 Australian Bureau of

    Statistics (latest relevant year)

    Labour Force Survey


    Australian Bureau of Statistics).

    Plus author’s calculations.







    103 Australian Bureau of

    Statistics (latest relevant year)

    Year Book Australia


    Australian Bureau of Statistics).




    105 Australian Bureau of

    Statistics (latest relevant year)

    Labour Force Survey


    Australian Bureau of Statistics).

    Plus author’s calculations.

    106 Australian Bureau of

    Statistics (latest relevant year)

    Year Book Australia


    Australian Bureau of Statistics).




    108 Australian Bureau of

    Statistics (latest relevant year)

    Labour Force Survey


    Australian Bureau of Statistics).

    Plus author’s calculations.

    109 Figure refers to Intensive

    Assistance off-benefit outcomes

    compared to

    Working Natio


    programs (1995-96) at constant

    2000 prices. Author’s calculation.





    GDP , 2007: $890 billion

    Population, 2005/6: 20.6 million

    Number of people in the labour force, 2005/6: 10.6 million

    Average weekly wage among adults in full-time work: $967.9

    Total spending on welfare, 2005/6: $42 billion

    Total spending on unemployment benefit, 2005/6: $4.4 billion

    Rate of per week unemployment benefit, 2005/6: $182.50

    Number of people unemployed, 2005/6: 514,000

    Number of unemployed people in 2005/6 out of work for one year or more: 264,000

    Number of days formerly unemployed people must be in work for their assistance to be considered

    successful: 91 until an interim period, 182 until final assessment


    Total spending on incapacity benefit, 2005/6: $7.3 billion

    Rate of per week incapacity benefit, 2005/6: $222.20

    Number of people on incapacity benefit, 2005/6: 712,000

    Total spending on lone parent benefit, 2005/6: $4.3 billion

    Rate of per week lone parent benefit, 2005/6: $220.20

    Number of people on lone parent benefit, 2005/6: 433,000

    Average cost of successful placement in the private or voluntary sectors, 2005/6: $19,750

    Average per-placement saving from use of the private and voluntary sectors, 2005/6: $11,639


    The experience of

    privatization of

    welfare services in


    by Jason Turner

    “?e welfare system as an institution is

    abhorred by society because it separates the

    receipt of income from the need to work.

    But why do we think of work as so necessary

    to legitimize income? In its unique

    capacity to enable individuals to have the

    opportunity to serve others by producing

    valuable goods and services, work fulfills a

    basic human need.Work connects individuals

    to larger society –and in order for

    welfare’ destructive influence to finally

    end, work and income must be permanently


    These words, written in 1995, laid out a

    fundamental premise of what was to


    Wisconsin Works

    , a welfare reform

    notable not only for its departure from the

    existing welfare system of income support,

    but also in that it was devised from scratch

    without reference to any existing law. I was

    part of the changes, so give some insider’s

    insights during the following account of

    the changes.

    In 1994 and 1995, the years when

    Wisconsin Works

    was being developed, the

    concerns of academics and policy makers

    about the rise of welfare dependency and

    its intractability were paramount. Since

    the onset of President Lyndon Johnson’s

    Great Society welfare dependency had

    grown and grown, sometimes pausing,

    never permanently retreating.

    Various reform efforts had been tried.

    These included experiments with versions

    of a negative income tax; required registration

    for job search; additional funds for

    education and training – all of which had

    failed to stem the rising tide of dependency.

    The challenge to the designers ofWisconsin

    Works was to devise a plan so powerful in its

    effects that it would be capable of stopping

    and reversing this trend.

    Wisconsin and the national income

    support system

    Government income support programs in

    America can be grouped into three categories

    – unemployment insurance, disability

    insurance and welfare.


    Unemployment insurance provides temporary

    income for individuals whose

    positions have been eliminated. Those

    fired or let go for good cause are not eligible.

    The amount of the benefit is set

    by the state up to a maximum and is

    based on the claimant’s work history and

    prior contributions to the fund through

    his employer. The maximum duration

    of benefits is typically sixmonths.


    1 A handful of states provide for

    “general assistance”, usually

    very low benefits available to

    singles, often concentrated

    among those with mental health

    or substance abuse problems.

    The experience of privatization of welfare services in Wisconsin


    • 35


    Disability insurance provides federally

    issued payments to individuals whose

    incapacity has been determined by the

    state to make the claimant incapable of

    providing for himself through earnings

    (parents may also claim disability payments

    on behalf of their children). The

    amount of these payments is not high,

    but is more than obtainable through




    is the term Americans use for

    the general income support program

    for low income families (named

    Aid to

    Families with Dependent Children


    until 1996, thereafter

    Temporary Assistance to Needy Families


    Eligibility is limited to families in

    which parents are living with their

    dependent children. Unlike parts of

    Europe, a young man or woman out of

    school with no dependents and no

    work history cannot walk into an office

    and ask for income support (although

    job centers offer low-cost or no-cost job

    training and education, plus help finding


    . As a practical matter, the

    overwhelming proportion of welfare

    recipients are lone mothers with one or

    two children, sometimes more.

    Monthly cash benefits usually range

    between $400 and $600 per month,

    with food stamps bringing the total up

    to $$650 to $800 per month.

    Wisconsin’s Governor Thompson was elected

    in 1986 on a platform of reforming welfare,

    and won an unexpected victory. Over

    the subsequent five years from1987 through

    1993 Thompson proposed changes to the

    welfare system almost once a year. These

    changes came in the form of experiments,

    called “waivers” of federal program rules.

    These programme change requests to the

    federal government would mostly be seen as

    modest by today’s standards, but they all

    struck a similar and consistent theme, namely

    that welfare recipients should work and

    accept personal responsibility.

    The state legislature, controlled by the

    opposition Democrat party, grew increasingly

    frustrated by the support for reform

    that the Governor had created through his

    series of proposals. Finally, the legislators

    struck back by passing a law which would

    completely abolish the existing



    The law provided that in its

    place, the Governor must propose a workbased

    alternative. Would the law calling

    for abolition of the entire cash safety net

    for the poor be vetoed by the Governor? If

    so, reasoned some of the legislators, the

    opposition could take away his claim as the

    agent of reform, and they would be free to

    make proposals of their own.

    Instead, Governor Thompson accepted

    the opportunity to design a work-based

    program from the ground up, unconstrained

    by existing law. He then set about

    his task in earnest, assembling a group of

    his staff to develop a proposal.

    The philosophical underpinnings of

    the Wisconsin approach to welfare


    The 1995 proposal resulting from the

    Governor’s planning group was unprecedented

    in two ways. First, it was a complete

    change from the existing approach

    rather than an attempt to work around the

    edges of existing law. As a result, the program

    was generated from a set of consistent

    philosophical principles which were mutually

    reinforcing. Some of these principles

    and their rationales are summarized below.

    1. For those who can work,

    only work should pay.

    There are both economic and practical reasons

    for tying income to work. First,

    experience shows that entitlements to

    income without work have unwanted

    effects on dependency. In addition, it is

    essential that parents understand they will

    always be responsible for supporting themselves

    and their families through work:


    Paying for success

    this influences behavior and motivation in

    ongoing constructive ways. Finally, experience

    shows that individuals without extensive

    work history are usually in a stronger

    employment position after one or two

    years of actual work (at any wage) than

    after a comparable period of work preparation

    through education and training.

    2. Begin with the assumption that everyone is

    capable of some kind of work.

    The best way to help an individual out of

    work get back into the labor force is to provide

    an actual work opportunity which

    matches their capabilities. This is contrary

    to many government subsidized “helping”

    programs which seek to identify barriers and

    limitations to work, and in so doing categorize

    and place individuals out of the reach of

    the workplace where they might very well

    have succeeded if given the opportunity.

    Only by testing the suitability of work

    through actual attempts to work can any true

    limitations which prevent full participation

    in the labor force be identified and resolved.

    3. Strengthening the ability of parents to provide

    for their children is a better approach

    than having the government intervene directly

    on their behalf.

    In well-meaning attempts to look after the

    interests of children, government has, over

    time, participated in many of the roles that

    were previously the exclusive responsibility

    of parents.

    There are many calls for government to

    take on still further responsibility for assuring

    the well-being of children. However,

    government cannot raise children, only parents

    can. Government can do the most by

    helping to put parents in a position to meet

    their responsibilities, not by taking away

    these responsibilities for itself.

    4. Measure the fairness of the new system by

    comparison with working families

    It is sometimes argued that a work-based

    welfare system will be unfair unless it can

    be shown that those formerly dependent

    on various benefit programs will continue

    to receive a comparable package while

    working. Others argue that it is unrealistic

    to expect work for wages unless such

    wages will guarantee a high enough standard

    of living to make work seem worthwhile.

    But self-sufficiency through work

    should be seen as an end in itself, quite

    apart from the package of benefits gained

    or lost as a result. More important is the

    relationship that those who are receiving

    welfare benefits have with comparable

    individuals who are working to support

    themselves and have not asked for assistance.

    5. Look to non-government organizations to

    deliver the program.

    It is axiomatic that government programs,

    authorized by legislation, must be overseen

    by government as an agent of the public

    interest. However, for too long government

    has been the assumed operator of the

    programs it devises. A more effective

    model is almost always for the government

    to set the ground rules and then let nongovernment

    entities actually operate programs

    under public oversight.

    The Wisconsin-Works program (or W2)



    there is no entitlement to

    cash assistance based on income alone.

    When an individual without work

    applies for help at a


    agency, their

    access to cash benefits is dependent upon

    their actions to help themselves become

    economically self-sufficient through private

    employment. All individuals participating

    in the program “earn” their

    income (cash assistance) through participation,

    measured by the hour, in one of

    four tiers of employment in private or

    public settings. These four are from

    highest to lowest, which make up an

    employment “ladder”.

    The Four Tiers of the Ladder:

    1. Unsubsidized Employment

    Individuals entering


    should always be

    guided to the best available immediate job

    opportunity in the private sector. The

    agency is geared and focused on matching

    every participant’s capabilities with the

    best work option available when they

    arrive, rather than diverting them to

    extended education and training.

    Unsubsidized employment pays the

    market wage (currently about $7.00 –

    $8.00 per hour) along with the earned

    income tax credit (EITC) which contributes

    up to $4000 per year depending

    upon family size and income.

    2 Subsidized Employment (Trial Jobs)

    Where an individual could succeed in private

    employment but lacks work experience

    or, in some instances, skills, the


    agency can provide wage subsidies to private

    employers for a temporary period in

    order to offset some of the additional costs

    of integrating a new employee into the

    worksite. In practice, Trial Jobs are rarely

    used because employers say that the economics

    favour hiring even unskilled workers

    directly without the modest Trial Job

    wage subsidy, and they prefer to stay away

    from the administrative tasks of collecting

    the subsidy.

    3. Community Service Jobs (CSJs)

    If an individual tries and cannot find private

    employment, the next option is to

    work in exchange for benefits, in either a

    government or non-profit organization

    which serves the public. Examples include

    simple office work or outdoor maintenance

    for government. These jobs are

    intended to be 30 hours per week, to be

    useful even with minimal training or background,

    and to offer the opportunity for

    individuals without a work background to

    learn-by-doing, i.e. to get them experience

    in supervised work and task orientation in

    an employment setting. An additional ten

    hours can be used for education and training,

    plus job search.

    In order to replicate the circumstances

    of the workplace, the W2 cash benefit of

    $673 per month is reduced (by a factor

    equal to $5.85, the minimum wage) for

    each hour of scheduled work missed.

    CSJs are intended to be temporary. Each

    assignment is generally not to exceed six or

    nine months. The by-product of labour in

    such a setting is intended to contribute to

    the community and fulfill the essential

    civic role of work – giving to others in

    exchange for wages.

    4. Transitional Jobs

    For those unable to perform independent,

    self-sustaining work even in a CSJ because

    of limitations such as mild disability or

    substance abuse, Transitional Jobs are

    available for participation in work or vocational

    training and other activity such as

    treatment, consistent with an individual’s

    capabilities. The grant amount is $628

    (less than that of a CSJ) for up to 28 hours

    of eligible activity, and up to twelve hours

    of additional education and training.

    Summary of Incentives for Participants to go


    In order to receive cash assistance, an individual

    enrolled in


    faces approximately

    the same number of hours dedicated to a

    community service job or a transitional

    job, as they would in full-time private

    employment. Leisure without obligations

    is no longer an option.

    At each step up the work ladder, from

    Transitional to unsubsidized employment,

    The experience of privatization of welfare services in Wisconsin


    • 37

    In practice, Trial Jobs are rarely used because

    employers say that the economics favour hiring even

    unskilled workers directly without the modest Trial Job

    wage subsidy

    2 The situation among counties

    outside of Milwaukee was much

    better, with most county agencies

    embracing the mission of

    economic self-sufficiency for its

    welfare population. But some

    objected to the state’s preferred

    “work-first” approach to achieving

    this objective rather than the

    traditional education and training

    model. The designer’s explanations

    and exhortations favoring

    the work-first approach and

    explaining its rationale, including

    the research evidence of its efficacy,

    were most influential when

    falling on willing county ears.

    However, even there the state

    had made progress in this

    regard by the introduction of

    “Work-First” pilots. These pilots,

    in which applicants for assistance

    looked for work even while

    their application was pending,

    proved successful in achieving

    early employment outcomes.

    Even better, the success of

    “work-first” was touted by the

    pilot counties themselves, leaving

    them to sell the idea rather

    than the suspect state.


    Paying for success

    the net after-tax income (which includes

    the earned income tax credit for those in

    private employment) increases. Therefore,

    participants have every reason to accept the

    highest employment option they qualify

    for, starting with unsubsidized employment

    if available, or if not to move up the

    ladder when the option presents itself.

    Supportive services such as subsidized

    child care or free medical care are no longer

    reserved for those enrolled in the welfare

    system, but rather to anyone on the

    income scale. Therefore, there is no longer

    an advantage to enrolling in welfare to just

    to receive these extra benefits.

    The Operational Considerations

    Driving Program Design and the

    Privatization of Serviced

    When implementing the work-based system,



    designers were faced with a

    dilemma. The organizational requirements



    are higher than that of running an

    income maintenance agency where the primary

    task is determining eligibility based on

    income (along with making referrals to separate

    job training or employment agencies).

    Previously, welfare operations were divided

    into two parts: the eligibility determination

    process on the one hand, and all other

    services such as job placement, training and

    subsidized child care on the other. County

    government always provided eligibility

    determinations directly, and would perform

    the other services either directly or through

    contracts to private service providers.

    The county-run welfare system could

    function satisfactorily where its central task

    was performing routine eligibility determinations.

    However, for more ambitious

    tasks the government has well-known limitations:


    It usually manages by process rather

    than outcome, i.e. projects are administered

    according to short-term needs

    rather than long-term goals.


    It lacks accountability due to the diffusion

    of responsibility;


    Its performance compromised by decisions

    subject in part to political considerations;


    It loses mission focus as energy is sunk

    into attempting to coordinate the multiple

    organizational units operating a

    given program (e.g. the state policy

    making function, the county administrative

    function, and the non-profit

    service delivery function);


    It has an inability to shed non-functioning

    operational units.

    In particular, Milwaukee, Wisconsin’s

    largest county (population one million),

    showed resistance to mission change or

    operational improvement. Like many large

    cities, Milwaukee’s welfare agency had

    weak management and a labor force rarely

    challenged to perform. In the period leading

    up to


    , the state had attempted to

    get the welfare agency to properly refer

    welfare recipients to private agencies for

    work or training; and when not participating

    in such training, to reduce cash benefits

    until recipients complied. The county

    agency exhibited limited bureaucratic

    capacity to carry out such multi-step

    processes. Furthermore there was not

    much interest among top management in

    the objective itself.


    In any event the requirements for


    went beyond the task of eligibility determination

    and referral for employment. An

    agency would have to act in such a way as

    to limit its own size and influence in the

    lives of participants and instead to empower

    them to achieve economic self-sufficiency.

    Through contract incentives, its designers

    believed they could achieve this.

    The two parts eligibility and job placement,

    were merged into one so that there

    would be a single organization wholly

    accountable for results. This organization,

    Wisconsin Works, owns the Job Center,

    provides eligibility determinations for cash

    The experience of privatization of welfare services in Wisconsin


    • 39

    benefits and gives work support, such as

    subsidized child care or “job access loans” (

    cash to help with enabling employment,

    e.g. for repair of an auto so the participant

    can drive to work). Most importantly, the

    role of the


    agency is to help applicants

    find the best available job opportunity at

    the time of application, and provide such a

    job itself as a last resort.

    Helping individuals with marginal work

    histories required closer connections

    among case-workers, employers and participants.


    case workers were paid higher

    salaries and carried far lower caseloads than

    traditional caseworkers (about 75 instead

    of around 300). They could be recruited

    from among the best of the existing case

    workers, but also from the outside.

    Personality characteristics were judged to

    be more important than longevity or a specialized




    agencies were designed to

    experience the maximum latitude to find

    their own solutions to achieving the desired

    outcomes; namely, employment and caseload

    reduction. In this regard state policy

    makers had struggled with some of the

    counties (not others) over the years before

    the implementation of



    Competition for vendor selection

    In order to maximize the likelihood of

    Wisconsin Works

    succeeding, its designers

    envisioned a competition for program

    operators, open to any and all organizations,

    county government included. As



    agencies, three of the four

    types of potential


    operators had different



    County government, particularly in

    smaller or rural counties, was in certain

    places the only organization with the

    infrastructure ready to go. Looking forward,

    in many instances counties had

    dedicated staff and supportive elected



    Non-profit organizations were sometimes

    well established and were already

    providing the kinds of services required


    Wisconsin Works

    – particularly

    those that were likely to be expanded,

    such as community service jobs, which

    rehabilitation agencies such as

    Goodwill already provided on a relatively

    large scale.


    For-profit businesses were already providing

    employment and training services

    and could be expected to offer

    dynamic energy and be more capable of

    assuming risk. However, in the period

    leading up to


    , for-profit organizations

    were only a modest factor in the

    overall inventory of operating providers

    so this sector would need to be developed.


    Quasi-government providers, in particular

    the “private industry councils”

    with governing boards appointed by

    the counties along with volunteers

    from the local business community

    were the least appealing as agents of

    change. They were usually highly

    politicized organizations that made

    grants to employment and training

    vendors, and were hobbled by internal

    conflicts on their boards and in no

    position to lead a dynamic in-house

    operating agency. Consistent with our

    open competition, they were not

    excluded from the RFP process.

    In an attempt to cajole the incumbent

    service providers into a force for change,

    their freedom of action was maximized to

    provide the clearest possible incentives for


    In the past, multiple organizations

    combined into consortia diffused responsibility,

    and required consensus to act.

    Poor performing partners could not be

    expelled because they were part of the

    management team. Further, states would

    ask for a multiplicity of happy results –

    but the achievement of one would often

    compromise the fulfillment of another.




    the state’s objectives were clear

    – to maximize family economic self-sufficiency

    through employment and to

    reduce the corresponding cash assistance

    caseload. Partnerships were permitted,

    but only ones with prime- and sub-relationships

    so that leadership roles would

    remain clear.

    Another consideration was size. A phenomenon

    was noted: the smaller the county

    agency, the higher the employment


    Of course, the welfare population in

    small counties is generally easier to serve,

    and the culture of self-reliance higher. But

    there were other differences too, having to

    do with span of control. In a big city welfare

    department a recipient might not see

    the same caseworker twice, something

    which leads to anonymity. Follow-through

    is less likely to occur face-to-face and more

    likely through a computer tracking program.

    Responsible recipients looking to

    connect with the department for help

    often will not get through to caseworkers

    by telephone, with voicemails piling up.

    Managers and supervisors manage by

    numbers, paper files and adherence to regulations,

    not through the exertion of personal


    For this reason Milwaukee was divided

    into six regions of about 6,000 welfare

    cases each, so that the state would contract

    with multiple vendors. Each region

    was drawn so that it would have, to the

    biggest degree possible, recipients of similar

    general characteristics. In addition to

    the expectation that smaller vendors

    would perform better, the scheme’s

    designers hoped that multiple vendors in

    Milwuakee would reduce the state’s risk,

    and that low performing vendors could

    be replaced with less disruption. Finally,

    they also hoped that smaller vendors

    becoming identified with their surrounding

    neighborhoods. Prospective vendor

    agencies were permitted to bid on more

    than one, and in the event one of the

    applicant vendors (Goodwill) was awarded

    two regions, each with its own agency


    Contract Incentives

    Rather than replacing the old



    with another one operated on the

    basis of the same regulatory “command

    and control” mechanism, the programme

    designers were looking to create powerful

    contract incentives which would be in

    alignment with the incentives that the participants

    themselves would experience.

    In the period prior to


    there was an

    experiment with a simple form of performance

    incentives for job placement and

    training agencies which had functioned

    better than expected.

    The vendors earned

    part of their allocations based on their job

    placement success and they went earnestly

    to work, increasing placements by 29% the

    first year. But what was equally remarkable

    at the time was the degree to which vendors

    had been misallocating effort and

    resources prior to the introduction of this

    outcome-based payment system. One vendor

    director said “I had thought we were

    supposed to be getting recipients general

    equivalency degrees” (remedial education

    degrees). He said this despite the fact that

    the state had been exhorting the vendors to

    change their emphasis to employment for

    some time. It thus seemed that words were

    deficient and that the right tools had finally

    been adopted.

    The initial contracts from September

    1997 through December 1999 were structured

    so that reductions in the caseload

    resulting from movement off



    assistance into employment would reduce

    costs which would then be shared with

    vendors as follows:


    The initial contract amount was set by

    county region by calculating the current

    total welfare expenditures in the

    most recent period (including cash

    3 For example, many states

    would ask their welfare training

    vendors to simultaneously

    increase exits due to employment

    and to maximize average

    wages at exit. These are both

    worthy objectives, but maximizing

    participant employment

    exits, including those with limited

    work experience, will depress

    average wages at exit (although

    not necessarily in the long run,

    as those in the labor market

    increase their wages over time).

    In this example it would be better

    for the state to give one

    assignment to welfare-to-work

    vendors (to get as many as possible

    into the labor market) and a

    second assignment to education

    and training vendors (to improve

    wages though skills upgrades

    among both the working and

    non-working population). The

    success measurements for each

    will differ, as will the corresponding


    4 For example, for every 100

    welfare cases in Milwaukee,

    there were 15 entered employments

    over a year, while the

    median mid-sized county

    obtained 26 and small counties


    5 In Milwaukee, the six job

    placement and training vendors

    were grouped into two teams, A

    and B. Each team had a lead

    vendor, and the proportion of

    referrals from the state to the

    respective teams was exactly

    proportionate to their respective

    total budgets for services. Since

    vendors from Team A had twothirds

    of the total job placement

    and training budget from the

    state, they were referred to all

    individuals with social security

    numbers whose last three digits

    were between 001 and 666, with

    the balance to Team B. This

    system of referrals was easy for

    desk clerks to learn, and at the

    end of the competitive period

    the state could easily determine

    employment outcomes by using

    the social security numbers to

    match against quarterly wage

    payments already present in the

    state data file.


    Paying for success

    benefits, employment services and

    administration), and reducing this

    amount by twenty percent. This total

    amount could be allocated by the

    agency according to its needs, after cash

    benefits were first paid to eligible participants.


    Agencies generating caseload reductions

    which generated savings beyond

    the total budgeted amount for the period

    could keep the first seven percent of

    the savings in the form of unrestricted

    income (or profit). Thereafter additional

    savings would be divided three


    1. Forty-five percent would revert to

    the state in the form of budget


    2. Forty-five percent could be

    allocated by the


    agency to their

    local community for expenditures

    in general support of low income

    families, with approval of these

    expenditures by the state;

    3. Ten percent of additional savings

    would revert to the


    agency in

    the form of unrestricted income

    (or profit).


    Agencies whose expenditures during

    the contract period exceeded the total

    allocation from the state would be obligated

    to provide the difference.

    The Contracting Process

    When the initial


    plan was

    made public in April 1995, it met with

    great interest. However, early discussion

    revolved around the provision which

    would open

    Wisconsin Works


    opportunities to any potential service

    provider – county, non-profit or for-profit.

    Since all state legislators have constituents

    employed in county welfare departments,

    almost all legislators heard from county

    employees concerned about the future of

    their jobs. From the perspective of the state

    welfare department, many or most of

    Wisconsin’s 72 county governments, especially

    in rural areas, were likely to do a

    good job administering


    , and the

    scheme’ designers did not want to jeopardize

    the overall future of the plan due to

    objections over which were unlikely to prevail.

    Looking to make the best of the situation,

    the designers created a right of first

    refusal for county governments which

    wished to operate


    and met certain performance

    objectives in the twelve months

    prior to contract selection. The performance

    objectives selected were directly related

    to ones which would make the


    agencies successful in helping their participants

    gain employment once the program

    went live in September 1997.


    response upon the publication of these

    measurement elements by county management

    and employees was electric. County

    elected officials and managers began in

    earnest to work on achieving the right of

    first refusal, setting up their measurement

    systems and organizing their agencies

    along the lines required. At the end of the

    measurement period all but about half a

    dozen counties had succeeded in meeting

    the requirements, with the most notable

    and expected exception of Milwaukee

    (where county elected officials had in any

    event by then decided not to compete for


    , reasoning that the fiscal risks outweighed

    the rewards).


    The state next contacted prospective

    vendors located outsideWisconsin in order

    to have a robust and competitive bidding

    market, especially within Milwaukee. At

    that time several big for-profit corporations

    specializing in information technology

    were actively considering entering the

    6 These performance objectives



    Reducing the welfare caseload

    by ten to twenty-five percent

    depending on prior results;


    Placing greater numbers of

    recipients into unsubsidized

    employment of at least 20

    hours per week, increasing

    over three measurement periods;


    Placing those not working in

    private employment into the

    equivalent of community service

    or other jobs, increasing

    over three measurement periods;


    Reducing expenditures for

    cash assistance;


    Credit given to counties which

    had cooperated with the state

    and performed well in the


    7 The right of first selection

    process showed that in certain

    circumstances government entities

    are motivated by financial

    incentive structures as well as

    private vendors are, and that the

    resulting improvements in operations,

    if any, should not be automatically

    assumed to derive

    exclusively from privatization.

    The experience of privatization of welfare services in Wisconsin


    • 41

    A phenomenon was noted: the smaller the county

    agency, the higher the employment rate

    welfare-to-work marketplace. They reasoned

    that their profitable IT businesses,

    already serving state human service agencies

    throughout the country, would be natural

    partners in gaining entry wherever

    they received contracts as welfare-to-work

    vendors. Companies considering such line

    extensions of their business at that time

    included Electronic Data Systems (EDS);

    Lockheed Martin IMS; Affiliated Computer

    Services (ACS); Deloitte & Touche;

    and IBM. However, although some of

    these companies already had welfare-towork

    contracts elsewhere, for various reasons

    none of these giants elected to bid on

    Wisconsin Works

    in Milwaukee or in the

    state (EDS partnered with Goodwill, an

    eventual Milwaukee vendor, but provided

    only IT services).

    It is possible that they were unnerved by

    the unusual state contract, in which vendors

    would be liable for making up shortfalls

    over the state budgeted amount

    (including shortfalls in cash assistance

    amounts) in the event they were not successful

    (smaller vendors, especially nonprofits

    agreeing to sign the state contract,

    perhaps reasoned that were they poised to

    go under, the state would have to bail them

    out or take over the program, which in fact

    was true although it never happened).

    One large for-profit that did submit a

    bid for several regions in Milwaukee was

    Maximus, a corporation which had pioneered

    the human service operations business

    in Los Angeles in the 1980s. Some

    smaller for-profits also submitted bids in

    Milwaukee and elsewhere in the state.

    After the RFP process contracts were

    awarded in Milwaukee to the following:


    Employment Solutions, a for-profit

    subsidiary of Goodwill Industries, a

    major provider of welfare-to-work services

    in Milwaukee and a long-time

    vocational rehabilitation agency with

    national presence. This agency was

    awarded two Milwaukee regions.


    Opportunities Industrial Corporation

    (OIC), an African-American community

    vendor which had provided welfare-

    to-work services in Milwaukee and

    was part of a national organization.


    United Migrant Opportunities Society

    (UMOS), a Hispanic community vendor

    which had provided welfare-towork

    services in Milwaukee.


    YW-Works, a for-profit subsidiary of

    the Milwaukee YWCA and two other

    organizations which had provided services

    in Milwaukee.


    Maximus, a for-profit corporation

    without previous contracts in


    Eight counties other than Milwaukee were

    served by five private agencies in the initial

    implementation contracts.

    The program results from the implementation


    Wisconsin Works

    As September 1997


    start date was

    awaited, the programme’s designers

    believed there would never be a more propitious

    set of circumstances boding for the

    experiment’s success:

    ?e plan was about to open its doors “n

    the street”almost exactly as planned, with

    virtually no significant design compromises

    along the way; the state economy was

    strong and general unemployment was

    going down; certain elements of the program

    had already been put in place in

    advance of implementation and the caseload

    was already moving in the right direction;

    the Right of First Refusal had organized

    the counties which would run the program

    around the program design; the

    Wisconsin public was well informed by the

    Governor and the media about the plan

    and was with it all the way; its natural

    opponents among foundation-funded advocacy

    organizations and academics at the

    University ofWisconsin were temporarily


    Paying for success

    subdued (even if not supportive they

    acknowledged the plan’s comprehensiveness

    and internal consistency); welfare recipients

    themselves had been exposed to and

    absorbed several years of media and public

    discussion about the importance of work;

    soon work in some fashion was to be the

    only option for receiving cash assistance;

    and for county and vendor operators alike

    the stakes were exceedingly high on both

    the upside and the downside.

    Events had created a “perfect storm”.

    The results were more powerful than

    anyone, including its most optimistic planners,

    had anticipated. Immediately after

    Wisconsin Works

    opened, the welfare caseload

    began to collapse. In the year prior to


    , Wisconsin had about 55,000 welfare

    cases statewide – this dropped to 11,000

    cases in the first thirty-six months of the

    program (thereafter rising slowly and then

    falling back to 11,000 in 2005)


    At the end of the second year the state

    looked back at all of the

    Wisconsin Works

    cases which had been enrolled in the program

    at some point and were now closed

    (32,000 cases or 81% of all who had

    enrolled). Of these, fully 76% had closed

    for earnings and income (9% had chosen

    not to roll over into the new



    from the old one, 8% had enrolled but

    then closed for failing to comply with


    program requirements such as attending

    work assignments, 4% no longer met basic

    eligibility requirements and 2% closed for

    other reasons). Among those working,

    about half were earning between $12,000

    and $18,000 per year (before additional

    refundable tax credits), about a third were

    earning less and about a fifth were earning


    (the 1998 poverty threshold was

    $13,001 for a family of three).

    As would be expected with large numbers

    entering the labor force, poverty

    dropped substantially in both Milwaukee

    and the rest of the state.

    The child poverty rate (a more sensitive

    measure than adult poverty) peaked at

    34% in Milwaukee in 1993, declining in

    the period before


    implementation, and

    then, in the first single year afterward,

    dropped a full six percentage points to

    20%. Thereafter it rose slowly along with

    the weakening of the national economy

    around 2002; by 2006 after the beginning

    of the recovery it had given back about half

    of the drop from the top (the state and

    national trends were much gentler, but also

    followed a similar curve).


    The Critics of Wisconsin Works

    WisconsinWorks is about ten years old, and

    has almost certainly been the most studied

    and analyzed program of welfare reform in

    the U.S. (in just the first three years the

    Urban Institute counted 53 studies completed

    or underway).

    There are numerous

    critics of


    and its work-first approach.

    Criticisms fall into two categories: shortcomings

    in program implementation, particularly

    in Milwaukee, and the continuation

    of poverty as a social problem. Issues

    related to the first of these criticisms will be

    discussed at greater length below.

    Wisconsin’s most vocal critics at first

    warned of serious consequences which

    could be anticipated from large numbers

    who would be unable to cope with the

    work requirements of


    , and so would

    drop- out or never apply. Many would end

    up homeless, resorting to crime or, worse,

    the neglect or maltreatment of their children.

    None of these alarming predictions

    have come to pass. Instead, the rate of

    homelessness has remained flat

    , crime is

    down 12% in the state since program


    , and incidents of child

    abuse and neglect have fallen inWisconsin

    as they have nationally.


    Of course, poverty has not been stricken

    from the face of the Wisconsin prairie.

    Even if every individual who enrolled in


    had found full-time work right away

    and kept their jobs forever, many of those

    8 Wisconsin Department of

    Workforce Development figures.

    Note that 9000 welfare cases

    without an adult (child-only) are

    excluded from both figures.

    9 Wisconsin Department of

    Workforce Development, Case

    Closure Study

    , Wisconsin

    Department of Workforce

    Development, 2000

    10 U.S. Census Bureau, see


    11 Urban Institute,

    Overview of

    Research Related to Wisconsin


    , Urban Institute, March


    12 As an example see University

    of Chicago Chapin Hall Center

    for Children,

    Issue Brief #107


    May 2006.

    13 Reliable statistics are difficult

    to obtain; anecdotal reports from

    author’s interviews with the

    director of Hope House,

    Milwaukee’s largest shelter, indicates

    that the total number of

    overnight stays has remained

    relatively constant since

    Wisconsin Works

    was introduced;

    fewer individuals are

    entering the shelter but this is

    offset by an increase in average

    length of stay. Also see

    Milwaukee Continuum of Care

    Steering Committee,

    Homelessness in Milwaukee


    July 2007 for a point in time survey.

    14 Wisconsin Office of Justice


    Crime and Arrests

    in Wisconsin 2006

    , Wisconsin

    Office of Justice Assistance, pp

    105, 2006

    15 See Urban Institute,


    Reform’s Effect on Child Welfare


    , Urban Institute,

    February 2001; University of

    Chicago, Chapin hall center for

    Children, Issue Brief #107, table

    5, May 2006.

    The experience of privatization of welfare services in Wisconsin


    • 43

    starting in entry level jobs would go

    through a period of income below poverty

    on the presumed way to wage increases

    over time. Yet wage growth has been

    stronger than academics predicted. An

    early study from the University of

    Wisconsin reported:

    In part because of the substantial caseload

    reductions that preceded the implementation

    ofW2, many of the first participants

    inW2 had low levels of education, substantial

    family responsibilities, and a history

    of reliance on welfare. Notwithstanding

    these barriers, we find higher levels of

    employment than have been found in other

    states, and substantial growth in employment

    and earnings over the short period

    considered. Average family income…

    increased, rising from about $12,000 in

    1998 to nearly $15,000 in 1999, and the

    poverty rate [among participants] fell from

    77 percent to 67 percent


    Looking over a longer period, for those

    who left


    at the end of 1999 and had

    full or part time employment in each of

    the following four years, average annual

    income increased by 22%. Of this group

    37% had exceeded poverty based on

    income alone, while 57% were above

    poverty after tax credits.

    17 18

    Privatization After Implementation

    While traditionally about half of

    Wisconsin’s welfare population lived outside

    of Milwaukee, after the sharp caseload

    reductions more than four fifths of the

    remaining small number of cases reside in

    the county, which is fully privatized, so

    that will be the focus of this section.

    However, it should be noted that in the

    balance of the state, the numbers of counties

    which have elected to privatize has

    steadily increased, from eight in the first

    contract, to 21 in 2005. Of the privatized

    counties, one is a medium sized city, three

    are suburban Milwaukee, and the other 17

    are rural.

    Within the rural areas ofWisconsin, the

    growth in the privatization of



    sense in part because there are so few

    remaining cases, and some counties no

    longer think of welfare as one of their main

    lines of business (superseded by foster care

    and adoption, medical assistance, distribution

    of food stamps, subsidized child care

    etc.). In the early years of



    in which funds were plentiful because of

    caseload reductions, many counties

    merged their related programs into “onestop”

    job centers with modern offices and


    Among the five contract vendors the

    state selected for Milwaukee, one might

    consider three of these high-performing

    (one of these got off to a bad start and

    recovered), one average, and one poor (the

    last is no longer in business).

    The transition

    between the old program and the

    onset of

    Wisconsin Works

    under the new

    privatized agencies, from those who experienced

    it, can best be described as an intensive,

    compressed and creative fervor.


    The new private agencies were not only

    operating under a completely new set of

    contract provisions, but also under a new

    mindset. Competition for good operations’

    directors was pronounced and in addition

    agencies responded by looking for line staff

    outside the traditional social work background.

    Some agencies advertised in sales

    and marketing magazines and sites because

    they were looking for employees comfortable

    interacting with businesses in a more

    professional environment and willing to

    accept contingent compensation based on


    In one high-performing agency the ratio

    of new hires to transitioned staff was high

    – about three to one. In others agencies it

    was lower. Of the older staff who had

    worked in county government, some transferred

    to other county human service jobs,

    some retired voluntarily, and some were

    Paying for success


    16 University of Wisconsin,

    Institute for Research on


    Child Support

    Demonstration Evaluation, Phase

    I Final Report

    , pp 2, 2001

    17 State of Wisconsin,

    Legislative Audit Bureau,

    Evaluation of the Wisconsin

    Works Program

    , table 27, April


    18 Another criticism is that

    Wisconsin’s work-first, or labor

    attachment model, would benefit

    from the addition of more hours

    of education and training, or

    even that education and training

    should substitute for community

    service work hours (note that

    non-work activities including

    training have been increasing as

    the program matures, but is

    intended to be an adjunct not a

    substitute for work). However,

    education and training as an

    intervention for the welfare population

    has been exhaustively

    studied over thirty years, including

    using extensive random

    assignment experiments, and

    has consistently been demonstrated

    to have barely detectible,

    or no detectible impact on

    wages or employment (one

    experiment in Portland Oregon

    with mixed work and training

    emphasis showed some


    19 This categorization is based

    on the judgment of the author of

    this paper, taking into consideration

    general performance over the

    entire contract periods, reputations

    within Milwaukee and the

    state, and the author’s own

    knowledge. The three high performers

    have been Goodwill

    (Employment Solutions), YWorks

    (YWCA lead agency); andMaximus.

    The average performer has been

    United Migrant Opportunity

    Society (UMOS) and the poor

    performer was Opportunities

    Industrialization Corporation (OIC)

    which is now out of business. The

    high performer which got off to a

    bad start and recovered is

    Maximus. The organization hired

    a poor director who was fired

    after a scandal, after which operations

    improved markedly. As of

    the date of this paper, the state

    has added new vendors and

    organized the contracts differently,

    which is not taken into consideration


    20 The descriptions which follow

    are based on interviews in

    January and February 2008 with

    Milwaukee agency directors and

    managers who were present at

    that time, and the author’s own


    hired by the new agencies as at-will

    employees. With competition for good

    staff, salaries increased. Former workers

    earned a salary of around $28,000 per year

    (1997 dollars) with sizeable benefit packages;

    new workers in one agency earned

    $35,000 to $38,000 with annual bonus

    opportunities of three to five thousand

    dollars, albeit with lower benefits. All

    agencies had their employees sign noncompete

    contracts, as they were concerned

    that the expense of training on this new

    program would be lost to competitors (in

    the event, no enforcement of these contracts

    ever occurred).

    In addition to hiring from the outside,

    agencies also directed their relationships to

    employers. All agencies paid special attention

    to assuring their job developers had

    solid connections with employers and

    access to jobs. One agency created a new

    title called “business service representative”

    which did no casework but whose

    job it was to take orders from employers

    and act as their representative within the

    agency in order to find the best employee

    fit. Business service reps would interact

    with caseworkers, each with their own

    bonus structure and incentives for job

    placement, but with differing internal customers.

    As pressure mounted prior to the start

    date of


    , managers and new staff

    worked late and the state helped by keeping

    its caseload computer system open

    nights and weekends. Each case transitioned

    on the computer from welfare to


    would take two hours; there were

    thousands of them to be changed. One

    agency set up a separate room with 20

    computers and a knowledgeable programmer

    to allow case workers to have quiet

    time for the transition. All this was occurring

    as new offices were being built and fitted-

    out (one agency built a new professional

    office from the ground up in the

    heart of a district which had once seen race


    Finding good middle managers is always

    key in government and human service

    organizations (and experience shows almost

    always more difficult than finding good

    agency directors or good line staff), and

    this task was made more difficult because

    new middle management hires were not

    familiar with the technical aspects of the

    job, nor the welfare rules, which had to be

    understood in order to explain to staff how

    to operate in the new environment. The

    absence of quality middle managers has

    been a major shortcoming of agencies

    which performed below expectations.


    In addition to their internal pressures,

    agencies had to go outside and to the public

    to assure that existing welfare recipients came

    into the offices and changed their enrollment.

    All agencies made special efforts, using

    door hangers, going to churches, being interviewed

    on television, and in some cases doing

    home visits if individuals did not respond to

    multiple letters. In the end, about 30%of the

    former welfare recipients declined to enroll in

    the new program (some were already working

    off the books, others might not be eligible,

    still others did not want to participate

    under the new work requirements).

    Work sites had to be prepared, and some

    agencies showed creativity here too.One purchased

    a plastics fabrication plant in order to

    provide easy work which would be fiscally

    self-sustaining, another used its own packaging,

    used clothing store, and commercial

    laundry service, still others used the county’s

    existing program with many outdoor assignments

    such as keeping parks maintained.


    Competition among vendors during this

    period was constructive. Each had its own

    geographic region, so in a sense they were

    not competing for “customers”. Every

    agency had its own performance goals and

    financial targets, but at the same time they

    did not want to be perceived as lagging.

    Once a month the five agency executive

    directors would meet, to hammer out common

    concerns (and, to a degree, to act as a

    counterweight to the powerful state agency).


    • 45

    The experience of privatization of welfare services in Wisconsin

    21 Good middle management is

    more important in government

    and quasi-government than in

    the private sector. There are

    more rules which must be

    observed, more direct supervision

    needed below because

    compensation is not directly tied

    to achievement, and more complex

    processes occur between

    top management directives and

    line staff implementation.

    22 However, the existing community

    service job inventory in

    Milwaukee operated by the

    county, which had been set up

    to accommodate single men and

    women under the recently terminated

    “general assistance” program,

    was insufficiently utilized




    Inter-agency work groups were formed

    around subject areas such as transitioning

    cases from one program to the other.

    The Successive Contracts

    During the first contract period the savings

    far exceeded the most optimistic expectations

    of the program’s designers. In

    Milwaukee, $319 million had been allocated

    for all expenditures including cash benefits.

    Of this amount 84% was spent, leaving

    $26 million for agency profits, $13 million

    for reinvestment in the community

    and $13 million to the state in the form of

    savings, just as provided for in the contract.

    Certain critics of


    have said that the

    first round of contracts did not contain performance

    criteria, but its designers certainly

    did not view it this way, nor did the vendors.

    TheseMilwaukee vendors were taking

    substantial risks in a completely unknown

    environment. Their performance would be

    measured by their employment success leading

    to caseload reductions, while leaving

    them with the operating room to get there

    in the best and most innovative way.

    Critics also claimed that agencies could

    simply reduce costs by denying service to

    applicants, or arbitrarily categorizing them

    as job-ready, thereby excluding access to

    community service jobs.

    This objection too, was unfounded.

    Agencies created procedures to move applicants

    through their system, and while individual

    cases might not be processed properly,

    no agency would risk denying service

    with the stakes so high (in addition, the state

    could penalize an agency $5000 for each

    instance of denied service, although it never

    found it necessary to invoke such a penalty).

    Nevertheless there were significant

    problems with vendor implementation.

    Two of the high-performing agencies billed

    hundreds of thousands of dollars improperly

    to the


    program as they chased new

    business outside the state and billed


    for employee work there.

    Both these

    agencies reimbursed these disallowed costs

    and made additional payments in compensation.

    However, one of these two agencies

    declined to bid on a third contract period

    as a result. Its loss was to be a great detriment

    to operations in Milwaukee.

    24 25

    In the face of bad publicity over the misallocation

    of funds, the state agency overreacted.

    In the third contract period the state

    added a surfeit of performance goals. It

    sharply reduced operating funding levels to

    below what was necessary, and permitted

    any performance bonuses to be earned

    only as a reimbursement for funds the

    agency had already expended, putting

    these agencies in a very serious financial

    risk were they to fail to meet the standards

    after the fact. In short, the state had shifted

    to the agencies almost all risks but without

    any opportunity for rewards.

    In addition to pressing its vendors beyond

    what was necessary for good management,

    the state made a second mistake. After the

    exit of the high-performing agency which

    had misallocated funds, there remained two

    of six regions without an operator. The state

    had failed to cultivate other potential

    replacement vendors from outside

    Milwaukee and so doubled the contracts of

    two of the existing Milwaukee vendors,

    which was acceptable as far as this went. But

    in the subsequent fourth contract period the

    state gave three of the six regions to the lowest

    performing vendor, reshuffling the deck

    once again and in a way which proved most

    unfortunate.The winning vendor’s executive

    director and others with whom he had ties

    were indicted for kickbacks in 2004 and sent

    to prison.

    The agency folded soon after.

    The Milwaukee operations are recovering

    from these setbacks. In the most recent

    contract period the state has at last brought

    in fresh vendors from outside the county,

    increased the contract responsibility of

    incumbent strong vendors, and management

    has significantly improved. In an

    indication that the state intends to devote

    more of its attention to


    , it

    Paying for success


    23 The agencies were Maximus,

    a for-profit vendor, and

    Employment Solutions, a forprofit

    subsidiary of Goodwill

    Industries of Southeastern


    24 The agency was Employment

    Solutions, which had contracts

    in two regions and performing

    exceptionally well.

    25 It should be noted that

    improper expenditures are not

    the exclusive province of the private

    sector. The state routinely

    audits county government

    expenditures and finds improper

    expenditures requiring reimbursement

    (for example a state

    audit of the Milwaukee County

    Hospital discovered such a high

    level of improper expenditures

    that the county resorted to selling

    the hospital, i.e. privatizing

    it). However, unlike government

    misallocations, private sector

    transgressions are considered

    more serious violations of the

    public trust by all concerned,

    and probably should be.

    26 The agency was

    Opportunities Industrialization

    Corporation of Greater

    Milwaukee, now defunct.

    has dedicated a new agency to the function.



    caseload is stable and low,

    and participant exit for employment continues

    at a high rate (though understandably

    slower than in the first early period

    after implementation).

    Lessons from the Wisconsin Works

    Privatization and Contract

    Management Experience

    Looking back from the vantage point of ten

    years, one sees that

    Wisconsin Works

    unleashed a wave of change on a scale unanticipated

    and unprecedented. Dependency

    was reduced far more than expected to the

    point that it is now considered only a secondary

    public policy issue. Former welfare

    recipients proved far more resilient than

    believed, took available jobs in large numbers,

    increased their wages over time and

    reduced the corresponding poverty rate. The

    state saved enormous sums as benefit payments

    to the formerly dependent were

    replaced by wages. The state legislature has

    re-allocated these funds in a variety of activities

    including programs for youth, literacy,

    nutrition, and domestic violence prevention.


    While the newly available funds were a

    welcome indication of program success it

    created a challenging political dynamic,

    with critics denouncing the profits earned

    by the private agencies. The state should

    have been better prepared to forthrightly

    make its case that program savings were

    only available because families were newly

    self-sufficient, and that a portion of the

    savings were being plowed back into the

    community through the reinvestment provisions

    or through state reallocations.

    The first lesson here is that program

    changes on the scale of


    , combined

    with privatization and new vendor contract

    incentives, will require swift iterative

    actions on the part of each actor, state and

    vendor community, in order to maintain

    an equilibrium and manage through the

    learning process.

    In addition, it is evident that the administration

    which took over after the election of

    2002 (Governor Jim Doyle, Democrat) lost

    focus on the program. In a familiar turn of

    events in every democracy, newly elected

    officials must stake out their mark distinct

    from their predecessor’s, and Wisconsin

    Works was indelibly linked to Governor

    Thompson, the previous post-holder.

    The new administration brought its

    own philosophy to governing, but in a way

    that was inconsistent and reactive. In

    2003 the state declared that the program

    should be participant-friendly and explicitly

    offer a greater array of services. It

    instructed the vendors to assure that all

    barriers to participation in the program be

    identified and resolved before reductions

    in benefits due to non-participation were

    taken, partly in response to program critics

    (due to the new rule’s ambiguity, vendors

    sometimes allowed inactive participants to

    receive cash benefits). Then, not long

    thereafter the state reacted to a resulting

    increase in the


    caseloads by again

    emphasizing early employment and work,

    leading to confusion on the part of agencies

    and vendors.


    The second lesson, then, is that the governing

    agency must maintain an interest in

    the state of the vendor marketplace and in

    the incentive signals it is sending. Vendors

    do not operate on autopilot, and will feel

    comfortable taking risks where they feel

    their footing is secure. Frequent significant

    changes in stated program objectives

    reduce performance; smaller changes at

    first are more readily absorbed.

    The state initially selected good vendors

    in Milwaukee (along with at least one poorperforming

    one) but it did not tend to the

    creation of a vibrant marketplace for the

    growth of competitive services there.

    People often take for granted that there will

    always be competitors and options in the

    private sector marketplace for goods and

    services. However the competitive marketplace

    environment for human service deliv



    • 47

    The experience of privatization of welfare services in Wisconsin

    27 Department of Workforce


    Letter from

    Secretary to State Auditor


    Department of Workforce

    Development, March 31 2001.

    28 State of Wisconsin,

    Legislative Audit Bureau,

    Evaluation of the Wisconsin

    Works Program

    , pp 98, April


    29 The Netherlands has done

    this, creating a new vendor community

    for job placement of several


    30 As an example of process

    measures that the initial contract

    did not measure but should

    have, is the weekly average

    number of work hours assigned.

    Lack of attention to this measure

    is in part responsible for its

    decline from 26.5 average weekly

    work hours in 1998 to 17.7 in

    2004. See State of Wisconsin,

    Legislative Audit Bureau,

    Evaluation of the Wisconsin

    Works Program

    , table 37, April


    31 Performance standards and

    measurements included required

    caseload ratios, caseworker training

    requirements, customer satisfaction,

    new required basic skills

    training for participants, wage

    increase measurements, job

    retention, full or part-time employment

    measures, among others.

    32 For instance, agencies allocated

    funds among program

    components differently, had variations

    in sanction rates, job

    placement and retention rates,

    performed differing participant

    assessments, used job access

    loans at different rates

    33 An estimated figure; Central

    Intelligence Agency (2008)


    2008 World Factbook


    Central Intelligence Agency).

    34 DeNavas-Walt, C.; Proctor,

    B.D.; and Smith, J. (2007)

    Income, Poverty, and Health

    Insurance Coverage in the

    United States: 2006


    States: US Census Bureau).

    ery remains insufficiently developed. The

    third lesson then, is that the state itself must

    actively create and nurture a competitive

    marketplace environment. It should recruit

    potential vendors from outside its borders

    even when no bids are active, it should conduct

    annual meetings with prospective vendors

    to discuss the program state of affairs,

    and it should consider setting aside a certain

    portion of each contract for new entrants

    into the marketplace.


    What about the nature of contract performance

    criteria themselves? The initial

    contract incentives in which agencies shared

    in reduced expenditures was powerful and

    outcome-based but did not give sufficient

    information about ongoing program activity

    levels such that thorough reports could be

    made to the public and policy makers.


    the other hand, in the third contract period

    the state created so many measurement

    standards that taken together they did not

    drive agency performance toward an understandable

    or sufficiently narrow mission


    The fourth lesson, then, is that

    the state should determine its most important

    policy objectives, organize its performance

    payments for these using outcome

    based measures (but a limited number of

    process measures too), and refrain from trying

    to achieve all conceivable desirable

    objectives through one contract.

    The fifth and final lesson from the

    Wisconsin experience is that the energy

    and creativity generated by private vendors

    in conjunction with powerful contract

    incentives were together essential to

    achieving the program success. During the

    most intensive and creative period of the

    Wisconsin Works

    transition, the new

    Milwaukee vendors, the other county

    agencies and the state worked cooperatively

    at a breakneck pace to get ready. Each

    organization knew its responsibility and

    mission objectives. Once operations began,

    each Milwaukee agency approached its

    objectives by operating their programs in

    different ways with differing emphases.


    And when all is said and done, including

    all the programs avoidable and unavoidable

    mis-steps, the lives of Wisconsin’s

    poor have been permanently improved.

    Paying for success


    35 United States Department of

    Labor, News: The Employment

    Situation: December 2006



    36 DeNavas-Walt, C.; Proctor,

    B.D.; and Smith, J. (2007)

    Income, Poverty, and Health

    Insurance Coverage in the

    United States: 2006


    States: US Census Bureau).

    37 Social Security

    Administration (2002)


    Security Bulletin, Statistical

    Supplement 2002


    States: Social Security




    39 US Department of Labor


    Evaluating State

    Unemployment Insurance (UI)

    Tax Systems Using the

    Significant Tax Measures Report

    (United States: US Department

    of Labor).

    40 United States Department of

    Labor, News: T

    he Employment

    Situation: December 2006





    42 Social Security

    Administration (2002)


    Security Bulletin, Statistical

    Supplement 2002


    States: Social Security


    43 Author’s calculations from

    Social Security Administration




    44 Social Security




    45 US Department of Health and

    Human Services, TANF Financial





    46 US Department of Health and

    Human Services (2006)

    Temporary Assistance for Needy

    Families Program (TANF):

    Seventh Annual Report to


    (United States: US

    Department of Health and

    Human Services).

    47 US Department of Health and

    Human Services




    United States


    GDP , 2007: $13.8 trillion


    Population, 2005/6: 296.4 million


    Number of people in the labour force, 2005/6: 152.7 million


    Average weekly wage among adults in full-time work, 2005/6: $735


    Total spending on welfare , 2005/6: $2.6 trillion37


    Total spending on unemployment benefit, 2005/6: $46.1 billion


    Rate of per week unemployment benefit, 2005/6: $275


    Number of people unemployed, 2005/6: 6.8 million


    Number of unemployed people in 2005/6 out of work for one year or more: 1.3 million


    Number of days formerly unemployed people must be in work for their assistance to be considered

    successful: varies by state.


    Total spending on incapacity benefit, 2005/6: $52.9 billion


    Rate of per week incapacity benefit, 2005/6: £237.21


    Number of people on incapacity benefit, 2005/6: 812,000


    Total spending on lone parent benefit, 2005/6: $11.3 billion


    Rate of per week lone parent benefit, 2005/6: $82


    Number of people on lone parent benefit, 2005/6: 1 million


    The labour market

    reform in Germany

    and its impact on

    employment services

    by Hilmar Schneider


    The German Labour Market Reform moved

    through several steps between 2003 and

    2005. It accompanied many changes in all

    areas of active and passive labour market policy.

    One of the major ideas of the so-called

    Hartz-commission, which was commissioned

    by the German government in 2002 to develop

    reform proposals, was to reduce unemployment

    by cutting the amount of time people

    spent out of work. It proposed placement

    procedures andmeasures to impair the receipt

    of unemployment compensation.

    Since many changes took place at the

    same time, it is difficult to disentangle the

    contribution of each of them. However,

    experts agree that the refoms, as a whole, significantly

    contributed to the sharp decrease

    in unemployment since 2005, although this

    still has to be proved in terms of econometric

    evidence. The number of people out of

    work dropped by one million between 2005

    and 2007, the biggest fall in Germany’s postwar

    history (see Figure 1).

    The German system of

    unemployment benefits and welfare

    Before 2004 there were three benefit levels:

    unemployment support (Arbeitslosengeld),


    • 49












    1960 1964 1968 1972 1976 1980 1984 1988 1992 1996 2000 2004

    reunited Germany



    Figure 1: Registered unemployment in Germany (thousands)

    Source: Federal Labour Agency

    unemployment aid (Arbeitslosenhilfe), and

    welfare (Sozialhilfe). Since 2005, unemployment

    aid has been abolished. Welfare

    is now called Arbeitslosengeld II.

    Workers (but not the self-employed or

    so-called ‘minijobbers’) are required to

    contribute to public unemployment insurance.

    The Federal Labour Agency (FLA)

    administers the scheme. They are entitled

    to receive support when out of work. The

    maximum duration of unemployment

    support is one year, depending on the

    number of months employed in the preceeding

    time period.

    For the elderly, the maximum duration

    of unemployment support is gradually

    extended. Until January 2006, it lasted for

    up to 32 months for job-seekers above the

    age of 57. But this was frequently used as

    an early retirement device. Since February

    2006, the maximum benefit period has

    been 18 months. From 2008 on it will

    again be 24 months.

    The level of unemployment support is

    not means tested and is only related to previously

    earned income. For job-seekers living

    without dependent children, it

    amounts to 60% of previous remuneration.

    For job-seekers living with dependent

    children, it is 67%.

    Up to 2004, job-seekers coming to the

    end of their unemployment support automatically

    qualified for unemployment aid.

    Although means tested, the latter was not

    subject to a time limit. Its maximum level

    was also attached to previously earned net


    For job-seekers living without dependent

    children, unemployment aid amounted

    to 50% of previously earned income.

    For job-seekers living with dependents, it

    came to 57% of previous earnings .

    Unemployment aid was paid by the FLA,

    but funded by federal taxes.

    Job-seekers without other sources of

    income are eligible for welfare. It is currently

    fixed at 347 euros per month

    nationwide (plus coverage of housing

    costs, which may vary regionally), for a

    single person. On average, this means

    around 650 euros per month for a single

    person and around 1,800 euros per month

    for a four-person household.

    Income from employment is largely

    deducted from welfare claims, something

    which causes a classic welfare trap for lowskilled

    workers by generating an implicit

    minimum wage. For a single worker it is

    around 10 euros per hour.

    In all, Germany is currently spending

    about 1.6% of its GDP on unemployment

    support and welfare. This figure doesn’t

    include spending on active labour market

    policies, which account for another 0.7%

    of GDP.

    Job-seekers subject to unemployment

    support have to notify their local office of

    the FLA. The same was true for job-seekers

    eligible for unemployment aid.

    Themunicipalities are responsible for welfare

    claims. Up to 2004, they had, in principle,

    to cover welfare costs on their own. In

    cases of financial weakness, however, they

    were supported by their federal states.

    Since 2005 joint offices of the FLA and

    local municipalities have been established

    (Arbeitsgemeinschaften) – a change that

    was meant to eradicate antagonism

    between local welfare authorities and the

    FLA (which had led to useless cost-shifting).

    Previously, municipalities gained by

    sending their clients to public employment

    programs for one year rather than investing

    in placement or training. In doing so, the

    workers gained sufficient eligibility for

    unemployment support during that year,

    and the municipalities moved them off

    their books for a while.

    The change was heavily opposed by the

    municipalities from the very beginning.

    They were upset because the government

    eliminated the former cost-shifting opportunity

    at the same time. Moreover, the

    municipalities feared losing even more

    autonomy by being forced into a joint

    operation with the FLA.


    Paying for success

    The labour market reform in Germany and its impact on employment services


    • 51

    The resistance of themunicipalities led to

    a compromise during the legislative process:

    69 of them were exempted from establishing

    a joint office. However, the performance

    of both organizations was to be evaluated

    before a final decision had to be made

    (experimental clause, § 6c SGB II).

    Meanwhile, the constitutional court

    decided that joint operations of federal

    institutions and municipalities was not in

    accordance with Federal Law so the change

    has to be abolished by 2010. The government

    now has to decide whether responsibility

    for welfare is to be completely given

    to the Federal State or the municipalities.

    The most likely outcome is a form of

    responsibility similar to the one used

    before 2005.

    The Reform of Placement Activities

    The reform itself was triggered by the socalled

    placement affair. In early 2002, it

    became evident that the number of placements

    claimed by the Federal Labour

    Agency (FLA) was remarkably biased

    upwards, because it used to accredit a large

    part of successful individual job search to

    itself. Up until then, the FLA had a de

    facto monopoly on placement activities.

    It turned out that only about 10% of

    the more than 90,000 employees of the

    FLA were actually engaged in placement

    activities, although placement was widely

    viewed as one of its major tasks. Not surprisingly,

    the process operated by the FLA

    became the subject of harsh criticism. As a

    consequence, privatization and re-organization

    of the schemes run by the FLA

    emerged as a suitable solution.

    Mainly in 2003, but also in 2002, several

    new placement instruments were introduced:

    In March 2002 the FLA started issuing

    placement vouchers (§ 421g SGB III).

    Each one entitles its owner to use a private

    placement agency after three

    months of unemployment (six weeks

    since 2005). If the agency successfully

    places a client, it receives an incentive

    payment from the FLA.

    The first payment is made at the beginning

    of employment, the remaining

    part only after six months of continuous

    employment. There are no market

    entry barriers to service providers.

    Originally, the value of the voucher was

    determined by the duration of unemployment.

    The first tranch could be

    encashed on the very first day of

    employment. This, however, induced

    fraud and deadweight loss. Thus, since

    2005, the first tranch is paid out only

    after six weeks. The total value of the

    voucher has also been fixed at 2000

    euros. Abuse fell substantially after the


    As an alternative to placement vouchers,

    the FLA, since 2002, has been

    allowed to contract-in third parties to

    carry out placement activities. After six

    months of unemployment, job-seekers

    may also now claim the use of these private

    agencies (§ 37 SGB III).

    Contrary to placement vouchers, the

    FLA has discretion over performancerelated

    bonus payments. Since 2003,

    private agencies have also been eligible

    for compensation if they are engaged in

    the placement of specific target groups

    like the elderly, youth, and long-term

    unemployed (§ 421i SGB III).

    In 2003, Public Temporary Work

    Agencies (PSAs) were introduced. They

    were meant to operate as the flagships

    of the reform (§ 37 c SGB III). Each of

    the 180 regional districts of the FLA

    had to establish at least one.

    The general idea was to make systematic

    use of the commonly observed

    attachment effect of temporary work.

    Since roughly one third of employees

    working on behalf of a temporary work

    agency are thought of as likely to be

    permanently hired by the lessee sooner

    1 See

    Le Grand J, Motivation,

    Agency, and Public Policy: Of

    Knights and Knaves, Pawns and


    , Oxford University

    Press, 2003


    Paying for success

    or later, it appeared to suggest the use

    of temporary work as a placement tool

    (with a special focus on hard-to-place

    unemployed). Therefore, PSAs were

    subsidized for hiring the hard-to-place

    unemployed and could also claim a

    bonus payment for each individually

    successfully placed at a lessee. The government

    intended to place 500,000

    people via the change.

    The main problem with outsourcing placement

    services arises from an unusual trilateral

    contract relationship between service

    provider, client, and paying authority (see

    Figure 2). A job-seeker makes use of a service,

    but does not have to pay for it. The

    service provider has to claim for bonus

    payments at the FLA, but the FLA can

    only indirectly judge whether this claim is

    justified. Because of a lack of cost control,

    such a constellation is vulnerable to serious


    According to contract management theory,

    the main problem is one of differences

    between principals and agents.

    The former

    is a representative of the public interest who

    wants to maximize public utility. The latter

    are public authorities like the FLA, external

    service providers and job-seekers.

    Within this setting, the FLA and service

    providers have to cope with their specific

    principal agent problem. Even if the head

    of each institution behaves in accordance

    with public interest, this may not necessarily

    be true of his employees. Each individual

    follows his or her own interest, which

    may mean he or she doesn’t follow the aims

    of the institution.

    Economically, this could be overcome

    via an internalization of economic risk, e.g.

    by cost-sharing between job-seekers and

    the FLA. This would stimulate job seekers

    to insist on service quality directly from

    the service provider and thus ensure that it

    behaves as the FLA wants.

    However, this was not implemented.

    Instead, FLA monitoring relies on information

    reported by service providers.

    Verification by the FLA is indirectly based

    on its records of the former job seeker and

    his or her time at a new employer.

    Evaluation results

    The reform act as a whole was accompanied

    by a Parliamentary resolution saying

    that each instrument had to be evaluated

    within the shortest possible time period

    (BT-Drs. 15/98). Based on this statement,


    Provider Client

    Federal Labour







    Payment Service

    Standard contract relationship:

    Trilateral contract relationship:

    Figure 2: Standard contract relationship vs. trilateral contract relationship

    a large scale research project was commissioned

    by the Federal Ministry of Labour

    and Economics, involving about 80

    researchers from various German research

    institutes and universities. The final

    reports were due in June 2006.

    External placement services were evaluated

    within module 1a “Reform of

    Placement Processes”, carried out by a

    research consortium under the leadership

    of the Social Science Research Center

    Berlin (WZB).

    Using a quasi-experimental

    approach that compared ex-post recipients

    of a placement voucher or participants

    in external placement activities to nonrecipients

    or non-participants, they mainly

    focused on individual effects. For each

    recipient a similar non-recipient was identified.

    Analogously, matched pairs were

    formed for participants and non-participants

    in external placement activities.

    Another module was devoted to macro


    The idea was to control for substitution.

    It could be that certain instruments

    have a positive impact on the individual

    level, but only at the expense of reducing

    the employment prospects of non-participants.

    In such a case, the overall effect on

    employment could end up in a zero-sum


    Unfortunately, however, the necessary

    aggregate data for placement vouchers and

    external placement activities are not available.

    It was therefore impossible to produce

    macro level results for the instruments

    under investigation.

    Independently of the research project

    commanded by the German government, a

    specific project on placement vouchers and

    external placement activities was carried

    out by the Centre for European Economic

    Research (ZEW) on behalf of the FLA.


    The focus of this study was micro-oriented

    and based on a matched-pair approach as


    Conditional independence is a central

    assumption of the matched-pairs method.

    This means that the treatment effect is

    always the same, given a set of characteristics

    of a certain individual. As a consequence,

    the treatment effect can be identified

    statistically by taking a sufficient number

    of pairs of individuals, where each pair

    is identical in terms of observable characteristics,

    except for the treatment.

    However, participants and non-participants

    may not differ systematically with

    regard to unobserved characteristics that

    affect the outcome. If this condition is violated,

    the pairs differ in more than only the

    treatment so it is not possible to ascribe

    differences in the outcome to the treatment


    Unfortunately, the validity of the condition

    of irrelevant unobservables can neither

    be tested nor proven. People could differ,

    for example, with regard to their general

    activity level, which is difficult to observe

    when based on usual characteristics like sex,

    age, education, profession, and the like.

    But if people who claim for a placement

    voucher are more active in general than

    non-claimants and if this activity level

    makes it more likely that claimants find a

    job, a comparison of claimants and nonclaimants

    in terms of job finding rates is

    biased by such unobserved a priori differences.

    It is therefore essential to make the

    best possible use of available observables.

    The two studies at hand were carried

    out thoroughly enough in that respect.

    However, the WZB study reports a

    remarkable incidence of deadweight loss

    under placement vouchers, which means

    that in a considerable number of cases,

    placements were declared, ex-post, as

    induced by a service agency, even though

    the match came about independently.

    Statistically, this is equivalent to a violation

    of the conditional independence assumption.

    Consequently, the results reported

    later on should be taken with caution.

    The recent quantitative importance of

    the three placement instruments under

    consideration can be found in Figure 3.

    Vouchers are playing a minor role. On

    2 Wissenschaftszentrum Berlin

    für Sozialforschung, Institut für

    Angewandte Sozialforschung,

    Evaluation der Maßnahmen zur

    Umsetzung der Vorschläge der

    Hartz-Kommission – Modul 1a:

    Neuausrichtung der Vermittlungsprozesse


    Bericht 2006 für

    das Bundesministerium für

    Arbeit und Soziales, 2007

    3 Rheinisch-Westfälisches

    Institut für Wirtschaftsforschung,

    Institut für Sozialforschung und



    der Maßnahmen zur Umsetzung

    der Vorschläge der Hartz-

    Kommission – Modul 1f –

    Verbesserung der


    Rahmenbedingungen und

    Makrowirkungen der Aktiven


    , Bericht 2006

    für das Bundesministerium für

    Arbeit und Soziales, 2007

    4 The results are published in

    Pfeiffer F, Winterhager H,

    “Vermittlungsgutscheine und

    Beauftragungen Dritter im

    Vergleich”, Zeitschrift für


    , 39 (3

    and 4), pp 425-445, 2006 and

    Pfeiffer F, Winterhager H.

    Selektivität und direkte

    Wirkungen von Vermittlungsgutscheinen:


    Befunde aus der

    Einführungsphase, Perspektiven

    der Wirtschaftspolitik

    , 7 (3), 399-

    419, 2006

    The labour market reform in Germany and its impact on employment services


    • 53


    Paying for success

    average, 5,500 people per month were

    placed via vouchers in 2006, with a slight

    increase in 2007. The number given refers

    to the number of payments for the first

    tranch of the voucher. Official numbers for

    the second tranch after six months are currently

    not available.

    Compared to the initialization period of

    placement vouchers, the recent figures

    show a sharp decline. In total, the FLA

    issued 1.4 million placement vouchers in

    its initial version until December 2004.

    8% of them were encashed by the end of

    2005. The following decline was a reaction

    to increasing evidence of the already mentioned

    abuse of placement vouchers.

    The average stock of people subject to

    placement by third parties culminated in

    late 2005, and was followed by a rapid

    decline afterwards. Out of an average stock

    of about 100,000 people in 2006, they

    were able to place about 46,000 clients.

    The average stock of job-seekers

    employed at a PSA amounted to less than

    4,000 in 2007. There were 12,600 in

    2005, and 6,000 in 2006. Here again, the

    tendency is still downwards. In 2006, PSAs

    were able to place 12,000 clients. However,

    they are clearly well short of the ambitious

    target of 500,000 set by the government.

    The overall number of placements

    encountered by the Federal Labour

    Agency in 2006 sums to 915,000 (compared

    to 7.4 million outflows per year, 3

    million of them into employment).

    According to the figures given above, the

    number of placements by private or voluntary

    services amounted to 124,000 in

    2006. It is thus fair to say that placements

    outside public employment services

    accounted for 13.6% of the total figure in


    Placement Vouchers

    Among the new placement instruments,

    the voucher is the only one that freely

    allows a job seeker to choose a service

    provider. In principle, this should induce

    competition among service providers and

    therefore allow for efficiency gains. As

    mentioned above, however, the divide

    between service recipience and payment

    responsibility may offset or even outweigh

    potential efficiency gains.

    Both available studies report a positive

    impact from placement vouchers. The

    WZB study, however, restricts this result to

    the second generation vouchers, while the

    ZEW study explicitly refers to the introductory









    03 03 04 04 05 05 06 06 07 07 08

    Third Parties



    Jun Dec Jun Dec Jun Dec Jun Dec Jun Dec Jun

    Figure 3: Quantitative Importance of placement instruments

    Source: Federal Labour Agency, Nürnberg; while the numbers for Third Parties and PSA refer to the stock of participants, the

    number of placement vouchers refers to the number of vouchers encashed after an employment period of six weeks.

    The labour market reform in Germany and its impact on employment services


    • 55

    According to the WZB report, the

    employment rate of placement voucher

    recipients was 7.7 percentage points higher

    than that of non-recipients. But the

    authors point out that matches induced by

    a voucher are less sustainable than other

    matches in terms of average employment


    This seems to be especially true for

    males in east Germany. The regional effect

    is so strong that it dominates the overall

    average, although the sustainability of

    matches in west Germany does not differ

    much between recipients and non-recipients

    of vouchers. The authors take this as

    an indication for a considerable amount of

    deadweight loss – especially in east


    The market for private placement isn’t

    transparent. There is no official institution

    rating or evaluating private and voluntary

    employment services. Responsibility for

    ratings is more or less at the discretion of

    local authorities. However, they have little

    information on external placement agencies

    – something which would allow them

    to conduct performance monitoring.

    This is different from the market for

    training providers. Since 2005, each training

    provider has had to be approved by a

    certifying authority according to a number

    of quality criteria. The only device of quality

    assurance in the private placement sector

    that exists is a code of conduct, which

    has been passed by the association of private

    placement agencies.

    However, this code is not very binding,

    so in the absence of sanctions area-wide

    quality assurance still needs to be solved.

    Only plain straightforward placement

    voucher fraud may lead to market exclosure

    by commercial inspectorates and possible

    legal action.

    Due to the absence of transparent monitoring,

    low-skilled workers are particularly

    likely to be disadvantaged, not only

    because they have more difficulties in finding

    a suited private placement agency, but

    also because they are not very profitable for

    private placement agencies. The potential

    for quality improvement by competition is

    widely corrupted by lack of market transparency.

    Placement by Third Parties

    Placement by third parties comes in several

    components: profiling, activation, job

    aquisition, and case management. This

    section, however, concentrates on the

    holistic approach to placement by third

    parties. This is the rule underlying § 37

    SGB III since 2006, but has been in practice

    since 2003.

    Judging by official announcements,

    placement introductions by third parties

    were not meant as a competitive measure,

    but as support for public placement activities

    instead. This points to a delicate issue.

    Since third parties have to be commanded

    by the FLA, it is unlikely that the FLA will

    make active use of an instrument that has

    the potential to shed bad light on its own


    If the FLA commanded third parties, it

    would probably try to shift hard-to-place

    clients. Service providers cannot reject certain

    clients that are assigned for a placement

    program by the FLA.

    Commanding third parties became

    accomplished via a centralized and nationwide

    tendering procedure launched by the

    FLA. Especially in the beginning, the

    advertised lots were relatively large and

    covered more than one, local FLA district.

    Sub-lots were then assigned to local labour

    agencies, who could assign clients of their

    choice to the program. Large-scale tendering

    was explicitly intended to destroy local


    The employment rate of placement voucher recipients

    was 7.7 percentage points higher than that of


    According to a 2003 report by the

    German Federal Court of Auditors, corruption

    in local districts of the FLA was a

    major source of exaggerated prices for

    placement services. Because of large-scale

    tenders, the resulting offers were indeed

    dominated by supra-regional service


    In the aftermath, prices deteriorated

    despite the fact that in 47 of 104 tenders in

    2004/2005 only one or no offer was made.

    However, from the second half of 2005

    onwards, lots became smaller again after

    heavy complaints by local labour agencies

    that organizational disorder was caused by

    new and unknown service providers.

    The culmination of the upheaval was

    marked in 2005, when the strongest competitor,

    holding a market share of 26%,

    went bankrupt. Since then, the tendering

    process has remained centralized, but now

    refers to certain local districts of the FLA.

    Interestingly, competition has not made

    much progress since then. In 2006, the

    FLA received only a single offer in two

    thirds of its auctions.

    The low profitability of the tenders is a

    major reason. Especially for contracts based

    on perfomance related bonus payments, the

    risk is taken by the service providers.

    In 2006, roughly two thirds of all contracts

    were based on performance-related

    bonus payments. For the rest, contractors

    could claim an additional general expense

    claim of 75 euros per client. In cases of a

    purely bonus-based contract, the proportion

    of successful placements must be

    unrealistically high in order to make the

    operation beneficial for a private service

    provider. Not surprisingly, this increases

    the incentive to concentrate on “good”

    risks in order to maximize revenue.

    Evidence for subsequent creaming

    effects is clearly supported in both evaluation

    studies. The development was accompanied

    by a deterioration in wages and a

    rise of precarious employment at private

    service providers. This has to be taken into

    consideration when interpreting the poor

    evaluation results for external placements.

    While established local service providers

    became threatened by bankruptcy, expanding

    inter-regional providers had to recruit

    inexperienced and low-paid employees.

    The overall service quality declined.

    Programs now typically run for six or

    twelve months. The typical lot size lies

    between 50 and 150 participants.

    Competition between local service

    providers is more or less non-existent. This

    makes it difficult to compare the performance

    of different service providers, because

    each comparison encounters the peculiarities

    of different local labour markets.

    Consequently, the incentive for service

    providers to increase their effort lags

    markedly behind a situation where several

    providers can be directly compared when

    competing within the same local area.

    The impact analysis carried out by the

    WZB does not support the expectation that

    private placement is more efficient than

    public placement. It clearly shows that external

    placement has no measurable impact on

    placement speed compared to direct placement

    by the FLA. On average, it usually

    takes about 4 months for participants as well

    as non-particpants to re-enter employment.

    However, in terms of sustainability,

    placement performance by third parties is

    even worse than that of the FLA. While

    the average placement duration of jobseekers

    placed by third parties is 12

    months, comparable job-seekers directed

    by the FLA stay in their jobs, on average,

    for 16 months. Third parties perform par-


    Paying for success

    While established local service providers became

    threatened by bankruptcy, expanding inter-regional

    providers had to recruit inexperienced and low-paid


    ticularly poorly when dealing with the elderly

    or in regions with high unemployment.

    Only males in East Germany are

    benefitting from external placement.

    However, the reason for this is unclear.

    According to the WZB study, the average

    costs for placement by third parties

    amounted to 600 euros in 2004 and 725

    euros in 2005. Unfortunately, no reference

    is available for non-participants. The

    WZB study incorrectly assumes zero

    placement costs for non-participants,

    which is only valid if non-participants

    received no alternative treatment at all.

    Since this is not the case, it is not possible

    to assess the cost-saving potential of external


    Placement by Personal Service Agencies

    PSAs are publicly subsidized temporary

    work agencies. Job-seekers assigned to a

    PSA get a regular work contract there. A

    PSA may also train its clients in order to

    improve their employability.

    Contrary to a commercial temporary

    work agency, however, the PSA has to

    employ its workers throughout the contract

    period of normally 12 months, no

    matter whether they can be lent out or not.

    Depending on the employability of a jobseeker,

    this introduces a risk for PSAs.

    The PSA business model is based on a

    mixture of bonus payments for successful

    permanent placement and case-related lump

    -sum payments to cover wages and training

    by the FLA. The establishment of a PSA is

    subject to a tendering process. The level of

    bonus payments and case-related lump-sum

    payments has to be offered by bidders.

    The hiring rate of PSAs is a field of obvious

    conflict. It turned out that PSAs, on

    average, hired less workers than agreed

    upon in the contract with the FLA. Because

    they want to reduce their risks, they are

    reluctant to hire too many workers.

    Critics argue that PSAs are more or less

    a subsidy for commercial temporary work

    agencies and that the instrument generates

    nothing but deadweight loss. This was

    especially true during the introductury

    phase of PSAs, when lump-sum compensations

    were relatively high on average.

    However, the argument even holds after

    prices have deteriorated. On average,

    lump-sum compensation is now 500 euros

    per case per month. Given the economic

    risk of not being able to lend out workers,

    there is not much left for additional activities

    like active placement or training.

    At first glance, it now pays for the FLA

    to send job-seekers to a PSA – the lumpsum

    compensation is less than the average

    cost of unemployment benefit. However,

    one has to take into account that average

    duration of unemployment is higher for

    employees of a PSA than for other unemployed


    Compared to the high expectations of

    PSAs at their introduction, they have been

    a complete failure. According to the evaluation

    results from the observation period

    2003-2004, the situation is dominated by

    a strong locking-in effect.

    Being employed at a PSA seems to lower

    job-search activities. Compared to a control

    group, the employment probability of workers

    in a PSA sixmonths after leaving the PSA

    is, on average, 6.4% lower than that of the

    control group at the same point in time. On

    average, members of the control group

    entered into employment one month earlier

    than their counterparts at a PSA.

    By the end of 2005, the government

    abandoned the clause that each local district

    of the FLA had to run at least one

    PSA. This should be viewed as the virtual

    end of the instrument. Its quantitatve

    importance is now close to zero.

    Conclusion – Efficiency Gains by

    contracting out placement services?

    By and large, the German experience with

    contracting-out of placement services is not

    encouraging. However, it isn’t evidence

    against the usefulness of competition and

    The labour market reform in Germany and its impact on employment services


    • 57

    privatization in general.The introduction of

    private placement services is related to existing

    public institutions that are not really

    interested in competition.

    Since the FLA can selectively assign job

    seekers to different agencies, the poor performance

    of private service providers may

    have been actively brought about by the

    FLA in order to maintain its power. In any

    case, the existing studies show a latent

    danger of deadweight loss when introducing

    external service providers. This is especially

    true for placement vouchers and


    To some extent, the failure of private

    placement services is because of an

    unsolved principal-agent problem. Service

    clients do not have strong incentives to

    insist on service quality from providers.

    Even if they did, they could hardly sanction

    poor quality. As an option for overcoming

    the lack of immediate quality control,

    models of cost-sharing between clients

    and the FLA should be considered.

    Besides this, it seems that competition

    didn’t improvement quality enough. This

    is especially true for placement by third

    parties. To a large extent, this may be due

    to a monopsonistic market structure where

    the FLA is the major service demander.

    There is nomarket driven price increasing

    mechanism in case of bad quality or no supply

    at all. Competitors try to enter the market

    via price dumping. Consequently, the

    profitability of private placement for hard to

    place clientele seems to be unattractive.

    The German experience thus suggests

    that price flexibility among service demanders

    is important. Competition should be

    organized by the best output at a given

    price level The winner could be rewarded

    with an extra lot at the same price level.

    Continuous tournaments among a number

    of competitors at various price levels

    would allow for an optimization of costs

    and benefits at mid-price levels.

    5 An estimated figure; Central

    Intelligence Agency (2008)


    2008 World Factbook


    Central Intelligence Agency).

    6 Destatis.

    7 Of population aged between

    15 and 65 years. Source: IWH.

    8 Gross wage figures in West

    Germany used. Includes part

    time jobs and ‘minijobs’. Source:

    Deutsche Rentenversicherung

    Bund: Rentenversicherung in

    Zahlen 2007.

    9 Bundesministerium für Arbeit

    und Soziales, Sozialbudget


    10 General rule for

    Arbeitslosengeld used:

    60%/67% of previously earned

    net income. Source:

    Bundesagentur für Arbeit.

    11 Count of people registered as

    unemployed at the Federal

    Labor Agency (who could be

    employed for up to 15 hours per

    week. Figure does not include

    individuals in active labour market

    programs, early retirement

    schemes and suchlike (exclusions

    which amount to roughly

    25% of the registered unemployed).

    Source: Federal Labor


    12 Please note, these figures are

    systematically underreported

    due to artificial interruptions of

    unemployment duration caused

    by sickness and suchlike (standardized

    OECD figures report an

    increase from 48 to 57%

    between 2002 and 2006).

    Source: Bundesagentur für


    13 Deutsche

    Rentenversicherung Bund:

    Rentenversicherung in

    Zeitreihen, p. 199.

    14 Includes partial payments.

    Source: Deutsche

    Rentenversicherung Bund:

    Rentenversicherung in

    Zeitreihen, p. 134.


    Paying for success


    GDP in 2007: $3.25 trillion

    Population, 2005/6: 82.3 million

    Number of people in the labour force, 2005/6: 54.8 million

    Average weekly wage among adults in full-time work, 2005/6: $827.33

    Total spending on welfare, 2005/6: $1 trillion
    (30.2% of GDP).

    Rate of per week unemployment benefit, 2007: $260.91.

    Number of people unemployed, 2005/6: 4.5 million (11% of population)

    Number of unemployed people in 2005/6 out of work for one year or more: 1.9 million

    Number of days formerly unemployed people must be in work for their assistance to be considered

    successful: 180.

    Total spending on incapacity benefit, 2005/6: $20.7 billion

    Rate of per week incapacity benefit, 2005/6: $237.63


    The creation and

    regulation of markets

    in employment services

    – Danish experiences

    by Thomas Bredgaard


    When trying to contract-out Public

    Employment Services (PES)

    , decision-makers

    are confronted with two intricate challenges:

    first, to create a market structure that resembles

    a “real market” and then, second, to

    establish regulatory mechanisms for monitoring

    and controlling market behaviour.

    Both of these tasks are fraught with

    important questions: can cost-efficient markets

    supply high-quality services? How can

    free-market competition be combined with

    public regulation? Can public regulation to

    reduce transaction costs and opportunistic

    behaviour go together with ambitions to

    reduce public bureaucracy?How can private

    service providers that are motivated by economic

    incentives be encouraged to cater for

    the long-term unemployed with severe barriers

    to labour market entry? How can a

    competitive market structure be combined

    with partnerships, policy learning and

    mutual trust between providers and purchasers?

    Contracting-out is not a panacea;

    there are several answers to these problems.

    The Danish experience in creating and regulating

    a market for employment services is

    reviewed in this essay.

    Since 2003 Danish


    have been partially

    contracted out to various “other

    actors” (private service providers, educational

    institutions, trade unions, unemployment

    insurance funds etc). The two

    phases of the implementation of the tendering

    model, the initial creation of a market

    structure parallel to the


    , and the

    subsequent regulation of the market

    through a new structure of local jobcentres

    in each municipality are reviewed here.

    It’s clear from the review that the contracting-

    out of


    is not responsible for

    Denmark’s current low unemployment

    rates, but rather, are a result of the historic

    economic boom experienced over the past

    few years. Furthermore, the problems in

    creating and regulating markets for former


    have not been resolved in Denmark,

    although important lessons can be learned

    by reviewing its experiences.

    Historical and institutional background

    In order to know where we are and where we

    are going, we need to know where we have

    come from. In international comparisons,

    Danish labour market regulation is quite

    unique. Even if the welfare state follows the

    Scandinavian traditions of universal services

    and benefit provision and the public sector is

    large in international comparison, the labour

    market has traditionally been characterised

    by limited state intervention.

    TheDanishmodel is “liberal” in the sense

    that the state does not, in principle, inter



    • 59

    vene in the voluntary collective bargaining of

    labour market organisations (trade unions

    and employers associations) and job protection

    for ordinary employees is weak.


    This has created a special relationship

    between a flexible labour market with high

    job-to-job mobility and a relatively generous

    social security system – often labelled

    Danish “flexicurity”.

    In the last couple of decades active

    labour market and educational policies for

    the unemployed have grown in importance

    as the third corner in a “golden triangle”.


    In it, active labour market policies have

    two functions: to motivate the unemployed

    to take and seek vacant jobs (a

    deterrence effect), and to help the unemployed

    that are not capable of finding a job

    on their own (a qualification effect).

    Denmark’s history of limited state intervention

    in its labour market shaped its system

    of job placements as well as its labour

    market policy. Until the late 1960s the

    “private” unemployment insurance funds

    (closely affiliated to the trade unions)

    mediated and allocated vacant job positions

    to their unemployed members as well

    as administering unemployment insurance.

    In 1969 a unified Public

    Employment Service (


    ) was created,

    among other things, to take over responsibility

    for job placements for the insured

    unemployed. However, the unemployment

    insurance funds remained somewhat active

    in allocating vacant jobs to their members

    through ‘parallel-broking’.

    The municipalities have become more

    active in unemployment policy since the

    late 1970s. Being responsible for benefit

    administration and service provision for

    recipients of social assistance (unemployed

    persons without insurance entitlements),

    they created make-work programs, education

    and training, especially for young people.

    They were also allowed, from 1991, to

    allocate jobs on the ordinary labour market

    to their target groups.

    After strong criticism of the


    ’ inability

    to service the ordinary labour market,

    the conservative-led coalition government

    decided in 1990 to ‘liberalise’ and modernise

    the PES. The liberalisation was

    meant to create a market for job placements

    through temp work agencies, unemployment

    insurance funds and trade

    unions. However, the market never materialised

    in full because of an “unlevel playing

    field” where the services of the PES

    remained free of charge. The PES subsequently

    maintained its position as the

    largest job provider for unemployed people.


    In 1994, a major labour market reform

    attempted to strengthen the position of the


    . Some of the main changes included a

    shift towards earlier and more individualised

    interventions in the unemployment

    spells and the introduction of new activation

    instruments (individual action plans,

    upgrading education and training, job

    rotation and leave schemes). The new activation

    policy was traded-off with a shorter

    unemployment benefit period (currently 4

    years) and a stronger criteria for labour

    market availability and sanctions.

    The implementation of the new active

    labour market policy was given to tripartite

    regional labour market councils in which

    representatives of the labour market organisations

    played a dominant role in policymaking.

    The new approach did not, however,

    halt the criticism of the


    for being

    inefficient, unresponsive and expensive.


    In 2002, a fundamentally new approach

    to PES reform was taken. The steering and

    organisation of labour market policy was

    changed substantially with the contracting-

    out of the


    , which was one of the

    main elements in the labour market reform

    ‘More people into employment’ (

    Flere i



    Rhetorically, the introduction of a

    “quasi-market model” was never framed as

    privatisation, contracting out or liberalisation

    of the


    , but merely as a wish to

    include “other actors” in service delivery.

    1 Jørgensen H,


    Cooperation and Conflict: The

    Policy-Making Process in


    , Edward Elgar, 2002

    2 Madsen PK, “Denmark:

    Flexibility, Security and Labour

    Market Success”,


    and Training Papers

    (International Labour


    , no 53, 1999;

    Bredgaard TF and Larsen F,

    “Quasi-Markets in Employment

    Policy in Australia, the

    Netherlands and Denmark”,

    CARMA Research Paper

    , no 6,


    3 Csonka A,

    Fri Formidling: Om

    Liberalisering af



    Socialforskningsinstituttet, 1992

    4 Bredgaard T and Larsen F,

    Udliciteringen af


    Australian, Holland og Danmark


    DJØF-forlaget, 2006; Bredgaard

    T, Larsen F and Møller LR,

    “Contracting out the Public

    Employment Service in

    Denmark: A Quasi-market

    Analysis”, in Bredgaard T and

    Larsen F,

    Employment Policy

    from Different Angles

    , DJØF

    Publishing Copenhagen, 2005


    Paying for success

    5 See Struyven L, “The New

    Institutional Logic of


    Employment Services

    ”, in

    Bredgaard T and Larsen F (eds),

    Employment Policy from

    Different Angles

    , DJØF

    Publishing Copenhagen, 2005

    6 Bredgaard T, Larsen F and

    Møller LR, “Contracting out the

    Public Employment Service in

    Denmark: A Quasi-market

    Analysis”, in Bredgaard T and

    Larsen F,

    Employment Policy

    from Different Angles

    , DJØF

    Publishing Copenhagen, 2005;

    Bredgaard T and Larsen F,

    Udliciteringen af


    Australian, Holland og Danmark


    DJØF-forlaget, 2006

    7 See Bredgaard TF and Larsen

    F, “Quasi-Markets in

    Employment Policy in Australia,

    the Netherlands and Denmark”,

    CARMA Research Paper

    , no 6,

    2007; Bredgaard T and Larsen F,

    Udliciteringen af


    Australian, Holland og Danmark


    DJØF-forlaget, 2006; Bredgaard

    T and Larsen F,


    Policy from Different Angles


    DJØF Publishing Copenhagen

    8 See Le Grand J and Bartlett


    Quasi-Markets and Social


    , Macmillan, 1993

    9 Le Grand J and Bartlett W,

    Quasi-Markets and Social Policy


    Macmillan, 1993

    The creation and regulation of markets in employment services – Danish experiences


    • 61

    This helped to secure broad political support

    – even from the trade unions, who

    came to consider themselves as service

    providers in this new market.

    The use of different types of service

    providers was, however, not new in itself:

    public authorities had previously contracted

    educational institutions and unemployment

    insurance funds in the implementation

    of labour market policies, especially in

    the delivery of education and training.

    But the split between purchasers and

    providers through binding and formal contracts

    and tendering rounds was new. A

    fresh “institutional logic” was introduced.


    The planning and delivery of employment

    services was delegated to independent

    service providers.

    All former restrictions

    on the duration, scope, target groups,

    average prices and types of activities were

    abandoned. Service providers winning

    contracts with the


    were allowed “freedom

    of methods” in selecting the most

    suitable path to reintegration of the unemployed,

    as long as they complied with the

    contract and the broad frames of the law

    on active labour market policy.

    The arguments for introducing tendering

    models in employment services vary

    according to the country of implementation.

    Nevertheless, there are several common


    The dominant ones are as


    Improved efficiency

    : open competition

    between independent service providers

    improves the efficiency of services

    (understood as quicker and higher

    employment effects). This creates more

    “value for money” (less public spending).

    Improved quality

    : service providers that

    target specific groups contribute to

    more innovative methods and tailored

    individual solutions, thus improving

    the quality of services. Freedom for the

    unemployed to choose a provider is

    also expected to improve the quality of



    : open tendering

    models are smoother, more flexible and

    responsive to changes in the business

    cycle and the needs of target groups.

    They thus create a less bureaucratic

    employment service – one where individuals

    are not clients but consumers

    and costumers.

    Before evaluating the Danish attempts at

    achieving these objectives, one should keep

    in mind that one is evaluating a “quasimarket”

    and not a conventional commodity




    implies that state provision is

    replaced with independent and competitive

    market providers.


    suggests that these

    markets differ from conventional ones in at

    least three ways: one, not all providers aim at

    maximising profits or are necessarily privately

    owned; two, demand in the market is

    often public and not private and, three, the

    choice of provider is often delegated to a

    third party (the purchaser). The preconditions

    for well-functioning quasi-markets in

    welfare policy are thus different from those

    of conventional markets. Le Grand and

    Bartlett use four evaluation criteria: efficiency,

    responsiveness, choice and equity.


    To achieve these criteria four preconditions

    must be fulfilled: one, a competitive market

    structure must be established (with many

    service providers as well as purchasers, easy

    entry and exit, and free pricing); two, accurate

    information on costs and qualitymust be

    available, together with low transaction costs

    to avoid moral hazard and adverse selection;

    three, providers must be motivated by economic

    incentives, and purchasers by the welfare

    of citizens; four, the creaming and parking

    of clients (inequality) must be avoided.

    Providers must be motivated by economic incentives,

    and purchasers by the welfare of citizens

    10 Newsletter from the Danish

    Ministry of Employment, no. 11,

    November 2004

    11 In absolute number, around

    22.000 unemployed individuals

    are currently referred to “other

    actors” during the year (app.

    14.000 full-time equivalents).

    12 Due to the national action

    plan (see below) unemployed

    individuals are transferred to

    external providers at a later

    point in time in their unemployment

    spell (typically after 6

    months), and they remain with a

    provider for typically one year.

    Compared to before, this leads

    to a reduction of the volume of

    persons referred to external

    providers. On the other hand,

    the intensity of activities has

    increased since providers have

    to provide longer-term and consecutive



    Paying for success

    The creation of the Danish


    When the Danish “quasi-market” had

    been in operation for nearly two years, the

    Minister of Employment proudly


    “I can’t think of any precedent in Danish

    politics when such a huge market has been

    created from scratch in such a short time. A

    market that, on top of everything, is

    already operating on a par with the PES

    […] Other actors are here to stay”.


    The Minister voiced one of the main

    parameters of success: the increase in the

    number of unemployed people referred to

    external service providers. Ironically, however,

    the size of the market has shrunk ever

    since the claim. As seen in the figure below,

    the number of insured unemployed

    referred to “other actors” peaked in early

    2005 with about 40 %. It currently stands

    at around 10 %

    ; so what happened?

    From its inception in 2003, the government

    set the first minimum target: that at

    least 15% of insured jobseekers should be

    referred to “other actors”. In addition, local

    authorities were asked to set minimum targets.

    The Minister of Employment could

    issue, but not dictate, guidelines for these

    targets, The tradition of municipal selfgovernance

    constrained him.

    Traditionally, local authorities have only

    sporadically used open competitive tendering.,

    They tended to only use external

    providers on an ad-hoc basis. In any case,

    the PES quasi-market was kick-started and

    grew constantly until its peak in 2005.

    A few contracts have been phased out

    since then, while the Minister of

    Employment issued an action plan (see

    below), which put a temporary halt to the

    conclusion of new contracts.


    new contracts were signed and gradually

    took effect in 2007. The market is thus

    expected to recover in the coming years.

    The new local jobcentres (an amalgamation

    of the


    and municipalities, cf.

    below) are also beginning to sign contracts

    with external providers, particularly for

    uninsured persons with severe barriers to

    labour market entry.

    These changes suggest that the “quasimarket”

    for employment services is

    dynamic and volatile. Between the first

    and second tendering rounds the composition

    of providers has shifted: private

    providers are winning market share at the

    expense of educational institutions and, to










    2003 2003 2003 2003 2004 2004 2004 2004 2005 2005 2005 2005 2006 2006 2006 2006 2007 2007

    1kv. 2kv. 3kv. 4kv. 1kv. 2kv. 3kv. 4kv. 1kv. 2kv. 3kv. 4kv. 1kv. 2kv. 3kv. 4kv. 1kv. 2kv.

    Figure 1: Percentage of unemployed referred to “other actors” (only

    unemployed individuals with unemployment insurance)

    Source: The Danish Labour Market Administration (www.jobindsats.dk.)

    a lesser extent, trade unions and unemployment

    insurance funds. Even as the

    market size shrunk considerably from

    2005, the same tendency applies; private

    providers are currently out competing the

    other types of providers (see table below).

    Efficiency, responsiveness, choice and

    equity are also central in the creation of a

    well-functioning “quasi-market”.


    (not only in Denmark) show that

    there are substantial transaction costs in

    establishing a quasi-market for employment

    services. Some will be transitional,

    but others, related to contract renewal,

    monitoring, supervision, benchmarking,

    and controlling market behaviour, are

    recurrent. There is one main problem:

    transaction costs tend to increase as the

    market becomes more competitive and


    While it is feasible to reduce transaction

    costs by entering into more stable provider

    and partnership relations (see the

    Australian Job Network

    ), this implies moving

    further away from the original intentions

    behind the competitive tendering


    In the first phase of contracting

    out (2003-2005) the transaction costs in

    Denmark turned out to be substantial –

    for some surprisingly so. The providers

    found that preparing and presenting tender

    bids required a disproportionate

    amount of resources. The administrative

    tasks, such as working out payment and

    documentation of jobseekers’ employment

    situation, also turned out to be quite


    not least because Denmark

    operated a regional tendering model with

    different tendering criteria in each region.

    Regional purchasers found it difficult to

    precisely describe the services required, and

    the call for tenders was therefore rather

    vague at times – and very heterogeneous

    across regions. This led to diverging interpretations

    of the services to be delivered,

    which again created extra work when

    attempts were made to settle differences

    over expectations.


    There is another dilemma in the competitive

    tendering model: transaction costs

    increase as jobseekers are given more


    of choice

    . “Quasi-markets” often

    depart from conventional commodity

    markets where the “consumer is the king”

    because choice is determined by public,

    not private, demand and, therefore, delegated

    to a third-party (the purchaser).

    A 2004 survey in Denmark found that

    67% of jobseekers did not feel they had

    had the opportunity to choose between


    They are often automatically

    allocated to a provider by the



    13 “Framework contracts” are

    pre-qualification contracts

    signed with external providers.

    In order to operate on the market,

    a service provider also

    needs to conclude a subsequent

    specific contract with purchasers.

    So “framework contracts”

    does not guarantee that

    any unemployed persons are

    referred to the provider.

    14 See Le Grand J and Bartlett


    Quasi-Markets and Social


    , Macmillan, 1993

    15 Considine M,


    States: The Public Management

    of Welfare-to-Work

    , Cambridge

    University Press, 2001; Sol E

    and Westerveld M,

    Contractualism in Employment

    Services: A New form of Welfare

    State Governance

    , Kluwer Law

    International, 2005

    16 Rambøll,


    Vedrørende Inddragelsen af

    Andre Aktører I



    Rambøll Management, 2004

    17 Rambøll, Erfaringsopsamling

    Vedrørende Inddragelsen af

    Andre Aktører I


    Rambøll Management, 2004

    18 Rambøll,


    Vedrørende Inddragelsen af

    Andre Aktører I


    Rambøll Management, 2004

    The creation and regulation of markets in employment services – Danish experiences


    • 63

    Table 1: Number of “framework contracts” and market shares in Denmark


    No. of Share in No. of Share in No. of Share in

    framework percentage framework percentage framework percentage,

    contracts 2003 contracts 2005 contracts 2007

    2003 2005 2007

    Private providers 94 51% 222 62% 83 79%

    Educational institutions 38 20% 59 16% 5 5%

    Trade unions and 25 13% 44 12% 3 3%


    insurance funds

    Consortia 19 10% 16 4% 13 12%

    Other 10 5% 19 5% 1 1%

    Total 186 99% 360 99% 105 100%

    Source: The Danish Labour Market Administration.

    19 National Labour Market


    Effekt af

    Indsatsen hos Andre Aktører:

    Foreløbig Opgørelse [New

    Frames for the Involvement of

    ‘Other Actors’],


    , 2005

    20 Rambøll,


    Vedrørende Inddragelsen af

    Andre Aktører I



    Rambøll Management, 2004

    21 Ministry of Employment,

    Handlingsplan for Brug af Andre


    Arbejdsmarkedsstyrelsen, 2004;

    National Labour Market


    Effekt af

    Indsatsen hos Andre Aktører:

    Foreløbig Opgørelse


    Arbejdsmarkedsstyrelsen, 2005


    Paying for success

    but rejection is possible if other “equally

    effective” providers are available (which is

    often not the case).

    As explained above, the Danish quasimarket

    model was introduced parallel to a



    system, which, in

    principle, should make it possible to compare


    outcomes and effects

    of the two systems.

    In practice, this is far from easy, as

    the first simple effect measurements have



    It has been far more difficult than anticipated

    to meet the conditions for establishing

    a well-functioning quasi-market for

    employment services. Admittedly, a market

    has been established with a relatively

    high number of service providers (at least

    until 2005) and purchasers, but the market

    is really more an expansion of what was

    already in existence than the establishment

    of a new quasi-market.


    It’s impossible to evaluate the efficiency

    of this market at present, either in terms of

    cost-effectiveness or employment effects.

    There have been high levels of information

    asymmetry and transaction costs. As a

    result, both transparency and freedom of

    choice have been reduced. Consequently,

    the authorities have introduced regulation

    to make both the quasi-market work better

    and to eliminate some of the unacceptable

    outcomes of the tendering model.

    Regulation of the Danish


    Although the Danish “quasi-market” had

    only been going for about 18 months, in

    spring 2005 the Minister of Employment

    issued a plan to rectify some of its worst

    aspects. It was the first example in Denmark

    of an attempt to re-regulate the market.

    Its main objective was to use external

    providers as part of a competence strategy

    rather than a resource one.

    Instead of relieving



    of standardised and administrative

    duties (like the “contact periods, see below),

    specialised external providers would be

    contracted for clearly defined target groups,

    where providers could “make a difference”.

    The action plan was also an attempt to

    create a more centralised and standardised

    market for employment services in light of

    the decentralisation of labour market policy

    to local jobcentres in each municipality

    (see below). The most important changes

    of the action plan were as follows:

    External providers only to be used

    where there is a clear employment

    rationale and for clearly defined target

    groups where they can

    make a difference


    The initial basic contact (after 3

    months) with jobseekers no longer to

    be contracted out to external providers.

    The use of external providers must be

    as smooth and un-bureaucratic as possible.

    Complicated (regional) payment

    models and contracts replaced by one

    simplified and centralised payment model

    and standard contracts.

    Financial incentives should be used to

    get jobseekers as quickly as possible into

    employment rather than using activation

    or educational activities (

    work first


    Performance payment must amount to

    75% (the remaining 25 % is paid as an

    intake fee). A successful outcome is

    measured as being in ordinary employment,

    education or adult apprenticeships

    for at least 13 weeks within the last 26

    weeks (transfer into private job training

    is rewarded with half the bonus). Pay can

    be adjusted to give a higher bonus for

    more challenging target groups such as

    long-term unemployed or immigrants.

    External providers are to take over


    financial responsibility

    for all activities

    It has been far more difficult than anticipated to meet

    the conditions for establishing a well-functioning

    quasi-market for employment services

    22 Rambøll, Erfaringsopsamling

    Vedrørende Inddragelsen af

    Andre Aktører I


    Rambøll Management, 2004

    The creation and regulation of markets in employment services – Danish experiences


    • 65

    during the contract period, and are

    obliged to ensure that jobseekers participate

    in activation activities for at least

    40% of the time.

    To support the tendering model, a

    national IT system was designed. There

    was to be more focus on





    of productivity.

    Comparable measurements of

    provider outcomes were to be established

    and published on the internet (at

    the time of writing, this module is still

    not operational).


    supervision and control

    of service

    providers to be strengthened by central

    administration as well as local jobcentres

    (random samples, unannounced

    visits, monitoring of contract compliance,

    outputs and outcomes).

    As indicated above, until 2005 external

    providers were mainly used to “relieve” the

    PES of administrative tasks; that is, as extra

    hands to help out with the contact periods.

    The “contact periods” were invented in

    2003 as a guarantee that the



    contact jobseekers at least every 3 months,

    and as an availability check on the individual

    unemployed. Previously, the PES had

    not been in contact with each unemployed

    person on such a regular basis, and the new

    contact periods, therefore, drained the

    resources of a


    -system that was already

    squeezed financially.

    Contracting-out this service to external

    providers thus provided financial and

    administrative relief for the


    , and soon

    became the dominant instrument for outsourcing

    to the market. By implication, the

    focus was largely on minimising costs and

    not maximising effects.


    With this approach, external service

    providers quickly developed into a flexible

    but extended arm of the


    , which did

    not promote specialisation, innovation and

    individualisation as intended. As the

    national action plan takes effect, this might

    change as external providers can no longer

    be used to handle purely administrative

    tasks (such as contact periods).

    However, at the same time, the financial

    risks of specialisation, innovation and individualisation

    are transferred from the public

    authorities to the providers. The payment

    model is tuned towards higher outcome

    bonuses, and providers have to

    finance all jobseekers’ activities themselves

    (instead of referring them to activities paid

    for by the authorities).

    One important advantage in the first

    Danish tendering round was that instead

    of relying on standardised national measurements

    and evaluations (since they were

    not available at that time), practically all

    purchasers used regular, informal meetings

    and dialogue with providers for mutual

    exchange of experiences and alignment of

    interests. It could be argued that the need

    for coordination and monitoring was partly

    solved by means of a culture of cooperation

    and dialogue.

    Whether it will be possible to preserve

    this culture in the wake of the action plan is,

    however, an open question. On the one

    hand, the involvement of external providers

    has been shifted away from administrative

    tasks to delivery of services where specialised

    providers can make a difference; on the

    other hand, it has introduced a much more

    outcome-focused payment model, something

    which may threaten the mutual trust

    needed for partnership and cooperation.

    In general, a competitive tendering

    model easily leads to conflicts of interest

    between principal (authorities) and agents

    (providers). As the principal wants to keep

    agents at arm’s length, expectations and

    outcome demands are communicated to

    the providers through tendering rounds

    and contracts. In principle, this secures

    transparency in the market and a level

    playing field for providers.

    But, on the other hand, most providers

    call for dialogue and cooperation. As they

    see it, discussion allows for corrections,

    improvements and harmonisation of interests,

    which may secure a more targeted and

    successful delivery of services. However, in

    the view of the authorities such cooperation

    might bias the process and favour

    some providers over others, and therefore

    not secure a level playing field.


    The new, centralised tendering regime

    and payment model may also have major

    implications for equal access to services

    and the fair treatment of jobseekers. So far

    creaming and parking

    has not been explicitly

    on the agenda in Denmark. This could

    be because such opportunistic behaviour

    does not occur; that information asymmetry

    means the authorities do not know

    what is happening on the ground; or that

    creaming and parking is taboo. Based on

    our data from interviews, we have found

    indications that all three explanations have

    some validity.


    It is also evident that the open prioritisation

    of target groups, something which is

    clear from a comparison with Australian and

    Dutch experiences, does not happen to the

    same extent in Denmark.

    However, there is

    a risk that creaming and parking will become

    more widespread with the new central payment

    model of 25% commencement fee

    and 75% performance-related pay and the

    transfer of economic risks upon providers.

    In general, quasi-markets have difficulties

    in catering for the needs of the jobseekers

    least ready for ordinary employment if

    outcomes are measured shortly after intervention

    (e.g. 6 months later). There are

    also problems when instruments are

    focused on short-term interventions (i.e.

    jobseekers are “turned in the door” which

    is the informal expression of the Danish


    ) and if economic risks are shouldered

    by the providers (i.e. providers motivated

    by economic incentives will be reluctant to

    invest in jobseeker where the return on

    investments is uncertain).


    The box below summarises the main

    characteristics of the Danish “quasi-market”

    model in employment services, as it

    has developed since its inception in 2003.

    23 Bredgaard T and Larsen F,

    “Implementing Public

    Employment Policy: What

    Happens When Non-Public

    Agencies Take Over?”,

    International Journal of

    Sociology and Social Policy

    , vol

    27, no 7/8, pp 287-300, 2007;

    Bredgaard TF and Larsen F,

    “Quasi-Markets in Employment

    Policy in Australia, the

    Netherlands and Denmark”,

    CARMA Research Paper

    , no 6,


    24 Bredgaard T and Larsen F,

    Udliciteringen af


    Australian, Holland og Danmark


    DJØF-forlaget, 2006

    25 Bredgaard T and Larsen F,

    Udliciteringen af


    Australian, Holland og Danmark


    DJØF-forlaget, 2006; Bredgaard

    T and Larsen F, “Implementing

    Public Employment Policy: What

    Happens When Non-Public

    Agencies Take Over?”,

    International Journal of

    Sociology and Social Policy

    , vol

    27, no 7/8, pp 287-300, 2007;

    Bredgaard TF and Larsen F,

    “Quasi-Markets in Employment

    Policy in Australia, the

    Netherlands and Denmark”,

    CARMA Research Paper

    , no 6,


    26 Bredgaard T and Larsen F,

    Udliciteringen af


    Australian, Holland og Danmark


    DJØF-forlaget, 2006; Bredgaard

    T and Larsen F, “Implementing

    Public Employment Policy: What

    Happens When Non-Public

    Agencies Take Over?”,

    International Journal of

    Sociology and Social Policy

    , vol

    27, no 7/8, pp 287-300, 2007;

    Bredgaard TF and Larsen F,

    “Quasi-Markets in Employment

    Policy in Australia, the

    Netherlands and Denmark”,

    CARMA Research Paper

    , no 6,



    Paying for success

    Service delivery systems Local jobcentres (one in each municipality) and external service providers

    Purchasers National labour market administration, employment regions (4), local jobcentre


    Levels of tendering National (service tenders), regional (state framework contracts), local

    (one or more jobcentres)

    Service providers Private consultancy firms, recruitment firms, temp agencies, educational

    institutions, unemployment insurance funds, trade unions

    Types of contracts Framework contracts (pre-qualification) and binding contracts (2+2 years)

    Payment model One national payment model

    75 % outcome payment (13 weeks uninterrupted ordinary employment with

    in last 26 weeks, transitions to ordinary education or adult apprenticeships.

    Half bonus if transition into private job training)

    25 % intake payment

    Providers have to finance all activities from their own budgets

    Price bids are given as average price for specified target groups for

    52 weeks

    Selection criteria

    Price (40-50 %)

    Other criteria (50-60%)

    Table 2: Summary characteristics of the Danish “quasi-market”

    for employment services

    The construction of new local jobcentres

    in each municipality has also been an

    important part of the 2005 plan. Until

    2007, Denmark had a two-tiered employment

    system: the


    had responsibility

    for services and benefits (together with the

    unemployment insurance funds) for people

    with unemployment insurance, while

    the municipalities remained responsible

    for people on various types of social security

    (social assistance, disability benefits,

    vocational rehabilitation etc.).

    However, when a liberal-conservative

    government came into office in 2001, the

    liberal Minister of Employment soon

    declared that the government was aiming

    for a more unified governance system (a

    ‘one-stringed’ system). It looked to municipalise

    the system by shifting the responsibility

    for, and target groups of, the



    new municipal jobcentres.


    It met fierce opposition, however, from

    its political opponents, and the labour

    market organisations (trade unions and

    employers association) that preferred the



    system in which they traditionally

    had major influence over labour market

    policy formulation, steering and implementation.

    The reform came to a standstill.

    However, a window of opportunity

    opened in 2004 when the government

    moulded the 275 municipalities into 98

    larger ones and transferred competencies

    away from the regional level. This gave

    impetus to governance reform in employment

    policy. After another round of heated

    political bargaining, the government found

    a narrow majority with the nationalist

    party (Dansk Folkeparti), and moved forward

    to implement a new local jobcentre

    structure from 2007.

    The new structure is not a full municipalisation

    as the government intended,

    but a rather incoherent organisational

    compromise. Each municipality now has a

    local jobcentre which is charged with

    (re)integration of the labour market.

    Benefit administration and sanctions

    remain with either the unemployment

    insurance funds (insured unemployed) or

    the municipal administration (uninsured


    There are two different types of jobcentres.

    In the majority of municipalities (77)



    and the municipal social and

    employment departments are working

    side-by-side in the same building. The


    is, however, still responsible for services

    and benefits for insured unemployed,

    and the municipalities for uninsured

    unemployed – which also means that there

    are two different executives: one from the


    and one from the municipality (quite

    similar to the German



    ARGE). The remaining 14 jobcentres are

    run exclusively by local authorities on a

    trial-basis (evaluation will take place in

    2010), which means that municipalities

    have also assumed responsibility for

    administering unemployment insurance

    benefits (cf. figure below).

    There are different ways of tendering in

    each level of this new governance system.

    The Minister of Employment can issue

    national tenders for specific groups which

    are administered by a new contracting unit

    in the national labour market administration.

    In September 2006 just such a tender

    was announced for two target groups: one,

    unemployed individuals with vocational

    education (e.g. bricklayers and hairdressers)

    or with a short to medium-term

    academic education (e.g. dental technicians,

    teachers, nurses) and, two, unemployed

    individuals aged 55+ (including

    27 In a longer-term perspective,

    the implicit objective was also to

    create a unified benefit system,

    i.e. to amalgamate unemployment

    insurance benefits and

    social assistance (and thereby

    also remove benefit administration

    from the unemployment

    insurance funds).

    The creation and regulation of markets in employment services – Danish experiences


    • 67

    After another round of heated political bargaining, the

    government found a narrow majority with the nationalist

    party, and moved forward to implement a new local

    jobcentre structure from 2007

    former executives). The national labour

    market authorities contract providers, but

    the local jobcentres are free to decide

    whether they want to make use of them

    (they sign a specific contract with each

    provider when they do).

    In between the national and regional

    level, tenders are made for unemployed

    academics. The tenders are national, but

    administered by the regional employment

    authorities. The local jobcentres are

    obliged to refer any unemployed academic

    in the municipality to the providers that

    have been certified to deliver these regional


    At the regional level, each authority can

    decide to make its own tender for target

    groups that are not covered by the aforementioned

    tenders. If they do so, the local

    jobcentres are obliged to use them.

    Even if the homogeneity of the target

    groups described above can be debated,

    there is, nonetheless, a clear division of

    labour between the national and regional

    authorities on the one side and the local

    jobcentres on the other. The former

    assume responsibility and contract out

    services for unemployed target groups

    with vocational and higher education

    (the easiest target groups), while the latter

    target groups with little or no education

    and training (the most difficult target


    The local jobcentres can – alone or

    together with other jobcentres – make

    their own calls for tender for their target

    groups. So even if the volume of the market

    has declined in recent years, it is clear

    that there are a number of different opportunities

    for contracting out in the new

    governance system, which is still in its


    In relation to the use of external

    providers in the local jobcentres, it is

    important to recall that the Minister of

    Employment can only dictate quantitative

    minimum targets for the state



    for insured unemployed. The municipalities

    are free to decide whether to contractout

    services to external service providers.

    Once the minister has set the targets

    for private provider involvement, it is,

    even in the public system, to a very large

    extent the public parts of the local jobcentres

    who, in cooperation with regional

    and local tripartite employment councils,

    manage the tendering process. Thus,

    there are currently two parallel systems

    for provision: the local jobcentres and private


    28 The national, regional and

    local employment councils (BER,

    RBR and LBR) exist to give

    advice to the respective political

    and administrative authorities,

    and assist in monitoring the

    results and effects of the local

    jobcentres. The councils are tripartite

    bodies composed of

    administrative/political representatives,

    representatives of the

    municipalities, and representatives

    of the labour market organisations

    (the latter have a majority

    of representatives in all three



    Paying for success

    Central government financing

    Monitoring results and effects

    Job centres

    Municipal financing





    Minister of Employment

    Administrative region Employment region




    BER: Central employment council

    RBR: Regional employment council

    LBR: Local employment council

    KB: Municipal government

    Figure 2: Organisational set-up of the Danish employment system


    Conclusion and policy challenges

    The Danish “quasi-market” for employment

    services has been in operation for

    nearly five years, and has gone through two

    distinctive phases of implementation.

    Firstly, it was created between 2003 and

    2005 – or, to be precise – an already existing

    market for reintegration of insured

    unemployed was substantially extended. In

    the second phase of implementation after

    2005, the market shrunk substantially as a

    consequence of the introduction of a new

    contract regime.

    In the new contract regime specialist

    providers are asked to “make a difference”

    for target groups rather than relieve the

    PES of administrative duties. The delivery

    of services is now taking place through a

    new structure of local jobcentres in each


    Judging by the Danish experience, it is

    difficult to create and regulate a market

    in employment services that lives-up to

    political expectations and the preconditions

    for a well-functioning market. The

    anticipated efficiency gains and cost-savings,

    which are spurring the introduction

    of quasi-markets, are still largely

    unknown and undocumented. Even if

    the Danish labour market is currently

    outperforming the majority of European

    countries on a range of labour market

    indicators (cf. appendix below), it is

    unwarranted to claim that the contracting-

    out of the


    accounts for the relative


    Instead, the quasi-market model creates

    a new type of employment policy, as well

    as new conditions for steering and regulating

    the labour market. Clouded in the

    “technical” language of improved efficiency

    and effectiveness, such changes are often

    neglected and depoliticised. Setting aside

    ideological motivations for choosing to

    contract out employment services, the

    Danish experience gives at least three reasons

    for decision-makers to introduce the

    quasi-market model.

    One, the open tendering model is well

    suited to support a politically intended

    shift in the substance of employment policy.

    Internationally, there is a trend towards

    a stronger work-first approach. This

    implies a shift away from both a traditional

    passive social security (where lack of

    demand is seen as the main cause of unemployment)

    and a human capital approach

    (where lack of competencies and qualifications

    is seen as the main cause of unemployment).

    The work first approach interprets

    unemployment as a result of inadequate

    (individual) economic incentives and

    motivation. Finding a job as quickly as

    possible, where any job is seen as a “good

    job”, is the objective.

    Implementing such a policy shift has,

    especially in Denmark, turned out to be

    difficult within existing public implementation

    structures. Policy intentions are distorted

    and skewed during the implementation

    process, especially by relatively

    autonomous local municipalities and case


    The quasi-market model offers

    a depoliticized and efficient structure for

    implementation of a work-first rather than

    human capital approach.

    Two, the unemployed are being prioritized.

    Tailoring initiatives and cost-efficiency

    is a positive formulation of this

    approach, but in reality it is a way to avoid

    spending disproportionate resources on

    unemployed people with few chances of

    ever finding a job, or the deadweight costs

    of putting resources into unemployed people

    capable of finding a job on their own.

    Private providers seem much more susceptible

    towards such creaming and parking

    than traditional public employment advisors.


    From a public policy perspective the

    result is, however, increasing inequality in

    service delivery, and inequality before the


    Finally, contracting out provides political

    decision-makers with better instruments

    to control the expenditures of

    29 Larsen F, Abildgaard N,

    Bredgaard T and Dalsgaard L,

    Kommunal Aktivering: Mellem

    Desciplinering og Integration


    Aalborg Universitetsforlag, 2001;

    Bredgaard T, Dalsgaard l and

    Larsen F, “An Alternative

    Approach for Studying Public

    Policy: The Case for Municipal

    Implementation of Active Labour

    Market Policy in Denmark”,

    University of Aalborg,

    Department of Economics,

    Politics and Administration,

    Working Paper

    , no 4, 2003;

    Considine M, Enterprising


    The Public Management

    of Welfare-to-Work

    , Cambridge

    University Press, 2001

    30 Also see Considine M,

    Enterprising States: The Public

    Management of Welfare-to-


    , Cambridge University

    Press, 2001

    The creation and regulation of markets in employment services – Danish experiences


    • 69

    employment policy. Admittedly, it might

    be more expensive to enter into contracts

    with non-public agents in the short-run,

    but it does give more flexibility (through,

    for example, changing the amount of services

    bought from providers according to

    fluctuations in unemployment rates). The

    problems of organisational restructuring

    are transferred from the public to the private

    sector. When unemployment figures

    are low it is probably much easier to fire a

    private provider than a public servant.

    Political decision-makers may have a

    variety of reasons for contracting-out public

    employment policy. One thing should,

    however, be an obvious conclusion from

    this analysis: choosing a quasi-market

    model involves much more than a discussion

    about “technicalities” in relation to

    how employment policy is implemented in

    the most effective and cheapest way. A

    quasi-market model (and its design) does

    have an important impact on the substance

    of employment policy and the possibilities

    of governing the labour market.

    Public decision-makers need to strike a

    balance between the dynamism of the market

    and the need to fulfil political and

    social objectives. On the one hand, authorities

    try to let the market work on its own

    premises as a means of improving efficiency,

    quality of services and innovation. On

    the other hand, interventions may be

    required to remedy the unintended outcomes

    of market behaviour; e.g., to reduce

    creaming and parking, transaction costs

    and underinvestment in education. Doing

    so, however, erodes the original market

    principles of competition, choice and

    autonomy. Striking the balance between

    the logic of the market and social goals is

    therefore inherently difficult – and basically

    not a technical or administrative issue –

    but a genuine political dilemma.

    Paying for success


    31 An estimated figure; Central

    Intelligence Agency (2008)


    2008 World Factbook


    Central Intelligence Agency).

    32 OECD (2007)



    (Paris: OECD).

    33 Danish Ministry of


    34 Statistics Denmark.

    35 Danish Ministry of


    36 Statistics Denmark.

    37 Danish Ministry of



    GDP , 2007: $310 billion


    Population, 2007: 5.4 million


    Rate of per week unemployment benefit, 2005/6: $500.35


    Number of people unemployed, 2005/6: 114,000

    (3.9% of population).

    Number of days formerly unemployed people must be in work for their assistance to be considered

    successful: 13 weeks in a given 26-week period.

    Rate of per week incapacity benefit, 2005/6: $526.39


    Number of people on incapacity benefit, 2005/6: 258,000


    Rate of per week lone parent benefit, 2005/6: $435.45


    It might be more expensive to enter into contracts with

    non-public agents in the short-run, but it does give

    more flexibility


    It’s the client, stupid!

    An active role for the

    client in Dutch

    employment services

    by Els Sol

    Historical background:

    marketization as part of a major

    welfare reform

    The Dutch began to reform their activation

    policies relatively early. They started

    to do so after seeing ever larger amounts

    of their population taking disability benefit

    despite many of them being fit

    enough to work. This phenomenon came

    to be known as the ‘Dutch disease’.


    The large number of disabled pensioners

    was a legacy of past policies. Previous governments

    had encouraged employers to shift

    mature-age workers with moderate disabilities

    from their payrolls on to the public disability

    pension system, which was integrated

    with workers’ compensation.

    Claimants received only limited help

    from the government in their search for

    employment. Job search requirements

    were relatively lax. Welfare recipients

    were seen as objects of policy rather than

    subjects who could actively contribute to

    increasing opportunities. The


    Employment Service



    ) delivered the

    government’s active labour market policies

    that were meant to lessen the problem.

    But it had a lousy image, small market

    reach, and, ultimately, little effect.

    The increase in demand for, and costs of,

    social security was initially met with

    retrenchment of the system. Expenditure

    was stabilised as a result, but there were no

    reductions in long-term unemployment.

    Further steps were taken once the government

    realised that its generosity could not be

    sustained unless the level of joblessness fell.

    A new ‘activating regime’ (



    ), inspired by a government thinktank,

    was subsequently introduced. The

    status quo was represented as a security

    net in which clients had become trapped.

    In the new system the same net would act

    as a trampoline in which people on benefits

    would be bounced back into the

    labour market.


    Unemployment and disability were no

    longer seen as the consequences of social

    and economic problems, but as the results

    of a social security system that could be

    used by individuals (employers, employees

    and benefit recipients) for their own benefit.

    Labour participation was to become

    the prime goal of social economic policy.

    The idea that full societal participation

    could be achieved outside the sphere of

    paid labour (through voluntary work, for

    example) was rejected.

    The activation and re-integration of

    claimants into the labour market (ideas

    inspired by the policy debates in the

    United States and United Kingdom) came

    1 Visser J and Hemerijck A,


    Dutch Miracle

    : Job Growth,

    Welfare Reform and Corporatism

    in the Netherlands, Amsterdam

    University Press, 1997

    2 WRR,

    Een Werkend

    Perspectief[A working


    , WRR, 1990


    to dominate Dutch policy thinking. Under

    several left-right ‘purple’ cabinets


    above Income’ became equivalent to the

    UK’s ‘Work First’ slogan. Employment was

    seen as the most important part in the

    attempts to solve several social and economic


    In 2007 the current Christian-Social

    Democratic cabinet acknowledged that for

    some, regular paid employment is a step too

    far and extended the targets to include social

    activation, but the ‘Work above Income’

    system still has its two original, and most

    important, components. The first one is a –

    relative to Anglo-Saxon countries – still generous

    social security system that encourages

    work and includes sanctions; an active

    labour market policy is the second.

    The latter rests on four elements: one,

    a general labour market policy; two, a

    ‘reintegration’ scheme which offers backto-

    work services to the unemployed and

    those most likely to lose their jobs; three,

    temporary labour market policies for specific

    groups (women and the old); and,

    four, sheltered employment for those

    whose productivity level is too low to (re-

    )enter normal employment.

    Since 2002 employment services have

    been delivered by companies and not-forprofit

    providers in a quasi-competitive

    system called the ‘re-integration market’.

    As a part of the broader welfare reform

    package a new governance system was set

    up to replace the five separate benefit

    agencies and the obsolete


    Employment Service

    . A negative evaluation

    of the

    Public Employment Service



    gave the government the chance to

    create what was to become a hybrid of

    public one-stop-shops and a private

    employment services market.

    In 2000 a government think-tank

    called ‘The Future of Labour Market

    Policy’ concluded that different labour

    market conditions called for deeper

    adoption of re-integration policies.


    report, ‘Aan de slag’ (Getting it done),

    recommended that the focus of reintegration

    policy be shifted from benefit

    dependence to help for movements from

    state support to the labour market. The

    report proposed the budgeting of welfare

    schemes; the use of placement results to

    steer benefit administrations; more flexibility

    in welfare-to-work and reintegration

    budgets; and the creation of a proper

    quasi-market for employment services.

    Private providers were seen as more effective,

    more client-oriented, and cheaper.

    3 Under Wim Kok between 1994

    and 1998, and 1998 to 2002.

    4 De Koning J

    et al


    Arbeidvoorziening in

    Perspectief[The PES in perspective]


    The Hague, 2005

    5 Parliament/TK, “Aan de Slag,”

    Parliament/TK, no 1 (27914),



    Paying for success

    Table 1 Chronological overview of the main reforms in governance 2001-2006

    Law Main features


    Structuur Uitvoering Werk en Inkomen

    (Structure Foundation of one unemployment benefit

    Implementation Work and Income)(SUWI) Agency (UWV)

    Foundation of one-stop shops (CWI)

    Abolition of the


    Obligation to tender all employment services

    Introduction of a market with for profit providers

    2003 Wet

    werk en bijstand

    (Work and Welfare Act) Decentralisation of welfare towards municipalities

    on a budget based system


    Wet werk en arbeid naar arbeidsvermogen

    Obligation of employers to offer employment

    (Act on work and income according to labour services to the disabled

    capacity for disabled)

    6 MvT, Parliament/ TK 28870, no 3

    7 For the disabled who are still

    under employment contracts

    with their employers, a


    market for employment services

    was created in 2006. The

    employer takes on the role as

    principal and buys employment

    services for his personnel from

    private (for profit) providers. A

    major incentive for employers to

    buy employment/outplacement

    services is the legal obligation to

    continue payments to a disabled

    employee under an employment

    contract for two years. This is a

    third market segment for the

    providers in the Dutch reintegration

    market, next to the municipal

    and UWV segment. Other

    than principals in the two first

    segments, employers do not use

    open tenders for employment

    services. As this new market

    segment in the reintegration

    market completely differs from

    the welfare-to-work market

    model in the UK where only

    transitions from benefit to work

    and are incorporated, and transitions

    from work to work are not,

    this article focuses on the other

    segments only.

    It’s the client, stupid! Dutch options for welfare to work


    • 3

    The report and its recommendations

    became the basis for the privatisation,

    agentification and decentralisation that

    were central to the institutional reforms

    that took place between 2000 and 2006

    (see table 1 for details).

    In line with the proposals, a new

    Structure for Work and Income




    was introduced in 2002. It replaced

    national employment assistance programs

    – including public sector jobs for

    jobless people and vocational training

    programmes – with a contestable market

    for employment assistance. Employment

    services for the most disadvantaged jobseekers

    were decentralised and contracted

    out to private and community organisations.

    Characteristics of the Dutch model

    Each of the main characteristics of the

    Dutch model of marketization are listed



    The Department of Social Affairs

    This government department is responsible

    for the general management

    of the programme but has delegated

    its role of gateway for the intake and

    referral of all social security benefit

    claimants to the local Centres for

    Work and Income (CWI), which are

    public entities. It has also given away

    its role as principal for the fully privatised

    system of employment service

    delivery for the long-term unemployed

    to the National Benefit

    Agency for the (short-term)

    Unemployed and Disabled (UWV),

    and to all 450 municipalities for the

    people on social assistance.

    Both the UWV and municipalities

    have been obliged to contract out.

    This rule has been enforced in order

    to create a proper market and to

    restrain municipalities from delivering

    their welfare-to-work services in

    house or through local public or nonprofit

    satellites (many of which have

    performed badly).

    But as a result of

    political pressure by the municipalities

    in 2006, this obligation lapsed for

    municipalities. In cases where municipalities

    do contract out they have to

    conduct tenders according to

    European norms (as does the UWV).

    Parliament wanted public tasks

    (intake, eligibility and sanctioning) to

    remain in public hands. In this

    model, private companies replace

    what the public sector previously

    offered the hardest to help claimants.

    The public Centres for Work and

    Income get the more standardised tasks

    of dealing with the mainstream benefit



    The Department of Social Affairs ultimately

    funds nearly all benefit and

    employment assistance schemes, but does

    not specify the form or content of contracts.

    The UWV and municipalities

    decide which employment services are to

    be offered to clients.


    Centres forWork and Income

    During the first six months of unemployment,

    claimants who are close to

    meeting the needs of available jobs are

    given help by the Centre for Work and

    Income (CWI). If recipients are still out

    of work after six months they are transferred

    to the UWV or municipality.

    People with significant difficulties in

    getting employment are immediately

    referred to the municipalities or the

    UWV and thereby to companies and

    not for profit providers. CWIs use an

    Private providers were seen as more effective, more

    client-oriented, and cheaper

    8 Teulings CN, Bovenberg A and

    van Dalen HP,

    De Calculus van

    het Publiek Belang

    , Ministerie

    van Economische Zaken, 2003

    9 Ministry of SZW,

    Naar een

    werkende reintegratiemarkt


    Ministry of SZW, 2002,



    Paying for success

    ‘Opportunity Meter’ to profile the people

    they work with. After being transferred

    to either the national UWV or a

    local municipality, claimants are offered

    intensive support, which is delivered by

    private providers in a quasi-market and

    is funded by the UWV employment

    services budget or the municipalities

    depending on the type of benefit.



    The Dutch municipalities are a separate

    public administration layer with

    their own social assistance budget,

    political responsibility and control.

    Since 2004 they have had complete

    policy freedom, so in each area can

    develop their own reintegration policies

    in return for big financial incentives.

    The Minister of Social Affairs is only

    responsible for the system of social

    assistance, which means he takes care of

    the judicial framework, the distributive

    budget code and monitor system.


    The UWV

    The relationships faced by the benefit

    agency are different. The Department of

    Social Affairs determines the instruments

    and the UWV implements them. Since

    the Structure for Work and Income Act

    (2001), the UWV has been steered and

    monitored on the basis of performance

    indicators which are related to the flow

    of claimants from unemployment or disability

    benefit to the labour market.

    There are no financial incentives.


    Private for and not-for-profit providers

    Dutch recipients of income support –

    except parents with young children and

    people with severe disabilities – are

    obliged to engage with the labour market.

    Individuals must maintain some

    sort of work in order to remain eligible

    for continued receipt of the benefit.

    Private employment providers thus

    have all recipients of income support as

    their clientele, i.e. those people in the

    social security systems for the unemployed,

    disabled and those on welfare.


    Client councils

    Jobseekers are represented by local and

    national client councils, which together

    form the National Client Council

    (LCR). The LCR is a statutory body

    whose main function is to represent the

    views of clients to the Minister as part

    of the RWI consultative mechanism.

    The LCR has had success in advocating

    individual client accounts for employment

    assistance (IROs, see below) and

    a network of independent advisors to

    help clients to locate and negotiate

    with reintegration providers.


    Board forWork and Income and Inspectorate

    forWork and Income

    Two organisations oversee the quasimarket:

    the Board for Work and

    Income (RWI) and the Inspectorate for

    Work and Income (IWI).

    The tripartite advisory Board for

    Work and Income (RWI) is responsible

    for the quality and transparency of the

    private market for employment services

    and for labour market and reintegration

    policies in general. A system of self-regulation

    (based on the subsidiarity principle


    is meant to induce transparency.

    The Inspectorate for Work and

    Income (IWI) is the second monitoring

    organisation. It tracks the legitimacy

    and efficiency of the provisions by the

    UWV and the legitimacy of the implementation

    of the WWB by the municipalities.

    Both organisations report to

    the Minister of Social Affairs

    The Department of Social Affairs established

    the main rules for evaluation of

    process effectiveness of the market in a

    report entitled ‘Towards a functioning

    reintegration market’ (

    Naar een werkende


    They are:


    Entrance and competition: no specific

    entrance criteria, a ‘sufficient’ level of

    competition, and a level playing field

    for FP and NFP providers.




    Public and open tender procedures.


    Payment by performance.


    High quality of service delivery.


    Access to services for all clients.


    Freedom of choice for all clients.

    The creation of a quasi-market and contracting

    out of employment services by

    benefit agencies are seen by the Dutch government

    as vital to the results-orientated

    system it wants to see. Both changes force

    principals to set concrete targets and formulate

    success criteria for providers.

    Because it brings policy and practice

    together, the role of principal makes

    municipalities and the unemployment

    benefit organisation more aware of what

    welfare-to-work schemes require. Competition

    between providers should uncover

    any new intervention techniques developed

    by providers. Furthermore, the existence

    of multiple principals – there are

    more than 450 municipalities – means

    there will be a laboratory for any changes.

    A Market for employment services in


    The Dutch market for employment services

    developed quickly. It used the experiences

    of already existing companies like

    temp work agencies, vocational training

    institutes and other specialized, mainly

    non-profit, organisations, and new commercial

    entrants. In the Netherlands the

    reintegration market quickly developed

    into a for-profit market.

    Currently, not-for-profit providers are a

    small proportion of the UWV market.

    They serve no more than 4 percent of it

    and tend to be specialised companies

    working with disabled people.


    exact figures for the whole market are not

    available. Historically, the Dutch have not

    had as great a role for the non-profit sector

    in this domain as the Anglo-Saxon countries

    have, due to better provision from the

    government itself.

    For an illustration of the effect of the

    introduction of the marketization for nonprofit

    companies in the city of Amsterdam,

    see table 2 below.

    Not much is known about the results of

    the non-profit providers, who are mainly

    active in the municipal market. There is no

    10 Koning P, “Maatschappelijk

    Verantwoord Reintegreren”,

    Economisch Statische Berichten


    no 4525 (92), pp 791-794, 2007

    It’s the client, stupid! Dutch options for welfare to work












    1996 1997 1998 1999 2000 2001 2002 2003

    profit non profit


    Table 2 Expenditure of reintegration budget in percentage of non -profit and

    for profit reintegration companies, Amsterdam, 1996-2003

    Source: Sol E and Hoogtanders Y, “The Road to Privatisation: The Amsterdam Case”, in Sol E, and Westerveld M (eds),

    Contractualism in

    Employment Services

    , Kluwer Law International, 2005

    11 Le Grand J,


    Agency and Public Policy: Of

    Knights and Knaves, Pawns and


    , Oxford University

    Press, 2003

    12 Koning P, “Maatschappelijk

    Verantwoord Reintegreren”,

    Economisch Statische Berichten


    no 4525 (92), pp 791-794, 2007

    13 Sol E, Castonguay J, van

    Lindert H and van Amstel Y,

    Work First Werkt: Op Weg naar

    Evidence Based Work First


    Divosa, 2007

    14 Kok L, Hollanders D and Hop


    Kosten en Baten van


    , SEO, 2006


    Paying for success

    monitoring of them on a national scale. It

    is often argued that not-for-profit

    providers may behave differently during

    market failures as they attract more intrinsically

    motivated employees who provide

    better services (and are often driven by

    equity concerns).

    If they receive donated

    labour, they might produce at lower costs,

    provide higher quality or focus on market

    segments that are less profitable.

    A Dutch empirical study

    found companies

    to be more active in selecting clients relative

    to other providers. They put more

    effort into encouraging potential clients to

    start a programme (which gets them extra

    payments). However, they do not place

    more people in work than not-for-profit

    providers. Other research into employmentbased

    early intervention strategies for social

    assistance claimants found better placement

    rates for profit-motivated providers than for

    municipal workplace companies.


    On a yearly basis the national public

    budget for reintegration amounts to two

    billion euros. It has been estimated that

    private providers account for around one

    third of public resources.

    Boaborea, their

    industry body, claims that only 17 percent

    of the budget is actually spent on their

    services, the rest being taken by the public

    administration (CWI, UWV, municipalities).

    The total number of employment

    services delivered annually in the period

    2002-2006 was 150,000, of which 19% of

    the services went to the disabled, 21% to

    the unemployed and 60% to social assistance


    Research over the period shows that the

    chances of a claimant getting intensive

    help is dependent on age (the young have

    a better chance) and duration of the benefit

    claim (the longer duration, the better

    (cumulative) probability). Men on social

    assistance have a 28% chance and women

    a 22% chance after one year, while unemployment

    benefit claimants face odds of

    33% and 32% respectively.

    Procurement and tender design

    There are a number of different models. At

    one extreme is a model of one prime contractor

    with a large contract for all benefit

    claimants and for a long period of time, at

    the other end is a system of multiple contractors

    for small groups and small, shortterm

    contracts. The Dutch chose the latter

    path in 2002 in order to address the risk of

    creaming and to give specialised (small)

    providers a level playing field. Providers

    were offered contracts to assist small

    ‘batches’ or ‘parcels’ of claimants with similar

    characteristics, specifying the form of

    help that should be given. The social benefit

    administration UWV contracted out

    by way of procurement of various cohort

    types on a national scale.

    In 2002 the agency contracted out as

    many as 53 cohort types and 474 individual

    cohorts. Over the years the benefit

    agency has decreased the number to just

    three groups (the disabled, young disabled

    and unemployed) and 20 types of individual


    Short contract durations (of one year)

    left major risks with providers and led to a

    high turnover of personnel among the

    providers. The substantial tender costs for

    providers were also a major difficulty.

    The cost and time involved in preparing

    tenders meant a heavy administrative burden,

    which initially built a high threshold for

    small, specialized providers, who lost territory.

    The introduction in 2004 of a kind of

    voucher by the benefit agency UWV, called


    Individual Reintegration Agreement


    below), eliminated this entrance barrier for

    small providers, who (re-)entered the market

    in large numbers. However, the municipal

    UWV contracts are awarded to providers on the basis

    of five criteria: experience, price, placement percentage,

    methods used and drop-out rates

    15 Mallee L and Mevissen J,

    “Re-Integratie in Nederland: Een

    Private Markt voor het Publieke


    Tijdschrift voor


    , no 2, pp

    106-119, 2007

    16 Vinke,



    , TNO,


    It’s the client, stupid! Dutch options for welfare to work


    market is still difficult for the small providers

    because of the red tape endured in tender

    procedures, the demands they face, the short

    timescales of the contracts, the payment

    structure (no cure, less pay), and the lack of

    guaranteed clientele.


    Between 2002 and 2008 the UWV

    organised seven tender rounds. Despite the

    great number of principals involved, in

    practice the majority of the market is in the

    hands of the benefit agency and four big

    cities (Amsterdam, Rotterdam, Den Haag,

    Utrecht). It is, therefore, a buyers’ market.

    UWV contracts are awarded to

    providers on the basis of five criteria: experience,

    price, placement percentage, methods

    used and drop-out rates. Initially,

    placement percentages were weighted

    highly. Depending on the target groups

    there are different weights for criteria. As

    the tenders up until now have been primarily

    based on price competition, it has been

    difficult for providers to stand out on quality.


    This led the UWV and municipalities

    to choose the lowest bidder for employment

    services, at the cost of service quality

    and the development of specialisms.

    Although the type of services was prescribed,

    in terms of content they were left

    at the discretion of the provider. When, in

    2003, the ‘no cure, no pay’ and ‘no cure,

    less pay’ models of payment were introduced

    – in order to give further incentives

    to providers – price competition threatened

    to lead to a race to the bottom. The services

    to be delivered during the bidding were

    of less importance. A cry for help from the

    former chair of Boaborea persuaded the

    UWV not lower prices any further.

    Contract design

    The first contracts ran for one year, but

    since 2004 the UWV and most municipalities

    have used longer contracts as a means

    to achieving more commitment and lower

    transaction costs. Lately, there has been a

    trend towards using framework contracts

    and longer contracts (of up to two years).

    They get rolled over if case managers

    (municipalities) and reintegration coaches

    (the UWV) are content with the service

    delivery and output. These public servants

    directly influence the number of clients

    providers actually get, thus giving them a

    bigger say in the market than before.

    The fee structure is geared to achieving

    outcomes. Retention in unsubsidised

    employment for at least six months is the

    key payable outcome for providers. In

    practice the duration can be shorter. The

    UWV uses the operational definition that

    a client has to have worked for two months

    and has an employment contract of at least

    four months. So a client still counts if after

    two months he drops out.

    Providers are either paid on a ‘no cure,

    no pay’ basis – for the relatively easy to

    help – or a ‘no cure, less pay’ basis for the

    difficult ones. Typically, the reward per

    client consists of a fixed payment when the

    action plan is approved (20% maximum),

    a fixed payment six months later (40-50%)

    and bonus payments in cases of job placement

    for at least two and six months (40-

    50%). In practice some providers park the

    hardest to help and cash the payment of

    the service inputs. There is only a fixed

    payment for the easy to place.

    Much attention has been focused on

    how to create a well functioning reintegration

    market (particularly around the design

    of transparent tender procedures with low

    thresholds in order to contract providers

    with the best price/product combination).

    As a result the focus of providers was on the

    bidding not on results, and on the means –

    packages of services, called ‘trajectories’ –

    and too little on the goal (outflow to work).

    A trajectory consists of some or one of the

    elements of diagnosis, guidance/ soft skills,

    vocational training and placement.

    Reintegration contracts generally did not

    offer providers sufficient funds to invest in

    the relatively costly forms of employment

    assistance such as training. As a result, the

    nature of reintegration services shifted from

    help for overcoming barriers to work

    towards assistance in the search for work.

    This is still a problem and probably more

    so now because the clients that remain are,

    because of the economy’s ability to create

    jobs for so many people, the least skilled

    and thus most difficult to help. One of the

    criteria for a well functioning reintegration

    market is the openness of service delivery to

    all (see above). The financial incentives of

    the Act on Work and Income has incentivised

    municipalities to seek quick wins in

    outflow, helping first and foremost the easy

    to help. The UWV, however, is more open

    to all clients, thanks to the IRO (see below).

    In all, since 2002 the UWV and the

    municipalities have gained experience in

    tendering, thus overcoming their teething

    troubles. Nowadays there is no excessive

    monitoring to ensure that providers are fulfilling

    their contractual obligations. Discontent

    with the lack of quality of service made

    the UWV start yearly auditing of providers.

    The lack of transparency in the Dutch

    model is a major problem. To create a

    threshold in the market and for providers

    to be able to distinguish themselves from

    other competitors, Boaborea developed a

    quality assurance system for the sector.

    There is currently a national

    quality mark

    for all providers (a ‘keurmerk’

    ) dealt with

    by a Foundation called Blik op Werk. Blik

    op Werk also developed a provider monitor

    with placement figures, satisfaction

    scores and other information for future

    customers and clients (the ‘reintegratiemonitor’).


    In 2006, the UWV

    developed a benchmark for itself that

    allowed it to assess providers with the best

    price-quality combination.

    IRO: The client as customer

    A major criticism of the contracts under

    the tender system was the ‘one size fits all’

    service and the lack of innovation and differentiation

    promised by the model’s

    designers. By making customers out of the

    immediate consumers of the service the

    government has countered part of the

    problem. The result has been more satisfied

    customers, more successful job placements,

    a proliferation of providers and an

    overhaul of the tender system.

    In 2004, the government introduced

    Individual Reintegration Agreements

    (IROs) between individual clients, the

    UWV and reintegration providers, following

    complaints by client organisations and

    Members of Parliament. The introduction

    was strongly supported by the National

    Client’s Council as a means to engaging

    clients in welfare-to-work activities and

    giving them greater control over the

    process. In the beginning the ministry was

    far from convinced that a personal budget

    for the client was necessary to provide

    them with freedom of choice, but pressure

    from parliament was intense, so it agreed.

    The core elements of the IRO arrangement

    for the client are the choice of

    provider and the opportunity to compose

    an activity plan. Both factors come close to

    making the client the principal, and thus

    to establishing a real market instead of a

    quasi one.

    The UWV buys the IRO in the name of

    the client and in line with his or her wishes.

    In only 2 percent of cases are IROs turned

    down. Every client that is diagnosed for intensive

    help can use the IRO, irrespective of

    whether or not a client needs guidance in it.

    Compared to the tender process, the

    IRO procedure offers more freedom of

    choice to the client, the provider more possibilities

    to deliver a ‘made to measure’

    service, and the reintegration coach/case

    manager more room to manoeuvre. For

    the differences between IROs and tendering

    see table 2.

    Between 2004 and March 2007, 66,455

    IROs started. 62 percent have been for

    people on unemployment benefit and 38

    percent for the disabled. IROs became

    popular overnight. This spectacular growth

    has now stabilised to 60 percent of all people

    on ‘trajectories’ by the end of 2006.

    17 www.blikopwerk.nl

    18 www.blikopwerk.nl

    Paying for success

    The price of an IRO service package

    (trajectory) is a factor 1.1 times more

    expensive for the disabled and 1.3 times for

    the unemployed. The placement percentages

    related to regular (tendered) IROs are

    1.3 to 1.6 times higher for the disabled and

    a factor of 1.1 to 1.3 times for unemployment

    benefit claimants. IROs can be negotiated

    to a maximum price of EUR 5,000

    each, but must be approved by the funding

    body, the UWV. The municipal market is

    reluctant to introduce similar budgets.

    The introduction of IROs has led to a

    dramatic expansion of the number of small

    providers in the market. Upon the introduction

    of IROs in 2004, the UWV had

    60 contractors. It now has 2,100. For optimal

    use of the IRO, clients need more

    information on providers than is currently

    available. The quality mark sets a threshold

    in the market, but there are still too many

    providers to choose from.

    The IRO has become such a huge success

    that it currently endangers the whole

    tendering system as the UWV is no longer

    able to refer the contractual amount of

    clients to providers and is unable to audit

    such a large amount of providers.

    The new purchasing system

    From 1 April 2008 the UWV (and thereby

    the Department of Social Affairs) introduced

    a completely new way of purchasing

    re-integration products and IROs.

    A socalled

    ‘purchase framework’ now determines

    how the agency will purchase products.

    Tenders for whole target groups will be a

    thing of the past. From now on the benefit

    agency will purchase for individual clients.

    There will no longer be price competition

    as the benefit agency will determine

    product prices. Competition is based on the

    quality of services of the provider. Providers

    are selected on the basis of minimal requirements

    (e.g. client satisfaction results, placement

    percentages). A provider has to meet

    these requirements in order to be on the

    potential providers list. The better results

    the more chance providers have of being

    able to continue to deliver services.

    A reintegration coach – an employee of

    the UWV – and the client determine together

    what is needed for the client to return to

    the labour market. If this is done, the client

    chooses the company that will deliver the

    services. In effect reintegration coaches are

    playing a larger role in the guidance of the

    clients, at the expense of the providers. Only

    in the case of the IRO will the provider be

    responsible for such guidance.

    The new rules are inspired by the

    Australian Star rating system, i.e. competition

    over quality is to replace competition

    over price. Providers that do not deliver the

    required standards will not be awarded any

    further contracts. The payment-by-placement

    results system applies to IROs.


    19 http://www.uwv.nl/zakelijk/reintegratiediensten/



    20 In percentages of total

    entrance per cohort for 2004.

    It’s the client, stupid! Dutch options for welfare to work


    • 9

    Table 2 Characteristics of UWV

    clientele, results and costs by

    IRO and by tender (2004-2007)

    IRO Tender

    Number 66, 3 ,323

    Client characteristics

    Status Disabled 3 % %

    Unemployed 62 % 3%

    Age Disabled 3 , % 39,

    Unemployed ,3% 2,9

    Education low skilled

    Disabled 2 %

    Unemployed 9%

    Price pro ‘trajectory’ in ? , 00 3, 0

    Disabled in ? , 30 ,320

    Unemployed in ? , 3 3, 30


    Disabled 0% 29%

    Unemployed % 3 %

    Total Cost per placement

    Disabled in ? ,02 9,9 0

    Unemployed in ? 6, 0 6, 0

    Source: APE,

    Vierde Voortgangsrapportage

    IRO, APE, 200

    The Act on Work and Assistance (2004),

    together with the strong financial incentives

    for municipalities and clients, has

    proven to be a major success – figures for

    social assistance are currently as low as in

    the seventies.

    However, with the limited

    data available it is difficult to assess the

    contribution of the new reintegration market

    to the effectiveness of the transition

    from welfare to work. A recent evaluation

    by the Department, called ‘Beleidsdoorlichting

    Reintegratie’, concluded that there

    is a ’small positive effect’ on the probability

    of making such a transition but that it

    does not, over the short-term, pay back in

    terms of benefit savings (see also effectiveness



    A major weakness of the model is the

    emphasis on short-term job outcomes and

    a lack of resources and incentives to invest

    in the most disadvantaged. A major

    strength of the system is the relatively

    strong emphasis on positive engagement

    with clients in the form of the introduction

    of IROs and the role of the Client

    Councils. The quality mark also helps to

    make the reintegration market more transparent.

    The flexibility and scope for experimentation

    at the local level are further

    strengths of the decentralised model.

    There is scope within the policy and service

    delivery framework for municipalities

    to switch to more investment and more

    intensive help for the many disadvantaged

    jobseekers attuned to local labour market

    needs. On the other hand the model suffers

    from a lack of professionalism among

    its municipal principals, from missing

    national standards regarding the quality

    of service provision, and from learning

    from practice. The lack of accurate

    national data on program effectiveness is

    also a problem.

    There has been little empirical research

    in to the net effectiveness of the privatised

    reintegration service in the Netherlands.


    The few studies that have looked into this

    area have concluded that the impact of

    welfare-to-work services is present but has

    been limited, and that employment services

    do shorten periods of unemployment.

    However this type of research cannot

    provide an answer regarding the relative

    effectiveness of private service delivery

    over public service delivery. Neither does

    it give much information on how to

    achieve a better price/result combination.

    Some reports conclude that (meagre)

    placement results are related to (low)

    prices and point the finger at the importance

    of a well qualified public principal

    for better results.



    The Netherlands has chosen a different

    model than the one proposed for Britain

    by David Freud. The Dutch system has big

    and small principals, including clients

    (IROs); a large amount of providers (some

    big and many small, but mainly for-profit

    in the unemployed and disabled sector); a

    mixture of for-profit and not-for-profit

    providers in the hard-to-help segment;

    national and local tenders (only using

    work budgets, no benefit ones); short-term

    contracts (one to two years); sustainable

    placement (six months); the use of quality

    marks for providers; partnership as the

    basis of cooperation between private

    providers and public organisations (chain

    management); and a weak culture of monitoring

    and evaluation.

    The Freud approach has as its main elements

    one national principal (the DWP);

    tenders with a small amount of big

    providers; the use of work and benefit budgets

    combined (‘DEL’ and ‘AMI’); a benchmark

    to focus competition on results (i.e. a

    star rating system); five year contracts; and

    obligations to guide clients for three years.

    Goal displacement is a danger in the UK

    model. In markets where there is only one

    buyer, providers tend to focus on the principal

    and his output indicators. This is a

    21 Bosselaar H, Bannink D , van

    Deursen D, and Trommel W,

    Werkt de WWB? Resultaten van

    Nieuwe Verhoudingen Tussen

    Rijk en Gementeen

    , Begroting

    SZW, 2007

    22 http://docs.minszw.nl/pdf/


    23 Heyma A, “De Effectivitiet

    van Reintegratieinstrumenten

    voor Arbeidsgehandicapten”, in

    Vos EL,

    Daadwerkelijk Effectief:

    Prestatiemeting van Reintegratie

    en Activering

    , pp 97-150,

    Plantijn/Casparie, 2005;

    Zwinkels W,

    Effectivitiet van Re-

    Integratie: Onbenutte Potenties

    van Privatisering

    , Tijdschrift voor

    Arbeidsvraagstukken, no 2, pp

    120-130, 2007; De-Graaf-Zijl M,

    Groot I and Hop JP,

    De Weg

    naar Werk: Onderzoek naar de

    Doorstroom Tussen ww, Bijstand

    en Werk, voor en na de SUWI


    , SEO, 2006; Hekelaar,

    Zwinkels and Braat,

    De Juiste

    Klant op het Juiste Traject: Een

    Onderzoek naar de Netto-

    Effectivitiet van het Rotterdamse

    Reintegratiebeleid voor het

    Ontwikkelen van Klantprofielen


    SWA, 2006; Kok L, Hollanders D

    and Hop J,

    Kosten en Baten van


    , SEO, 2006

    24 De Groot I, Kok L and Gueller


    Kwantitatief Effect WWB


    SEO, 2007; Hekelaar, Zwinkels

    and Braat,

    De Juiste Klant op

    het Juiste Traject: Een

    Onderzoek naar de Netto-

    Effectivitiet van het Rotterdamse

    Reintegratiebeleid voor het

    Ontwikkelen van Klantprofielen,

    SWA, 2006; De-Graaf-Zijl M,

    Groot I and Hop JP,

    De Weg

    naar Werk: Onderzoek naar de

    Doorstroom Tussen ww, Bijstand

    en Werk

    , voor en na de SUWI

    Operatie, SEO, 2006


    Paying for success

    risk in a market where the relationship

    between output indicators and net effectiveness

    is necessarily indirect. Star ratings

    do not offer a sufficient solution for this

    problem as experiences in the Australian

    market show.


    In the Dutch model, the introduction of

    the individual contract IRO and multiple

    buyers/principals are safeguards against

    this drawback. IROs focus on the client as

    the real customer and on matching

    through direct quality control by the

    client. This incentivises providers to deliver

    custom made services.

    The large amount of providers in the

    Dutch market increases the risk of lack of

    transparency in the market. A quality mark

    for providers and a national provider’s

    guide must safeguard against this danger.

    Providers must use the quality mark to distinguish

    themselves from competitors.

    The weak spots in the Dutch model

    are the short time horizon of the contracts

    which endanger the quality of services

    offered in the market, the noncommittal

    attitude in professional guidance

    by the Department of Social Affairs and

    – despite the large amount of study

    reports – the lack of solid evaluation studies.

    Obviously there are differences between

    more and less marketized systems for the

    delivery of public employment services.

    Shortcomings, however, are not univocal

    to one or the other model of marketization.

    At this time no conclusive evidence

    from research is available. The choice for a

    largely for profit market delivery system or

    a hybrid remains first and foremost a political


    25 Thomas M, “A Review of

    Developments in the Job


    Parliament of Australia

    Research Paper

    , no 15, 2007-


    26 An estimated figure; Central

    Intelligence Agency (2008)


    2008 World Factbook


    Central Intelligence Agency).

    27 Centraal Planbureau


    nlinfo.html); Centraal Bureau

    voor de Statistiek



    28 Centraal Planbureau


    nlinfo.html); Centraal Bureau

    voor de Statistiek




    29 Centraal Planbureau



    30 Centraal Bureau voor de








    32 Ministerie van Sociale Zaken

    en Werkgelegenheid








    33 Centraal Planbureau


    nlinfo.html); Centraal Bureau

    voor de Statistiek




    34 Centraal Bureau voor de

    Statistiek (http://www.cbs.nl/nl-







    35 UWV (http://www.uwv.nl/).-

    36 Centraal Bureau voor de






    It’s the client, stupid! Dutch options for welfare to work


    The Netherlands


    GDP, 2007: $2.7 trillion


    Population, 2005/6: 16.3 million


    Number of people in the labour force, 2005/6: 7.4 million


    Average weekly wage among adults in full-time work, 2005/6: $841.3


    Total spending on welfare, 2005/6: $143.3 million


    Total spending on unemployment benefit, 2005/6: $6.8 billion


    Rate of per week unemployment benefit, 2005/6: $1, 359 (maximum)


    Number of people unemployed, 2005/6: 483,000


    Number of unemployed people in 2005/6 out of work for one year or more: 200,000


    Number of days formerly unemployed people must be in work for their assistance to be

    considered successful: 6 months



    Total spending on incapacity benefit, 2005/6: $16.3 bilion


    Rate of per week incapacity benefit, 2005/6: $1, 359 (maximum)


    Number of people on incapacity benefit, 2005/6: 958, 340


    Total spending on lone parent benefit, 2005/6: $6.6 billion


    Rate of per week lone parent benefit, 2005/6: $338.6


    Number of people on lone parent benefit, 2005/6: 365,000

    Paying for success


    37 Ministerie van Sociale Zaken

    en Werkgelegenheid








    38 Centraal Bureau voor de








    39 Centraal Bureau voor de









    40 Ministerie van Sociale Zaken

    en Werkgelegenheid








    41 Centraal Planbureau




    Table I Type of public private partnerships created under the Dutch model

    Public market Public market Private market

    Type of clients Unemployed and disabled Social Assistance Sick and disabled

    with rights, without claimants employees, with employment

    employment contract contract

    Intake and referral of clients Centre for Work and Centre for Work and Companies specialised

    Income Income in working conditions

    Income maintenance Employee Insurance Municipalities Firms

    Agency (UWV), Means tested Contribution-based

    Contribution based Contribution based

    Direct job placement CWI CWI Reintegration companies

    Case Management/ Re-integration Re-integration Re-integration companies

    Provision Employment Services companies companies and in-house

    Sanctions Employee Insurance Municipalities Firms

    Agency (UWV)

    Tendering Employee Insurance Municipalities Firms (not obligatory)

    Agency (UWV) (not obligatory)

    Voucher UWV Municipalities

    Individual Reintegration Personal Reintegration

    Agreement (IRO) Budget (PRB