Making Workfare Work: The US Experience — Interview with Lawrence M. Mead

New York University’s Professor Lawrence M. Mead discusses workfare and income inequality in light of the US experience after a decade of welfare reform.

Date Posted

1 Oct 2007


Issue 3, 14 Oct 2007


Singapore has put in place a Workfare Income Supplement scheme as an additional social safety net for low-income workers. What can we learn from the US experience in work-for-benefit programmes?

It is important to distinguish American-style work enforcement from Singapore’s workfare. The two initiatives are markedly different: in Singapore, workfare refers to the supplementation of wages, while in the US, workfare entails requiring people who receive aid to work.

The national Personal Responsibility Act of 1996 accelerated welfare reform. The Act meant that welfare was no longer a legal entitlement. However, there was no real reason to abolish the entitlement since the intention was to condition aid on work. Another excessive aspect of US welfare reform is the time limit enforced which only allows recipients to be on aid for a maximum of five years. This was unnecessary since reform was not aimed mainly at ending dependency, but instead at encouraging people to work.

One state, Wisconsin, even insisted on a universal and immediate work requirement, which was more severe than required. Wisconsin’s strong administrative capacity, however, meant that welfare reform met with success in the state. Washington was another state that saw the smooth running of welfare programmes post-1996. Like Wisconsin, Washington had a strong tradition of good government.

US welfare reform was aimed mainly at raising work levels. Reducing dependency was secondary. The idea that the poor should be entirely non-reliant on the Government is never going to happen and it is even unnecessary. In reality, even the middle class is reliant on the Government. Strict self-reliance is not a meaningful goal. Instead, the goal of reform should be to achieve joint responsibility between citizens and the Government. The poor have to help themselves as much as others help them.

The focus on work rather than reducing dependency meant that welfare benefits were not generally cut. Most of the states which control benefit levels actually raised them in the sense that work incentives were increased. That is, recipients were allowed to keep more welfare if they worked. The combination of earnings with remaining welfare could raise incomes substantially. The cost of work incentives along with the childcare and healthcare provided to support working meant that welfare did not save money for the Government, contrary to what some commentators expected. Rather, spending increased.

Helping the poor move out of poverty also means that there is an attempt to close the widening income gap between the rich and the poor. What should governments do to deal with the challenge of inequality?

Inequality is a real issue but less important than may appear, for two reasons. First, the causes are not well understood. We do not know why inequality is rising. Some scholars find that the trend is related to the decline of labour unions, to the rising payoffs to higher education, and to a lesser degree, to growing trade and immigration, but these factors account for only part of the trend. Second, concern about inequality is mostly confined to academics. The ordinary person cares less about it and, as a consequence, politicians do not grant it much attention. A more pressing concern among ordinary Americans is whether they can secure a job and get ahead on their own. They are not particularly offended by the rich.

There appears to be greater political concern about inequality in Europe than there is in the US. This seems to be the case in Singapore as well. The political class in Singapore is concerned about inequality in a way that is not true in the US. Here the problem is a more serious political question.

If governments choose to tackle this problem, it is clear that one reason for growing inequality is disorganisation at the bottom of society and falling work levels among the poor. Welfare reforms and other efforts can help to raise work levels at the bottom. In the 1990s, the massive increase in earnings among the low-income as a result of welfare reform did slow trends toward inequality considerably. In fact, measures like this might even reverse the trends and reduce inequality. Hence, getting the lower-income population to work more steadily is the best thing governments can do to retard inequality.

Does making welfare recipients work for a reward facilitate upward mobility, for example, through increased earnings? Are there also other benefits or costs of working that should be taken into consideration?

Work does to some extent translate into social mobility. Recipients who go to work typically experience some gains in income. That gain is less, admittedly, in states paying high welfare benefits, as the recipients lose more income when they leave welfare for work. But even in these states they are generally better off than they were before, at least in terms of income.

One can argue that the gains in income may not necessarily offset the strains and difficulties that come with working. Some academics have concluded that while mothers who work have higher incomes, they also have higher expenses than non-working mothers. Thus, they may not be better off than their non-working counterparts. However, welfare recipients typically also receive monies from other sources such as families or partners to whom they may not have been married. When they go to work they usually get more from these sources than they did while on welfare, so again they are usually better off than they were on welfare.

Clearly there is progression of earnings over time, provided a mother keeps working steadily. These gains are lost, however, if mothers drop in and out of work, as many do. Welfare recipients who work continuously usually get out of poverty within a few years.

Do the children of welfare recipients do better in future if their mothers work? Evaluations suggest that some young children do better in school if their mothers have to work, while some adolescents are more likely to get in trouble after school if their parents are working and unavailable to supervise them. But all these effects are small. In the longer term, we can only presume that it is good for the children to see their mothers get off welfare and start to work. Furthermore, the children may receive more stimulation in childcare than they would if they had stayed at home with a mother who does not work. As a consequence of having been in a stimulating environment, children of recipients may perform better in schools.

In the US, former welfare recipients are usually better off because they have additional rewards such as childcare support and the Earned Income Tax Credit (EITC),1 a wage subsidy for low-paid workers much like what Singapore is instituting. This subsidy on earnings is different from traditional welfare since it is conditioned on work. While EITC does not spur people to work, as welfare work requirements do, it does make them better off if they work.

However, welfare recipients have not utilised this benefit to the fullest. Currently, they can receive this benefit for any amount of work they engage in. Many work in dribs and drabs and so do not use the credit as much as they could. One reason for that is that EITC subsidises even very short working hours. It sets no minimum of hours before you get the subsidy. Some welfare reform experiments require clients to work 30 hours a week to get subsidised. This is to ensure that they get the maximum benefit.

Enforcing a threshold also has another good effect: it offsets the possible income effect that may be produced by receiving the subsidy. That is, the subsidy might permit people to cover their financial needs by working fewer hours than before, which is undesirable. Some policymakers are hesitant about increasing the EITC, as has been proposed, because this income effect might reduce rather than increase working hours among the poor. Instituting a 20- or 30-hour threshold for EITC would be one way to forestall this.

Many welfare reform programmes in the US have been carried out with the assistance of non-profit organisations and private agencies. Do you think this is a good strategy for success?

Initially I was hostile to sub-contracting of welfare reform programmes to non-profit bodies and profit-making companies. I thought this would complicate the structures and cause confusion among welfare recipients. However, there are many instances where contracting has played a constructive role in implementing reform, especially in implementing job search requirements for welfare recipients. It was the difficulty of building the job search aspect into welfare that caused many states to turn this task over to private agencies. Generally contractors have proven more constructive than I once thought, and I say now that we should consider this strategy. However, I still have doubts about extreme privatisation, such as in Australia where they abolished the public job service and now have a maze of private agencies.

One question this strategy raises is whether it cuts back on the Government’s role in welfare reform. At one level, it might seem that the Government is heading in that direction when it turns to private agencies. That is one reason why privatisation may be impolitic. In reality, however, welfare reform is closely controlled by the Government even where private agencies are used. The benefits and the work requirements are always defined by elected officials, either at the federal or state level, and the funding is public.

Do you think that the workfare model would see success even in times of economic downturn when jobs are scarcer?

Whether or not poor people are expected to work by law has more effect on dependency than economic upturns and downturns. In determining whether welfare recipients work, the unemployment rate remains important, but it is secondary to enforcement pressures. Even during a downturn, jobs are frequently available and there is a great deal of hiring and firing. The immigration rate shows that low-level jobs are still widely available. Job seekers may be unemployed for a little longer than if the economy was thriving. The more skilled probably face a tougher time because they seek higher-paying jobs, and these are less readily available.

In the US, people stay on welfare mainly because they are not even looking for work, not because jobs are unavailable. To be unemployed, you have to be available for work—that is, you have to have entered the labour force—but have not yet accepted a job. Most people who enter the labour force find jobs so quickly that they are never recorded as unemployed at all. Similarly, those who leave jobs are seldom left jobless—they also leave the labour force and so again are not classified as unemployed. Most welfare recipients are not in the labour force and are not even looking for work. That is a much bigger problem than unemployment.

Unemployment as a problem is enshrined in our collective memory in the US because of the Great Depression of the 1930s, when jobs clearly were lacking and a quarter of the labour force was unemployed. Joblessness has been far less of a problem since then. It is a problem mostly for the higher reaches of the income hierarchy. They tend to have expectations of getting certain kinds of jobs, with conditions they think are optimal. So they do not accept just any job and unemployment for this group tends to be longer-lasting. In contrast, among the poor, the overwhelming problem is not employment or even unemployment but just getting into the workforce and staying there.

Is there a proper time to implement welfare reform? Are there certain conditions that should be in place before making workfare a permanent feature of the social safety net?

In the US, welfare reform did not come too early since we had a history of prior experimentation going back as far as the 1960s. In fact, there was already a work dimension in welfare policies as early as 1961. It increased in intensity up until the late 1980s when the Family Support Act was enacted. The enforcement of work in the US did not begin with the Personal Responsibility Act of 1996, although that legislation accelerated the process. The earlier experimentation taught that enforcing work in welfare was possible and how best to do it.

In Singapore, the Government is not acting too soon because the Workfare Income Supplement is not so radical as to change everything. Furthermore, it is designed in such a way that ongoing adjustments are possible. The Workfare Bonus Scheme, which preceded the Supplement, might be an even better idea. These proposals respond to your welfare problem, which is mostly about low-income workers and not, as in the US, getting people to work.

While it is not too soon to act, it is also critical to research the effects of the current bonus system and the Workfare Income Supplement, to see their actual effects and for the purpose of fine-tuning.

Singapore might want to learn from the US model, but bear in mind that every government that reforms welfare has to generate its own consensus about how to do that. In the US, every state made its own decisions about reform, although within federal rules. There was less borrowing from other states than many had expected. Even the more influential states, such as Wisconsin, had little influence on how welfare reform occurred in other states.

Singapore will have to justify what it has done based on its own experience. The example of Wisconsin is worth only so much to you! Only after a series of experiments will you know how best to help your low-wage workers and get support for your policies.


Lawrence M. Mead is Professor of Politics at New York University, where he teaches public policy and American government. Professor Mead is an authority on the problems of poverty and welfare in the US. Among academics, he was the principal exponent of work requirements in welfare—the approach that now dominates US national policy. He is also a leading scholar of the politics and implementation of welfare reform, subjects on which he has written several books and over a hundred publications. These works have helped shape welfare reform in the US and abroad. Professor Mead visited Singapore as a Distinguished Visitor of the Centre for Governance and Leadership (CGL), Civil Service College, where he met with senior policymakers and delivered a New Insights Lecture on “Welfare Reform in America: Successes and Limitations”. This article was extracted from an interview conducted by Dr Theresa Devasahayam, CGL Senior Researcher, on 23 June 2007.


  1. The Earned Income Tax Credit (EITC), sometimes called the Earned Income Credit, is a refundable federal income tax credit for low-income working individuals and families. Congress originally approved the tax credit legislation in 1975 in part to offset the burden of social security taxes and to provide an incentive to work. When the EITC exceeds the amount of taxes owed, it results in a tax refund to those who claim and qualify for the credit. To qualify, taxpayers must meet certain requirements and file a tax return, even if they did not earn enough money to be obligated to file a tax return.

Back to Ethos homepage