Welfare reform

Among the issues that won’t be addressed in detail in the current election campaign is the case for, and against, welfare reform along the lines adopted in the US. Although the Howard government has made various changes aimed at increasing ‘mutual obligation[1]’, there has been nothing approaching the reforms to the main pure welfare program in the US, then called Temporary Aid for Needy Families, and received primarily by single-parent families. Among those who still think substantial measures of microeconomic reform are needed, welfare reform along US lines is at the top of the list, and Wisconsin, where Governor Tommy Thompson slashed welfare rolls, is generally held out as the model.

Although there are various rationales for welfare reform, the only one I think worth considering is the claim that welfare perpetuates poverty. While relieving immediate distress, it is argued, welfare encourages a culture of dependence that perpetuates poverty. Against this, I’d put the argument that what matters most in preventing dependence is the availability of good jobs, and that a government commitment to full employment is what is needed for a genuinely mutual or reciprocal obligation to work.

There’s not likely to be a convincing and rigorously defensible empirical resolution of this debate any time soon. However, one telling piece of anecdotal evidence is worth a dozen regressions[2], so I thought I’d check out the Wisconsin example. What I found is a report with the headlines

Poverty rate hits 10-year high

State’s struggles also evident in growing number of uninsured: 11%

The rise in the poverty rate[3] is attributed mainly to the loss of manufacturing jobs, rather than to adverse effects of welfare reform. This is consistent with my general view that we need to look harder at employment and unemployment. But reform is clearly playing a significant role

We have gone since 1996, when Pay for Performance hit, from distributing 1.5 million pounds of food to 10 million on an emergency basis,” Tussler [executive director of Hunger Task Force of Milwaukee[ said, referring to the reform requiring work or job-seeking in exchange for welfare benefits.

I’m not claiming that this example proves that welfare reform increases poverty. But it’s hard to see how Wisconsin can be regarded as a successful exemplar of welfare reform when the poverty rate is higher than it was before the main phase of reform, and still rising.

fn1. As many have pointed out previously, the current government’s notion of mutual obligation involves drastically reducing its own obligations, while increasing those place on benefit recipients

fn2. Irony tags were clearly needed here, but Textile doesn’t support them

fn3. This is an absolute poverty measure, based on a poverty line set in 1965, and adjusted since then only for inflation. Of course, it’s not absolute by comparison with third world countries, and it’s about twice the income that was considered to constitute poverty 100 years ago.

26 thoughts on “Welfare reform

  1. John

    I don’t quite follow your argument. On the one hand you say that the rise in poverty in Wisconsin is due to the loss of manufacturing jobs and not to welfare reform. But then you say that Wisconsin cannot be a good exemplar of reform when poverty is higher after reform than before. The second statement seems to ignore the logic of the first.

    The problem in assessing any welfare policy is the absence of a good counterfactual. That is, it is impossible to hold all other factors equal before and after the reform. So we are inevitably left with the problem of having to control for other factors when assessing the effects of reform. The problem with the Wisconsin example has always been that while it was enjoying its success in the mid to late 1990s, the US was riding the tide of improving labour market conditions, which assisted the transition from welfare to work. With the more recent poor performance of the US labour market, the poverty statistics are now looking a lot worse than a few years ago.

  2. It seems that the programs are designed to move people off welfare and not reduce poverty. If you take that as being the main point of the programs (to reduce the welfare roll) then they are working fine. Welfare reform is not implemented to alleviate poverty. The the driving factor behind welfare reforms is political. It has nothing to do with economics.

    Hence why welfare reforms do not greatly affect the poverty rate.

  3. John,

    I would be interested in your reader’s election thoughts on the concept of the “guaranteed adequate income” (GAI) or Basic Income Guarantee (BIG) as advocated by Milton Freedman (on the right) and George McGovern and Brazilian government (on the left).

    In Australia, the Australian Green party has long recommended investigation into a GAI as the one policy that is by its very nature guaranteed to dramatically reduce and perhaps even eliminate poverty without substantially altering the beneficial aspects of a market economy.

    BIG is a proven, efficient, effective, and equitable solution to poverty – so why do the Old Parties dismiss this as “kooky” without even responding to the arguments that the Greens are trying to raise in the community – and why doesn’t the Murdoch Microphone (Australia’s very own China Daily)dare to ever “expose” this policy to the Australian people?


  4. More later in reply to JQ, but for now a reply to Mark without giving the backing (I’ll give that later too).

    My enquiries suggest that GAI is long term equivalent to both negative income tax (the “five economists’ plan”) and to Professor Kim Swales’ GST-oriented approach I look into at my publications page (follow the links from my home page). That is, equivalent if set at the same levels.

    The differences come in during the transitions and from the differences in levels. GAI is set far too high to be sustainable; it would need to be set as a guaranteed INadequate income to work – but then, its transitional problems catch too many people in the works. That’s partly because its transition is far too long. Atrue GAI has to hope that people won’t lose the work ethic before fiscal drag turns it into an inadequate (top up) income, and that almost certainly can’t work without casualties and might not work at all (i.e., it would be killed by transitional costs).

    A good idea that won’t work, like a child’s drawing of a spaceship with the fuel tank under the seat.

  5. John,
    I speak as a total non-expert on these things; excuse me if my question seems naive.

    Isn’t it a bit dangerous to compare us with the U.S. when it comes to welfare?

    For example, isn’t the make up of sole parents in Australia is different. I saw some stats on this once from a government department (sorry, about a year or so ago, and I can’t remember enough details to relocate) and they were anything but the sort of image of sole parents presented in the media and certainly different from the image of sole parents presented in the U.S. media. I seem to recall the stats indicated most sole parents here were mostly divorcees etc. rather than teen mums, their average time on welfare is short (2-3 years??), they tend to have a higher participation rate (e.g. many work part time as well as get welfare) than even the unemployed.

    I was also involved with some market research a few years ago that included single parents on welfare as a research group and their attitude was (surprisingly) anything but welfare dependant – they were certainly well motivated and actively pursuing moving off welfare, doing courses to improve skills etc. Don’t know how much that has changed now.

    I also wonder if that same applies for the unemployed.

    If our demographics and attitudes are different, then shouldn’t we be tailoring our policies to fit our own?

  6. Mark, my argument is that, if welfare reform were an effective way of reducing poverty, we would have seen beneficial effects sufficient to offset a fairly mild recession. Conversely, as you observe, it now seems clear that the beneficial effects imputed to welfare reform were mainly due to the strong economy (or at least that the success of the program depended on a strong economy.

    Saint, I agree that there are big and important differences, but it’s still important to look at the US example.

    I’ll try and talk more about GMI and other issues soon.

  7. Causality is always hard to estabilish. However the worst way to go in estabilishing causality is via anecdotes .One regression is worth millionty millions of anecdotes . Otherwise just leave it to the economic barbarians and consign millions to a dwarfed existence. The first point is to elaborate a reasonable model of the phenomena : rates of transition from the welfare pool to work ; rates of part-time employment and study in the welfare pool .i.e. preparation for work . A lot of these considerations as ever are “non – economic” because so called tradional economics is such a poor model for studying societies .

  8. If welfare reform has no effect on poverty, but costs billions less — then surely it is a good thing?

    I like the negative income tax, wrote my honours thesis on it, and it is a part of the platform of the liberal democratic party (www.ldp.org.au). However, I don’t believe it can work as both revenue and benefits neutral to all people. It’s the same old welfare triangle of incentives-cost-level. One of these has to be sacrificed. Those on the left would sacrifice cost (make the govt pay more), those on the right would sacrifice the level (make the recipients live on less – or get a job) … but nearlly all economists agree that we shouldn’t sacrifice the incentives. Currently there is an EMTR of 60 or 70 (and sometimes around 100%)… and this needs to chance.

    In my opinion this (EMTR) is the biggest outstanding economic issue in Australia.

  9. The ideological assumption of “welfare reform” agendas is that long-term welfare recipients (other than age pensioners) are lacking in social capabilities and in the capacities required to be responsible self-determining adults. The stereotype is of the unemployable couch potato.

    This assumption can be questioned on a number of grounds. An important one in Australia is that casualisation of the labour market means that many people in paid employment don’t earn enough for long enough or on a sufficiently regular basis to be able to exit the welfare rolls. This phenomenon is particularly pronounced in particular occupations, industries and regions. As a result many of the so-called “long term unemployed” are really long-term under-employed or long-term peripherally employed.

    Far from being dysfunctional or incapable, many people in this predicament are highly capable, sometimes with multiple qualifications and aspirations for more. They are usually doing all they can to obtain a more secure career footing, but have to deal with economic and political circumstances beyond their control.

    What is the effect of “welfare reform” on these groups of people? In the Australian regime of “mutual obligation” and mandatory Intensive “Support”, it can mean disruption of useful self-initiated activies such as study or voluntary activity by time-wasting and stressful dealings with Job Network providers, first to determine which (if any) version of finger-painting for grown-ups they should be press-ganged into under the rubric of Intensive “Support”, and then (if they can’t talk their way out of it), the finger-painting.

    Under the even more regressive regime in some US states which is favoured by Peter Saunders of the CIS, it would mean mandatory ***full-time*** work-for-the-dole after six months on the welfare rolls. This would be inimical to remaining in paid casual work on the same level as previously (as they would not be on-call for their employer whilst they were out painting rocks for the dole), and catastrophic for self-initiated efforts to genuinely improve one’s employability through study, etc.

    Ceteris parabus, welfare reform of this kind would perpetuate and worsen poverty for many Australians.

  10. I’ve wondered how much it would concentrate the politician’s minds if they could be shamed into setting a new top marginal tax rate set at the maximum of the current top rate and the highest of the Effective Marginal Tax Rates – starting say at
    backbencher’s salary.

  11. The Dec 2002 issue of the Australian Journal of Labour Economics dealt with Welfare and the Labour Market. An article by Uni of Wisconsin’s Barbara Wolfe looked specifically at her home state’s experience.
    The article explains that the state’s program was set up to ensure that all working age people with income at or below 115% of the poverty line have work, and do work. Those who do not have standard jobs have state-subsidised private sector jobs, or government funded community jobs, or are kept busy on approved community activities. These people are required to work at least 30 (but usually 35) hours per week. The ones in standard or subsidised private sector jobs are eligible for Earned Income Tax credits. Health insurance and child care are available for those below an eligibility cut off.
    The US states began getting a block grant from the Federal government when the previous assistance regime was terminated in 1996, and presumably it is this grant which funds the subsidised and government-created jobs.
    Commenting generally about the US developments, Wolfe indicates ‘the reform has clearly increased self sufficiency, on average, among former recipients. Many are working. Many
    have increased their earnings over time. Income inequality has increased and this is not surprising. Second, continued participation in benefit-in-kind programs is sensitive to implementation and program requirements [ie have to make sure people take up the benefits to avoid destitution]..Third, school age and younger children appear on average not to have been harmed..Alternatively, adolescents may be showing signs that some of them are indeed worse off [through lack of supervision]’
    Wolfe makes it clear that the program is considerably more expensive in the short run than cash assistance, but suggest taxpayers are more willing to pay more to support the poor if they are seen to be working for their keep [ie become deserving poor]
    The issue of the journal has an intro by Chapman and Gregory, and interesting articles by Peter Dawkins, Patricia Apps, and [the real] Peter Saunders.

  12. The Hawke government introduced the JET scheme based on a USA equivalent with a few changes to humanise the process and in recognition that Sole parents have unique issues. This was the attempt to prevent children living in poverty by 1990 and there has been little credit given to the success of the program in the early years.

    Under the current government it has more compulsion and is probably less effective as it relies more on Job Network services which are targeted to those who have more flexibility in their lives.

    Sole parents have to do the job of raising children which still means a lot of time and effort. Forcing sole parents with children under the age of 16 off payments will result in a great many more social ills so we may save money on welfare but have to spend it on police and social workers instead.

    Most sole parents who are not harassed by ex partners or mentally ill will try and balance work and family so that their children don’t live in poverty. They need to have access to childcare which has always been a prime part of the JET program.

    Welfare reform has been driven by the belief that private enterprise can do it better. This is questionable in an industry where there is very high staff turnover and the emphasis is on compulsory attendance at interviews and where jobseekers who find their own job have the government paying funds to the Job Network based on their results.

    To reduce Welfare costs there needs to be a great deal more thought about how people make choices and what choices there are and how the work/life/Family balance can be met.

  13. Saint – I’m seem to remember the very recent figure in OZ is less than 3% of single parents are teens (or it might even be under 20yo).

    I’m sure you are also aware that the teen single parents arent the year 12 girls from Methodist Ladies College.

  14. All this bypasses the idea that welfare is nothing more than legislated altruism; if funding of welfare was voluntary, no-one would have a problem with it, and I think it would be a bloody sight better administered and audited. Why is it that no-one ever questions the right of a welfare state to exist? BTW- the old chestnut about available work is cobblers- demand is far in excess of supply. Just locally, one of our contract carriers has had to drop all it’s work for the rest of the week due to having four trucks idle, and no takers to fill the vacancies.
    Current payment levels don’t provide much incentive for those not inclined to work; I have supported the idea of a national insurance scheme (voluntary) rather than the bottomless pit of dipping consolidated revenue.

  15. Welfare reform does not rest on the assumption that the unemployed are couch potatoes. It amazes me how little people on the left actually understand about their ideological enemy — straight from the Tim Blair school of ignorance. “He who knows only his own side of the case, knows little of that” – J.S.Mill.

    Both of the Peter Saunders are real… though I agree that only one is worth listening to. http://www.cis.org.au

    The fact that people who can’t get welfare seem to magically find more jobs shouldn’t be dismissed as voodoo science and an evil ploy from zionist conquistadors. It’s worth understanding. And to pretend that incentives don’t change behaviour is odd to say the least. And stupid to say more than the least.

  16. one telling piece of anecdotal evidence is worth a dozen regressions
    No it’s not, and as a social scientist you should be ashamed to think so. A proponent of US welfare reform would say that the counterfactual is that poverty would have got worse even faster. Your anecdote is no rebuttal of that assertion – but the right dozen regressions might be.

    The great bulk of impact evaluations using more rigorous approaches do indded find welfare-to-work caused net reductions in poverty, though it was nowhere near as influential as the jobs boom and the EITC. This is not Australian-style ‘program evaluation’ – the evaluators are at arms length from the politicos and in many cases started with very different priors.

    Australian opponents of this approach do better to argue that ‘welfare’ in a US context goes overwhelmingly to never-partnered black mothers – a very different group from most of our (divorced or separated) sole parents, let alone most of our other welfare recipients. Also it’s predicated on strong demand for unskilled workers – but neither I nor you would advocate the (anti-union, ultra-low wages) policies needed to generate that strong sectional demand.

    Also, the Yanks have modified the Orshansky poverty line in the last few years, mainly to take account of more up to date consumption behaviour and changes in relative prices (see here for details). Be cautious of statements about movements in US poverty levels over the last 10 years as it may not be comparing apples with apples; there’s not much left of the original 1965 real value approach. It’s still budget standards and Engel curve based though (ie an absolute line linked to costs of “essentials”).

  17. John, your link requires registration, and I don’t do registrations except through Bugmenot.
    However even that doesn’t work.
    So I’ll assume that the article covers much of the same ground as the Joyce Report
    The section on “What we have learnt”, and Wisconsin are instructive.
    I think the most important fact to remember when reading the report, is the fact that “poverty” generally relates to single female parents. Women who marry and stay married generally avoid the poverty trap.

  18. It worked the first time (both at work and at home), but asked me to register when I went back for a second look.

    DD, what does the Wisconsin experiment have to do with Welfare to Work?

  19. Frances – I wouldn’t be surprised if it was 3% or less. But it’s hard to fight perceptions with facts(!).

    I have observed a few things in the course of my sojourn on this earth *grin*: if you want to make headlines bash the unemployed, single parents, indigenous Australians and migrants. This is about the only area I think where our pollies actually influence public opinion.

    I’m sorry I don’t get the reference to the Methodists. I think however Howard was raised as one….am I getting warm?

  20. James –
    “Welfare reform” in the US context involved two approaches – the EITC and subsidised childcare to deal with financial incentives and intensive ‘help and hassle’ – some states (including Wisconsin)use the govt as employer of last resort. Both approaches are explicitly directed at pushing people into a job, any job; they are welfare-to-work programs. Wisconsin was the most enthusiastic pioneer of the second apporach with its ‘Wisconsin Works’ program.

  21. Thanks, DD. I asked bacause there wasn’t anything in the linked story about employer subsidies, childcare assistance, tax credits and so on -things you would normally associate with a welfare to work program. But I have done some research now and, as you say, Wisconsin does have a comprehensive program, and is indeed famous or at least infamous for it.

    If poverty increased sharply because of a local manufacturing slump, it may be unfair to judge Wisconsin’s program by the poverty statistics. But it reflects badly on the design of the American welfare system as a whole if each state has to bear the brunt of a regional shock when times are prosperous overall.

  22. Tinkering with Health, Howard on Women
    Looking at my referrals list makes the site look like a subsidiary to John Quiggin Enterprises. In order to show that this site still maintains editorial independence, I will take my proprietor to task over his Welfare Reform posting. I’m not quite s…

Comments are closed.