A conservative template for welfare reform.

As a result of remarkably successful policy shifts in both Canada and the United States, there has emerged a conservative template for welfare reform--focus on early intervention, make work pay for families with children, and target workfare carefully and sparingly.

These policies can improve the lives of the poor by reducing poverty and increasing employment for those formerly dependent on welfare. As a bonus, they cost less. In this essay, I review the most important conservative developments in welfare reform, and conclude with two case studies--Ontario and Alberta--that provide empirical evidence for my conclusions.

Welfare and employable individuals

The key problem with welfare programs in both Canada and the United States at the beginning of the 1990s was that too many employable individuals found themselves trapped in a cycle of dependency on welfare. In fact, because Canada's provincial welfare programs embraced single individuals, single parents and two-parent families with children, while their American counterparts were generally restricted to single parents, the problem of employable individuals trapped in welfare was much more acute in Canada. In Alberta, for example, nearly two thirds of all people on welfare were classified as "employable" in 1993, while in Ontario, fully three quarters of the welfare caseload was classified as "employable" in 1995.

Another problem was the large number of new applicants to welfare programs. Using Alberta as an example again, 15 per cent of welfare recipients in the early 1990s had never been on welfare before. A rise in what was intended as short-term use, particularly by young employable individuals, was the most important reason for this increase. In effect, these recipients were using welfare not as a program of last resort but as a second unemployment insurance program. Despite the short-term intentions behind this strategy, it was a worrying phenomenon because nearly every study shows that a first stint on welfare dramatically increases the likelihood of being on welfare again at some point in the future, and a troubling number of those returnees become trapped in a cycle of dependency.

The most insidious trap is the perverse incentive of benefits that exceed market earnings. This incentive became a reality as welfare rates shot upward in an attempt to be more "compassionate" to the poor. In particular, benefit levels for families with children, even if set at basic subsistence levels, were more than a single breadwinner could earn in the workforce. High benefit levels attracted even more people onto welfare programs by making it more difficult for them to find jobs that paid more than welfare, and then kept them there for the same reason. The largest gaps between welfare rates and low-income market rates appeared in Quebec, the prairie provinces and Ontario in the early 1990s.

The conservative response

The conservative template for dealing with these challenges tries both to keep people off welfare in the first place and to wean those who have become dependent from the system. Strategies for preventing the onset of welfare dependency focus on welfare applicants rather than welfare recipients. (1) Known in the parlance as "early intervention," this approach includes lower benefits, tighter administration, screening and time limits. Time limits, of course, are also part of the strategy to wean people from welfare, along with a properly constructed workfare system and an appropriate system of earnings supplements to families with children. I will discuss each of these strategies in turn.

Lower benefits: The first key to early intervention is lowering benefit levels so that they do not exceed the wages of low-income working people, thus ending the perverse incentive to better one's situation by leaving the work world in favour of welfare dependency. This is fairly straightforward in the case of single individuals without children, for whom one can simply ensure that welfare benefits are slightly below the pay of a full-time minimum wage job. A larger problem is families with children: even a subsistence level of income support for families with two or more dependents is usually higher than one person could earn at minimum wages. I address this sticky issue below.

Tighter administration and screening: The second key to early intervention is administrative reforms that ensure welfare benefits are targeted only to those who need them. Conservatives sought to reduce the intake of short-term users--those who were using the system as a form of unemployment insurance--by tightening eligibility, particularly for young employable individuals. Reforms required that employable applicants exhaust all other means of support, including active labour market programs, before applicants qualified for welfare benefits.

Over time there emerged a move to more objective approaches to reduce the intake into welfare programs. In 1997, for example, the OECD recommended that countries adopt a formal screening program to segment applicants into categories based on their employability. (2) In the United States (particularly Wisconsin Works) and Australia, screening systems employ formal profiling tools based on objective characteristics. For example, Australia's Job Seeker Classification Instrument collects information on 18 dimensions, including demographic characteristics, qualifications, labour force experience, family status, location, mobility, personal disadvantages and disabilities. (3) A score derived from these factors determines job seekers' eligibility for particular levels of service. The central idea is to keep job-ready individuals out of passive welfare programs.

In most jurisdictions, screening is done informally by caseworkers, an approach that allows administrative laxity over time. A formal pattern allows welfare caseworkers to make what might be considered harsher judgements in the short term to prevent more people from becoming dependent on welfare in the longer term.

Time limits: Time limits force applicants to answer up front the question of whether the receipt of welfare benefits is truly their last resort--if they experience a stint on welfare today, that may preclude their use of welfare programs in the future. It also addresses repeat use of welfare programs that seriously erodes the human capital of employable people.

Canadian provinces (British Columbia is a recent exception) are among a minority of jurisdictions that provide indefinite assistance to employable individuals. (4) The vast majority of countries impose time limits of from three months to five years on the receipt of cash benefits by employable clients, and in recent years those limits have been shortened. Time limits on the receipt of benefits have been a successful strategy for reducing the attractiveness of welfare, and have contributed to the reduction in caseloads in the United States. (5) In addition to adding incentives to stay off welfare, they obviously get people off the rolls once the specified time period has elapsed--which brings us to the issue of "weaning."

EARLY INTERVENTION IN THE FORM OF lower benefits, tighter administration, screening and time limits dealt with the biggest problem facing welfare programs in the early 1990s--the preponderance of young, employable, single individuals who were wasting their human capital through repeated stints on welfare. However, these measures did not adequately address two other problems: the growing number of people for whom welfare became a way of life and the tricky issue of setting welfare benefits for families with children. Workfare is the conservative approach to dealing with the first problem--though conservatives have shown a disturbing tendency to prescribe it as a solution for all welfare problems. Earnings supplements address the second issue.

Workfare: Workfare is best thought of as a policy solution to a transitional problem. The need for workfare arose because of the growing number of people whose lives had been ruined by poorly designed and administered welfare programs. By the early 1990s there were thousands of potentially employable individuals whose human capital had been seriously eroded by repeated use of welfare. The purpose of workfare is to adjust the price of hiring hard-to-employ individuals. That is done by increasing the value (or relative price) of the individual's potential contribution through training or a community placement or by lowering the cost of hiring the individual through employment subsidies. The proper target for workfare is therefore people who could not have found work on their own. This brings us back to effective screening.

If there is no effective screening to determine who requires workfare, those who do not need assistance may end up being the ones getting it. The problem is exacerbated by the incentive for caseworkers to show good results by pushing easier-to-employ individuals through employment assistance programs--the practice of "creaming." International evidence suggests that employment assistance programs such as training or employment placement programs are most effective when they are well targeted to high-need individuals. There is also a great deal of evidence that providing employment assistance beyond job searches to employable youth is ineffective. (6)

Workfare should therefore be targeted only to high-need individuals who have become trapped in welfare programs, who are employable but have lost important life and work skills that a workfare program can restore. There probably is an argument for an ongoing program of workfare, though I would argue that any such program should be very small and limited only to so called "hard cases": people who, with a well-designed welfare program, are on the fine line between employable and not employable.

Earnings supplements--making work pay: As indicated above, welfare benefits for families are often higher than what these families can earn in the labour market, thus creating perverse incentives to get on welfare and stay there. Earnings supplements can "make work pay" by boosting the return from work so that it exceeds that of receiving welfare. Earnings supplements are best directed at families with children (single- and dual parent), with the supplement often tied to the number and age of the children.

In a C.D. Howe Institute study, John Richards discusses earnings supplements in detail and raises a number of design and implementation issues. (7) First of all, earnings supplements must "feel" like wages. People who receive wages have very small time lags between work and wages. Earnings supplements must therefore minimize the time between work and receipt of the earnings supplement.

In addition, the supplement must not become an integral part of the "welfare bureaucracy." For those receiving welfare, lowering the clawback rate on social assistance benefits could achieve a similar effect to the supplement. However, there are two important drawbacks to this approach. It forces all eligible families onto welfare to receive the supplement, and it forces people to deal with the pervasive government intrusion that characterizes welfare bureaucracies. The bottom line is that the income they receive should feel more like part of their wages than a form of welfare payment. Finally, the phase-in rate should be designed so that it boosts wages over as large an income range as possible.

Case study 1: The "Cardinal reforms" in Alberta

The welfare reforms most consistent with this template are those implemented by the former Minister of Family and Social Services in Alberta, Mike Cardinal. (8) Cardinal's reforms began in 1993 when 7 per cent of Alberta residents were on welfare. The province's welfare program was poorly administered, benefits were overly generous and increasing numbers of young people were getting trapped in a cycle of dependency (just under half of recipients were single, and almost two thirds were under 34 years old).

In conformity with conservative early intervention principles, Alberta reduced the intake of recipients by tightening eligibility, particularly for young employable individuals; required that employable applicants exhaust all other means of support, including active labour market programs, before granting welfare benefits; and enforced these administrative rules by setting targets for each local welfare office. As a result, the number of people on welfare fell by about 35,000 (one third of the 1993 caseload in Alberta) as a result of lower intake. Alberta also reduced benefit levels, insisting that they not exceed the wages of low-income working Albertans. This meant a 19 per cent reduction for single employable individuals, with 12 to 13 per cent reductions for other demographic groups. Mandatory workfare was not part of the Alberta reforms. The closest Alberta got to workfare was three small programs targeted at long-term welfare users--at a cost per participant of four times the benefit rate for single employable individuals.

The impact of Alberta's early intervention was substantial. The province reduced inflows into welfare by 60 per cent within a year. (9) The overall reduction in welfare use from 7 to 2 per cent of the population was not the result of pushing people out but of making it tougher to get in.

The impact on the incidence of low income--sometimes called poverty--was equally impressive. In 1993, 14 per cent of Alberta's population fell below Statistics Canada's Low Income Cut-Offs (LICO), measured after taking into account all taxes and government transfers. By 1998, the number of Albertans living below the LICO fell to 12.5 per cent--a drop that is two and a half times larger than the drop in Canada as a whole over the same period.

These are impressive numbers, but they do not tell the entire story. The real story is what happened to those people who no longer relied on welfare to pay the bills. In my study of the Alberta welfare experiment, I looked at the employment rates for individuals under the age of 25, a group that made up a majority of the reduction in welfare clients in Alberta. (10) I found that for the first time in nearly two decades, employment rates for young Albertans kept up with economic growth in Alberta after 1993. In fact, the number of jobs created for young people in Alberta during the two years following welfare reform exceeded the number of jobs created for young people in the rest of Canada--even though Alberta made up only 10 per cent of the Canadian population.

An even more impressive result was discovered in a remarkable set of data recently drawn out of the 1996 census by C.D. Howe Fellow John Richards. (11) Richards compared labour force data in Calgary and Edmonton to Winnipeg, Regina, Saskatoon, Toronto and Montreal. He found that employment rates in Alberta cities were over 7 percentage points higher, and unemployment rates 2.1 percentage points lower, than in these other Canadian cities. Even more interesting, Richards dug down to the neighbourhood level, dividing neighbourhoods into poor and non-poor. He found that Alberta's employment rate in poor neighbourhoods was 8.5 percentage points higher, and unemployment rates 4.7 percentage points lower, than in other Canadian cities.

These results bolster my earlier findings that the increase in employment rates for Albertans under 34 was large enough to explain the reduction in welfare rolls. In short, earlier evidence suggested that Alberta had discovered the most impressive job creation program for youth that the country had ever seen. The twist that Richards has added is that Alberta has found the most impressive job creation program for poor youth that the country has ever seen. And just to placate the critics, it wasn't the economy that did the job. Economic growth in Alberta was weaker following the 1993 welfare reforms than during the late 1980s. Yet during the late 1980s, the number of welfare recipients in Alberta increased.

In short, Alberta's reforms to tighten eligibility and reduce benefits resulted in a two-thirds reduction in welfare recipients, a reduction in poverty and an increase in the number of poor people who found work. On top of these benefits, the reforms saved Alberta taxpayers millions of dollars.

Case study 2: The Mike Harris reforms in Ontario

Between 1980 and 1995, Ontario went from having the lowest percentage of its population on welfare and relatively stingy welfare benefits to having the highest proportion of its population on welfare and benefits far above that of any other province. (12) In 1992 employable individuals and students made up more than three quarters of the General Welfare caseload.

The key element in the Harris government's reforms was a mandatory workfare program for employable individuals. This program has been poorly targeted, is exceedingly expensive and has been a bureaucratic nightmare. (13) It is a good example of the adage that "workfare is a word some conservatives use to justify bigger government."

Ontario also reduced welfare benefit levels. Although these benefit reductions were long overdue, Ontario's benefits as a ratio of low-income wages remain 10 per cent above the levels in Canada's three other large provinces, suggesting that there is room for further reduction. (14) Other than these reductions in benefit levels, the Ontario reforms ignored Alberta's "early intervention" model. There were few changes to eligibility rules, and workfare recipients had to be on welfare for four full months before being eligible for workfare. (15)

What were the results of the Ontario reforms? Ontario had 11 pet cent of its population on welfare in 1995 when it initiated its reforms. It now has 7 per cent of its population on welfare--the level Alberta started with in 1993, and the current national average. And Ontario's reforms have been very expensive. Ontario spends 40 per cent more per welfare client than Alberta does--$5,660 per client versus $4,050 in Alberta (which is still higher than the national average of just under $4,000).

Apart from modest and insufficient benefit reductions, Ontario did not focus on early intervention. Instead, it introduced a poorly targeted workfare program. There has been some reduction in welfare use, but the number of Ontarians on welfare remains frighteningly high. And to top it off, Ontario has the most expensive welfare program in the country.

Summary

Despite the incredible record of conservative welfare reforms in Alberta and the United States, not all governments (even conservative ones) appear ready to embrace the conservative template for welfare reform--focus on early intervention, make work pay for families with children, and target workfare carefully and sparingly. This template has proven to be a way to reduce welfare use, poverty and the cost to taxpayers and increase jobs for the poor. It is hard to imagine a more attractive outcome for everyone involved.

Notes

(1) See Organization for Economic Co-operation and Development, Labour Market Policies that Work: Policy Brief (Parts: OECD, 2001); Douglas J. Besharov and Peter Germanis, "Welfare Reform--Four Years Later," Public interest, no. 140 (Summer 2000); Human Resources Development Canada, Lessons Learned: Reconnecting Social Assistance Recipients to the Labour Market, Final Report (Ottawa: HRDC, 2000); Ross Finnie, Low Income Poverty Dynamics in Canada: Entry, Exit, Spell Durations and Total Time (Ottawa: Human Resources Development Canada, 2000): John P. Martin, What Works among Active Laborer Market Policies: Evidence from OECD Countries' Experiences (Paris: Organization for Economic Cooperation and Development, 1998); Kenneth J. Boessenkool, Back to Work: Learning from the Alberta Welfare Experiment (Toronto: C.D. Howe Institute, 1997): John Richards, Retooling the Welfare State: What's Right, What's Wrong, What's to Be Done (Toronto: C.D. Howe Institute, 1999); John Richards, Neighbors Matter: Poor Neighborhoods and Urban Aboriginal Policy (Toronto: C.D. Howe Institute, 2001).

(2) See Organization [or Economic Cooperation and Development, Enhancing the Effectiveness of Active Labour Market Policies: A Streamlined Public Employment Service (Paris: OECD, 1997); HRDC, Lessons Learned; Martin, What Works.

(3) See Government of Australia, Job Network Evaluation, Stage One: Implementation and Market Development and Stage Two: Progress Report (Canberra: Department of Employment, Workplace Relations and Small Business, 2000-2001).

(4) See David W. Kalisch, Tetsuya Aman and Libbie A. Buchele, Social and Health Policies in OECD Countries: A Survey of Current Programs and Recent Developments (Paris: Organization for Economic Cooperation and Development, 1998).

(5) See Robert Rector and Sarah E. Youssef, The Determinants of Welfare Caseload Decline (Washington: Heritage Foundation, 1999); Besharov and Germanis, "Welfare Reform."

(6) See OECD Labour Market Policies that Work; HRDC, Lessons Learned; Martin, What Works; Kenneth J. Boessenkool and William B.P. Robson, Ending the Training Tangle: The Case against Federal Provincial Programs under El (Toronto: C.D. Howe Institute, 1997).

(7) Richards, Retooling the Welfare State.

(8) See Boessenkool, Back to Work; Richards, Neighbors Matter.

(9) Boessenkool, Back to Work.

(10) Ibid.

(11) Richards, Neighbors Matter.

(12) See D.M. Brown, "Welfare Caseload Trends in Canada," in John Richards and William Watson, eds., Helping the Poor: A Qualified Case for "Workfare" (Toronto: C.D. Howe Institute, 1995), pp. 59-64.

(13) See Kenneth J. Boessenkool, What Ontario Could learn from Alberta's Welfare Experiment (Toronto: University of Toronto, Centre for the Study of State and Market, 1998).

(14) See Boessenkool, Back to Work.

(15) See Boessenkool, What Ontario Could Learn.

Kenneth J. Boessenkool is President of Sidicus Consulting Ltd., an economic and public policy consulting firm. He is the author of Back to Work: Learning from the Alberta Welfare Experiment (1997) and What Ontario Could Learn from Alberta's Welfare Experiment (1998).