In 1966, the Canada Assistance Plan was launched with the goals of "the prevention and removal of the causes of poverty and dependence" and "the provision of adequate assistance to persons in need." Almost a quarter of a century later, with more than 3 million Canadians still living in poverty, it is clear that the welfare system has failed to achieve these goals.
This is despite, or because of, the proliferation of different federal and provincial income security programs: besides the CAP, which matches provincial welfare spending, there is the Guaranteed Income Supplement, Old Age Security and the Canada Pension Plan for the elderly, the child tax credit and family allowance for parents, unemployment insurance for the unemployed, earned income supplements for the working poor, and many others.
Years of ad hoc tinkering have produced a tangle of overlapping, uncoordinated programs that serve to perpetuate the very poverty they were intended to erase. Hence a wide range of economists and welfare experts have come to support rolling most of these into a single, unified support system, under the generic heading of a "guaranteed annual income."
The problems with the present system are well known. While it is no longer necessary to show that one is blind or widowed to qualify for assistance, there remain a battery of eligibility categories and needs tests, administered at the discretion of the welfare officials, at the expense of the dignity and often the civil rights of the beneficiary. The complexity of the system, it has been said, redistributes income not so much from rich to poor, as from the ignorant to the well-informed. Many fall through the cracks.
More serious still are the problems associated with the level and structure of benefits. Formerly, practice alternated between two polar opposites, depending on the program: either benefits were available to everyone equally, regardless of income; or they were abruptly withdrawn once the recipient started to earn income above a certain amount. Too much money went to those who don't need it, and not enough went to those who do.
The Conservative government's move to tax back pensions and family allowances from the rich attacked the first. But the second, and worse, problem remains. Those who depend on social assistance too often find themselves trapped: if they earn an extra dollar in income, they lose that dollar -- or sometimes even more than a dollar, when more earned income means disqualification from a program altogether -- in welfare benefits.
The working poor are in effect being taxed at a rate of 100% or more. That many quite sensibly choose to stay on welfare is not because benefits are so generous as to make work unnecessary. It is rather that they are withdrawn so sharply as to make work impossible. This is made worse by the interaction of several programs at once, whose benefit schedules and eligibility requirements will rarely have been attuned with one another. In the jargon of economists, recipients experience a "discontinuity in their budget line." In more popular usage, they are caught in the "poverty trap."
In the mid-to-late 1980s, a raft of important policy documents appeared in support of fundamental reform: the 1984 Quebec White Paper on the Personal Tax and Transfer Systems, the 1985 Macdonald Royal Commission, the 1986 Newfoundland Royal Commission on Employment and Unemployment, the 1987 Forget Commission on Unemployment Insurance, and the 1988 Ontario Social Assistance Review Committee, along with numerous other reports from private and social agencies. Nearly all of them point, directly or indirectly in one direction: the guaranteed annual income.
The idea is to replace most if not all existing existing income support programs with a single, more or less unconditional payment, assuring that no person would fall below a certain floor income. At the same time, support would be gradually withdrawn as earned income rose, in such a way that funds were targeted on those most in need, without presenting the working poor with confiscatory implicit marginal tax rates.
The Macdonald Commission, for example, recommended that the federal Guaranteed Income Supplement, Family Allowance, Child Tax Credit, married and child exemptions, social housing and CAP funds be pureed into a Universal Income Security Program (UISP). This would provide $2,750 to each adult under the age of 65, plus $750 for each child (the first child in a single parent family would get the full adult allotment).
A family of four with no outside income would thus receive $7,000 annually in monthly installments ($9,000, if the personal exemption were also eliminated), plus whatever the provinces provided. This would be taxed back at a relatively gentle rate of 20%, meaning that a family with $35,000 would no longer receive any federal benefits.
Contrast that with the situation at present. The Quebec White Paper estimated that a welfare recipient with spouse and two children who takes a job at minimum wage (then $4.00) could expect to keep only 85 cents an hour, after taxes and changes in benefits -- an implicit marginal tax rate of 79%. This is in a province that, unlike others, has a small income supplement for the working poor. And it does not include the loss of benefits in kind -- subsidized housing, drug plans, etc. -- provided to welfare recipients.
A person in similar circumstances in Toronto in 1983, according to a study for the Macdonald Commission, faced an implicit marginal tax rate that started at zero, shot up to 75%, climbed to 100%, fell back to 28%, bounced up to more than 80%, dropped as suddenly to 30%, and thereafter rose to 40%, all before he had earned $20,000.
The total effect on labor supply of these sorts of disincentives is vast. The Quebec study found that the number of employable people on welfare had increased ten times as quickly as the number of unemployables between 1973 and 1983. Reform of the system, it calculated, would increase employment in the province by as much as 67,000 full-time jobs.
Curiously, most of the early studies of the GAI fretted over the possible work disincentive of giving out an unconditional income grant. Comparing present circumstances with the sort of GAI schemes envisaged, it is clear where the greater disincentive lies: not in giving money to people whether they work or not, but in taking it away from them when they do.
All the same, those two variables -- the basic payment, and the rate at which it is withdrawn -- remain the parameters defining the limits of any GAI scheme. At too high a level of benefits, and too low a tax-back rate, the system will, whatever its effects on the work ethic, bankrupt the treasury. But if one makes the tax-back rate too steep, in an effort to target benefits to the truly needy, then work disincentives become a problem again. Lowering the level of benefits would make a more gradual tax-back rate affordable. But that might leave those who simply can't work in a desperate way.
The GAI, in which a flat sum of money is allocated to each household, then taxed back, is conceptually identical to the negative income tax, in which households are paid a percentage of however much their income falls below a certain break-even threshold (above which the tax turns positive). This can be seen by imagining what the NIT would be for those with zero income. As income rises above that base amount, equivalent to the GAI, the payments would be gradually reduced, just as if they had been taxed back.
In this light, it is clear that "universality" is a side issue. The GAI is technically universal, since it goes to everybody, while the NIT is technically targeted, since it varies according to income, but they amount to the same thing. The GAI/NIT makes social assistance income-tested, by integrating the tax and transfer systems into one, at the same time as it smooths the process of withdrawal, by letting people keep more of their welfare benefits as they begin to work and earn money again.
Although the negative income tax was first suggested by Milton Friedman, the GAI has traditionally been associated with the left. George McGovern was thought a dangerous radical in 1972 for suggesting a modest guaranteed annual income. In recent years, however, it has been embraced by the right. Today the constituency for the GAI ranges across the political spectrum, from the Fraser Institute to the United Church of Canada.
Of late, however, some sections of the left have begun to turn against the idea. Beyond the purely ideological -- sociologist Marjorie Cohen confessed in This Magazine that she had once favoured the GAI, until she saw an editorial endorsing it in the Financial Post -- their concerns range from the effects on market wages of the arrival of workers newly released from the poverty trap, to the inadequacy of the GAI in dealing with the culture of poverty: while the GAI removes the barriers to the transition from welfare to work, the Ontario committee argued, it does not actively encourage it, either.
Others on the left worry that conservative advocates of the GAI have a hidden, budget-balancing agenda. But the GAI can be a fiscal tonic without cutting benefits. The intent of rationalization is not to reduce the level of payout, but to prevent the sorts of perverse work incentives that impede the transition from welfare to work. Cost savings would come not from reducing help to those in need, but from reducing the number in need.
Cutting back the welfare bureaucracy is another potential gain, depending on how far the rationalization is carried out. The difference between the amount actually spent on welfare programs, and the amount it would take to lift every household above the poverty line, is called the "poverty wedge." It measures the amount of leakage to the bureaucracy. In Canada, it is estimated at $x billion, or xx% of all welfare spending.
The benefits of a GAI, in short, are these: simplification, lower costs, securing the safety net, and eliminating the poverty trap. To these might be added a fifth: by its automatic nature, the GAI transforms welfare from a charitable handout or paternalist program to an entitlement of citizenship, a benefit we receive as members of the club. Moreover, with a GAI in place, other social policy reforms become more feasible. The Forget commission on unemployment insurance stressed the importance of using the savings from UI reform to increase income support for the working poor. Likewise, a minimum income guarantee might replace current legislation setting minimum wages.
One reason we have never had a fullblown guaranteed annual income in Canada is the difficulty in getting the various levels of government involved on board. An attempt was made in the mid-1970s, with the federal-provincial Social Security Review, but foundered on the necessity to constrain provincial tinkering with the tax code. But fiscal difficulties, especially those of the federal government, combined with changes in the economic and demographic structure of society, give new urgency to the reform imperative, especially in the context of overall tax reform.
The likelihood of what the economist Tom Courchene calls a "big bang" reform of welfare is nonetheless slim; the provinces and Ottawa are instead proceeding on their own separate tracks. The least one can hope for is that they will try to ensure they are not working at cross-purposes. More optimistically, once Ottawa has reformed its own component of the welfare system, it may encourage the provinces to move along the same lines.
Indeed, the increasing use of tax credits delivered through the federal income tax system already gives the federal government a greater presence in social assistance, nominally a provincial responsibility. The introduction of the refundable child tax credit in 1978, the refundable sales tax credit in 1986, and the conversion of many tax deductions into tax credits in 1987 may be signs of a GAI coalescing out of the mire of the federal tax code.
Ontario, meanwhile, has begun implementing the proposals of the 1988 Social Assistance Review Committee, which envisaged GAI-type reforms, although technically eschewing the GAI itself. Some provinces already have modest income supplements for the working poor in place.
These disparate measures fall well short of a comprehensive GAI. In the end, however, whether a full-blown GAI is adopted is less important than that its underlying principles, especially that of preserving work incentives by maintaining support for the working poor, be respected.