Skip to main content

Brian Lee Crowley

Of the many complex problems that tax the ingenuity of civilized societies, how to make social welfare work is one of the most difficult. The welfare system is always a careful balancing act between keeping people from destitution while not undermining self-respect, destroying the incentive to work, or bankrupting the taxpayer. All welfare approaches are more or less unsatisfactory compromises between these values.

The election of left-leaning governments in both Alberta and Ottawa has once again brought to the fore an allegedly simple and elegant solution: a guaranteed annual income (GAI) or a "negative income tax." A GAI would theoretically allow us to eliminate the many complicated programs that constitute the welfare state, thus allowing us to downsize a big and expensive bureaucracy. No more programs to make sure the less well-off are housed, receive medical care, have a basic income, get child benefits and so forth. No more stigmatizing of the poor and vulnerable.

Instead everybody would be entitled to a basic income, with a clawback ensuring that only those who really needed the money would get to keep it and could spend it as they wished instead of having social workers and bureaucrats second-guessing them.

It's not just in Canada that this idea is being discussed. In Finland the government is looking at such a GAI, replacing all existing welfare benefits with a monthly payment of €800 ($1,191) to every adult. The projected annual cost: €47-billion, roughly equivalent to Finland's total welfare spending.

However simple and elegant these schemes can appear in theory, in practice they are fraught with difficulty. For example, welfare systems and their effects on behaviour are not static, but dynamic. People adjust their behaviour over time in accordance with the incentives of the system. We saw this in Canada before the welfare reforms of the 1990s. In Ontario, for example, with every recession the number of people on welfare went up.

Nothing surprising there. But, discouragingly, in the subsequent economic recoveries the number of welfare recipients never declined. Instead of being a roller-coaster, rising and falling with the economic cycle, welfare became a one-way escalator. By the mid-nineties one Ontarian in eight was getting welfare, at a time when Ontario was still our wealthiest province.

Something similar has been observed repeatedly in analyses of the impact of EI on Atlantic Canada. Whatever the threshold needed to qualify for full benefits in terms of hours or weeks worked, the amount of work accumulated by the region's claimants magically peaks at the exact number needed to be fully "stamped up," in the local vernacular. In other words employers and workers (and sometimes provincial governments through make-work programs) conspire together to push the maximum number of people onto the federally-funded program by ensuring they get enough work to get the full entitlement, and no more.

So people respond to incentives. The incentives inherent in a GAI, which is by definition universal and therefore cannot distinguish between the indolent and the unfortunate, the motivated and the venal, the disabled and the dysfunctional, reward behaviour most societies find undesirable while punishing ones that are to be encouraged.

One way to prevent the GAI from being a trap for those who lack a strong work ethic is to set the benefit level quite low, which is likely also the only way to make it politically palatable for the majority of taxpayers who have to work to fund a benefit available to all with no expectation of any kind of behaviour in return. That has a couple of consequences.

First, the income guaranteed will likely be at bare subsistence level, which will not please advocates for the poor.

Second, since the whole idea is that we will have eliminated the plethora of social welfare programs, we will no longer have the means to direct higher levels of targeted aid to those whose circumstances warrant it, such as the disabled. We can restore such programs, but then there go the cost savings from dismantling the welfare bureaucracy. Moreover, there are lots of programs, such as EI and workers compensation, that play a different role and couldn't be replaced by a GAI.

Nor is it a simple thing to administer a GAI through the tax system. Taxes are only paid annually and on an individual basis, whereas benefits would need to be paid at least monthly and to take account of family or household circumstances to avoid terrible inequities in their distribution. And that's not even raising the issue of the disparities in the cost of living in different regions.

One might almost think that H.L. Mencken had the GAI in mind when he wrote, "For every complex problem there is an answer that is clear, simple, and wrong."

Brian Lee Crowley is the managing director of the Macdonald-Laurier Institute, an independent non-partisan public policy think tank in Ottawa.