Before implementing a basic income there are questions the government must resolve, such as what would be the effect on other anti-poverty programs?

The drumbeat for a basic income continues. Canada’s business sector has recently promoted the idea, and reaction to the cancellation of the Ontario Basic Income Pilot Project has been almost uniformly negative, with a few exceptions. The idea of a basic income — which guarantees a minimum income to all, regardless of employment status — has always looked good at first glance.

It seems intuitive that a transfer of income from those better off to those less well off would reduce poverty and treat people fairly. Other commonly identified outcomes include a reduction in food insecurity, an increase in health, increased participation in society and greater educational attainment. Indeed, a basic income could offset much of its cost by lowering long-term health and social costs.

But it is important to enter any significant economic policy shift with eyes wide open, especially one projected to have an annual net cost of $43 billion. Devilish details abound and must be seen to, sooner rather than later.

Moral hazard

Moral hazard, that evocative term popularized by Kenneth Arrow, the Nobel laureate in economics, usually refers to how participants behave in health insurance and other insurance markets. A moral hazard can arise when insurance protects people against the consequences of their behaviour, creating an incentive for them to take greater risks. A basic income program is really a society-wide insurance scheme.

To take a personal example of moral hazard, some years ago, I purchased an SUV with, as the sales agent informed me, “state-of-the-art all-wheel-drive.” I would not need winter tires since this car had the surefootedness of a mountain goat — at least that was my reading of the sales brochure. And indeed, so it seemed, until the first episode of snow and freezing rain, when I slid uncontrolled into the ditch. The penny dropped and I slowed my driving. Then I purchased snow tires, which improved my traction, and my confidence soared. I increased my driving speed, negating the advantage of the snow tires, and again hit the ditch. I had responded to my perception of greater safety by taking more risks — twice.

Policy analysts have long recognized the potential for safety nets to change people’s behaviour. Economic theory predicts that when offered cash support, recipients tend to work less — an example of moral hazard. The income maintenance experiments of the 1970s, including the Manitoba Basic Annual Income Experiment, known as Mincome, focused on measuring withdrawal from the labour market. They all found a relatively small reduction in work over the life of the experiments. Women in two-parent households with young children tended to reduce their paid work more than men, an unsurprising finding in the social context of 40 years ago.

It may be easy to dismiss concern about work reduction as a conservative obsession with the value of work and the nobility of toil, save for three points. First, no research has measured the long-term behaviour of workers who receive a basic income. None of the basic income experiments of the 1970s persisted longer than four years, and most recently the Ontario pilot project lasted barely a year. The long-term reduction in work could be smaller, especially if government removes other income-tested programs such as rent subsidies, but we do not have enough data to say either way.

Second, the less people work, the costlier a basic income program becomes, to compensate for lower income from wages. Projections for the cost of a basic income usually do not account for reduction in work by recipients, again because we do not know how big a reduction would occur.

Third, policy research has focused exclusively on recipients, often implying a simplistic dichotomy between the low-income beneficiaries and the taxpayers with supposedly much higher incomes — the 1 percent who will fund the basic income. But not only the uber-rich would be tapped to pay for the program: it would be funded by the taxes of all working Canadians whose income is above the eligibility cut-off. For the Ontario Basic Income, the cut-off was $34,000 in earnings for a single person. It is important to find out whether people working full-time or at several jobs, and earning just above this line, would respond to the introduction of a basic income by working less so that they could qualify. In order to be successful, a basic income program needs broad political acceptance, and it requires working Canadians to keep paying taxes at an increased level.

Complications of implementation

Another wrinkle is that most analysts, proponents and opponents alike, look only at income as the basis for eligibility, forgetting that wealth is also a factor in economic well-being. Many Canadians have significant equity in their homes. How to integrate this large source of wealth, among others, into eligibility determination is a major question in the introduction of a basic income.

One of the selling features of a basic income is the low cost of using the income tax system to both determine eligibility and make payments. Basic income is commonly delivered as a negative income tax: after annual tax returns are filed, those whose income is found to fall below the threshold receive payments instead of paying taxes.

Yet the income tax system does not track wealth well. Mincome did adjust payments to account for self-reported wealth, including home equity. But it would be costly to track personal wealth across the entire population. Doing so would require data sharing between the federal government and municipal property tax systems, to estimate the value of real estate, as well as collecting information on pension plans and investments. If we do not track wealth as part of calculating eligibility for a basic income, then we risk paying it to people living in million-dollar homes. This does not seem right.

Income fluctuations are yet another challenge to implementing a basic income. Taxation operates on a calendar year, and in Canada, all those liable for income tax must file a return by April 30. Imagine someone loses their job on January 1, 2019. The April 30, 2019, return would not reflect this income loss, and if the person remains jobless until January 2, 2020, the renewed income would not be recorded until the April 2021 return. This is an extreme example, but it is far from unusual for low-income households to experience variations in income that are not recorded by CRA until tax filing. Some whose employment or other income declines could wait several months to be accepted into the program. Others whose income rises, but who still choose to receive basic income payments, could experience painful clawbacks the next time they file a return and the government demands repayment.

With a basic income, CRA would need to implement more complete tracking systems to assess eligibility. All of this is possible and commonly done by social assistance systems, but the considerable additional arrangements and costs are not typically noted by proponents of a basic income.

A basic income program would be easy for the CRA to run if there were a list of all the low-income persons who would be eligible. But that list does not exist. One reason is that many households do not file income tax, so CRA does not know about them. Social assistance is a tax-free benefit and low-income persons with casual employment often do not file simply because they owe so little tax. CRA would need to allow “walk-ins” who would petition to receive the basic income, and there would be administrative costs to verify their claims of eligibility.

CRA would probably also need to boost its efforts to detect fraud. Given the likely cost of a basic income program, and the need to assure taxpayers that only truly eligible persons are receiving the benefits, CRA would likely widen the scope of its reviews and audits. Increased oversight of basic income recipients seems inevitable.

A final complication counters the common expectation that a basic income could support the dismantling of the welfare state. With a basic income, would rent subsidy programs persist, or property tax relief, or a host of other anti-poverty programs directed to low-income Canadians? If these measures are not needed, the savings would be substantial and the net cost of a basic income becomes quite low. Milton Friedman certainly viewed this prospect as a major advantage of a negative income tax. But would government be able to take this path?

The political costs would be high. For example, in Ontario recipients of social assistance (Ontario Works and the Ontario Disability Support Program) receive supplementary health benefits, most importantly for prescription drugs. Fierce opposition would arise to any move to eliminate supplementary health benefits for social assistance recipients moving to a basic income. But if governments do follow through on withdrawing anti-poverty programs, the result could be ironic: basic income recipients would be paying full price for goods and services that were subsidized for them before they signed on for a basic income.

Advocates of a basic income tend to gloss over behavioural changes and administrative challenges with this program. Moral hazard may lead both recipients and those whose income is just above the eligibility level to work less, and the resulting drop in tax revenues would increase costs. Wealth is an often-neglected element of economic well-being, but no one has a clear answer for how wealth should factor into the calculation of eligibility for a basic income. And, far from eliminating administrative costs, a basic income may require increased expenditure to verify and monitor eligibility. Finally, and most important, it is unlikely that government will be able to roll back anti-poverty programming.

Government must work these issues out well before implementing a basic income. Adjustment after the fact risks creating unfairness, increased scrutiny and higher taxes. The resulting political disenchantment would undermine the potential of a significant step in the evolution of social policy.

Photo: Shutterstock, by William Potter


Do you have something to say about the article you just read? Be part of the Policy Options discussion, and send in your own submission. Here is a link on how to do it. | Souhaitez-vous réagir à cet article ? Joignez-vous aux débats d’Options politiques et soumettez-nous votre texte en suivant ces directives.