Guaranteed Income

The experiment that could end welfare

By Evelyn Forget

Last December the UK press published a story that was echoed by media around the world. Finland was about to embark on an experiment to determine whether a guaranteed annual income (GAI) could reduce the bureaucracy and expense of the welfare state while enhancing human dignity. The story followed reports that the Dutch city of Utrecht would launch its own Basic Income experiment in 2016. The Swiss will soon test a similar proposal in a referendum. As these high-income countries tentatively explore the feasibility of a GAI, low- and middle-income countries as diverse as Brazil, Mexico, South Africa, Malawi, Kenya and India are launching experiments.

In Canada some premiers and the mayors of Calgary, Edmonton, Oshawa, Kingston, Sarnia and Thunder Bay have called for some form of GAI. During PEI’s last provincial election, all four party leaders advocated a GAI. Naheed Nenshi, Calgary’s visionary mayor, made anti-poverty activists sit up and take notice last May when he wondered whether the Alberta NDP might make “some significant changes to the taxation system that would help us manage poverty in a brand new way… a negative income tax, maybe.” Edmonton Mayor Don Iveson has called for twin GAI experiments in Alberta’s two biggest cities, saying the evidence in favour of a guaranteed income is “overwhelming.” Alberta Finance Minister Joe Ceci himself has written in support of a GAI.

The Canadian Medical Association, the Canadian Public Health Association, the Canadian Association of Social Workers, Food Secure Canada and Community Food Centres Canada have supported the idea of a GAI. In Ontario 194 physicians signed a letter to the Minister of Health and Long-Term Care. The Ontario Public Health Association and the Association of Local Public Health Agencies also registered support. This spring Ontario announced it will conduct a basic-income pilot, and Senator Art Eggleton tabled a motion encouraging the federal government to do the same.

A guaranteed annual income is a single, universal, unconditional cash benefit delivered through the tax system.  Recipients spend the money as they see fit. No employment or job training requirements are imposed. It is designed to replace a big part of the existing welfare state. A GAI operates like a refundable tax credit and is intended to help low-income households. Families with no other source of income would receive a level of support based on family size; as incomes rise benefits would fall, until, at some predetermined income, the GAI would disappear entirely. The reduction in benefits under a GAI would be less than proportional to the increase in income, meaning that people who work would be better off than those who don’t. This is different from most existing income assistance (i.e., welfare) programs, under which people are effectively punished for working. Their welfare benefit is reduced at the rate of a dollar lost for every dollar earned.

Canada already has evidence of the benefits of a GAI. Between 1974 and 1978 Canada conducted an experiment to determine what would happen if everyone were offered an income-tested GAI. MINCOME (“minimum income”) took place at two sites. In Winnipeg a subset of the population was randomized into a treatment group that received the GAI and a control group that made do with existing programs. The town of Dauphin, population 10,000, was chosen as a “saturation site,” in which everyone had access to the same program: If income fell below a certain level, the family would receive top-ups. Benefits were reduced by 50 cents for every dollar a family received from elsewhere. Controls for Dauphin were other, nearby communities.

Making a mother debate through a Plexiglas window with a welfare agent about her kids’ needs is beneath us as a society.

MINCOME was a partnership of the provincial and federal governments. In 1976, however, the provincial NDP government of Ed Schreyer fell and was replaced by a PC government led by Sterling Lyon, who wasn’t keen on his predecessors’ special project. Federally, the minority Trudeau Liberals were increasingly distracted by oil price shocks, persistent inflation, unemployment and high interest rates. Over the course of the experiment, political support waned. MINCOME was unceremoniously shut down in 1979, with no significant analysis completed. Boxes of data changed hands several times, ending up in a federal archives, where they attracted little interest for three decades.

Shortly after the experiment ended, the scientific director of MINCOME, Derek Hum, and economist Wayne Simpson salvaged some of the labour market data from the Winnipeg sample. Their research showed that the GAI didn’t cause adults with full-time jobs to work significantly less. Two groups of workers, however, did reduce their work hours. Married women, at a time when women were first entering the workforce in large numbers, used the GAI to “buy” longer maternity leaves (the law at the time sanctioned only four to six weeks of leave). Adolescent boys—“young, unattached males” in the jargon of the day—also worked less.

Intrigued by these findings, I went looking for the data a few years ago—and ran into difficulty. I followed the traces of MINCOME and eventually tracked the data lode down. It was more than a bit overwhelming to find it still in paper format, packed in boxes from floor to ceiling in a Winnipeg warehouse. Some 1,800 cardboard boxes held the periodic surveys of family members, short interviews on special topics, and intake forms, all mixed with blank surveys, keypunch cards and information booklets. There was no electronic database.

The second difficulty arose from my university’s institutional ethics board, which oversees research to ensure it meets appropriate standards. In the 1970s social scientists did not ask subjects to sign consent forms or waivers for future contact. The ethics board therefore decided I should preserve family privacy by not contacting participants.

The two most obvious ways to get at the experiment’s results were closed to me, but the “young, unattached males” who worked so much less during the experiment were on my mind. If they weren’t working, what might they have been doing instead? I called a friend at Manitoba’s Department of Education and he spent a few days entering enrolment figures from the 1970s into spreadsheets. Bingo. As I suspected, high school enrolment and graduation increased significantly in Dauphin during the experiment. While I’d been forbidden to contact participants, no one prevented me from writing stories for the local newspaper or doing radio interviews. I invited people who wanted to talk—participants, interviewers, the general public—to call.

Conversations with participants in the experiment confirmed what I’d suspected: young men in low-income families were often under pressure to become self-supporting, but MINCOME had allowed some families to support their sons a little longer. A lucky cohort of young men in Dauphin had unexpectedly been able to graduate high school in a time and place where that was unusual. While their older siblings went into agriculture or manufacturing—jobs that paid well in the 1970s but have since dwindled—the MINCOME kids graduated and, in some cases, went on to college or university. These young, unattached males reduced the hours they worked in the short run but had greater opportunities over their lives and could provide better opportunities for their own children.

People were also healthier. Fortunately Manitoba had implemented universal health insurance in 1971 and collected a great deal of data to ensure hospitals were properly funded. An unintended benefit is that the province can now reconstruct individuals’ medical histories. Privacy is ensured, but I was able to access the aggregate outcomes for Dauphin residents and compare them to other Manitobans of the same age and sex living in similar towns. Relative to the control group, MINCOME recipients were associated with an 8.5 per cent reduction in hospitalizations and fewer visits to family doctors, largely due to fewer mental health complaints.

MINCOME took place 40 years ago. Why such enthusiasm now for a dramatic change in the way we deliver social assistance?

Canada is currently worse than many other countries at addressing poverty. A 2013 UNICEF study ranked Canada 15th out of 29 rich countries in terms of our children’s material well-being. One in seven people in Canada lives in poverty, while one in eight struggles to put food on the table. Roughly 20 per cent of single mothers and 15 per cent of single elderly people live in poverty. People with disabilities are twice as likely as others to be poor. Some 13 per cent of children grow up in poverty, but this rises to 40 per cent for indigenous children. Forty per cent of northern households and 62 per cent of children living in the north do not have enough to eat.

Existing welfare benefits vary widely between provinces and even between programs within a province. Income assistance often carries with it the expectation that claimants prove they have searched for work. Eligibility requirements are often so punitive that people must own almost nothing before they can apply. Single mothers still face questions about whether they’re in a relationship, or who the father of a child is, so he can be forced to pay support.

Disability income support (such as Alberta’s Assured Income for the Severely Handicapped) is usually higher than general income assistance, but carries with it the necessity of continually providing paperwork certifying ongoing disability. Completing the forms can be difficult, costly and demoralizing. Most physicians charge patients for the service; medicare doesn’t pay for it. Individuals and their physicians are forced to portray capabilities in the worst possible light to ensure approval.

Moving between systems is a nightmare. For example, income assistance on First Nations reserves is the responsibility of the federal agency Indigenous and Northern Affairs Canada, while the provinces pay off-reserve costs. However, jurisdiction is never clear when people move on and off reserve several times or where complex family relationships mean that some children have First Nations status while their siblings don’t, or where some children live with extended family. All of these issues engender a costly and inefficient bureaucracy and expose recipients to the insecurity of discretionary decisions. People can’t predict with certainty the benefits they’re entitled to, and governments must employ many people to ensure regulations are enforced.

More than a third of Canadians—those over 65 and those under 18—already have access to some form of guaranteed income. Canadians over age 65 are entitled to Old Age Security (OAS) and Guaranteed Income Supplement (GIS), which together have significantly reduced poverty among seniors.

Children under 18 have access to the recently enhanced Universal Child Care Benefit and National Child Benefit Supplement. Child benefits are paid to custodial parents based on the number of their children. These benefits have been shown by Mark Stabile and Lauren Jones of the School of Public Policy and Governance at the University of Toronto to improve child and maternal health, and children’s educational outcomes. Those who worry that GAI recipients might make poor decisions can take heart from research showing that parents spend child benefit money on better nutrition, school supplies and housing.

In contrast to the young and old, Canadians between the ages of 18 and 64 are solely at the mercy of an inconsistent patchwork of government programs. At a time when so many jurisdictions are exploring the feasibility of a Basic Income or GAI, Canadian policymakers should extend such a model to working-age adults here too.

The Canadian experiment with a GAI demonstrated its benefits: longer maternity leaves, better educational outcomes, better nutrition, better physical and mental health, less pressure on other social programs, no significant evidence that primary earners work less, and considerable evidence that the women and children who do work less spend their time in socially beneficial ways. Nevertheless MINCOME took place 40 years ago. Why is there such enthusiasm now for a dramatic change in the way we deliver social assistance?

Those who advocate replacing welfare with a guaranteed income point to four main advantages: a better response to a changing labour market; less bureaucracy; a less expensive, streamlined system; and an opportunity to empower recipients while providing work incentives.

The labour market has undergone dramatic changes in recent decades. Rather than long periods of employment in a single industry or even with a single employer, a growing proportion of the labour force finds itself facing contract employment interspersed with periods of unemployment or underpaid work such as internships. Employers invest less in training, shifting the cost to employees and to society. Guy Standing, a prominent UK economist, has documented the growth of precarious employment and calls for a Basic Income to allow people to build lives while managing unemployment, underemployment, job insecurity and the need to continually retrain.


MINCOME recipient Amy Richardson shows photos of her family in 1970s Dauphin. (Benjamin Shingler)

The most important distinction between a GAI and existing income assistance is the regulatory framework. Hugh Segal, a former Progressive Conservative senator and long-time advocate for GAI, has noted the several hundred different rules that can disqualify or affect the income of a welfare recipient in Canada. If income assistance carries with it the necessity of demonstrating job search activity, then a bureaucracy must exist to ensure that people are searching for jobs.

This has several consequences. An expensive bureaucracy must enforce rules. Excessive regulation also means recipients are at the mercy of frontline discretion. People who design and deliver programs value discretion because it means frontline workers can meet sudden and unexpected needs. Recipients, however, are more likely to tell stories of differential treatment and unanticipated denials. In Segal’s words, “The notion that we would make a mother debate through a Plexiglas window with a welfare agent about how much she needs to raise her child is beneath us as a society.”

Policymakers in Finland and Utrecht both cite simpli-fication as a reason to explore a GAI. A simple system is easier and less expensive to deliver. People with similar needs are more likely to receive similar treatment. As attractive as this is to policymakers, however, it makes some social justice advocates nervous. They point to the hard-fought benefits that people with disabilities currently receive and suggest that a GAI would threaten those.

A GAI isn’t a silver bullet to replace all other benefits. Best thought of as a replacement for welfare—income assistance and the basic living expenses portion of disability supports—a GAI would augment other programs that address specific needs. Many people might still require temporary help with addictions, or affordable housing or childcare.

The most significant benefit of a GAI, however, is its ability to empower individuals. The person who receives the  income decides how to spend it. Individuals can plan ahead and invest in themselves (by upgrading their education, for example) if they know they have a GAI to fall back on. One of my favourite souvenirs of the old MINCOME project is a picture of a well-used pickup truck. One of the Dauphin families ran a market garden and relied on an old truck to take produce to market. When the truck broke down, they were out of business. MINCOME, however, allowed them to purchase a used truck that still sits in the farmyard as a relic of that period of their lives.

Another story was told to me by a librarian who had just retired. She had been a recipient of mother’s allowance—the name for welfare at the time—when MINCOME was introduced. As a single mother she was frustrated because she’d been denied the opportunity to take job training while collecting welfare. When she transferred to MINCOME, that decision was hers alone. She soon had a part-time job that eventually became full time, and when I spoke with her she had just retired and was justifiably proud of having modelled a different kind of life for her two daughters.

A GAI empowers people. For example, they can plan ahead and invest in their education.

There are challenges in introducing a gai, but some difficulties cited by skeptics are imaginary. They note correctly that a GAI subsidizes low wages and conclude that the introduction of a GAI would drive down wages. In fact the opposite outcome is likely. No one has to work to collect a GAI, and therefore potential employees would have the power to refuse low-paid or dangerous work.

Others claim a GAI would worsen the position of women by allowing them to remain outside the workforce. Every society relies on caregiving work to function, much of it historically done by unpaid women. A GAI would allow such people to be paid and to save for their futures.

Still other voices claim a GAI would be inflationary. Every time the rental allowance in existing welfare packages is raised, landlords raise their rents. Some critics imagine a GAI would have a similar effect. But a rental allowance can only be spent on rent, so an increase in the allowance automatically triggers an increase in rent. In contrast, a GAI is spent at the discretion of the beneficiary.

Both the National Child Benefit and the GIS were introduced without causing a general increase in prices. A GAI is redistributive in the truest sense; it takes money from the well-off, who are more likely to save it, and gives it to the poor, who are more likely to spend it. The effects of a GAI are spread across many different commodities. The increase in expenditure would lead to higher employment, wages and consumption, as more goods are produced to meet the new demand.

By decreasing poverty we won’t have to spend so much on social programs. The 2012 report “Poverty Costs: An economic case for a preventative poverty reduction strategy in Alberta,” for example, estimates the annual costs of poverty to Alberta—through higher use of health, justice and social services—at $7.1-billion to $9.5-billion. The data is backed up by frontline service workers such as Gary Bloch of St. Michael’s Inner-City Health Program in Toronto, who spend much of their time treating the consequences of poverty. “We see chronic conditions, including complications, showing up in patients’ 20s and 30s that we would normally expect to see in those aged 50 and above,” says Bloch.

The Public Health Agency of Canada estimates that 20 per cent of all healthcare costs are associated with poverty, and virtually any condition is exacerbated by poverty. At a time when Canada already spends more than $60-billion a year on hospitals, the 1 in 10 Canadians who can’t afford to fill their prescriptions often show up at hospitals as costly ER admissions. Addressing the upfront causes of poor health through a Basic Income would go far toward constraining rising healthcare costs. Many other social programs, such as special education or child services, also address issues exacerbated by poverty.

The most realistic challenges to a GAI are upfront cost and the difficulty of introducing it in a federal system in which provinces have authority over welfare. Is it feasible to introduce a GAI in a federal system in which provinces are responsible for adult benefits? Co-operation between levels of government is the best way to ensure an orderly transition, but the relationship hasn’t always been easy. The federal government can introduce a GAI without provincial co-operation, in the same way that the National Child Benefit was introduced and expanded. Most provinces pay for welfare, at least in part, with money transferred from the federal government. Transfers depend, among other things, on the poverty rates in the particular province. The introduction of a GAI by the federal government would reduce poverty rates and thus reduce subsequent transfers. Over time the GAI would form a greater proportion of adult benefits, while separate provincial programs would shrink.

Provinces could also lead change. Streamlining adult benefit programs, reducing regulations and extending benefits to the working poor can be done at a provincial level.

David Northcott, long-time director of Winnipeg Harvest Food Bank, likes to tell the story of Big Bill. Big Bill was a volunteer at Winnipeg Harvest and also a client. He lived in a downtown hotel and received provincial income support, which paid his rent directly to his landlord. He struggled with many issues, including basic hygiene; he couldn’t always use the shower because it was often broken. He wore all his clothes all the time because he worried about theft.

One day he came to work at Harvest clean, shaved and with no need for food. He’d shopped at the grocery store and cooked for himself in his own kitchen in the apartment he’d just rented. Why the sudden difference? Big Bill had turned 65. With access to OAS and GIS, he was transformed. Less than $16,000 a year changed his life, and almost certainly cost the system no more once his food bank use, rental subsidy, income assistance and incidental use of other social programs are taken into account.

It’s easy to argue for giving a guaranteed annual income to children or aged widows. It’s more challenging to make the case for extending the benefit to working-age people. But the evidence shows that the effects can be as profound.


Evelyn L. Forget is an economist and professor in the department of Community Health Sciences at the University of Manitoba.



Beyond Belief

Dianne Macaulay sits at a picnic table in front of the brand new, $42-million St. Joseph’s Catholic High School at the northeastern edge of Red Deer. Across the road, new tracts of suburban housing and retail centres gradually take over slow-rolling farmland. Built to accommodate 900 students, the school will ...

The Problem with Charity

Christmas is coming, snow is flying and bells are jingling around Salvation Army kettles. Albertans will be personally generous, not only by filling the kettles, but in countless other ways: volunteering at food banks, filling hampers for the needy, donating toys, writing cheques to homeless shelters. Of all provinces, Alberta ...

Overhauling AISH

Little did Ralph Klein expect to ignite a firestorm when he joked at a 2004 election rally that two women had been “yipping” at him about AISH (Assured Income for the Severely Handicapped) payments. “They didn’t look severely handicapped to me,” he said. “I’ll tell you that for sure. Both ...