Income support and security architecture


A short history of income security policy fragmentation in Canada

Current Divisions





Financing the Transition


Social policy has a significant impact on the functioning of the food system.  Education, health care, housing, employment and income all affect how individuals, communities and nations interact with food.  All these issues are addressed on this site, but here I focus on how creating a health – promoting and sustainable food system will require progressive changes in Canada’s  income security architecture.  As discussed under Problems, food insecurity, a lack of income is a main source of food insecurity in Canada. Those with no income are at the greatest risk of food insecurity, followed by those on social assistance, those earning a wage or self-employed, and those on pensions and other retirement income (Tarasuk and Mitchell, 2020).  However, there is a relationship between some forms of income, race and degree of protection from food insecurity. For example,  "Black and white households have similar probabilities of food insecurity if they are reliant on social assistance or child support, but the effect of a reliance on seniors’ income sources differs by race. White seniors are more protected. The probability of food insecurity among Black households reliant on seniors’ incomes is the same as that of white working households reliant on wages/salary or self-employment." (Dhunna and Tarasuk, 2020). In subsequent work Dhunna and Tarasuk (2021) concluded that, "Black households had significantly higher predicted probabilities of food insecurity than their white counterparts across all main sources of household income except child benefits and social assistance."

Poverty doesn't just have individual and family costs.  Many of our social, cultural, economic  and health processes are negatively affected.  Government spending on poverty is already enormous and only partially effective.    In 2007,  the money required to bring everyone to the poverty line was estimated at $12.3 billion, at least half of what poverty was already costing us (National Council of Welfare. 2011). Homelessness alone is estimated to cost $7 billion annually in shelter, health care, social services and corrections (Gaetz et al., 2013).  Health care system additional costs associated with poverty of the lowest income quintile were estimated from 2007 data to be $7.6 billion (Ontario Association of Food Banks, 2008). Those in the lowest quintile experience significantly higher disability, mental health problems, and chronic conditions (Lightman et al.,  2008).

The current approach  reflects the fact that we largely fail to focus on poverty prevention and social determinants of health, but rather attempt to mitigate poverty once it exists.  We treat poverty as primarily an economic issue, rather than seeing it as a human rights and right to food issue. A new poverty prevention approach means net costs far below gross costs, given the savings that will result in other spending areas.

A short history of income security policy fragmentation in Canada

The absence of an integrated joined up approach to income security has some parallels with the  food system.  Prior to Confederation, social welfare was either: 1) based on the French model, delivered by the church, church related organizations and committees; 2) influenced by British poor laws, as in NS and NB, whereby townships collected money for the indigent and set up asylums and other related institutions; 3) based on English common law in Upper Canada without poor laws at first, so that municipalities had to apply to the provincial administration for charters to run social welfare institutions (only later did initiatives  reflect more the poor laws); or 4) served by private charity in most rural areas.  There were relatively minor state expenditures on social welfare (Moscovitch and Albert , 1987).

The BNA Act of 1867 gave provinces responsibility for hospitals, prisons, asylums and charitable institutions.  The federal government had jurisdiction over special groups, First Nations, war veterans and federal prisons.  Welfare was hardly mentioned and considered the domain of the provinces, subsequently delegated by them to the municipalities.  The need for health and welfare provisions grew with industrialization, but the Constitution gave the provinces more authority in this arena, unless it could be clearly demonstrated that there was a national need.  Provinces, and especially the municipalities, didn’t have the resources to deal with the growing need (Guest, 2003).

Piecemeal improvements began shortly after Confederation.  Starting in 1874, Ontario began providing operating grants to charities (previously it had helped them only with capital costs). Budgets expanded rapidly in the late 19th century, with urbanisation, growth in manufacturing, trade unions, and investments in capitalist reproduction such as education and health care.   Money was available for Children’s Aid Societies and other charities.  Governments were responding to the social ferment of the period (Moscovitch and Albert , 1987).

Such social pressures continued into the early 20th century.  Cash assistance programs began in 1916, including Mother’s Allowance , a product of advocacy from women’s groups, and needs-tested entitlements started around the same time (see for a timeline, Stapleton, 2017). By 1930, most provinces had such measures, plus old age pensions (for those over 70), workers compensation (so that workers could not sue their employers) and war-related pensions.  But relief remained largely a local issue.  However, the federal and provincial governments were paying more of the local bill through the Bennett years (1930-35).  Bennett’s New Deal in 1935 included Unemployment Insurance (UI).  Ultimately, the courts decided this terrain was provincial (1939) and the measure was rescinded.  However, a constitutional amendment in 1941 allowed Prime Minister King to develop a federal UI program and to also change income tax (Moscovitch and Albert , 1987).

King did bring in changes from 1941, but largely to appease and manage the growing influence of the Co-operative Commonwealth Federation (CCF).  The war years produced many reports, including Marsh (1943), on developing a comprehensive approach to social welfare which had been missing up to that point.  Marsh was the first to clearly link food and social policy, in that he recognized that having a nutritionally adequate diet is central to identifying a suitable minimum income. Wages, he felt, did not reflect basic family needs, including a nourishing diet. He produced a comprehensive income architecture that built upon the existing programs of the day, but unfortunately, only some of Marsh’s recommendations were implemented.  And In the post – war period, among those pieces that were in place,  much was dismantled, including public housing stock sold off through the newly created Canadian Mortgage an Housing Corporation (CMHC).  Government restricted public expenditure on housing in favour of private money.  This went on until the 60s.  Canada didn’t really put in place the legislative and fiscal architecture for a welfare state until the 60s, substantially behind many countries in Europe that started in the 40s and 50s (Moscovitch and Albert, 1987).

At least since the 1980s, the federal government and some provinces have been promising to end poverty, especially child poverty, largely through refinements to the social assistance system.  This has met with limited progress.  In  many ways the situation has worsened because of the removal of federal conditions on the health and social transfers.

Current divisions

Today, Canada’s income security programs are primarily conditional, based on employment, including Employment Insurance (and often training and retraining for the job market), Canada and Quebec Pension Plans, and certain tax breaks including those for RRSPs. A few are consistent with a Guaranteed Annual Income approach, discussed under Substitution. We have a hodge podge now, a mix of public and private, voluntary and mandatory, that leaves many holes.  The federal, provincial, municipal governments and First Nations are all involved.  As the BC Expert Panel on Basic concluded, it is not really a system, but rather a diverse and complex network of programs.  They identified close to 200 BC and federal government programs of relevance, involving at least 20 departments and 35 points of access (Green et al., 2020). That short to medium term illness depends on your workplace plan and that self-employed people rely on private disability insurance indicate some of the current public private split realities. A parallel system exists with payments into pensions.  It is odd that we have workers compensation for injury, but nothing for illness except through employers.  Regulatory and programmatic changes have aggravated the situation, particularly requiring more weeks to qualify for Employment Insurance (EI) (formerly UI), and benefit and eligibility changes to social assistance. Less than 40% of the unemployed in Canada are currently eligible for EI and the associated job training opportunities, down from 47% in 2006 (Curry, 2015).  To receive publicly funded social assistance, you have to exhaust most, if not all, your private financial resources (Stapleton, 2008a,b,c) and then you have no resilience to absorb shocks and make the transition off welfare back into the workplace. Disability benefits are very uncoordinated (Stapleton, 2016).  The consequence is that some 70% of social assistance recipients are still food insecure with 29% falling into the severely food insecure category (Tarasuk et al, 2014). And food banks in most provinces report that at least half their clients are on social assistance.

The federal government role includes tax policy /  income support programs (e.g., Canada Child Benefit, GST Tax Credit, and  Working Income Tax Benefit) and other taxation instruments to generate revenue and equity. It provides pensions (Canada Pension Plan, Old Age Security and Guaranteed Income Supplement).   It transfers money to the provinces via the Canada Social Transfer to support their income support measures (over 80% of all spending on income security programs is federal). With its responsibility for Inuit, First Nations and Metis people, newcomers and person with disabilities, it has some economic development programs.

The provinces also collect taxes, offer economic development programs and provide social assistance.  Municipalities collect property tax and other modest revenue generating taxes and often share responsibility for funding and delivering social assistance programs with the provinces.  A co-ordinated, integrated, comprehensive architecture remains elusive.  What is needed is a joined up income security architecture that fits with a joined up food policy.


A complex and deeply entrenched reality such as poverty and income precarity will not disappear overnight. It requires progressive elimination, a transition approach.  Consistent with joined up food policy, the first  stage transition elements  expand coverage, increase benefits, and remove perverse disincentives.  The substitution stage involves filling some of the holes in the overall structure of the system.   The redesign stage offers an integrated comprehensive architecture that uses current experiences and realities to identify a possible future approach that would more successfully serve the purposes of social welfare.

Financing the transition

In 2013, all three levels of government spent some $177 billion on social assistance (Statistics Canada). And the COVID pandemic has provoked billions more in emergency benefits for both workers and employers (a projected $57 billion for workers in 20-21 according to federal budget 2021). So, clearly there's money available with which to do gradual reallocations.  Different proposals on new approaches have different cost estimates, but the GAI is likely the most expensive, with different designs estimated to cost typically in the $30-50 billion range.  The BC Expert Panel report (Green et al., 2020), despite some questionable GAi design assumptions and either - or scenarios, suggests that a GAI only makes sense in terms of effectiveness and cost if implemented nationally as part of a comprehensive strategy with multiple elements (see Redesign).

The Parliamentary Budget Officer (2018) made national estimates using the negative income tax model of the aborted Ontario pilot with a design to eliminate social assistance payments (Ontario Works and Ontario Disability Support Program).  Payments were set to bring participants to 75% of the low income measure, regardless of employment or educational status. Individuals received up to $16,989, with couples receiving up to $24,027, and those with disabilities having a top up of $500 per month. Payments were reduced 50 cents on each dollar from employment, and dollar for dollar on EI and CPP, but other benefits were not clawed back.  For the national calculations, the PBO assumed that GST credits and Working Income Tax Benefits would also be replaced by the GAI.  Such a program, with some limitations based on the modeling employed, would have a net annual cost of around $44 billion with approximately 7.5 million people receiving benefits.

A key question, as the PBO analysis implies,  is whether the GAI is additive or partially replacing existing social assistance programing. If a federal GAI program could be a primary replacement for basic provincial welfare programs, significant savings would result at the provincial level.  The approach to GAI outlined under Substitution could be set at the basic level of OAS/GIS and could be gradually implemented by age category.  It would have net costs (if eliminating basic provincial social assistance) of about 50% of gross expenditures for a single person (Emery et al., 2013).  Lower provincial tax credits would also increase savings (PBO, 2018). As such it would be a partial replacement for disability programs, but not all of them and would be less costly than the PBO estimates.  Such provincial savings would also then likely result in reductions in the federal CST which would allow for reallocations to other social expenditure areas.

The PBO (2021) supplemented this interpretation with a distributional analysis. They found very modest income declines for middle and upper income earners associated with the loss of certain tax programs, and a 17% increase in income for the lowest income quintile.  Based on the Market Basket Measure, poverty rates were reduced 49%.  Impacts on labour force were small.  The gross costs would be $85 billion to start.

Typically, programs with the fewest conditions are the easiest and cheapest to administer. "In 2013, the total cost of operating the OAS program was 0.3% of the total annual program cost. In contrast, the total administrative cost of Employment Insurance, a program with a high degree of gate-keeping and extensive eligibility criteria, was 8.1% of the total annual program cost. This is significant when one considers consolidation and its cost-saving potential" (Lammam & MacIntyre, 2015). Administrative architecture is fundamentally already in place so transition costs will be low.  The tax system may require some adjustments, including to the basic exemption, with implications for the provinces as well, particularly Quebec which operates on its own much of what the CRA does in other jurisdictions.  Some of the gross expenditure would also be returned in income tax (Kennelly, 2017).

Pasma and Regehr (2019) have proposed 3 models of GAI with a basic income level of $22,000 ($31,113/couple) and that roll in many existing programs (e.g., some tax credits, benefits and supplements) and would largely eliminate poverty, with higher net costs, but also wider benefits. All three rely on federal and provincial government participation. Models one and two are income-tested with reduction rates of 40% and model 3 is a universal demogrant option.  Model 1 focuses on adults 18-64, and retains the Canada Child Benefit (CCB) and OAS/GIS..  Model 2 includes seniors and rolls in OAS/GIS.  Model 3 pays identical benefits to all adults over 18 and retains the CCB.  None of the models include supplements for those with disabilities.  Differential changes based on model are required to the tax system, particularly increases to personal rates for higher income earners and changes to thresholds. Corporate tax rates would be increased to help finance the GAI.  All models are  partially financed by rolling in provincial social assistance which is eliminated, and changes to provincial taxation.  In their analysis, all GAI options can be paid for with the changes outlined.  Model one costs $134 billion, model 2 $187 billion and Model 3 $638 billion.  In addition to largely eliminating poverty, these changes could also have positive regional economic development impacts associated with higher median incomes. This comprehensive analysis shows how GAI can be financed with multiple changes to multiple processes and clearly the transition process would have to be well designed.

There is also evidence that income transfers can reduce deep food insecurity and that the highly food insecure are higher users of health care services, $2322 per year more in health care costs relative to those in  food secure households (Tarasuk et al., 2015; Brown and Tarasuk, 2019), so such interventions can likely reduce health care utilization.  Forget (2011) found 8.5% reductions in hospitalization rates in the Mincome experiment (see also Goal 3, Integrating food as health promotion into health care). Although the Ontario pilot program in four communities was cancelled after only a year or so, self-reporting by participants suggests improved food security, mental  and physical health, and lower health care utilization (Ferdosi et al., 2020). There may also be savings related to increased schooling, and reduced food bank and homeless shelter use (Emery et al., 2013).

So, although the gross costs are significant, based on design specifics the net costs to governments may be substantially lower.  However, the challenge for different levels of government, and different departments within the different levels, will be that the units with increased expenditures will not necessarily be the ones generating savings.  This will require some high level reconsideration of transfers within and across levels.

Although in no way ideal, but to the federal government's credit, the COVID-19 emergency showed how a basic income benefit can quickly be rolled out on a massive scale, and with adjustments as evidence of effectiveness demands. The federal government programme Canada Emergency Response Benefit (CERB) is being replaced with a potentially permanent Canada Recovery Benefits (CRB, 3 programs including Sickness and Caregiving) that targets those not covered by EI.  Although still limited, it has more GAI-like features, including payment floor, wider eligibility than EI and  higher permitted thresholds of income from other sources, including permitting full-time employment.  All this tells us that the transition process does not need to be long and drawn out.