Saskatchewan Assured Income for Disability:
A Study in Unintended Consequences

Saskatchewan Assured Income for Disability (SAID), a program aimed at people with severe disabilities, has already expanded well past its intended target, and continues to grow. Those drawn to the program face massive disincentives and very limited economic prospects, while program costs continue to rise. Saskatchewan disability policy can be much better than this.

Read the full article below or click the button to download a PDF.


January 23, 2019

Saskatchewan Assured Income for Disability:
A Study in Unintended Consequences

By Rick August


The Saskatchewan Assured Income for Disability (SAID) program was intended as an income support program for people with severe and prolonged disabilities. As we shall see, however, the program already reaches a wider population than this, and it continues to spread as time passes.

The result is a larger, more costly and—for its beneficiaries—a more confining welfare system, but it may also signal a basic change in the dynamics of welfare dependency in Saskatchewan.1 The goal of this paper is to explore some of the characteristics of SAID that have caused it to expand its reach, to identify undesirable consequences of this expansion, and to suggest ways to shift Saskatchewan disability policy towards a more effective and inclusive model.

For the sake of clarity, let me set out my own standards for disability policy. I believe that it is possible for a community to offer security and support where needed, but still let every individual make the most of his or her human potential. This is my model of inclusion of people with disabilities.

Measured against this standard, SAID succeeds in providing slightly more economic comfort than the program it replaced. It fails with respect to opportunity, however, by enticing more people into welfare and keeping them there through massive disincentives to work and participate in the community.

As we shall see, SAID is increasing welfare costs and caseloads in Saskatchewan, in both respects quite dramatically. This alone should be cause for concern.

It may also be transforming welfare more fundamentally, from a protection against economic hardship, to a health-related income replacement program—one with ill-defined boundaries and an ever-expanding scope. If trends in this direction continue, there will be far-reaching consequences for Saskatchewan.

Because of the resources it consumes and the vested interests it has created, SAID will be hard for any future government to reform. The key to positive change, in my view, will be to articulate, and stay true to, a strategic vision of real economic and social inclusion of people with disabilities.


Some would describe welfare as a necessary evil for a developed country like Canada. It is intended to act as the last layer of protection against economic risk for citizens of a market-based society. The existence of welfare, however, inevitably creates a tension between social protection and work incentives.

Work has to remain worthwhile for a society to survive. Even fairly meagre welfare benefits, however, have proven to attract some adults away from work, and over the long term, the existence of a welfare option allows some citizens to plot a workless life course.

There is no easy way to resolve this problem. Most governments have opted for compromise, trying to balance minimal benefit adequacy with token efforts to encourage work, hoping the dependent population will remain small and politically uncontentious. Sometimes, however, this awkward welfare stasis is upset when governments make policy errors, or fail to respond to changes in the environment.

One example of environmental change occurred from the 1970s through the 1990s. Changes in family norms and labour market structure resulted in at least one in twelve Saskatchewan households on welfare.2

Fortunately, this situation triggered a sensible policy response. Changes to benefit structures and stronger in-work supports helped Saskatchewan reduce its welfare dependency from over 8% to less than 4% by the late 2000s.

Dependency is now rising again in Saskatchewan, in my view primarily because of SAID. If the experience of other jurisdictions holds true, this upward trend in dependency will continue unless a policy change intervenes.

Understanding disability

We know there are people who are too disabled to work. The problem is, apart from a small group of profoundly disabled individuals, we can’t accurately predict in advance who they are.

The life impact of a given disabling condition varies greatly based on personal characteristics, treatment processes, motivations and other individual circumstances. How a disability affects an individual is also influenced by the community environment and by public policy.

Most people are supportive of government efforts to improve the lives of people with disabilities. As always, however, the devil is in the details. How government supports people with disabilities is the key question.

Good disability policy requires, in my view, an accurate understanding of the concept of disability, in at least two important respects.

First, disability is not a binary issue. People are not either disabled or non-disabled. Human beings vary greatly in their capacities and potential, regardless of disability. At the same time, virtually every individual has some disabling element in his or her life. Some are minor, some severe, and many are in-between. Some elements are almost intractable, some are fairly easily managed. Disability is a continuum, not a discrete state.

Second, disability and employability are two separate issues. Some individuals with severe disabilities make their way very well in the world, others less so. It depends, again, on individual characteristics and circumstances, on community accommodation, and on the policy environment.

SAID is misaligned with both of these concepts. It treats disability as a binary or discrete state. Applicants are required to prove incapacity at an essentially arbitrary threshold to be considered “disabled”.
Once they do so, they are exempted from work expectations and placed within a benefit structure that carries with it, as we shall see, the most extreme of work disincentives. To qualify for SAID is to be consigned, in effect, to the status of permanent unemployability.

SAID caseloads and costs

The SAID program was initially rolled out to a population of 2,700 individuals living in institutional settings. A decision was announced in 2012 to expand SAID to independent living arrangements, with a new caseload target of 9,700.3

As Figure 1 shows, SAID has already over-run this target by about 50%. The average monthly caseload of the SAID program in 2017-18 was 14,988, with every expectation this will continue to rise.


Is this too many? Evidence strongly suggests it is.

Legacy welfare included a category of cases considered disabled for program purposes. This category held about 13,000 households at the time SAID was established in 2009. An internal study of this group had earlier estimated that only about 60%, or about 7,800 cases, actually had any significant functional impact from their disabling condition.4

Figure 2 shows the relative proportion of SAID cases compared to general welfare. SAID now represents 43% of Saskatchewan welfare cases, on a rising trend.


Figure 3 shows the trend in welfare dependency before and after SAID, and the proportions of dependency contributed by SAID and general welfare. It can be seen that most of the increase in welfare dependency in recent years can be attributed to SAID.


Naturally, rising caseloads and dependency rates are reflected in costs (see Figure 4). The overall welfare bill in 2008-09 was $235 million. By 2017-18 it had risen to $446 million. This is a 90% increase in nominal dollars, and a 64% increase in constant 2008 dollars. The SAID program alone in 2017-18 cost $224 million, equivalent to 95% of the entire welfare budget in 2008-09.5


Why is the SAID caseload increasing?

• Messaging

Welfare is very sensitive to political context. SAID was treated by the government as a showcase initiative under the slogan of making Saskatchewan “the best place in Canada to live for people with disabilities”.6 The message was, in effect, that to be on SAID was a good thing.

This outlook was reflected in remarks of officials and ministers as the SAID caseload grew. The Social Services Annual Report for 2012-13 called rising SAID caseloads an “achievement”, and the following year the Annual Report noted approvingly that “[g]reater than anticipated growth likely reflects the enrolment of new clients who were not previously recipients of any provincial assistance program.”7

In other words, the program was considered a success if it attracted more individuals with disabilities into the welfare system. This view seems not to have changed since 2013. The finance minister, in a recent statement, reiterated her view that higher welfare costs were an achievement of her government.8

• Questionable design choices

A number of design decisions were made in the development of SAID that are, to say the least, unusual for a last-resort program. Regulations lay out an extensive list of resources that can be ignored as income, including trust assets and trust income. This means that individuals that have been endowed, presumably by affluent parents, can continue to receive full public subsidy regardless of personal wealth. Major alternative income sources like the Guaranteed Income Supplement are still ignored for most recipients.

Changes made in 2017 raise a different sort of concern. Presumably in response to caseload pressure, a decision was made to exclude new applicants from SAID with deemed eligibility for any amount of Old Age Security (OAS), the basic federal pension program.9 This appears to be a blanket exclusion of seniors from SAID, regardless of their income or disability status.

• Diagnosis-based eligibility

Disability programs tend to lose focus when they use medical diagnosis as the main driver for eligibility. While diagnosis is necessary for medical management of disabling conditions, it is an unreliable predictor of the effects of disabling conditions on daily life.

The use of medical diagnosis also shifts the power of qualification, in part, away from the program administration and into the hands of medical professionals, who generally view themselves as accountable to the patient rather than to public policy. To put it plainly, physicians can be used to stretch the boundaries of benefit systems.10

This problem is avoidable if the benefit system focuses not on the disabling medical condition, but on its functional impact on daily living. There were admirable efforts in the early design of SAID to base the program on functional impacts rather than diagnosis, but these seem to have been abandoned in favour of a diagnosis-driven approach.11

• Advocacy

Disability advocacy can be a powerful force to expand program eligibility. Advocacy pressure is always aimed at qualifying more people, and thus it tends to expand the boundaries of a binary program like SAID. The result is a system that serves more and more people who are, on average, less and less disabled, at greater and greater cost to the public.

Advocacy pressure also makes governments reluctant to undertake needed reforms of programs like SAID. Because of public sympathy for people with disabilities, any aspect of public policy related to their well-being is politically sensitive. Changes to SAID so far have been greeted with sharp criticism, which politicians are understandably reluctant to resist.

• Demography

There have been fairly significant demographic changes in Saskatchewan in recent years. As Figure 5 shows, in the decade before 2017 Saskatchewan gained more young children and more young working-age adults, but also quite a few more people in the 50-69 age groups.12


Older age groups can be expected to have more health problems. Particularly in a diagnosis-driven program, these can often be reframed as disability. The caseload trends suggest that SAID may have laid itself open to becoming a substitute for employment and pension income among the late middle-aged and elderly.

• Enticing people away from work

SAID base rates are at least 70% higher than general welfare, and various add-ons can make the differential much higher. The relative rewards from welfare and work are always an issue in program design because, in the long run, they affect individual choices. To illustrate, let me express welfare rates in terms of hours of work at minimum wage.

Depending on individual characteristics, a single person’s net income from general welfare would be equivalent to before-tax annual earnings of from $8,000 to $12,000, which would represent between 14 and 20 hours work per week at minimum wage.13 The equivalent gross earnings to match minimum SAID benefit would be about $20,000 per year, or about 35 hours per week.

Not everyone can work, and suitable jobs are not always available for those who can, but when work is possible, public policy should support it. This is not just for reasons of fiscal responsibility. It matters how people get their money. Research validates what we know from our own experience, that people with an economic role and purpose are, in general, happier and healthier.14

SAID reflects a philosophy of increasing economic comfort through welfare. Contrast this approach, if you will, to the admirable roots of the SARCAN social enterprise, which set out to take employment of people with disabilities beyond sheltered workshops and closer to the mainstream economy.15

Looking at current SARCAN operations from the viewpoint of a consumer, it seems to me that SARCAN’s ˋreal jobs’ approach is struggling to compete with the enriched welfare of SAID as a life choice for adults with disabilities.

• Extreme work disincentives

SAID is relatively generous by welfare standards, but most Saskatchewan citizens would not consider an annual net income of $15-20,000 per year as their highest lifetime aspiration. As we shall see below, someone who qualifies for SAID is unlikely ever to do better.

Moving from welfare to paid employment depends on a lot of things, but one critical factor is that work has to make economic sense. Figure 6 shows the results of another type of analysis of the economic effects of employment, again comparing a SAID recipient to someone on general welfare.

This analysis uses a concept called the effective marginal tax rate, which is the percentage loss from gross earnings to taxes and benefit reductions as a person moves up the income scale. The gray bars are the general welfare recipient, the black bars SAID.16


As can be seen, the marginal tax rate challenge is significant enough for the general welfare recipient. There is little or no net return on work until the individual reaches about $16,000 in annual earnings, which is equivalent to about 28 hours per week work at minimum wage.

That is bad enough, as disincentives go. The situation for the SAID recipient, however, is much worse. Effective marginal tax rates are as high as 130% over a long range of earnings, which means that a SAID recipient in those earnings ranges actually sustains a $300 net loss of income by earning $1,000 more.

The earnings threshold to escape this extreme disincentive is reached at the equivalent of about 43 hours per week at minimum wage, or about $12.50 per hour for a full 40-hour work week. The individual on SAID has to jump from no earnings to nearly $30,000 per year, in a single step, in order to benefit from work. This is only likely in very exceptional circumstances.17

SAID’s expansion was predictable

SAID’s caseload growth could have been easily predicted, because it is consistent with past experience in other such programs.18 Figure 7 shows the caseload trend of what appears to be the political model for SAID, Alberta’s Assured Income for the Severely Handicapped (AISH) program.

Ten years into its existence, AISH had about 16,000 cases. In March 2017 it had 57,000. As Figure 8 on the following page shows, this program has changed the vector of dependency in Alberta—AISH is now essentially the go-to welfare option for the province. Saskatchewan appears to be heading in the same direction.


AISH is certainly not the only example of disability caseload inflation. The proportion of the US population covered by the Social Security Disability Benefit, for instance, has risen from 1.3% in 1970 to over 3.2% today. Canada Pension Plan Disability beneficiaries have also risen, from 0.36% of the population in 1981 to 0.92% in 2017.19

SAID-type programs were embraced by European social democracies as they built their post-war welfare states. Holland, for example, found that its disability benefits, eventually costing over 3% of GDP, had become more about hiding unemployment than supporting real economic inclusion. The Organization for Economic Co-operation and Development (OECD) has for many years strongly advised governments against passive disability benefit programs like SAID.20


The future of dependency

This paper argues that the SAID program has created a number of immediate and serious problems for Saskatchewan:

  • It burdens the taxpayer with a large and ever-growing welfare cost;
  • It expands the economically dependent proportion of the population;
  • It acts as a poverty trap for beneficiaries;
  • It deprives the economy and communities of human potential;
  • It absorbs the fiscal resources that could be applied to better disability policy.

These specific outcomes are concerning in themselves, but the trend in SAID points to an even more worrying problem. To illustrate, consider once again the example of Alberta and the AISH program (Figure 8).

Some years ago, Alberta made a considerable show of tightening its general welfare program, with apparent success. But looking at the overall welfare picture in Alberta, we see that Alberta has not really tamed welfare, it has merely shifted more of it behind a screen of administratively-framed disability.

Saskatchewan appears to be headed for the same fundamental shift in the welfare paradigm. We are in transition from a program primarily aimed at relieving economic distress, to one which is driven by medical diagnosis. With an aging population, it becomes more and more possible—and more likely, given the incentives built into the SAID structure—to repackage economic circumstances as disability.

As in Alberta and elsewhere, we can expect the scope of SAID to keep spreading wider, since all the internal and external forces acting on the program push in that direction. Persons assessed as disabled will soon be a majority of welfare cases in Saskatchewan. Paid at preferential rates, effectively exempt from employment expectations and constrained by powerful work disincentives, disability beneficiaries will become a true permanent dependent class.

What should government do?

We can do better than this. I believe an effective disability policy should support both economic security and opportunity. Following are, in outline, some suggestions for principles to be followed in designing a disability supports system that meets these goals:

  • The program goal should not be enhanced welfare, in the manner of SAID, but the displacement of welfare by better alternatives. Whatever the presenting problem, there is almost always a better solution than welfare.
  • To approach this goal, welfare should become, once again, a single unified program that addresses only universal human needs of food, clothing and shelter.
  • Universal needs should be addressed at the same benefit levels for all beneficiaries, without discrimination on the basis of perceived worthiness.
  • Non-universal needs, including disability needs, should be addressed outside of welfare, so that supports neither add to the depth of dependency of people who receive them nor discourage personal effort to participate in work and community.21
  • Disability supports should be based on functional impact of disability, not medical diagnosis.
  • Because disability needs are so specific and variable from person to person, disability assessment and supports should be individualized.
  • From the perspective of economic inclusion, a program that is targeted by income-testing is better than welfare-style needs-testing. The traditional income-testing model is not, however, suitable for disability supports, where cost levels vary widely. A two-dimensional program design is needed.22
  • Government should provide basic income security to people with disabilities, but should focus new disability resources on efforts to support people in the mainstream—real members of communities, real independence, and where possible, real jobs.

The policy I am advocating is a competing vision to SAID. Unfortunately, a better policy option must also compete with SAID for public funds. The transition will be difficult. The best path forward might in fact be what I call strategic incrementalism—that is to say, incremental change, but with a consistent strategic direction driving each measured step.

Difficult yes, but possible.

© R.J. August, 2019


1 In this paper I use the term welfare to describe last-resort income support programs, which currently include the Saskatchewan Assistance Program (SAP), the Transitional Employment Allowance (TEA) and, since 2009, Saskatchewan Assured Income for Disability (SAID). SAP and TEA I treat as a single entity that I call, in their current role, general welfare, and in their pre-SAID role, legacy welfare programs.
Excluded entirely from this analysis are federally-funded welfare programs on reserves, which operate in a distinct economic, social and policy environment. On-reserve welfare dependency, recently estimated to me by federal officials at 41% in Saskatchewan, is a subject well worthy of its own discussion.

2 Saskatchewan dependency by population peaked at just over 8% in 1996. Dependent households represented over 10% of all dwellings in the province.

3 Ministry of Social Services, Annual Report, 2011–12.

4 Functional impact of disability is an important concept in good disability policy. Functional impact assessment is an approach designed to determine what real-world barriers to inclusion are created by each individual’s disabling condition. Functional impact assessment lays the basis for measures to limit or neutralize these barriers.

5 All dollar amounts are from public accounts. Constant dollar calculations are based on the Saskatchewan All-Items Consumer Price Index.

6 Ministry of Social Services. Annual Report, 2013-14, page 7.

7 Ibid.

8 Donna Harpauer. Letter to the Editor: Response to Meili Equalization Op-Ed. Regina Leader Post, July 10, 2018.

9 Saskatchewan Assured Income for Disability Policy Manual, June 2018. Seniors remain eligible for legacy welfare programs.

10 This is the reason many benefit administrations like private insurers and workers´ compensation programs use their own medical assessment personnel.

11 Most references to disability impact were expunged from the SAID regulations in 2016.

12 Statistics Canada. Table 17-10-0005-01. Population estimates on July 1st, by age and sex.

13 At this writing the Saskatchewan minimum wage is $11.06 per hour. The gross earnings equivalents to non-taxable welfare benefits are determined through the author’s integrated tax-benefit model.

14 For examples of research on this topic, see Gordon Waddell & Kim Burton, Is Work Good for Your Health and Well-being? London: The Stationery Office, 2006. This study was commissioned by the UK Department for Work and Pensions.

15 For out-of-province readers, SARCAN is a recycling operation formed by the Saskatchewan Association of Rehabilitation Centres in 1988. Government policy places SARCAN in a central role in Saskatchewan’s rather extensive beverage container recycling system, and government provides it with both direct and indirect subsidies.

16 The effective marginal tax rate is an analytical tool used in designing programs to minimize negative program interactions. The comparison is between a single person on SAID and a general welfare recipient in the “not fully employable” category.

The analysis is a product of the author’s integrated tax-benefit model. Program and tax effects are measured at increments of $1,000 annual gross income. Included are SAP, SAID, Canada Working Income Tax Benefit, Canada GST Credit, Saskatchewan Low-Income Tax Credit, CPP and EI payroll taxes, and Canada and Saskatchewan personal income taxes.

17 It is important to note that earnings exemptions in welfare do not help with this problem—they actually make it worse, because they push the program’s reach farther up the income range. Earnings exemptions encourage minor earnings but discourage substantial earnings. Fixing work disincentives cannot be done within a strictly welfare paradigm. One approach that can work is to make the welfare program as basic and compact as possible, and to pair it with non-welfare earnings supplements.

18 See my paper,
Paved with Good Intentions: The Failure of Passive Disability Policies in Canada.

19 Open Data Canada

20 A good summary of OECD advice to governments on disability policy is contained in Transforming Disability into Ability: Policies to Promote Work and Income Security for Disabled People (OECD, 2003). See also my paper, Paved with Good Intentions, referenced above.

21 The recently shuttered Saskatchewan Rental Housing Supplement, an income-tested benefit with disability components, was an example of a program offering an alternative to welfare housing support. Its phasing out is a significant step backwards for Saskatchewan anti-poverty efforts.

22 The traditional income-testing model sets a one-size-fits-all rate, then reduces benefits as income rises. In a program like child benefits, setting a single rate for basic needs is reasonable, but in disability support, costs can vary to extremes. A two-dimensional test would vary benefits both upward by needs, and downward by income. Minor examples of this approach exist in current programs. In fact, the province’s drug plan in its early days was designed along these lines.


Contact Rick August for more information about this article.

Contact Us

We're not around right now. But you can send us an email and we'll get back to you, asap.

Not readable? Change text.