Policymakers have struggled for decades with the problem of designing an income support system that provides an adequate safety net while promoting economic self-sufficiency.

Government safety net programs like income assistance (IA) pit one of these objectives against the other; any increase in the generosity of the program directly reduces the incentives to work and leave the program. Several recently tested policies may have shown a way out. By providing extra income supplements only to working parents, these policies have encouraged welfare recipients to work, increased their income, and benefited their children. A drawback of most of these policies, however, is that they cost the government money. This report describes results from a policy that may be different in this regard: the Self-Sufficiency Project (SSP) for welfare applicants.

Conceived and funded by Human Resources Development Canada (HRDC), managed by the Social Research and Demonstration Corporation (SRDC), and evaluated by MDRC and SRDC, SSP offered a temporary earnings supplement in the form of a monthly cash payment to single parents who had been on income assistance for at least one year and who left income assistance for full-time work. The supplement was paid on top of earnings from employment for up to three years, as long as the person continued to work full time and remained off income assistance. While collecting the supplement, the single parent received an immediate payoff from work; for a person working full time at the minimum wage, total income before taxes was about twice her earnings.

To measure the effects of its supplement offer, SSP was designed as a social experiment using a rigorous, random assignment research design. In the SSP “applicant study,” a group of more than 3,000 single parents in Vancouver and the lower mainland of British Columbia who had recently been approved to receive income assistance were selected from the IA rolls. One half of these people were assigned at random to a program group, which was offered the opportunity to receive SSP supplement payments, while the remainder formed a control group. Those assigned to the program group were informed that if they stayed on welfare for a full year, they would become eligible for the SSP earnings supplement. Those who did stay on income assistance for a year were then told that they could receive the SSP earnings supplement if they stopped  receiving income assistance and began working full time (30 or more hours per week) during the second year after random assignment.

An earlier report compared outcomes for the program and control groups through 30 months and found some remarkable results. SSP increased employment, earnings, and income for recent IA applicants, but it did so without costing the government extra cash transfer payments. In other words, the earnings supplement was paid for by reductions in welfare payments and by the higher payroll and income taxes that resulted from the earnings generated by the program’s work incentive. This report extends the results of the earlier report through four years after random assignment and argues that SSP continued to have substantial effects, though the effects declined somewhat over time.

The major findings of this report are as follows:

• By supplementing earnings to make work pay, the Self-Sufficiency Project (SSP) substantially increased employment. During the fourth year after random assignment, the program group worked full time for 5.4 months on average, compared with 4.3 months for the control group, for an impact of more than one month. The SSP supplement offer also increased the average number of months of employment per year from 6.0 months for the control group to 6.7 months for the program group.
• By requiring full-time work, SSP substantially increased earnings. Because SSP required people to work full time to receive the supplement (and because SSP had such a large effect on employment at high-wage jobs), it generated large increases in earnings. During the fourth year after random assignment, program group members earned $11,950, compared with $10,333 for control group members — an increase of $1,617 for the year, or about 16 per cent over the  control group level of earnings.
• SSP led to no increase in net public transfer payments. One of the most remarkable findings of the earlier report on applicants was that the supplement offer paid for itself through higher taxes on the earnings generated by the program. This result continued to hold with longer follow-up. During the last six months of the follow-up period, program group members received $129 per month in supplement payments. At the same time, program group members received $69 less in monthly income assistance (IA) payments than control group members, on average, and they paid an estimated $56 per month more in payroll and income taxes than control group members. The combination of increased tax revenues and reduced IA payments more than offset the cost of SSP supplement payments.
• Most employment resulting from SSP was stable. A number of other studies have found that earnings supplements have encouraged or allowed welfare recipients to find stable employment. SSP’s supplement offer for welfare applicants is no different. The program increased the proportion of people who ever worked full time by about 10 percentage points, but it increased the proportion who stayed at full-time work for more than a year by about nine percentage points.
• SSP reduced poverty by a substantial amount. SSP encouraged people to work using the “carrot” of financial incentives, not the “stick” of reduced welfare benefits. As a result, SSP’s large effect on earnings reduced by more than six percentage points the proportion of families below Statistics Canada’s low income cut-off. However, the program’s effects on poverty were substantially lower than at the 30-month point (when it reduced poverty by 11 percentage points), and a majority of program group members still had income below the low income cut-off.
• The pattern of SSP’s effects for applicants changed over time in an expected way. The program’s impacts on employment, earnings, and income peaked near the beginning of the third year, after all people who were going to initiate supplement payments by finding full-time work had done so. The program’s effects on employment, IA use, and income declined during the third and fourth years, however, as control group members became more likely to go work. This “control group catchup” was expected, since SSP was expected to encourage some people who would have gone to work without the supplement offer to go to work sooner.