How an Earnings Supplement Can Affect Union Formation among Low-Income Single Mothers

Welfare payments to single mothers in Canada with children had opposite effects in different provinces on marriage and cohabitation decisions, suggesting social norms overrode economic incentives.


The relationship between economic circumstances and marital decisions is complex. As welfare policies can create incentives for individuals to either marry or remain single, policymakers have been trying to better understand how welfare assistance may impact union formation (marriage or partnership). For example, income eligibility requirements for standard welfare programs may be a disincentive to marry because a spouse’s added income would reduce or eliminate eligibility for welfare support. Welfare policies may also influence family structure through their effects on employment and income. This paper examines the impact of the Canadian Self-Sufficiency Project (SSP) on marriage and cohabitation. The SSP differs from mainstream welfare programs in that the recipient must work a minimum of 30 hours per week to be eligible to receive assistance. Also, the SSP offers more generous support than traditional welfare ($30,000 in New Brunswick, $37,000 in British Columbia, etc.). This study examines the difference in marriage and cohabitation patterns among single mothers who were randomized to either receive the SSP or remain on a traditional welfare program.


Offering an earnings supplement to single mothers in place of welfare altered rates of marriage and cohabitation, but whether the effects were negative or positive varied by province.

  • In New Brunswick, the supplement increased both marriage and cohabitation by 4.3 percentage points. In contrast, in British Columbia the program decreased the rate of marriage by 2.5 percentage points and had no effect on cohabitation.
  • In New Brunswick, increases in marriage occurred across a large variety of subgroups: the youngest age group; those who were living in rural areas, not working at baseline, or with relatively short welfare histories; those with only one child. No group experienced a negative effect of the SSP on rate of marriage.
  • In British Columbia, the SSP decreased marriage for single mothers aged 40-49, those who were living in urban areas, those who were not working at baseline, and those who were non-French speaking or native born. It did not have a statistically significant effect on other population subgroups.
  • The program also affected the timing and length of relationships. In New Brunswick, it encouraged marriage (but not cohabitation) earlier in a relationship and increased the length of time in a partnership, while it delayed couple formation and reduced the duration of relationships in British Columbia.

In short, earnings supplements did impact union formation, but the local context appears to have significantly influenced the positive or negative direction of the relationship. More research needs to be conducted as to why the SSP had an opposite effect; however, the authors hypothesize that men may be attracted to women’s employment and earnings when the local economy is depressed, and that local norms and values may have an influence on marriage rates and single parenthood.


The study uses data from the Canadian Self-Sufficiency Project (SSP), which was implemented between 1992 and 1998 in British Columbia and New Brunswick. The SSP provided an earnings supplement to single parents for up to three years if they left mainstream welfare and worked full-time. Single parents who had been receiving welfare for at least one year were eligible for the study. A total of 4,648 single mothers were randomly assigned to receive or not to receive the earnings supplement. Although the entire supplement-offer group was eligible to receive the earnings supplement, not all participants met the requirement to find full-time employment within one year.

Baseline data was collected through surveys at the time of random assignment, while follow-up data was collected in several rounds over 36 months. Data from administrative records supplemented the survey data.

Related GAP Studies