The Impact of Conditional Cash Transfers on Marriage and Divorce
Conditional cash transfers to women, aimed at increasing children’s school enrollment and family health in Mexico, had modest effects on their marriage or divorce rates.
Conditional cash transfers –payments to individuals in exchange for certain actions, such as children’s school attendance and preventive health care visits – are becoming an increasingly popular instrument of development policy and poverty alleviation programs. Such programs often channel resources to the female heads of households in order to ultimately benefit children. By targeting women, conditional cash transfer programs support gender equity and women’s empowerment, but they may also have unintended consequences on marriage and divorce. They may change spousal relationships by improving a family’s financial stability or by fostering women’s economic independence. As growing evidence demonstrates that parental divorce and growing up in a single-parent household may have negative long-term effects on children, including education levels, mental health, and socio-economic status during adulthood. In 1998, the Mexican government began the PROGRESA Program, which targets more than 2.6 million women in poor, rural communities in Mexico with cash transfers that are contingent upon children’s school attendance, family health checks and adult women’s participation in health clinics. This study evaluates the short-run impact of the cash transfers on marriages among low-income households.
The overall impact of PROGRESA, a large-scale conditional cash transfer program in Mexico, on marriage dissolution and formation was modest, although the effects were larger for certain subgroups of the population.
- The overall share of women in marital unions did not change as a result of the PROGRESA Program. In both the treatment and control groups, the village-level proportion of women who were either married or cohabiting remained around 95% during the study’s entire two year follow up period.
- Among families who were intact at baseline, the two-year separation rate was approximately 0.8% for eligible partners in the treatment group and 0.47% for eligible control group couples. Thus, the treatment group experienced a slight increase of 0.32 percentage points (67%) in martial dissolution over the two years of follow up.
- Women of indigenous background experienced the highest rates of marital dissolution, with an increase of 0.62 percentage points as a result of the program, with no changes observed among non-indigenous couples.
- New union formation effects were concentrated among younger women. Women in the program who were separated or divorced between the ages 16-35 experienced an increase in new cohabitation or marital union rates of 3.4 percentage points after 1 year and 3.9 percentage points after 2 years, with no effects observed among older women.
Ultimately, this seems to suggest that the PROGRESA program impacted how marriages are formed and “unformed,” consistent with economic theory. Younger women may have been more likely exit existing and enter new marriages, given that their access to programming may give them more prospects in the “marriage market.”
This study uses data from the PROGRESA Program in Mexico. PROGRESA was launched by the Mexican government in 1997 and is one of the largest conditional cash transfer programs in existence. The program provides cash transfers to the mothers of over 2.6 million children conditional on school attendance, family health checks, and women’s participation in health clinics. PROGRESA was designed to target the rural poor and the amount of money received through the transfers is on average equivalent to about 10 percent of an eligible family’s expenditures. Communities eligible for the program were randomly assigned to two groups. The treatment group started receiving benefits in March-April 1998, while the control group was phased into the program in November-December 1999.
Extensive baseline interviews were conducted in October 1997, and several follow-up survey rounds were carried out in 1998 and 1999. Each survey was a community-wide census that included information on household demographics, income, expenditures and consumption, and the marital status of each person 8 years of age or older. The author of this study restricts the analysis to a sample of eligible households with mothers between the ages of 16 and 55. The sample contains 15,376 households.