The Old Lady Who Swallowed a Fly

My daughters are amused to no end by the classic children’s song about the old lady who swallows a fly and thinks she’ll die. To remedy the situation, she swallows a spider to catch the fly, a bird to catch the spider, a cat to catch the bird, and so on, until eventually the old lady swallows a horse and, well, meets an untimely death.

In economics, the phenomenon described by this song is known as the theory of the second best: if it’s not possible to achieve the optimal conditions required for solving a problem, the second best solution may have unintended consequences. Getting back to the story of the old lady, she died precisely because she swallowed a menagerie in order to fix something as harmless as having eaten a fly. In this post, I want to share some thoughts—a bit more serious than the story of the old lady—about another application of the theory of the second best.

A few weeks ago I attended the Inter-American Conference on Social Security (CISS) in Mexico to participate in a very interesting seminar. Three of the six presentations made that day reflected on the impact that child care services in Brazil and Mexico have made on the labor force participation of mothers. My reflection is about a comment made by Gerardo Esquivel from the Colegio de México (College of Mexico), which I thought about during the almost two hours it took me to return to the airport.

Gerardo commented that he’s yet to hear a good justification for the effort made by many researchers to quantify the impact of child care on the labor force participation of women. More or less, he asked, “Even if these services have no impact on female labor participation, is this a reason to dismiss them? What if they actually have an impact on other variables, such as child development or the wellbeing of mothers, who can study thanks to this service?” I should clarify that Gerardo himself acknowledged that his thoughts were not based on some sort of contempt for the public policy goal of increased female labor participation. Yet, are child care services the right policy tool to achieve this goal?  Or is this a case similar to that of the old lady?

Gerardo’s reflections lead us to wonder to what extent do the lack of child care services explain the lower labor participation rate of poor young women in Latin America and the Caribbean. Chances are that it restricts participation but is not the main cause of the problem. On the contrary, the obstacles that these women must overcome to enter the labor market are numerous and related to their lack of abilities (education, training, job skills) and the characteristics of labor markets in the region (inflexible, a lot of informal employment). But, what do recent studies that assess the impact of child care services on female labor participation tell us about the topic?

1. In Mexico, Estancias Infantiles (government-subsidized day care centers) have a greater impact on the employment outcomes of women who did not work prior to entering the program. However, an impact evaluation of the program prepared by Gustavo Ángeles and others shows that these women represent just 23% of the beneficiaries, as the rest already used other child care services for their kids before entering the Estancias Infantiles program.
2. Using data from household surveys, Mercedes Mateo and Lourdes Rodríguez found that low-cost child care services that participate in the Estancias Infantiles program increase the likelihood that a mother with children ages 0-6 years will seek or accept employment. Furthermore, it appears that those mothers who were already working prior to using the service trade in more unstable care arrangements for the service offered by the Estancias Infantiles. Gabriela Calderón examines the same subject but using different data. In addition to confirming the positive effects of the program on female employment, she found that the program improved the employment quality of these mothers and reduced the time they spend parenting.
3. In Brazil, Ricardo Paes de Barros and others analyze data from Rio de Janeiro, finding that free, public child care services had a significant impact on the employment of mothers (an increase of 36% to 46%), which resulted in a 16% increase in household income. The number of hours worked by women who were already working prior to using this service did not increase.

This evidence is consistent with the hypothesis that the restrictions preventing poor women from participating in the labor market are partly related to child care access, but there are probably other important factors that also affect the decision. With this question settled then, Gerardo’s thoughts on the topic make even more sense. Those of us who attended the CISS conference agreed that much more research is needed on the economics of child care services in Latin America. And I also came back thinking it’s time we address the issue from different angles.