Two recent books reveal an internal debate about the value of childcare and women's work at the Inter-American Development Bank. Impact evaluations show home visitation programs are cheaper and better for kids than center-based childcare. But a new volume argues the cost-benefit calculation may change once impacts on women, and not just children, are added to the equation. 

The Early Years: Child Well-being and the Role of Public Policy by María Caridad Araujo, Martín Ardanaz, Edna Armendáriz, Jere R. Behrman, Samuel Berlinski, Julian P. Cristia, Yyannu Cruz-Aguayo, Luca Flabbi, Diana Hincapie, Analía Jalmovich, and Sharon Lynn Kagan, Florencia López Bóo, Ana Pérez Expósito, Norbert Schady; edited by Samuel Berlinski & Norbert Schady. Inter-American Development Bank, 2015.

Cashing in on Education: Women, Childcare, and Prosperity in Latin America and the Caribbean by Mercedes Mateo Díaz and Lourdes Rodriguez-Chamussy. Latin American Development Forum. World Bank and Inter-American Development Bank, 2016.

On Wednesday the Inter-American Development Bank (IADB) launched a new book extolling the benefits of subsidized, public childcare to increase female employment in Latin America. I read the book already sympathetic to its conclusions, but as the external discussant at the book event my main role was to express surprise at the Bank's sudden reversal: as of 2015 The Economist was reporting that the Bank opposed big public daycare programs.

Those earlier reports referred to a major study published by the IADB in 2015 on early-childhood development in Latin America, titled The Early Years, which combined a small but exciting body of new experimental research with a broader empirical review of early-childhood programs in the region. A major contribution of that volume, edited by Samuel Berlinski and Norbert Schady with a long list of illustrious coauthors, was to weigh the evidence behind three alterative early-childhood program varieties:

  1. Home visits (i.e., parenting training)
  2. Preschool (part time, with an educational focus)
  3. Daycare (full time, focused on custodial care that frees up parents' time to work)

The book provides a very clear framing of careful empirical research. And if, like me, you're skeptical of the value of training people how to parent better, the results here might surprise you.

Home visits offer bigger learning gains at lower cost compared to daycare

The Early Years reached a fairly unambiguous conclusion (see in particular chapter 6): the cognitive benefits to children from home visits are clearly larger and more robust than from daycare. The gains from some of the home visitation programs are almost staggering, exceeding a full standard deviation increase in two cases. Meanwhile, the few rigorous studies of daycare in Latin America showed small (0.2 sd) gains, or in one case, a negative impact on children's cognitive development.

Home visits vs. Day care programs

The Early Years, CC BY-NC-ND 3.0 IGO

Admittedly, some caveats are needed. Five of those nine gray bars, including some of the biggest effects for home visitation programs, all come from quite small studies in Jamaica, with sample sizes ranging from 140 children all the way down to just 39 kids. Similar to recent debates about the US preschool literature, we might worry about basing too much policy on a small number of small sample studies which may have unique characteristics or be context dependent. Beyond the positive Jamaica home visitation studies though, the Ecuador study by Rosero & Oosterbeek (2011) compares daycare and home visits head-to-head, with big effects for the latter (0.55 sd) and negative effects for the former.

Overall, this evidence base led The Early Years authors to conclude in 2015 that home visits and pre-school are a better investment than day care. Making a few heroic assumptions about the long-term economic value in the labor market of early childhood cognitive gains, The Early Years estimated that home visits have a benefit-cost ratio in three Latin American countries of around 3, preschool even higher at 3 to 5, and daycare at a significantly lower ratio closer to 1.

Home visits vs. Day care programs vs. Preschool

The Early Years, CC BY-NC-ND 3.0 IGO

Given that of those three options, only daycare programs really provide full-time childcare, this pointed to an uncomfortable trade-off (my clarifications in square brackets, based on use of these terms in the book):

Governments that seek to improve child development and facilitate the incorporation of women into the labor market may face a difficult trade-off. On the one hand, in the region, parenting and [part-time] preschools have had the greatest impact on child development; however these programs are unlikely to impact female labor supply significantly. On the other hand, [full-time] daycare programs can make it easier for women to work; however, unless they are of high quality [i.e. expensive], daycare programs will not benefit children.

And while the authors of The Early Years were acutely aware of the gender issues here—some of the volume's authors also wrote some of the most relevant empirical studies on childcare and female labor force participation in Latin America—the benefit-cost calculations and the policy recommendations focused on children's outcomes, acknowledging but largely setting aside implications for mothers' careers.

Counterpoint: daycare isn't just for the kids

The new book launched at the IADB this past week can be read as an oblique dissent from the earlier volume, or at the very least, a significant change of emphasis. (It's also a very useful, data rich reference for anyone interested in gender issues in the region.)

In Cashing in on Education: Women, Childcare, and Prosperity in Latin America and the Caribbean, Mercedes Mateo Diaz and Lourdes Rodriguez-Chamussy explicitly bemoan the fact that earlier work on childcare in the region has focused only on the gains to children, and ignored the benefits to women. Rather than children’s skills, they focus on the role of childcare in expanding female labor force participation—and hence aggregate economic productivity.

Some context is relevant here. Latin America has abysmal female labor force participation rates. Perhaps unsurprisingly in the land of machismo, women often don't work. The numbers Mateo Diaz and Rodriguez-Chamussy present are striking.

Adult men in Latin America are almost universally in the labor force. But across Latin America and all its various sub-regions—Andean, Southern Cone, and Central America—only 50-60 percent of women are working or searching for work. While that number has risen in recent years, it's only barely converging at present.

Labor force participation

Cashing in on Education, CC BY 3.0 IGO

So how much exactly does low female labor force participation cost Latin American economies? There’s no agreed methodology for how to answer this question. Internationally, economists calibrate macro models to account for the misallocation of talent (Hsieh et al 2013), often yielding large implied losses. But these estimates hinge on a long series of complex assumptions. The new IADB volume takes a much more straightforward route, simply calculating foregone income based on the education levels of women out of the labor force.

That simply calculation implies losses that range from 3.5 percent of GDP in Mexico to 16.8 percent of GDP in Honduras.

Productivity loss from women staying out of market

Cashing in on Education, CC BY 3.0 IGO

Childcare works

More to the point, there is reasonably strong evidence that childcare matters for women’s employment. Mateo Diaz and Rodriguez-Chamussy review the modest but growing number of rigorous studies that have looked at the impact of various forms of daycare and preschool on female labor force participation, with specific focus on Latin America.

They find fairly robust, positive impacts of meaningful economic magnitude. At the upper end, effects are quite large: Attanasio and Vera-Hernandez (2004), for instance, find that access to low-cost childcare in Colombia raised the probability of employment from 12 to 37 percent. Large effects on female employment have been documented in Brazil, Mexico, and Ecuador. 

Impact of childcare policies on female labor outcomes

Cashing in on Education, CC BY 3.0 IGO  

Interestingly, these effects appear somewhat larger than the modest or null effects found in some studies in the US or Norway, where baseline female labor force participation rates are much higher. Mateo Diaz and Rodriguez-Chamussy note that effects in Chile also appear ambiguous or non-existent, which may reflect greater baseline female employment there as well.

A not-so-tidy conclusion

Cashing in on Education makes clear that there are often big effects from free daycare on female employment. But are they big enough to justify the costs?

It would make for a tidy resolution to this debate (and to this blog post) to say that The Early Years showed home visits are superior when judged solely on the basis of children's cognitive gains, but Cashing in on Education reverses this conclusion by incorporating effects on women as well. That would be a nice teachable moment about the need to go beyond comparative cost effectiveness (focusing on one outcome across several interventions) to a more comprehensive cost-benefit analysis (recognizing multiple outcomes, and using monetary values to add up the benefits from each).

Reality is not so tidy. It's not obvious that even the benefits in terms of female labor force participation are large enough to offset the cost of custodial daycare programs. Consider the results from Colombia, which is one of the "best buys" reviewed in Cashing in on Education, with big benefits and relatively low costs. Nevertheless, a quick and dirty calculations suggests a 25 percentage point jump in women working (as cited in the book) multiplied by a monthly income for Colombian women of about $700 in PPP terms (per SEDLAC) fails to overcome the reported program cost per child per month of about $200 in PPP terms (0.25*$700/$200 <1). Even if you add that to the cognitive benefits from daycare in Colombia cited in The Early Years, the results still fails to surpass the benefit-cost ratios for home visits or preschool (see the table above). 

There may be important benefits that this quick-and-dirty calculation ignores, and that don’t receive much attention in either book. For instance, the employment benefits in my example are static, while just this month a new paper in the American Economic Review documents the long-lasting wage effects of fertility on women's wages. Building in those kinds of dynamic effects could easily tip the scales in favor of child care in a richer cost-benefit analysis.

More fundamentally though, gender inequality per se matters, not just as an input into aggregate economic output—something which all the authors of both books would seem eager to sign up for, but isn’t clearly reflected in the conceptual framework of either book. Even as it makes the case for subsidized daycare, Cashing in on Education concedes a lot by framing the gains from childcare and even female labor force participation in terms of GDP, with arguably inconclusive results. An alternative view—articulated at last week’s even by the World Bank’s deputy chief economist, Ana Ravenga—is that free, universal access to public childcare promotes both equality and a norm about gender roles that we may value intrinsically.

Perhaps that can be the focus of the third IADB volume on the topic.