Chapter 10B. Costa Rica
- Kalpana Kochhar, Sonali Jain-Chandra, and Monique Newiak
- Published Date:
- February 2017
- Anna Ivanova, Ryo Makioka and Joyce Wong
Costa Rica ranks low on economic participation and opportunities for women, despite the high educational attainment of women. Costa Rica boasts a number-one ranking in the World Economic Forum’s Gender Gap Index on the subcomponent of women’s educational attainment, reflecting a large gender education gap where women outperform men (Figure 10.1). Nonetheless, it ranks 105th out of 142 countries in the same index on the subcomponent of economic participation and opportunity for women. This poor ranking reflects the gender wage gap and low female labor force participation, which is much lower than in other emerging market economies even where the male participation rate is at almost the same level. In particular, the female labor force participation rate is 10 percentage points lower than the average for Brazil, Chile, Colombia, Mexico, and Peru (known as the LA-5; Figure 10.2). The differences are particularly pronounced among professional and technical workers. Stagnating female labor force participation rates in Costa Rica over the past decade are all the more surprising given a pronounced increase elsewhere in Latin America during this period.
Figure 10.1.Educational Enrollment by Gender in Costa Rica
Sources: World Bank, World Development Indicators database; and IMF staff estimates
Note: LA-5 = Brazil, Chile, Colombia, Mexico, and Peru.
Figure 10.2.Emerging Market Labor Force Participation Rates
Sources: World Bank, World Development Indicators database; and IMF staff estimates.
Notes: LA-5 = Brazil, Chile, Colombia, Mexico, and Peru. CAPDR = Central America, Panama, and Dominican Republic.
Higher female labor force participation could help raise productivity and spur growth in Costa Rica, given women’s high levels of education. It could also help mitigate the impact of a shrinking workforce in the face of forthcoming demographic pressures. Indeed, Costa Rica has the highest life expectancy among Latin American and Caribbean countries (79 years versus 75 years for the region as a whole), and as a result, the percentage of Costa Ricans ages 65 and above is expected to double from 6.5 percent in 2010 to 14.1 percent by 2030.
To better understand possible actions that could be taken to raise female labor force participation in Costa Rica, this analysis addresses the following questions: (1) What are the main determinants of female labor force participation rates? (2) Why are these rates relatively low in Costa Rica compared with the LA-5 countries? (3) Is Costa Rica different from other upper-middle-income countries, and if so, why?
One potential explanation for Costa Rica’s relatively low female labor force participation rates is its status as a middle-income country. The literature finds a U-shaped relationship between the level of economic development (for example, GDP per capita) and female labor force participation (Goldin 1994; Figure 10.3). When a country is poor, women work out of necessity, mainly in subsistence agriculture or home-based production. As income rises, activity shifts from agriculture to industry, where jobs are further away from the home, which makes it is more difficult for women to juggle their responsibilities for home production and children with a market job.
Figure 10.3.The Relationship between Economic Growth and Female Labor Force Participation
Sources: Goldin 1994; and IMF staff estimates.
Note: LA-5 = Brazil, Chile, Colombia, Mexico, and Peru.
As education levels rise, fertility rates fall, and social stigma weakens, women shift into the services sector, which appeals more to women’s comparative advantages (Rendall 2010). At the household level, these changes can also be described with a neoclassical labor supply model: as the husband’s wage rises, there is a negative income effect on the supply of women’s labor. Once wages for women start to rise, however, the substitution effect increases the incentives for women to increase their labor supply, and eventually this effect dominates the negative income effect.
Income level does not explain everything, however. There is a large variation in female labor force participation rates even among upper-middle-income countries. For example, countries with similar GDP per capita as Costa Rica have female labor force participation rates ranging from 20 to 80 percent, suggesting the importance of factors other than income. In fact, Costa Rica actually enjoys several conditions found to be associated with higher rates, including a low fertility rate, high educational attainment, and a services-dominated production structure. (Bloom and others 2007; Klasen and Pieters 2015; Gaddis and Klasen 2014).
Costa Rica’s Economy
Costa Rica has a relatively large services sector but low overall investment—factors considered important in determining female labor force participation. In 2012, the services sector accounted for almost 70 percent of total GDP, with manufacturing and agriculture each accounting for about half the remainder. This share of services is relatively high compared with other countries with similar levels of income per capita, such as Malaysia and Thailand where the size of the services sector is close to 50 percent. On the other hand, investment in Costa Rica has been relatively low at about 20 percent of GDP during 2010–14, ranking behind other emerging market economies (Figure 10.4).
Figure 10.4.Total Investment, 2010–14
Sources: IMF, World Economic Outlook database; and IMF staff estimates.
Notes: LA-5 = Brazil, Chile, Colombia, Mexico, and Peru. CAPDR = Central America, Panama, and Dominican Republic.
The number of internet users per 100 people in Costa Rica is at the LA-5 average. Labor market efficiency rated at 4.5 out of 7, a little above the LA-5 average but lagging the world maximum. Interestingly, the fraction of urban residents in Costa Rica is lower than in other LA-5 countries, at 74 percent versus 81 percent, despite its relatively small size, high level of GDP per capita, and high educational attainment.
As in many advanced economies, female labor force participation rates in Costa Rica differ significantly between women who have children and those who do not, particularly for women between the ages of 20 and 40, which is also a prime age for accumulating work experience. There is a similar but larger difference in labor force participation between married and unmarried women, and the differences reach nearly 20 percentage points for women between the ages of 20 and 40. Both of these trends have also been extensively documented for the United States (see, for example, Attanasio, Low, and Sanchez-Marcos 2008).
In an analysis using microdata from a household survey, many of the usual drivers of female labor force participation rates identified in the literature are shown to be operative in Costa Rica: education, marital status, and urbanization are important drivers for female labor force participation (Table 10.2). Higher educational attainment, ownership of cell phones and computers, and living in an urban area are positively and significantly associated with higher female labor force participation. These results indicate the importance of information and physical ability to reach jobs. As shown in the literature, being married has a negative and significant association with female labor force participation, as do the presence in the household of young children and the elderly, although to a lesser degree.
|Dependent Variable||All Women||All Married Women|
|Independent variable||Dummy on labor force participation|
|With children ages||–0.005||–0.097***||–0.104***|
|With children ages||–0.055***||–0.058***||–0.061***|
|With elderly > ages||–0.027*||–0.018||–0.051|
|Log head income||–0.044***|
|Region fixed effects||Yes||Yes||Yes||Yes||Yes||Yes|
Cross-country regression results are mostly consistent with the microdata findings (Table 10.3). In particular, the importance of investment in infrastructure, the presence of children proxied by higher fertility rates, higher education levels, and internet access is confirmed by the cross-country regressions. However, in contrast with the microdata evidence for Costa Rica, the share of urban residents has a negative and marginally significant coefficient. Intuitively, urbanization may have two opposing effects on female labor force participation: although increased access to services jobs helps increases female participation, the need for urban women to commute may impair their availability to work compared with women in rural areas where work is much closer to home. These factors, combined with the importance of the services sector and higher education levels for women in Costa Rica (both of which tend to cluster jobs in urban centers) and the relatively low levels of urbanization in Costa Rica, help explain why urbanization has positive and significant effects on labor force participation for Costa Rican women.
|Dependent variables||Independent variable: Female labor force participation rate|
|Log GDP per capita||–46.918***||–52.519***||–54.902***||–55.646***||–69.790***||–66.596***||–92.947***||–113.19***|
|(Log GDP per capita)2||2.511***||2.757***||2.845***||2.944***||3.638***||3.525***||4.623***||5.963***|
|Fertility rate per 100||–1.881||–1.949||–2.267*||–2.171||–0.831||–0.695||–2.051|
|Share of urban residents||–0.102*||–0.116||–0.105||–0.135||–0.203|
|Share of female secondary education||0.023||0.003||0.086||0.039|
|Share of female tertiary education||0.182***||0.066||0.147**||0.342**|
|Share of male tertiary education||–0.228***||–0.127||–0.091||–0.416**|
|Labor market efficiency||9.012***||9.424***|
|Invest in transportation and télécoms||1.461*|
Differences in investment explain a large portion of Costa Rica’s lower female labor force participation rates as compared with other LA-5 countries (Figure 10.5). Low investment in telecommunications and transportation are the most important factors. Another is total GDP per capita. Finally, the contribution of the residuals to the difference between Costa Rica and the LA-5 is negative and could reflect factors that are not captured by the model, possibly including social stigma about women working and cultural elements.
Figure 10.5.Contributors to Female Labor Force Participation in Costa Rica and Five Other Latin American Countries
Source: IMF staff calculations.
Note: Height of bar indicates the contribution to female labor force participation in Costa Rica minus the contribution in other LA-5 countries (Brazil, Chile, Colombia, Mexico, Peru).
Policies to Close the GAP
Policies to close the female labor force participation gap in Costa Rica vis-à-vis the LA-5 could include higher investment and measures to support working women with children. One obvious choice would be to increase investment, not only in physical infrastructure but also in promoting the development of information technology and telecommunications.
One factor that constrains investment is implementation capacity. Policies to improve implementation could thus serve not only to increase female labor force participation but also to take advantage of the large pool of educated women in the country. Costa Rica has a relatively low fertility rate—1.8 children a woman, compared with 2.5 for Panama—and in 2013, the fertility rate in Costa Rica reached the lowest in its history. These low fertility rates combined with low female labor force participation rates could signal a weak system of childcare, whether public, private, or family-based, although other explanations such as cultural norms are also possible.
2013. “On the Origins of Gender Roles: Women and the Plough.” Quarterly Journal of Economics128 (2): 469–529..
2008. “Explaining Changes in Female Labor Supply in a Life-Cycle Model.” American Economic Review98 (4): 1517–52..
2007. “Fertility, Female Labor Force Participation, and the Demographic Dividend.” NBER Working Paper 13583. National Bureau for Economic Research, Cambridge, Massachusetts..
2014. “Economic Development, Structural Change, and Women’s Labor Force Participation: A Reexamination of the Feminization U Hypothesis.” Journal of Population Economics27: 639–81..
1994. “The U-shaped Female Labor Force Function in Economic Development and Economic History.” NBER Working Paper 2707. National Bureau for Economic Research, Cambridge, Massachusetts..
International Monetary Fund (IMF). 2015. “Costa Rica Selected Issues and Analytical Notes.” IMF Country Report 15/30. Washington.
2012. “Do Labor Market Opportunities Affect Young Women’s Work and Family Decisions? Experimental Evidence from India.” Quarterly Journal of Economics:1–40..
2015. “What Explains the Stagnation of Female Labor Force Participation in Urban India?” The World Bank Economic Review1–30..
Population Reference Bureau. 2014. “Life Expectancy Gains and Public Programs for the Elderly.” Today’s Research on Aging30.
2010. “Brain versus Brawn: The Realization of Women’s Comparative Advantage.” Unpublished..
World Economic Forum. 2014. The Global Competitiveness Report 2014–2015.Cologny: World Economic Forum.
Tico Times. 2014. “Costa Rica Reports the Lowest Fertility Rate in Its History.” March25. http://www.ticotimes.net/2014/03/25/costa-rica-reports-the-lowest-fertility-rate-in-its-history
Evidence from Microdata
We first estimate a model containing many of the drivers of female labor force participation identified in the literature, using a 2012 Costa Rican household survey. The following regression is run using Costa Rica’s household survey (the Encuesta Nacional de Hogares, ENAHO) of 2012:
where prim_second_edui, second_tertiary_ed ui, and more_than_tertiary_ed ui are dummy variables for the woman i’s final educational attainment level, and urba ni, marrie di, cellphon ei, and compute ri are dummy variables for the location of the household in urban area, household being a married couple, and house hold having a cell-phone. kid_0to6., kid_6 to 12i, and old_morethan_70i are equal to one if a household has a member in these categories, respectively. log(headincom e)i is the log of income of a household head. Regional fixed effects are also included.
In order to understand the differences between the main drivers of female labor force participation in Costa Rica and those of other countries, cross-country data is examined next. A panel is constructed for 184 countries from 1990 to 2013, mostly using World Development Indicators complemented by labor market efficiency data from the Global Competitiveness Report. The following regression is estimated following Bloom and others 2007:
where log(GDPCapit at) and log(GDPCapit at)2 control for the countries’ GDP per capita levels, FLFPirt is the female labor force participation rate for country i in region r at year t, fertilityit is the fertility rate, and internetit is the number of internet users per 100 people. sharefemalesecondaryeduit, sharefemaletertiaryeduit, and sharemaletertiaryeduit are the ratios of total female (male) enrollment for secondary and tertiary education levels to the total female population (male population). urbanit is the percentage of urban residents out of the total, labormarketqualityit is an indicator for labor market efficiency, and investmentit is the log of investment in telecommunications and transportation with private participation.1 Dummies include the regional dummy δr, the year dummy γt, and the year-region dummy μrt. Error terms εirt are clustered at the country level.
Cross-country regression results are consistent with many of the findings in the microdata. Results for these regressions are reported in Table 10.2. The importance of investment in infrastructure, the presence of children proxied by higher fertility rates, higher education levels, and internet access are all supported by the cross-country regression results. These factors have also been found to be important in the literature (see, for example, Jensen 2012 and Klasen and Pieters 2015). Investments in transportation and telecommunications have positive and significant coefficients, as do coefficients on internet access. In the latter case, the effect is stronger when labor force participation of women under the age of 25 is considered, suggesting perhaps the importance of technology for the younger cohorts. The share of female tertiary enrollment also has positive and significant coefficients, whereas that of male tertiary educational attainments is negative and statistically significant, in line with the results from microdata on the impact of husband’s earning capacity on female labor force participation. Fertility rates, which serve as a proxy for the effect of children on women’s decision to work, also have negative and marginally significant coefficients.
In addition, cross-country regressions also help shed light on the importance of development levels, urbanization, and labor market efficiency for female labor force participation. First, the polynomial of the log of GDP per capita is statistically significant and generates the well documented U-shaped relationship between female labor force participation and the level of economic development (see, for example, Goldin 1994 and Gaddis and Klasen 2014). The polynomial fit is quite good and Costa Rica is located at the bottom of the U shape. Second, measures of labor market efficiency are positively and significantly related to labor force participation rates. Finally, and in contrast with the microdata evidence for Costa Rica, the share of urban residents has negative and marginally significant coefficient. Intuitively, urbanization has two contradictory effects on female labor force participation. While increased access to services jobs helps female labor force participation, the need to commute may impair women’s availability to work when compared with rural areas where women work much closer to home. These factors, combined with the importance of the services sector and higher education levels of women in Costa Rica (both of which tend to cluster jobs in urban centers), together with relatively low levels of urbanization in Costa Rica, may explain why urbanization has positive and significant effects on female labor force participation in Costa Rica specifically.
A version of this analysis was previously published as IMF 2015.
Investment in telecoms with private participation is the value of telecom projects that have reached financial closure and directly or indirectly serve the public, including operation and management contracts with major capital expenditure, greenfield projects, and divestitures. Investment in transport with private participation is the value of transportation projects that have reached financial closure and directly or indirectly serve the public, including operation and management contracts with major capital expenditure, greenfield projects, and divestitures.