Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Economic cycle and deceleration of female labor force participation in Latin America

Economic cycle and deceleration of female labor force participation in Latin America We study the behavior of female labor force participation (LFP) over the business cycle by estimating fixed effects models at the country and population-group level, using data from harmonized national household surveys of 18 Latin American countries in the period 1987–2014. We find that female LFP follows a countercyclical pattern—espe - cially in the case of married, with children and vulnerable women—which suggests the existence of an inverse added-worker effect. We argue that this factor may have contributed to the deceleration in female labor supply in Latin America that took place in the 2000s, a decade of unusual high economic growth. Keywords: Economic cycle, Female labor force participation, Latin America JEL classification: J22, J16, N3 1 Introduction the goals of reducing poverty and income inequality Over more than 50  years female labor force participa- in the region (Parada et  al. 2017). It could also imply a tion (LFP) has been increasing markedly and steadily in stagnation of the global labor supply, given the absence almost all regions of the world. Latin America has not of significant changes in male labor force participation been an exception: about 70 million women entered the (Beccaria et al. 2015). labor market during the second half of the 20th century, The strong and rather unexpected economic expan - reflecting the increasingly important role of women in sion that Latin America experienced in the 2000s, well the region (Chioda 2011). However, since the early 2000s above the long-term trend, has been pointed out as one the growth of female LFP has slowed down significantly of the likely driving factors of the slowdown in female in most Latin American countries and it has even come LFP (Gasparini and Marchionni 2017). Better macroeco- to a halt in some of them. While the growth of female LFP nomic conditions can affect the entry of women into the was on average 0.91 percentage points a year between labor market at least in two different ways, with effects in 1992 and 2002, it slowed down to 0.35 points a year opposite directions. On the one hand, a better economic between 2002 and 2014. This deceleration has occurred context can encourage women to enter the labor market, despite both a large increase in the level of schooling of as they face more favorable labor conditions and higher women and a decline in fertility levels, two factors that wages (substitution effect). On the other hand, a bet - should favor the increase in female labor supply (Gas- ter economic scenario may imply lower unemployment parini and Marchionni 2015; Beccaria et al. 2015). and higher earnings of male partners and the expansion The recent slowdown in the growth rate of women’s of social safety nets, two factors that alleviate the pres- labor supply may affect their prospects for empowerment sure on other family members, especially female spouses, in society (World Bank 2012), postpone the reduction to look for a job, and hence negatively affect their LFP of gender gaps in the labor market, and even undermine (income effect). This latter channel, a version of the typi - cal added-worker effect, could have been more relevant for women in vulnerable households in Latin America, *Correspondence: serranojqn@gmail.com since (i) their labor supply is more elastic to income Departamento de Economía, FCE-UNLP, Oficina 322, Calle 6 No 777, shocks, either coming from earned or unearned income, 1900 La Plata, Argentina Full list of author information is available at the end of the article © The Author(s) 2019. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Serrano et al. J Labour Market Res (2019) 53:13 Page 2 of 21 and (ii) their households were the most benefitted by the two groups: those that depend on individual decisions economic changes in the 2000s. and preferences, and those that are out of the individu- In this paper we use a large dataset of microdata from al’s control. The first group includes the decisions asso - harmonized household surveys of all Latin American ciated with human capital investment (education) and countries to provide evidence on the empirical links family formation (marriage, fertility). The second group between labor force participation and the cycle and trend comprises the returns in the labor market (e.g. gender components of GDP, and use the results to argue about wage gaps), household technologies (availability of home the likely driving factors behind the observed decelera- appliances, electricity and internet connection), health tion in female LFP in the region. In particular, and unlike technologies (contraceptive methods), cultural factors Gasparini and Marchionni (2017) that carry out an analy- (religion, gender discrimination), and public policies sis based on observations at the country level, we estimate (taxes, cash transfers, children and elderly care services, fixed effects models of LFP based on a panel that follows maternity and parental leave). The variables that we are groups of individuals defined according to their level of most interested in are those related to macroeconomic schooling and age in each Latin American country over conditions, which are part of the second group of factors, the period 1987–2014. Working at a population-group such as the trend and cycle components of GDP. level allows us to incorporate a broad set of regressors, A strand of the literature studies the role of macroeco- combining variables linked to the macroeconomic con- nomic conditions by estimating the effect of recessions and text (cyclical and trend components of GDP) with others economic crises on the aggregate rate of female LFP. In par- related to demographic characteristics and public policies. ticular, these works try to assess the relevance of the added- We find that whereas female LFP is positively associated worker effect (AWE). The concept was originally conceived with the trend component of GDP, it has a countercyclical by Woytinsky (1940) and later developed by Ashenfelter behavior: large short-term expansions of GDP, beyond its (1980), Heckman and Macurdy (1980), Lundberg (1985) long-term increasing trend, are associated with a reduc- and Maloney (1987). The AWE refers to the entry into tion in female labor supply. The evidence is consistent the labor market of household secondary workers, usually with an inverse added-worker effect: better economic wives, after a transitory reduction in household income conditions for primary workers cause a delay in secondary (e.g. after the household head becomes unemployed). workers’ entrance into the labor market. This relationship u Th s, the relevance of the AWE with regard to female LFP is stronger for married women (either in formal or con- relies on a typical income effect that arises at the household sensual unions) with young children, who often act as sec- level in the context of a unitary model that assumes that ondary workers in their households, especially those with women are secondary workers and that leisure is a normal low educational attainment, living in rural areas, and from good. In the context of life cycle models with no liquidity low-income families. These results are consistent with the constraints, the AWE is expected to be negligible as long as hypothesis that the exceptionally high economic growth the loss in income due to unemployment is small compared in Latin America in the 2000s is a relevant factor behind to the primary worker’s lifetime earnings (Lundberg 1985). the deceleration in female LFP, which was particularly In addition to the income effect represented by the intense among vulnerable married women. AWE, unfavorable economic perspectives in the labor The rest of the paper is organized as follows. We market may lead to a substitution effect known in the lit - begin in Sect. 2 by briefly reviewing the literature on the erature as the discouraged worker effect, which operates in determinants of female LFP in Latin America. Section  3 the opposite direction. If the latter is outweighed by the describes the data sources that are used in the study. In Sect.  4 we present preliminary evidence on the associa- tion between the economic cycle and the recent decelera- Busso and Romero Fonseca (2015) apply a meta-analysis of the determinants tion of female LFP in the region. Section  5 describes the of women’s labor supply and estimate bivariate models with country fixed empirical strategy and Sect.  6 reports and discusses the effects to assess the role of each factor in explaining the evolution of female results. In Sect. 7 we conclude with some final remarks. LFP in Latin America. Their results suggest that the positive long-term trend is driven primarily by the expansion of health and household technologies and by the gradual change in cultural factors. Furthermore, traditional fac- 2 Background tors such as the increasing female educational attainment and the decreasing Disentangling all the factors that may account for the fertility rates have also contributed to the long-term expansion of female par- ticipation in the labor force and vice versa. Other works reach similar conclu- observed patterns in female labor supply is not an easy sions using aggregate decompositions, such as Peña et al. (2013) for Colombia task, since several potential driving forces are simultane- or Gasparini et al. (2015) for several countries in Latin America. ously at play. Busso and Romero Fonseca (2015) and Chi- In turn, the effect can be boosted if the increase in time away from the labor market of primary workers reduces the opportunity cost of carrying oda (2011) present a broad conceptual framework, which out market activities for secondary female workers, through the substitution divides the possible determinants of female LFP into of tasks within the household. Serrano et al. J Labour Market Res (2019) 53:13 Page 3 of 21 AWE, female labor force participation should exhibit a In contrast, in their study for Latin American coun- countercyclical behavior. tries between 1965 and 1987, Cox Edwards and Roberts In practice, the strength of these effects depends on (1994) show that the AWE is significant for low-income the relevance of other factors, such as the availability countries, although not for richer countries, such as of alternative strategies to cope with negative income Argentina or Chile. In a more recent work, Bhalotra and shocks (child labor, unemployment insurance, etc.) and Umaña-Aponte (2010) use a panel data set for 63 devel- the existence of imperfect credit markets together with oping countries for the period 1986–2006, and find that liquidity constraints (Mankart and Oikonomou 2017; the relationship between female employment and growth Garcia-Perez and Rendon 2016). Consequently, empirical is negative on average for Latin America and Asia, but studies for developed countries, such as the U.S. or the positive for Africa. We improve the empirical strategy United Kingdom, find small added-worker effects (Cul - of these earlier studies by using more disaggregated and len and Gruber 2000; Stephens 2002; Prieto-Rodriguez better-quality data (more Latin American countries, with and Rodriguez-Gutierrez 2003; Bredtmann et  al. 2017), longer and more comparable time series), by including or even no evidence of its existence (Layard et  al. 1980; more controls, and by focusing directly on female labor Maloney 1991). Instead, in developing economies such force participation instead of the employment rate. as those of Latin America, the AWE may be larger due In this paper we deepen the analysis of the main to the lack of unemployment insurance benefits, the fact hypothesis in Gasparini and Marchionni (2017): the that many households face financial restrictions, and the deceleration of the growth rate of female labor supply role of women as secondary workers, reinforced by solid could be related to the strong economic expansion expe- family structures with a strong attachment to traditional rienced by the Latin American countries in the 2000s. gender roles, and low levels of women’s skills and edu- The better economic scenario, which resulted in an cational attainment within some population groups. In improved labor situation especially for the less skilled addition, given the importance of the informal sector in male workers, could have encouraged an inverse added- the region, entry and exit barriers in the labor market are worker effect on their female partners. In other words, relatively low, which facilitates changes in female partici- women, especially the most vulnerable, may have decided pation (Basu et al. 2000; Maloney 2004). to postpone their entry into the labor market due to a Several empirical studies analyze the validity of the AWE lower pressure to search for a job. This paper overcomes hypothesis in the region. Martinoty (2015) uses the col- some methodological limitations and at the same time lapse of the Argentina’s convertibility regime as a natural deviates in some directions from the analysis in Gasparini experiment to evaluate the effect of changes in husbands’ and Marchionni (2017), besides extending the period of labor situation in the labor participation decision of their analysis. Their estimations are based on cross-country wives, finding evidence of a statistically significant AWE. panel data and include as unique regressors the cyclical Similar results are found by Cerrutti (2000) and Paz (2009) and trend components of GDP. Instead, in this paper we also for Argentina in the 1990s and 2000s, respectively, estimate multivariate models combining variables asso- Fernandes and Felicio (2005) for Brazil, and Parker and ciated with the macroeconomic context (the abovemen- Skoufias (2004) for Mexico. On the other hand, McKen - tioned trend and cycle components of GDP) with others zie (2004) and MacKenzie (2003) find no evidence on the related to demographic characteristics and public poli- presence of AWE when studying household strategies to cies. Among the regressors, we include the coverage of compensate the negative shocks from the financial crises conditional cash transfer programs, proxied by the pro- of 2002 in Argentina and 1995 in Mexico, respectively. portion of beneficiaries in the population. In addition, The literature that explores the dynamics of female we build a panel dataset disaggregating the adult popula- labor supply with aggregate level data provides more tion into groups defined by education and age for each mixed results, with some papers that even report a procy- country, which significantly increases the cross-section clical behavior for developed countries (Tachibanaki and variability. In this way, our models allow for a better iden- Sakurai 1991; Darby et al. 2001; Lee and Parasnis 2014). tification of the partial correlations between labor force participation and each of its covariates, as we control for fixed effects by country and by population group, among other variables. Regarding this issue, there is a variety of recent literature, both theoretical and empirical, which tries to reconcile the differences between estimates of elasticities of female labor supply with respect to wages, based on micro data 3 Data sources or aggregate data. For instance, Attanasio et  al. (2015) estimate a life-cycle Our analysis is mostly based on microdata from house- model to explain female labor supply in the United States, trying to bridge the discrepancies between micro and macro estimates. Among other results, they hold surveys, which are part of the Socioeconomic Data- find that the aggregate elasticities of labor supply vary throughout the eco - base for Latin America and the Caribbean (SEDLAC), nomic cycle, being stronger during recessions. Serrano et al. J Labour Market Res (2019) 53:13 Page 4 of 21 a project jointly developed by CEDLAS at Universidad restrict the sample to adults between 25 and 54 years old. Nacional de La Plata and the World Bank. Household Labor behavior of younger individuals is more related to surveys are not homogeneous across countries and in education decisions while the labor supply of older peo- some cases not even for the same country over time. ple mostly depends on the relevance and dynamics of the Given such heterogeneity, careful survey processing is pension system. necessary to ensure as much comparability as possible of estimates among countries and years. This is precisely 4 The economic cycle and female labor force one of the advantages of using the SEDLAC database, participation in Latin America where microdata are harmonized using similar defini - The deceleration since the early 2000s, after a strong tions of variables for each country/year and a consistent increase during the previous decades, has been a major and documented protocol (see SEDLAC 2014). In this change in the dynamics of female labor supply in Latin paper, we use SEDLAC microdata of 18 countries (the 17 America. Figure  1 shows that the deceleration is robust countries in continental Latin America and the Domini- to the group of countries chosen. While the growth of can Republic). Table  8 in the Appendix describes the female LFP was on average 0.91 percentage points a year corresponding surveys and the years they cover. between 1992 and 2002, it slowed down to 0.35 points a Our econometric estimations are based both on a non- year between 2002 and 2014. Even though the contrast balanced panel dataset of those 18 countries over the between the growth rates in both periods is not similar period 1987–2014, and on a more disaggregated panel in all Latin American countries, it is significant in most dataset that follows 9 groups of women defined accord - of them and sufficiently widespread to be visible in the ing to their age and education for each one of the 18 regional average. In some cases there are even signs countries. However, to compute the descriptive statistics of stagnation (Fig.  5 in the Appendix). Unlike women’s, that we show in Sect.  4, we use a smaller sample of only men’s labor supply is much higher and more stable over 15 countries for the shorter 1992–2014 period. With this time. Thus, the recent deceleration in the growth rate of restricted sample we build a balanced panel using lin- female LFP delays the closing of the gender gap in labor ear interpolations and extrapolations, for which we take participation. information from the most proximate surveys. Table 9 in Gasparini and Marchionni (2015, 2017) show that the the Appendix presents a schematic summary of the com- deceleration was stronger for women that are in more position of both the non-balanced and balanced panels. vulnerable conditions, especially those with low edu- All the statistics at the country or population-group level cation, living in rural areas, and who are married and are computed using the corresponding sample weights. with children. Vulnerable women usually have a weaker However, in order to describe the situation for the whole attachment to the labor market, and thus they are more region we use simple averages across countries instead prone to enter and exit the labor market depending on of population-weighted averages, to avoid that highly the economic situation inside and outside their house- populated countries, such as Brazil and Mexico, drive the holds. The fact that the deceleration in female LFP since results. The demographic, social and labor variables are the early 2000s was especially intense among vulnerable obtained from the SEDLAC microdata. The rest of the women suggests that changes in the macroeconomic variables, such as per capita GDP, some institutional and context could have played an important role. political variables, or the coverage of social programs are taken from alternative sources (e.g. the World Develop- ment Indicators from the World Bank, or CEPALSTAT). Table 10 lists the variables used throughout the study, their Female labor supply strongly and persistently expanded since the 1960s in definitions and the corresponding sources. Latin America (Chioda, 2011). An additional clarification before moving on to the 7 The slowdown is also evident when grouping the countries by sub-region next section: following the literature on labor supply, our (South and Central America) or by their initial levels of female LFP. analysis focuses on prime-age people. In our case, we Argentina, Bolivia, Brazil, Chile, Costa Rica, Ecuador, Honduras, México, Panama, Paraguay and Venezuela experience a deceleration in the growth rate in female LFP since the decade of the 2000s. We focus on labor force participation since most of the action seems to have taken place in that margin. Gasparini and Marchionni (2015) find that Most of the household surveys included in the sample are representative “Changes in hours of work for female workers were not large, not very dif- at the national level. The exceptions are Uruguay before 2006 and Argentina, ferent between decades, and not significantly different from those of males. where surveys cover only the urban population, which, however, represents Likewise, changes in unemployment seem to have been small and with no more than 85% of the total population. significant gender differences. These patterns reinforce the claim that the This group excludes Dominican Republic, Colombia and Guatemala, for dynamics of labor force participation are among the most noticeable labor which there are no comparable national household surveys previous to year phenomena with a clear gender dimension over the last decades.” Chioda 2000. (2011) and World Bank (2012), among others, share this conclusion. Serrano et al. J Labour Market Res (2019) 53:13 Page 5 of 21 GDP per capita Female LFP (left) 1994 1998 2002 2006 2010 2014 Fig. 1 Female labor force participation in Latin America. Source: own Fig. 2 Female labor force participation and economic growth. calculations based on microdata from national household surveys. Source: own calculations based on microdata from national Women aged 25–54. Unweighted means of Latin American countries. household surveys, per capita GDP (in PPP-adjusted constant US$) The series of 2 countries includes Argentina and Brazil. The series of from WDI. Women aged 25–54. Unweighted means of 15 Latin 9 countries adds to the previous Bolivia, Chile, Costa Rica, Honduras, American countries Mexico, Uruguay, and Venezuela. The series of 15 countries adds to the previous Ecuador, El Salvador, Nicaragua, Panama, Paraguay, and Peru. The series of 18 countries includes the previous 15 countries labor conditions for the primary workers improve during plus Colombia, Guatemala, and the Dominican Republic a period of strong economic expansion, female secondary workers could feel discouraged from participating in the labor market. The incentives for women do not necessar - In fact, the macroeconomic conditions in Latin Amer- ily imply an exit from the labor market, but a postpone- ica were significantly different in the 2000s compared to ment of the decision to enter the market, for instance, to the previous decade. While the average GDP grew at an allocate more time caring for their children or the elderly annual rate of 1.9% between 1990 and 1999, it grew 3.5% at home. Also, the better economic conditions inside and per year on average between 2000 and 2014, despite the outside the household can relieve the pressure on women effects of the 2008 global financial crisis. In addition to acting as secondary workers to take any kind of job, the higher growth rates, the 2000s were characterized by allowing them to wait until they find a job that fits their more macroeconomic stability than previous decades. preferences. Figure  2 shows that the strong increase in per capita Therefore, if the AWE outweighs the substitution GDP in the region over the 2000s occurred in coinci- effect, female LFP should exhibit a countercyclical behav - dence with the deceleration of female LFP. As discussed ior. In this sense, it is important to notice that the AWE in Sect.  2, the effect of economic growth on women’s should be much more relevant for women from more labor supply could be either positive or negative depend- vulnerable households. In fact, as discussed above, poor ing on whether it is the substitution or the income effect women, with low educational attainment and with young that prevails. On the one hand, an improved macro- children are more likely to act as secondary workers, as economic context can encourage female LFP through a they have a weaker attachment to the labor market and substitution effect, but on the other hand it can depress their labor decisions are more sensitive to the economic female labor supply because the better economic situa- situation in their households (Michalopoulos et al. 1992; tion in the household alleviates the pressure on women Kimmel 1998; Eissa and Hoynes 2004; Naz 2004; Tamm to look for a job outside the house, allowing them to 2009). It is precisely these women who have benefitted postpone their entry into the labor market. most from the economic expansion in the 2000s, through As above mentioned, the latter argument is a vari- improvements in the employment rate and earnings of ant of the hypothesis of the added-worker effect (AWE), men in their families. generally used to explain the increase in female LFP in Figure  3 illustrates this point by showing the evolu- response to unemployment shocks and the drop in fam- tion of some labor variables for high-skilled relative ily income during economic recessions. Conversely, as to low-skilled prime-age men. The hourly wage ratio between men with high and low education has substan- tially shrunk since the early 2000s. This fact suggests that Figure  6 in the Appendix shows the GDP per capita series for each one of in addition to the generalized increase in real wages, the the countries under analysis. 45 50 55 60 65 70 8000 9000 10000 11000 12000 13000 Serrano et al. J Labour Market Res (2019) 53:13 Page 6 of 21 Fig. 3 Unemployment and hourly wage ratios between men with high and low education. Men aged 25–54. Source: own calculations based on microdata from national household surveys. Note: the ratios are defined as high education/low education, where low education = less than secondary complete, high education = tertiary complete. Unweighted means of 15 Latin American countries economic improvement was greater in poor than in non- poor households. Also, there has been a pronounced decline in the unemployment rate of unskilled men, which fell from 6.1% in 2002 to 3.7% in 2014, in contrast with a more stable behavior of the unemployment rate for the skilled men, which fluctuated around 4%. Fig - ure  3 then highlights the existence of a potentially rele- vant added-worker effect especially for more vulnerable women who, given the positive assortative mating, are the ones likely married to low- education workers. Fig. 4 Evolution of public transfers in Latin America. Source: a own An additional factor that could explain the deceleration calculations based on non-contributory social protection programs database, Social Development Division, ECLAC; b own calculations in female LFP is the income effect due to an increase in based on microdata from national household surveys. a shows public transfers to families, especially from conditional unweighted means of 17 Latin American countries: Argentina, Bolivia, cash transfer programs (CCTs), which have strongly Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, expanded in Latin America since the early 2000s both El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama, in terms of number of beneficiaries and amounts trans - Paraguay, Peru, and Uruguay. b Women aged 25–54, unweighted means, monetary and non-monetary public transfers. Unweighted ferred. Figure  4a shows that the coverage of CCTs in the means of 15 Latin American countries region, measured as the percentage of the total popula- tion that are beneficiaries of these programs, grew from 2.7 to 18% between 2002 and 2010, and remained fairly CCTs in Mexico, Brazil and Colombia cover around 50% stable over the following years. CCTs consist of transfers of the poor, while coverage in Uruguay reaches 80%. to poor families with children. The transfers are usually The potential effects of CCTs on female LFP are ambig - monetary and conditional on households investing in uous. On the one hand, there may be a negative effect children’s human capital (education, health, and nutri- operating through three different channels. For women tion). Currently, almost all the countries in Latin America who had the urgency to get a job due to the difficult eco - have some kind of conditional cash transfer program, nomic situation in their homes, CCTs can provide the reaching a large fraction of the poor population. For instance, according to Stampini and Tornarolli (2012) CCTs programs are not the only policy tool for poverty alleviation. For instance, non-contributory pensions have strongly expanded in the region during the 2000s, adding another source of non-labor income for the more vulnerable households. Serrano et al. J Labour Market Res (2019) 53:13 Page 7 of 21 economic relief that allows them to delay their entrance on an unbalanced panel of 18 Latin American countries into the labor market. Moreover, since women are typi- over the period 1987–2014. cally the recipients of the cash transfer, they may perceive Y = α + βln(GDPpc) + η + μ ct c ct (1) ct the subsidy as earned income in exchange for their efforts to ensure compliance with the conditionalities associated Y = α + β cycle + β trend + η + μ ct 1 ct 2 ct c ct (2) with the program, which reduces their available time to where the left-hand-side variable Y is the LFP rate for ct engage in market activities while encouraging the tra- women aged 25 to 54 for country c at year t . In model ditional division of gender roles within the household (1) we include in the right-hand side the logarithm of the (Garganta et  al. 2017). Finally, the beneficiaries of CCTs real per capita GDP, while in model (2) we use its cyclical may believe that in order to continue to be eligible for the and trend components, estimated through the Hodrick- program, they should work less to remain poor. All these Prescott filter. The country fixed effects η capture both three channels would involve a negative effect of CCTs observable and unobservable factors that vary across on female LFP. On the other hand, CCTs may have posi- countries but that are fixed over time, avoiding potential tive effects on female LFP. If conditionalities require that sources of omitted-variable bias. children go to school, they may induce women to allocate As for other covariates of LFP, it is difficult to identify more time to market activities as they do not need to use statistically significant relationships given the small num - it for childcare anymore. This could also imply that chil - ber of countries and the limited variability that many social dren have less time to work, which could reduce house- and economic indicators exhibit over time. To gain more hold income and increase the demand for earned income sample variability we build a new panel with data at the inside the family (Busso and Romero Fonseca 2015). population-group level, where the groups are the result Figure 4b shows the evolution of labor participation of of combining three levels of educational attainment (low, women according to whether their households receive medium and high) and three age ranges (25–34, 35–44, or not public transfers, which include CCTs and other and 45–54). This latter panel then follows these nine popu - non-contributive pensions, either monetary or in kind, lation groups of women in each of the 18 countries over and exclude retirement pensions. While labor force par- the period 1987–2014. Based on the population-group ticipation rates are similar for the two groups of women level panel, we propose the specification in model (3): over the 1990s, since the early 2000s a gap develops due to a decline in labor supply of women from beneficiary Y = α + β cycle + β trend + β um gct 1 ct 2 ct 3 gct families. This fact is consistent with the hypothesis of the (3) + β cct + ϕX + θQ + η + ν + μ 4 ct ct gct c g gct negative income effect discussed above. According to the meta-analysis carried out in Busso and Romero Fonseca The left-hand-side variable Y is the LFP rate for women gct (2015), CCTs generally have a negative effect on female in group g for country c at year t . As in models (1) and LFP and hours worked, even though several studies for (2), the right-hand side includes the cyclical and trend countries in the region find the effect to be not statisti - components of per capita GDP for each country and year. cally significant. To assess the relevance of the added-worker effect we We have argued that there are several factors that could include the male unemployment rate of group g in coun- explain the deceleration in female LFP in Latin America, try c of year t um as an additional explanatory vari- gct and it seems that one of the most important is related to able. If the hypothesis of the inverse added-worker effect the economic cycle, even though the evidence presented is valid, we expect that as we add this factor, the absolute so far is merely descriptive. The relationships described value of the estimated coefficient of the cyclical compo - could be economically irrelevant or not statistically sig- nent of per capita GDP decreases. Also, in order to evalu- nificant. Therefore, in the following sections we deepen ate the relationship between the conditional cash transfer into the analysis of the role of the economic cycle on programs and the changes in female LFP, we include the women’s labor participation decisions to assess the extent coverage of such programs for each country and year to which macroeconomic conditions can account for the ( cct ). X are other controls that vary across countries ct ct recent slowdown in the rate of growth of women’s labor and through time, such as the share of the value added supply in the region. of the service sector in the GDP and the share of rural 5 Empirical strategy In order to study the dynamics of women’s LFP over the Our basic specification does not control for year fixed effects, but we per - economic cycle, we use a set of panel data models that form some robustness checks where we do, finding that the main results hold. See the discussion at the end of Sect.  6.2 and Table  6. Since GDP varies by allows us to control for some potential sources of bias. country and by year, it is not possible to control for year fixed effects sepa - We start with the simple models in Eqs. (1) and (2), based rately for each country. Serrano et al. J Labour Market Res (2019) 53:13 Page 8 of 21 6.1 Analysis at the country level population, and Q are covariates that also vary across gct We start with the simple models at the country level in the groups g , such as education, fertility, marital status, Eqs.  (1) and (2), as in Gasparini and Marchionni (2017). age of children, average wage of women, and gender wage Table 1 shows the OLS estimation results of country fixed gap. Tables 10 and 11 in the Appendix provide definitions effects models of labor force participation based on an and descriptive statistics of the variables. η and ν are c g unbalanced panel of 18 Latin American countries in the fixed effects by country and by group, respectively. 1987–2014 period. The dependent variables are female We estimate models (1), (2) and (3) for women, men, LFP (columns 1 and 2), male LFP (columns 3 and 4), and and the gender ratio (men/women), in order to explore the gender ratio in LFP (men to women, in percentage, differences in the effect of the economic cycle by gender. in columns 5 and 6). As explanatory variables we include The estimation results are presented and discussed in the logarithm of real per capita GDP, and alternatively its Sect. 6. cyclical and trend components. Although in model (3) we control for a quite large set The results in Table  1 suggest that changes in GDP are of potentially relevant factors, including the unobserved positively related to female LFP: a 10 percent increase in heterogeneity that varies across countries or across per capita GDP is associated with an increase in female groups but is fixed over time, it may still be insufficient LFP of 2.26 percentage points (pp.) on average. When and other potential sources of bias may persist. For decomposing per capita GDP, we find a statistically sig - instance, measurement error in the cyclical component nificant relationship both with the cyclical and the trend of GDP per capita could lead to attenuation bias of our components, but with opposite signs. Whereas the trend estimator of β . However, as we will see in the next sec- component is associated with an increase in female LFP tion, all our results indicate a strong and statistically sig- (2.56  pp.), the short-term movements are countercycli- nificant countercyclical behavior of female LFP. In any cal: a 10 percent increase in the cyclical component of case, we carried out robustness checks to other alter- the GDP is associated with a reduction in female LFP of native measures of the cyclical and trend components approximately 2.17  pp. This latter result is consistent and the results hold. Also, the reciprocal relationship with the hypothesis that the unusual strong economic between the business cycle and female LFP may bias our growth experienced by many Latin American coun- OLS estimates towards zero. This would be the case if, for tries during the 2000s contributed to the deceleration in instance, female LFP positively affects economic growth women’s labor supply. In fact, a back-to-the-envelope- (Tsani et  al. 2013; Elborgh-Woytek et  al. 2013). We calculation with the results from Table  1 suggests that address the concerns regarding this potential source of the cyclical component of the GDP growth of the Latin endogeneity by applying an instrumental variables strat- American economies over the 2000s may account for egy to identify the causal effect of the cyclical component 43% of the deceleration in female LFP. of GDP on female LFP. We propose to use exports prices In contrast with the case of women, changes in per as instruments for the cyclical component. The valid - capita GDP or its cyclical and trend components are not ity of this instrument is discussed in the next section. For associated with male labor supply, which shows very lit- the estimation, we use the two-stage least squares (2SLS) tle variation over the period under study. As a conse- procedure, considering the cyclical and trend compo- quence, the labor force participation gap between men nents as endogenous variables. and women is negatively correlated with the GDP trend, but positively correlated with the cyclical component. 6 Results In sum, the estimation results from models in Eqs.  (1) In this section we present the main results of our anal- and (2) indicate that female LFP increases with economic ysis. We start by discussing the models at the country level, then expand the analysis to a panel at the country- group level and we end by providing instrumental vari- ables estimates to reinforce the credibility of our results. 15 Our results are consistent with those of Bhalotra and Umaña-Aponte (2010), who find a countercyclical pattern of female LFP in Latin America and Asia. Using a similar regression framework, we find that the employment For the Hodrick-Prescott filter we use a smoothing parameter of 100 rate is also positively related to the GDP per capita growth for women but (Hodrick and Prescott 1997). We checked the sensitivity of our results to not for men. The effect of the trend component is positive and particularly changes in the smoothing parameter of the Hodrick-Prescott filter and to the strong for women, while the cycle is especially strong for males. In addition, use of three other filters: the Baxter-King band pass filter (Baxter and King unemployment is negatively related to GDP, and the effects of both cycli- 1999), the band pass filter (Christiano and Fitzgerald 2003) and a Butterworth cal and trend components are negative and statistically significant for both filter (Gómez 2001). In all cases, our results hold, i.e., we find a countercycli - genders. In turn, wages increase as GDP expands for both men and women. cal and statistically significant behavior of female labor participation. Results An increase in the trend component of per capita GDP is associated with available upon request. reductions in the gender wage gap, while short-term expansions widen it. We are thankful to one referee for suggesting the use of this instrument. These results are available upon request. Serrano et al. J Labour Market Res (2019) 53:13 Page 9 of 21 Table 1 Models of labor force participation Women Men Ratio men/women (1) (2) (3) (4) (5) (6) Log GDP per capita 22.6*** − 1.1 − 73.0*** (2.79) (0.87) (14.41) Cyclical component − 21.7*** 2.3 61.9*** (6.05) (1.58) (16.25) Trend component 25.6*** − 1.3 − 82.2*** (2.86) (0.93) (14.83) Observations 304 304 304 304 304 304 Countries 18 18 18 18 18 18 R-squared 0.56 0.66 0.04 0.06 0.49 0.57 Latin American countries, panel 1987–2014. Adults aged 25–54 Fixed effects (by country) OLS regressions. Unbalanced panel of 18 countries. Dependent variable: in columns 1 and 2 (3 and 4) is female (male) labor force participation as percentage of women (men) aged 25–54; in columns 5 and 6 is the LFP ratio (men/women) expressed in percentage. Log per capita GDP is the logarithm of real gross domestic product per capita. Cyclical and trend components of GDP are obtained by applying the Hodrick-Prescott filter to the log of per capita GDP. For detailed data definitions and sources, see Table 10 in the Appendix. Robust standard errors clustered by country in parentheses. ***Significant at 1% level, **5%, *10% development, reducing the gender gap in labor supply; families or for women living alone. Finally, the results in however, short-term economic expansions are associated columns 9 to 13 suggest that women with young children with a decline in the entry of women into the labor force, are more prone to react to improved economic condi- and thus with a widening of the gender gap. tions than those with older children or no children at all, We already suggested that an inverse added-worker a result again consistent with the inverse-AWE story. effect (AWE) could be the mechanism behind these As discussed in Sect. 4, another implication of the AWE results. Although we cannot provide causal evidence argument is that we should observe a stronger reaction with the data at hand, we can explore whether some pre- to the cyclical movements in economic activity for the dictions of this hypothesis are consistent with the evi- group of more socioeconomic vulnerable women. To this dence. In what follows we explore two implications of aim we estimate the female LFP models by education, the inverse AWE argument. First, since the mechanism household size, per capita family income, marital status, operates through adjustments within the household, the area of residence, and family types. We also build a vul- results should differ across women from different house - nerability index based on the principal components of hold arrangements. In particular, we expect the sensi- these variables. According to this index, vulnerable indi- tivity of LFP to cyclical fluctuations of GDP to be higher viduals are those with a very low educational attainment, for women living with her spouse and with young chil- living in rural areas, with many children, and with low dren, since, as argued above, they are more likely to act incomes. Table  3 shows the estimation results of model as secondary workers within the household. The results (2) by educational group, and for the vulnerable and non- in Table 2 are consistent with this expectation. The coef - vulnerable population, defined as quintiles 1 and 5 of the ficient for the cyclical component of GDP is higher (in vulnerability index, respectively. absolute value) for women living with a spouse (“mar- According to our results, the negative relationship ried”); in fact, the coefficient is negative but statistically between the cyclical component of per capita GDP and non-significant for the group of single women. Accord - female LFP is stronger for less educated, more vulner- ingly, the ratio in LFP between married and single women able women. Whereas the coefficient associated with falls (i.e. the gap in LFP between these groups becomes larger) as the cyclical component of the GDP increases (column 3). We also divide women into “head of household” and other members. We When classifying women by their family structure (col- use two alternative definitions for “head of household”: (i) status self-reported umns 4 to 8), the negative effect of the cyclical component in the survey, or (ii) family member with the highest earnings (economic defi - nition). In both cases the coefficient of the cyclical component is much higher of the GDP is significant only for women in two-parent (in absolute value) for group of women who are not heads of household. In families. The effect vanishes for women in single-parent fact, in the economic definition the coefficient is non-significant for female household heads. The results for other population groups are consistent with those pre- We are thankful to one referee for making this observation. sented in Table 3 and are available upon request. Serrano et al. J Labour Market Res (2019) 53:13 Page 10 of 21 Table 2 Models of female labor force participation by household structure Married or single Family structure Married (M) Single (S) Ratio M/S Two parents (T) Single‑parent (S) Single‑person (P) Ratio T/S Ratio T/P (1) (2) (3) (4) (5) (6) (7) (8) Cyclical component − 24.8*** − 8.8 − 23.3** − 24.0*** − 9.9 − 4.9 − 22.6*** − 23.6** (7.34) (8.01) (8.04) (6.63) (5.94) (7.70) (5.19) (9.07) Trend component 27.0*** 11.1*** 26.1*** 26.4*** 13.8*** 16.2*** 23.2*** 19.3*** (3.2) (3.15) (4.79) (2.80) (2.09) (3.86) (3.24) (2.89) Observations 268 268 268 304 304 304 304 304 R-squared 0.657 0.260 0.399 0.639 0.453 0.224 0.535 0.289 Youngest child is 0–5 (G1) 6–17 (G2) No children (NC) Ratio G1/G2 Ratio G1/NC (9) (10) (11) (12) (13) Cyclical component − 29.1*** − 24.0*** − 17.5** − 15.3* − 21.8* (8.07) (6.57) (7.62) (8.55) (12.16) Trend component 26.1*** 27.4*** 28.2*** 6.1 5.5 (3.19) (2.51) (3.26) (3.93) (3.78) Observations 303 303 303 303 303 R-squared 0.606 0.634 0.636 0.080 0.059 Latin American countries, panel 1987–2014. Women aged 25–54 Fixed effects (by country) OLS regressions. Unbalanced panel of 18 countries. Dependent variable: female labor force participation as percentage of women aged 25–54; and ratios of female LFP ratio between groups expressed in percentage. Cyclical and trend components of GDP are obtained by applying the Hodrick- Prescott filter to the log of real per capita GDP. For detailed data definitions and sources, see Table 10 in the Appendix. Robust standard errors clustered by country in parentheses. ***Significant at 1% level, **5%, *10% trend component, particularly for the more vulnerable/ the cyclical component is − 23.6 for the group of women less educated groups. with less than complete secondary schooling, it halves (− 11.3) for women with a degree from a tertiary institu- 6.2 Analysis at the country‑group level tion. In fact, that coefficient is negative and significant in In order to gain more sample variability, we estimate a regression for the ratio of LFP between low and high- model (3) based on a more disaggregated panel dataset educated women, suggesting a widening gap between where the units of observation are population groups these groups over the expansive phase of the business defined in terms of schooling and age for each country cycle. Columns 6 to 8 report the results for the vulner- and year. This also allows us to include other poten - ability groups defined above. Whereas the coefficient tial drivers of female LFP, besides the cyclical and trend associated with the cyclical component is negative (− 9.0) components of GDP, such as the male unemployment but not statistically significant for non-vulnerable women rate and the coverage of CCT programs. We also control (top quintile of the vulnerability index), the coefficient for other potential covariates of female LFP: factors that for the most vulnerable group is much larger in absolute are jointly determined with the labor supply decisions value (− 26.0) and highly significant. (education, marriage, fertility), and factors that are likely Unlike the results in Table  3, male labor supply does exogenous to the individuals (gender income gap, cost of not seem to move in line with changes in per capita GDP: childcare and eldercare services, women’s income, share the coefficients of the LFP regressions are much smaller 21,22 of the services sector in GDP, rural population). for men and non-statistically significant in most cases. Accordingly, the labor force participation ratio between men and women is positively related to the cyclical com- It should be noted that including women’s wages and a proxy of the impor- ponent of per capita GDP and negatively related to the tance of the services sector contributes to controlling for factors related to the labor demand, which in turn are correlated with the business cycle. There - fore, these controls would likely capture the fact that recessions or short-term economic expansions can generate compositional changes in the productive structure, thus affecting female LFP. The correlations of each of these additional controls with female LFP The results for men and for the ratio men/women (not reported) are avail - have the expected signs according to the empirical literature (see Busso and able upon request. Romero Fonseca 2015). Serrano et al. J Labour Market Res (2019) 53:13 Page 11 of 21 Table 3 Models of female labor force participation by educational attainment and vulnerability Education Vulnerability Low (L) Medium (M) High (H) Ratio L/M Ratio L/H Vulnerable (V) Non‑ vulnerable (N) Ratio V/N (1) (2) (3) (4) (5) (6) (7) (8) Cyclical component − 23.6*** − 16.9** − 11.3*** − 15.5*** − 19.9** − 26.0*** − 9.0 − 22.4* (7.12) (7.33) (3.30) (5.16) (7.29) (7.97) (6.04) (10.67) Trend component 22.1*** 10.3*** 10.1*** 20.8*** 18.9*** 29.1*** 13.5*** 29.8*** (2.63) (3.51) (2.94) (4.43) (2.73) (3.32) (3.14) (6.40) Observations 304 304 304 304 304 304 304 304 R-squared 0.557 0.219 0.260 0.351 0.397 0.520 0.235 0.256 Latin American countries, panel 1987–2014. Women aged 25–54 Fixed effects (by country) OLS regressions. Unbalanced panel of 18 countries. Dependent variable is female labor force participation as percentage of women aged 25–54. Columns show estimations of the models dividing individuals by different levels of educational attainment and quintiles of a vulnerability index; and ratios of female LFP ratio between groups expressed in percentage. Low education = less than secondary complete; medium education = secondary complete or tertiary incomplete; high education = tertiary complete. Vulnerable = individuals who are in quintile 1 of a vulnerability index based on the principal components of level of educational attainment, marital status and number of children in the household. Cyclical and trend components of GDP are obtained by applying the Hodrick- Prescott filter to the log of real per capita GDP. For detailed data definitions and sources, see Table 10 in the Appendix. Robust standard errors clustered by country in parentheses. ***Significant at 1% level, **5%, *10% Table  4 presents the estimation results of alternative Table  5 shows the results of estimating model (3) specifications of model (3) with fixed effects by coun - with the labor force participation ratio between men try and by group. As a general result, we find that the and women as the dependent variable. Cyclical expan- sign and statistical significance of the coefficients associ - sions are associated with a widening of the gender gap: a ated with the cyclical and trend components of GDP are 10-percent increase in per capita GDP is associated with robust across specifications. The absolute value decreases an increase in the gender gap that ranges between 3.2 pp. as we incorporate additional regressors into the model. and 5.8 pp. across specifications. CCTs coverage is statis - For instance, according to the basic model with controls tically significant only in specifications that include the (column 2) a 10-percent increase in the cyclical compo- full set of controls. A 10-pp. increase in CCTs coverage nent of per capita GDP is associated with a fall in female is related to a widening of the gender ratio in labor supply LFP of 2.17 pp., whereas the estimated fall is reduced to of about 1.8 pp. 1.72  pp. when we include male unemployment, CCTs Table 6 reports a range of specifications of models and coverage and the full set of controls (column 8). different estimation methods as robustness checks. For Our estimates suggest that men’s unemployment rate instance, we use pooled OLS, a first-difference estima - may explain part of the countercyclical behavior of female tor, the Arellano and Bond (1991) GMM estimator for a LFP. Indeed, when the model includes that variable, the dynamic model, and a random-effect estimator. We also coefficient associated with the cyclical component falls estimate model (3) adding year fixed effects to the coun - significantly. The partial correlation between male unem - try and group fixed effects. We add controls for educa - ployment and female LFP is always positive and statisti- tional group trends, and additional variables that refer cally significant in the regressions with controls. When to cultural or legal factors, although in these cases the men’s unemployment rate increases 1 pp., women’s labor number of observations falls. We also restrict the sam- supply rises around 0.21 pp. This result is consistent with ple to countries with available data for more than 10 and the added-worker effect and could explain part of the 20 years. In all cases, the main results hold: the counter- negative association between female LFP and the busi- cyclical behavior of female LFP seems a robust result. ness cycle. On the other hand, the coefficients for CCTs Of course, we are not controlling for all possible factors coverage are negative in the models including controls affecting female LFP and then our results could be biased (columns 4 and 8) although small in magnitude and never due to omitted variables. One of these variables could be 24 25 statistically significant. the expansion of pre-school and child-care coverage. Unfortunately we were not able to build such a variable for all countries spanning over the period analyzed in the The tables with the coefficients for all the variables in the regressions are available upon request. In part this result may be due to little variability in the data: we have data on CCTs coverage by country-year and not by group. We are thankful to one referee for this observation. Serrano et al. J Labour Market Res (2019) 53:13 Page 12 of 21 Table 4 Models of female labor force participation (1) (2) (3) (4) (5) (6) (7) (8) Cyclical component − 17.2*** − 21.7*** − 13.1*** − 20.3*** − 13.5** − 18.2*** − 10.1* − 17.2** (5.24) (5.08) (4.24) (5.28) (5.44) (5.99) (4.79) (6.39) Trend component 15.5*** 7.3** 15.8*** 10.0*** 15.3*** 8.6** 15.4*** 11.2*** (2.79) (3.25) (4.17) (2.67) (2.68) (3.27) (3.97) (2.64) CCTs coverage 1.3 − 2.0 1.2 − 1.6 (4.97) (3.22) (4.70) (3.30) Male unemployment 18.4 20.9** 15.5 21.0** (11.03) (9.12) (11.65) (9.72) Additional controls No Yes No Yes No Yes No Yes Observations 2736 2537 2511 2321 2669 2476 2445 2261 R-squared 0.837 0.854 0.831 0.851 0.845 0.858 0.839 0.856 Latin American countries, panel of education and age groups, 1987–2014. Women aged 25–54 Fixed effects (by country and by group) OLS regressions. Unbalanced panel of 9 groups in 18 countries. Dependent variable: female labor force participation as percentage of women aged 25–54. Cyclical and trend components of GDP are obtained by applying the Hodrick-Prescott filter to the log of real per capita GDP. CCTs program coverage as share of population who are beneficiaries (not available for Venezuela). Unemployment rate for men in each education and age group. Additional controls: average years of education, average number of children, share of married women, share of women in charge of old person, average age of children in household, female hourly wage, hourly wage ratio (men/women), service sector value added as share of GDP, rural population as share of total population. For detailed data definitions and sources, see Table 10 in the Appendix. Robust standard errors clustered by country in parentheses. ***Significant at 1% level, **5%, *10% paper. However, we do not expect the inclusion of that 2000s seems to have been largely attributed to improved variable to be a serious threat to our main results. For exports prices for the region. Since Latin American econ- countries for which data is available, the rate of growth omies are small, international prices can be considered in pre-school coverage was either constant or increasing exogenous and then largely unaffected by domestic vari - in the 2000s, and then it can hardly account for a decel- ables such as LFP. eration in female labor force participation. In any case, Table  7 shows the results using the exports price our results also hold for women without children and index for each Latin American country computed by with children older than 6, who were not affected by an the United Nations Conference on Trade and Develop- increase in pre-school and child-care coverage. ment (UNCTAD) as instrumental variable for GDP. In By all means, all the models estimated so far may all specifications, independently of whether we consider potentially suffer from endogeneity bias. Although we are the cycle or the trend as endogenous variables, the esti- aware that we cannot rule out all endogeneity concerns mated coefficient for the cyclical component of per capita with the data at hand, in the next section we carry out an GDP continues to be negative and statistically significant, additional robustness exercise to reinforce the credibility although larger than the fixed effects estimates from sub - of our results. section  6.2. This suggests that, if the instruments were valid, our previous results suffer from attenuation bias, 6.3 Instrumental variables so they could be interpreted as a lower bound of the true With the aim of alleviating the concern for endogene- effect of the economic cycle on female LFP. ity we use instrumental variables and estimate model (3) through two-stage least squares (2SLS). In order to 7 Concluding remarks instrument the cyclical and trend components of the In this paper we explore the relationship between female GDP we follow Besley and Pearson (2008), Brückner and labor force participation and the trend and cycle compo- Ciccone (2010) and McGuirk and Burke (2016), among nents of GDP based on fixed-effect models using panel others, and use exports prices. There is a large literature that discusses the close relationship between exports prices and the cyclical component of economic growth in Latin America (e.g. Erten and Ocampo 2013; Ocampo Prices are constructed using UNCTADstat Commodity Price Statistics, 2017). In particular, the large upswing in GDP over the international and national sources and UNCTAD secretariat estimates. Alternatively, we also used terms of trade and the international prices of the main commodities exported by each country as instruments for GDP In a sample of 6 countries with data from the early 1990s, the share of chil- (UNCTAD data). When using these instruments, we again find a negative dren aged 3 to 5 attending pre-school increased a rate of 1.1 pp per year in the and significant coefficient for the cyclical component of per capita GDP 1990s and 1.2 pp in the 2000s. (results available upon request). Serrano et al. J Labour Market Res (2019) 53:13 Page 13 of 21 Table 5 Models of relative labor force participation (male/female) (1) (2) (3) (4) (5) (6) (7) (8) Cyclical component 49.8*** 58.1*** 41.8** 56.4*** 32.0* 45.4** 26.4 45.0* (13.90) (16.71) (14.45) (18.86) (15.39) (19.01) (17.05) (21.81) Trend component − 53.2*** − 17.2 − 58.5*** − 28.0* − 56.7*** − 19.3 − 62.2*** − 30.6** (10.63) (14.81) (15.35) (14.27) (11.27) (14.59) (15.79) (12.90) CCTs coverage 9.9 17.3* 12.0 19.2* (19.36) (8.54) (19.53) (9.36) Male unemployment − 99.5** − 74.8** − 93.9* − 77.9** (42.65) (27.72) (44.98) (28.55) Additional controls No Yes No Yes No Yes No Yes Observations 2736 2537 2511 2321 2669 2476 2445 2261 R-squared 0.719 0.753 0.714 0.750 0.728 0.761 0.723 0.759 Latin American countries, panel of education and age groups, 1987–2014. Adults aged 25–54 Fixed effects (by country and by group) OLS regressions. Unbalanced panel of 9 groups in 18 countries. Dependent variable: labor force participation ratio (men/ women) expressed in percentage. See notes to Table 4 for more details. Robust standard errors clustered by country in parentheses. ***Significant at 1% level, **5%, *10% Table 6 Models of female labor force participation Dependent variable Obs. Female LFP (%) LFP ratio (men/women) Cyclical comp. Trend comp. Cyclical comp. Trend comp. (1) Base model − 17.2*** 15.5*** 49.8*** − 53.2*** 2736 (5.24) (2.79) (13.90) (10.63) (2) Pooled OLS − 14.1* − 0.6 35.3 2.0 2736 (7.46) (0.53) (22.12) (1.56) (3) Differences − 8.4** − 2.6 19.7* 12.4 2052 (3.70) (4.27) (10.92) (12.70) (4) Dynamic panel model/1 − 5.7*** 0.0 24.8*** − 0.0 1773 (2.16) (0.11) (6.61) (0.31) (5) Random Eec ff ts − 13.4*** 6.1*** 33.0*** − 11.8*** 2736 (2.60) (0.63) (7.73) (1.70) (6) Year fixed effects − 14.0** 8.2* 39.2** − 32.5* 2736 (6.35) (4.38) (17.86) (15.68) (7) Year fixed effects + interaction − 14.0** 8.2* 39.2** − 32.5* 2736 year and educational groups (6.42) (4.42) (18.04) (15.84) (8) Additional controls 1 − 21.7*** 7.3** 58.1*** − 17.2 2537 (5.08) (3.25) (16.71) (14.81) (9) Additional controls 2 − 19.9*** 8.7 42.5*** − 21.8 936 (4.81) (5.23) (11.21) (15.66) (10) Only country > 20 periods − 17.5** 19.2*** 55.7** − 64.9** 1224 (4.97) (4.50) (18.17) (18.23) (11) Only country > 10 periods − 17.7*** 15.9*** 50.7*** − 54.2*** 2646 (5.28) (2.82) (14.13) (10.80) Latin American countries, panel of education and age groups, 1987–2014. Adults aged 25–54 (1) Base model specification is the same presented in column 1 of Table 4 for female LFP and Table 5 for the LFP ratio (men/women). (2) Pooled OLS regression with standard errors clustered by country in parentheses. (3) First difference estimator with standard errors clustered by country in parentheses. (4) Arellano and Bond (1991) GMM estimator with robust standard errors; we instrument for cycle component using a double lag. (5) Random-effect estimator with robust standard errors in parentheses. (6) Base model adding year fixed effects with standard errors clustered by country in parentheses. (7) Specification (6) + interactions between education groups and year dummies. (8) Base model specification with control variables; these results are the same presented in column 2 of Table 4 for female LFP and Table 5 for the LFP ratio. (9) Specification (8) + additional controls such as percentage of married women using modern contraceptive methods, an indicator of legal abortion, percentage of women with a washing machine, percentage of non-practicing catholic and an index of labor market regulations. (9) Base model specification restricting the sample to countries with available data for 20 years or more. (11) Base model specification restricting the sample to countries with available data for 10 years or more. ***Significant at 1% level, **5%, *10% Serrano et al. J Labour Market Res (2019) 53:13 Page 14 of 21 Table 7 Models of female labor force participation FE 2SLS Cyclical component of GDP Cyclical and trend endogenous components of GDP endogenous (1) (2) (3) (4) (5) (6) Cyclical component of GDP per capita − 17.2*** − 21.71*** − 36.5*** − 28.7** − 34.1** − 32.4*** (5.24) (5.08) (11.36) (11.52) (12.67) (10.88) Trend component of GDP per capita 15.5*** 7.32** 15.9*** 7.9** 16.6*** 11.8* (2.79) (3.25) (2.79) (3.35) (2.55) (5.98) First stage for the cyclical component of GDP per capita Exports prices—cycle 0.170*** 0.178*** 0.168*** 0.180*** (0.037) (0.040) (0.038) (0.040) Exports prices—trend − 0.047*** − 0.022*** − 0.020*** − 0.010 (0.013) (0.007) (0.007) (0.008) F test 12.97 19.21 9.903 11.59 Overid. test (p value) 0.481 0.372 Additional controls No Yes No Yes No Yes Observations 2736 2537 2736 2537 2736 2537 Fixed effects OLS and instrumental variables 2SLS. Latin American countries, panel of education and age groups, 1987–2014. Women aged 25–54 Robust standard errors in parentheses ***p < 0.01, **p < 0.05, *p < 0.1 Columns 1 and 2 report the fixed effects (by country and by group) OLS estimates from columns 1 and 2 of Table 4. Columns 3 to 6 report the fixed effects (by country and by group) 2SLS regression results. Unbalanced panel of 9 groups in 18 countries. Dependent variable: female labor force participation as percentage of women aged 25–54. Instrument variables: cyclical and trend components of the UNCTAD´s exports price index. Over-identification test refers to the Hansen J statistic, where the joint null hypothesis is that the instruments are valid. See notes to Table 4 for more details. Robust standard errors clustered by country in parentheses. ***Significant at 1% level, **5%, *10% data from harmonized national household surveys for imply a loss of productivity, making women less likely to all Latin American countries. We find that female LFP work in the future, regardless of the macroeconomic con- is positively related to the trend component of per cap- ditions. Furthermore, it could mean a strengthening of ita GDP—long-term effect—and negatively related to the traditional gender roles in the household, negatively the cyclical component—mostly related to short-term affecting the perspectives of women to participate in the shocks. This latter link is stronger for vulnerable women, labor force in the long term. with low educational attainment, married, with young Acknowledgements children, and in low-income households, which is con- This paper is based on the research that the authors carried out within the sistent with an inverse added-worker effect. We believe project “Medición de las diferencias de género en las habilidades, las limita- ciones de la familia y las preferencias de carrera”, Gender and Diversity Division, these results may shed light on an intriguing fact: the Inter-American Development Bank. It is also a follow up of Joaquín Serrano’s significant deceleration in female LFP in Latin America dissertation at the Master’s Program in Economics at Universidad Nacional de in the 2000s, a decade of exceptionally high economic La Plata, in turn based on evidence from a recent book edited by two of the authors (Gasparini and Marchionni, 2015) with the support of IDRC-Canada. growth. We are grateful to Andrew Morrison, Monserrat Bustelo, Claudia Piras, Luana Our results have nuanced implications in terms of well- Ozemela, Guillermo Cruces, Matías Busso, Carlos Lamarche, Lorena Garegnani, being. On the one hand, the deceleration in female LFP Inés Berniell, seminar participants at AAEP (2016), Network of Inequality and Poverty (UNGS, 2014), IDB ( Washington-DC, 2015), LACEA (2017), and to the may reflect the fact that in a more favorable economic editor of the journal and two anonymous referees for valuable comments and context some women are no longer bound to enter the suggestions. labor market to take precarious low-quality jobs. How- ever, as suggested by Gasparini and Marchionni (2017), staying out of the labor market during some time could Serrano et al. J Labour Market Res (2019) 53:13 Page 15 of 21 Authors’ contributions Competing interests All authors actively participated in the design and implementation of the The authors declare that they have no competing interests. study. All authors conceived and designed the analysis; JS collected the data; MM, LG and PG contributed data and analysis tools; JS performed the Appendix statistical analysis. All authors wrote the final manuscript. All authors read and approved the final manuscript. See Tables 8, 9, 10 11 and Figs. 5, 6. Funding Not applicable. Availability of data and materials The raw data corresponds to national household surveys, whose microdata sets are publicly available. The specific processed datasets used in the study are available from the corresponding author upon request. Table 8 National household surveys used in this study Country Survey name Acronym Surveys used Argentina Encuesta Permanente de Hogares Puntual EPH 1992–2003 Encuesta Permanente de Hogares Contínua EPH-C 2003–2014 Bolivia Encuesta Integrada de Hogares EIH 1992, 1993 Encuesta Nacional de Empleo ENE 1997 Encuesta Contínua de Hogares ECH 1999, 2000 Encuesta de Hogares EH 2001, 2002, 2005, 2007–2009, 2011–2013 Brazil Pesquisa Nacional por Amostra de Domicilios PNAD 1988–1993, 1995–1999, 2001–2009, 2011–2014 Chile Encuesta de Caracterización Socioeconómica Nacional CASEN 1987, 1990, 1992, 1994, 1996, 1998, 2000, 2003, 2006, 2009, 2011, 2013 Colombia Encuesta Continua de Hogares ECH 2001–2005 Gran Encuesta Integrada de Hogares GEIH 2008–2014 Costa Rica Encuesta de Hogares de Propósitos Múltiples EHPM 1989–2009 Encuesta nacional de hogares ENAHO 2010, 2012–2014 Dominican Rep. Encuesta Nacional de Fuerza de Trabajo ENFT 2000–2014 Ecuador Encuesta de Condiciones de Vida ECV 1994, 1995, 1998, 1999, Encuesta Nacional de Empleo, Desempleo y Subempleo ENEMDU 2003–2014 El Salvador Encuesta de Hogares de Propósitos Múltiples EHPM 1991, 1995, 1996, 1998–2014 Guatemala Encuesta Nacional sobre Condiciones de Vida ENCOVI 2000, 2006, 2011 Honduras Encuesta Permanente de Hogares de Propósitos Múltiples EPHPM 1991–1999, 2001–2013 Mexico Encuesta Nacional de Ingresos y Gastos de los Hogares ENIGH 1989, 1992, 1994, 1996, 1998, 2000, 2002, 2004–2006, 2008, 2010, 2012, 2014 Nicaragua Encuesta Nacional de Hogares sobre EMNV 1993, 1998, 2001, 2005, 2009 Medición de Nivel de Vida Panama Encuesta de Hogares, Mano de Obra EMO 1989, 1991 Encuesta de Hogares EH 1995, 1997–2012 Paraguay Encuesta de Hogares (Mano de Obra) EH 1990 Encuesta Integrada de Hogares EIH 1997, 2001 Encuesta Permanente de Hogares EPH 1999, 2002–2014 Peru Encuesta Nacional de Hogares ENAHO 1997–2014 Uruguay Encuesta Continua de Hogares ECH 1989, 1992, 1995–1998, 2000–2014 Venezuela Encuesta de Hogares Por Muestreo EHM 1989, 1992, 1995, 1997–2011 Source: own elaboration Serrano et al. J Labour Market Res (2019) 53:13 Page 16 of 21 Table 9 Composition of the panel datasets used in this study argbol brachl colcri domecu slvgtm hndmex nicpan pryper uryven 1992 x~ x~ ~x 1993 ~x ~~ ~~ x~ ~ 1994 ~~ ~~ ~~ x~ ~ 1995 ~~ ~~ ~x 1996 ~~ ~~~x ~ 1997 ~~ ~~ ~ 1998 ~~ 1999 ~~ ~~ 2000 ~~ ~~ ~ 2001 ~~ ~ 2002 ~~ ~ 2003 ~~ ~~ 2004 ~~ ~ 2005 ~ 2006 ~ 2007 ~~ ~~ 2008 ~~ 2009 ~ 2010 ~~ ~~ 2011 ~~ 2012 ~~ x 2013 ~~ x 2014 x x The shaded cells correspond to the available surveys, which constitute the unbalanced panel that we use in the econometric estimates. The cells marked with the ~ and x signs are interpolated and extrapolated data, respectively, used to compute the descriptive statistics for the Latin American average in Sect. 4. Note that the Latin American average excludes Colombia, Dominican Republic, and Guatemala Serrano et al. J Labour Market Res (2019) 53:13 Page 17 of 21 Table 10 Description and sources of variables used in this study Variable Description Source Female labor force participation Female labor force participation as percentage of National household surveys women aged 25–54 Male labor force participation Male labor force participation as percentage of National household surveys men aged 25–54 Labor force participation ratio (men/women) LFP ratio (men/women) expressed in percentage National household surveys Log GDP per capita Real gross domestic product per capita (log) WDI Database, World Bank CCTs coverage Share of total population who are beneficiaries of CCTs beneficiaries: own elaboration based on data conditional income transfer programs from ECLAC. Population: WDI Database, World Bank Male unemployment Unemployment rate for men aged 25–54. National household surveys Years of education Average of years of education (log) National household surveys Number of children Average number of children (log) National household surveys Married women Share of married women aged 25–54 National household surveys Women in charge of old person Share of women who are in charge of old persons National household surveys (> 70 years old) Age of children Average age of children in household National household surveys Hourly wage gap (men/women) Gender ratio of average hourly labor income of the National household surveys main job, PPP adjusted. (men/women) Female hourly wage Hourly labor income of the main job of women, National household surveys PPP adjusted Service sector (value added) Value added of service sector (share of GDP) WDI Database, World Bank Rural population Rural population as share of total population WDI Database, World Bank Index of exports prices Value index of exports (f.o.b.) converted to U.S. UNCTAD dollars and expressed as a percentage of the average for the base period (2010) Table 11 Mean of main variables LFP Employment Unemployment GDP (log) CCTs coverage Service sector Rural pop. (value added) Women Men Women Men Women Men Argentina 67.1 94.1 62.7 89.6 6.6 4.8 16.0 27.4 62.9 8.4 Bolivia 73.8 97.0 71.0 94.9 3.7 2.2 4.8 26.8 50.2 31.9 Brazil 71.2 92.4 66.4 89.0 6.8 3.7 12.4 26.5 70.8 14.6 Chile 64.3 92.3 60.2 87.6 6.3 5.1 16.4 4.1 61.7 10.8 Colombia 72.9 96.1 66.2 91.1 9.2 5.2 9.5 10.1 58.0 23.8 Costa Rica 62.2 94.9 56.9 89.6 8.4 5.6 11.0 3.6 69.4 24.1 Dominican Rep. 57.8 90.3 55.1 88.4 4.6 2.1 8.7 23.4 66.9 21.9 Ecuador 63.3 96.7 60.8 94.5 4.0 2.3 8.5 17.4 51.8 36.5 El Salvador 61.3 92.8 59.8 89.3 2.5 3.8 6.7 2.8 61.9 33.7 Guatemala 50.1 96.5 49.4 95.3 1.4 1.3 6.2 24.7 59.6 48.9 Honduras 57.6 94.8 54.1 91.3 6.0 3.7 3.9 44.2 59.8 45.9 Mexico 58.5 96.2 57.0 91.8 2.5 4.5 14.7 22.5 62.2 21.0 Nicaragua 62.7 95.5 59.3 91.7 5.4 4.0 3.6 2.6 54.2 41.5 Panama 66.4 96.6 63.3 94.2 4.6 2.5 12.6 3.3 69.6 33.7 Paraguay 69.5 95.4 65.6 92.4 5.6 3.1 6.6 8.4 50.6 40.6 Peru 79.4 95.3 77.6 93.7 2.2 1.6 7.8 10.5 57.6 21.7 Uruguay 79.6 96.0 74.5 93.0 6.4 3.1 14.0 7.9 64.2 4.8 Venezuela 68.0 95.4 63.0 90.0 7.4 5.7 15.6 42.1 11.2 Latin America 65.9 94.9 62.4 91.5 5.2 3.6 9.9 15.6 59.6 26.4 The table shows averages over the period available for each country. Labor force participation (LFP) and employment as percentage of adults aged 25–54. Unemployment rate for adults aged 25–54. Real GDP in logs. CCTs coverage as the percentage of total population who are beneficiaries. Value added of Service Sector as percentage of GDP. Rural population as percentage of total population Serrano et al. J Labour Market Res (2019) 53:13 Page 18 of 21 Fig. 5 Female and male labor force participation. Latin American countries. Source: own calculations based on microdata from national household surveys. Adults aged 25–54 Serrano et al. J Labour Market Res (2019) 53:13 Page 19 of 21 Fig. 6 Per capita GDP. Latin American countries. Source: own calculations based on data from WDI Database. Note: per capita GDP in thousands of PPP-adjusted 2011 US$ Serrano et al. J Labour Market Res (2019) 53:13 Page 20 of 21 Author details Erten, B., Ocampo, J.A.: Super cycles of commodity prices since the mid-nine- Centro de Estudios Distributivos, Laborales y Sociales (CEDLAS), IIE-FCE, teenth century. World Dev. 44, 14–30 (2013) Universidad Nacional de La Plata, and CONICET, La Plata, Argentina. Depar- Fernandes, R., Felicio, F.: The entry of the wife into the labor force in response tamento de Economía, FCE-UNLP, Oficina 322, Calle 6 No 777, 1900 La Plata, to the husband’s unemployment: a study of the added worker effect Argentina. in Brazilian metropolitan areas. Econ. Dev. Cult. Change 53(4), 887–911 (2005) Received: 29 August 2018 Accepted: 21 September 2019 Garcia-Perez, J.I., Rendon, S.,: Family job search and wealth: the added worker effect revisited, working papers 16–34, Federal Reserve Bank of Philadel- phia (2016) Garganta, S., Gasparini, L., Marchionni, M.: Cash transfers and female labor force participation: the case of AUH in Argentina. IZA J. Labor Policy 6, 10 (2017). https ://doi.org/10.1186/s4017 3-017-0089-x References Gasparini, L., Marchionni, M.: Deceleration in female labor force participation in Arellano, M., Bond, S.: Some tests of specification for panel data: Monte Carlo Latin America. Econ.-LACEA J. 18(1), 197–224 (2017) evidence and an application to employment equations. Rev. Econ. Stud. Gasparini, L., Marchionni, M. (eds.): Bridging gender gaps? The rise and decel- 58, 277–297 (1991) eration of female labor force participation in Latin America. Center for Ashenfelter, O.: Unemployment as disequilibrium in a model of aggregate Distributive, Labor and Social Studies, La Plata, Buenos Aires, Argentina, 1 labor supply. Econometrica 48(3), 547–564 (1980) edition. https ://goo.gl/yeEQZ M (2015) Attanasio, O., Levell, P., Low, H., Sánchez-Marcos, V.: Aggregating elasticities: Gasparini, L., Marchionni, M., Badaracco, N., Serrano, J.: Characterizing female intensive and extensive margins of female labour supply. NBER Working participation changes. In: Gasparini, L., Marchionni, M., (eds.) Bridging Papers 21315, National Bureau of Economic Research, Inc (2015) gender gaps? The rise and deceleration of female labor force participa- Basu, K., Genicot, G., Stiglitz, J.E.: Unemployment and wage rigidity when labor tion in Latin America, vol. 1, chapter 4, pp. 199–260. Center for Distribu- supply is a household decision. Working papers 00–10, Cornell University, tive, Labor and Social Studies, 1st edition (2015) Center for Analytic Economics (2000) Gómez, V.: The use of butterworth filters for trend and cycle estimation in Baxter, M., King, R.G.: Measuring business cycles: approximate band-pass filters economic time series. J. Bus. Econ. Stat. 19(3), 365–373 (2001) for economic time series. Review of economics and statistics, MIT Press, Heckman, J.J., Macurdy, T.E.: A life cycle model of female labour supply. Rev. 81(4) (1999) Econ. Stud. 47(1), 47–74 (1980) Beccaria, L., Maurizio, R., Vázquez, G.: El estancamiento de la tasa de partici- Hodrick, R., Prescott, E.: Postwar U.S. business cycles: an empirical investigation. pación económica femenina en Argentina en los 2000 s. Mimeo (2015) J. Money Credit Bank. 29(1), 1–16 (1997) Besley, T.J., Persson, T.: The incidence of civil war: theory and evidence. Working Kimmel, J.: Childcare costs as a barrier to employment for single and married paper 14585, National Bureau of Economic Research (2008) mothers. Rev. Econ. Stat. 80(2), 287–299 (1998) Bhalotra, S.R., Umaña-Aponte, M.: The dynamics of women’s labour supply in Layard, R., Barton, M., Zabalza, A.: Married women’s participation and hours. developing countries. IZA discussion papers 4879, Institute for the Study Economica 47(185), 51–72 (1980) of Labor (IZA) (2010) Lee, G.H., Parasnis, J.: Discouraged workers in developed countries and added Bredtmann, J., Otten, S., Rulff, C.: Husband’s unemployment and wife’s labor workers in developing countries? Unemployment rate and labour force supply: the added worker effect across Europe. ILR Review, pp. 1–31 participation. Econ. Model. 41(C), 90–98 (2014) (2017) Lundberg, S.: The added worker effect. J. Labor Econ. 3(1), 11–37 (1985) Brückner, M., Ciccone, A.: International commodity prices, growth and the out- Maloney, T.: Employment constraints and the labor supply of married women: break of civil war in sub-saharan Africa. Econ. J. 120(544), 519–534 (2010) a reexamination of the added worker effect. J. Hum. Resour. 22(1), 51–61 Busso, M., Romero Fonseca, D.A.: Determinants of female labor force participa- (1987) tion. In: Gasparini, L., and Marchionni, M., editors, Bridging gender gaps? Maloney, T.: Unobserved variables and the elusive added worker effect. Eco - The rise and deceleration of female labor force participation in Latin nomica 58(230), 173–187 (1991) America, volume 1, chapter 6, pp. 199–260. Center for Distributive, Labor Maloney, W.F.: Informality revisited. World Dev. 32(7), 1159–1178 (2004) and Social Studies, 1st edition (2015) Mankart, J., Oikonomou, R.: Household search and the aggregate labour Cerrutti, M.: Economic reform, structural adjustment and female labor force market. Rev. Econ. Stud. 84(4), 1735–1788 (2017) participation in Buenos Aires, Argentina. World Dev. 28(5), 879–891 Martinoty, L.: Intra-household coping mechanisms in hard times: the added (2000) worker effect in the 2001 Argentine Economic Crisis. Working paper, Chioda, L.: Work and family: Latin American & Caribbean Women in Search of a working paper GATE 2015-05 (2015) New Balance. World Bank, Washington, DC (2011) McGuirk, E., Burke, M.: Economic shocks and varieties of conflict: global prices, Christiano, L.J., Fitzgerald, T.J.: The band pass filter. Int. Econ. Rev. 44(2), real income and local violence in Africa. Miemo. SSRN: https ://ssrn.com/ 435–465 (2003) abstr act=27762 63 (2016) Cox Edwards, A., Roberts, J.: The effects of structural adjustment on women McKenzie, D.J.: How do households cope with aggregate shocks? Evidence in Latin America. In: Horton, S., Kanbur, R., Mazumdar, D., (eds.) Labor from the Mexican Peso crisis. World Dev. 31(7), 1179–1199 (2003) markets in an era of adjustment, vol. 1. Washington, DC: World Bank, 1st McKenzie, D.J.: Aggregate shocks and urban labor market responses: evi- edition (1994) dence from Argentina’s financial crisis. Econ. Dev. Cult. Change 52(4), Cullen, J.B., Gruber, J.: Does unemployment insurance crowd out spousal labor 719–758 (2004) supply? J. Labor Econ. 18(3), 546–572 (2000) Michalopoulos, C., Robins, P., Garfinkel, I.: A structural model of labor supply Darby, J., Hart, R., Vecchi, M. (2001), Labour force participation and the business and childcare demand. J. Hum. Resour. 27(1), 166–203 (1992) cycle: a comparative analysis of France, Japan, Sweden and the United Naz, G.: The impact of cash-benefit reform on parents’ labour force partici- States, Japan and the World Economy, 13(2), pp. 113–133. Dell, M., pation. J. Popul. Econ. 17(2), 369–383 (2004) Jones, B. F., and Olken, B. A: Temperature Shocks and Economic Growth: Ocampo, J.A.: Commodity-led development in Latin America. In: Alternative evidence from the Last Half Century. American Economic Journal: Macro- pathways to sustainable development: lessons from Latin America, economics 4(3), 66–95 (2012) International Development Policy series No. 9 (Geneva, Boston: Gradu- Eissa, N., Hoynes, H.: Taxes and the labor market participation of married cou- ate Institute Publications, Brill-Nijhoff ), pp. 51–76 (2017) ples: the earned income tax credit. Journal of Public Economics, Elsevier Parada, C., Alejo, J., Gasparini, L.: Distributive implications of changing 88(9–10), 1931–1958 (2004) female employment. Revista Cuadernos de Economía, in press (2017) Elborgh-Woytek, K., Newiak, M., Kochhar, K., Fabrizio, S., Kpodar, K., Wingender, Parker, S., Skoufias, E.: The added worker effect over the business cycle: P., Clements, B.J., Schwartz, G.: Women, work, and the economy; macro- evidence from urban Mexico. Appl. Econ. Lett. 11(10), 625–630 (2004) economic gains from gender equity, IMF Staff Discussion Notes 13/10, Paz, J.A.: El Efecto del Trabajador Adicional: evidencia para Argentina International Monetary Fund (2013) (2003–2007). Cuadernos de economía 46, 225–241 (2009) Serrano et al. J Labour Market Res (2019) 53:13 Page 21 of 21 Peña, X., Bernal, R., Amador, D.: The rise in female participation in Colombia: Tsani, S., Paroussos, L., Fragiadakis, C., Charalambidis, I., Capros, P.: Female fertility, marital status or education? Ensayos sobre política económica labour force participation and economic growth in the South Mediter- 31(71), 54–63 (2013) ranean countries. Econ. Lett. 120(2), 323–328 (2013) Prieto-Rodriguez, J., Rodriguez-Gutierrez, C.: Participation of married World Bank.: The effects of women’s economic power in Latin America and the women in the European labor markets and the “added worker effect”. J. Caribbean. Technical Report 21315, Poverty and Labor Brief—LAC region Socio-Econ. 32(4), 429–446 (2003) (2012) SEDLAC.: Socioeconomic Database for Latin America and the Caribbean. Woytinsky, W.S.: Additional workers and the volume of unemployment in Technical report, CEDLAS-World Bank (2014) the depression. Pamphlet series 1, Washington: Social Science Research Stampini, M, Tornarolli, L. The growth of conditional cash transfers in Latin Council, Committee on Social Security (1940) America and the Caribbean: did they go too far? IZA Policy Papers 49, Institute for the Study of Labor (IZA) (2012) Publisher’s Note Stephens, M.: Worker displacement and the added worker effect. J. Labor Springer Nature remains neutral with regard to jurisdictional claims in pub- Econ. 20(3), 504–537 (2002) lished maps and institutional affiliations. Tachibanaki, T., Sakurai, K.: Labour supply and unemployment in Japan. Eur. Econ. Rev. 35(8), 1575–1587 (1991) Tamm, M.: Child benefit reform and labor market participation, p. 97. Ruhr Econ. Pap., No (2009) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal for Labour Market Research Springer Journals

Economic cycle and deceleration of female labor force participation in Latin America

Loading next page...
 
/lp/springer-journals/economic-cycle-and-deceleration-of-female-labor-force-participation-in-XXNtuxuBiv

References (38)

Publisher
Springer Journals
Copyright
Copyright © 2019 by The Author(s)
Subject
Economics; Labor Economics; Sociology, general; Human Resource Management; Economic Policy; Regional/Spatial Science; Population Economics
ISSN
1614-3485
eISSN
2510-5027
DOI
10.1186/s12651-019-0263-2
Publisher site
See Article on Publisher Site

Abstract

We study the behavior of female labor force participation (LFP) over the business cycle by estimating fixed effects models at the country and population-group level, using data from harmonized national household surveys of 18 Latin American countries in the period 1987–2014. We find that female LFP follows a countercyclical pattern—espe - cially in the case of married, with children and vulnerable women—which suggests the existence of an inverse added-worker effect. We argue that this factor may have contributed to the deceleration in female labor supply in Latin America that took place in the 2000s, a decade of unusual high economic growth. Keywords: Economic cycle, Female labor force participation, Latin America JEL classification: J22, J16, N3 1 Introduction the goals of reducing poverty and income inequality Over more than 50  years female labor force participa- in the region (Parada et  al. 2017). It could also imply a tion (LFP) has been increasing markedly and steadily in stagnation of the global labor supply, given the absence almost all regions of the world. Latin America has not of significant changes in male labor force participation been an exception: about 70 million women entered the (Beccaria et al. 2015). labor market during the second half of the 20th century, The strong and rather unexpected economic expan - reflecting the increasingly important role of women in sion that Latin America experienced in the 2000s, well the region (Chioda 2011). However, since the early 2000s above the long-term trend, has been pointed out as one the growth of female LFP has slowed down significantly of the likely driving factors of the slowdown in female in most Latin American countries and it has even come LFP (Gasparini and Marchionni 2017). Better macroeco- to a halt in some of them. While the growth of female LFP nomic conditions can affect the entry of women into the was on average 0.91 percentage points a year between labor market at least in two different ways, with effects in 1992 and 2002, it slowed down to 0.35 points a year opposite directions. On the one hand, a better economic between 2002 and 2014. This deceleration has occurred context can encourage women to enter the labor market, despite both a large increase in the level of schooling of as they face more favorable labor conditions and higher women and a decline in fertility levels, two factors that wages (substitution effect). On the other hand, a bet - should favor the increase in female labor supply (Gas- ter economic scenario may imply lower unemployment parini and Marchionni 2015; Beccaria et al. 2015). and higher earnings of male partners and the expansion The recent slowdown in the growth rate of women’s of social safety nets, two factors that alleviate the pres- labor supply may affect their prospects for empowerment sure on other family members, especially female spouses, in society (World Bank 2012), postpone the reduction to look for a job, and hence negatively affect their LFP of gender gaps in the labor market, and even undermine (income effect). This latter channel, a version of the typi - cal added-worker effect, could have been more relevant for women in vulnerable households in Latin America, *Correspondence: serranojqn@gmail.com since (i) their labor supply is more elastic to income Departamento de Economía, FCE-UNLP, Oficina 322, Calle 6 No 777, shocks, either coming from earned or unearned income, 1900 La Plata, Argentina Full list of author information is available at the end of the article © The Author(s) 2019. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Serrano et al. J Labour Market Res (2019) 53:13 Page 2 of 21 and (ii) their households were the most benefitted by the two groups: those that depend on individual decisions economic changes in the 2000s. and preferences, and those that are out of the individu- In this paper we use a large dataset of microdata from al’s control. The first group includes the decisions asso - harmonized household surveys of all Latin American ciated with human capital investment (education) and countries to provide evidence on the empirical links family formation (marriage, fertility). The second group between labor force participation and the cycle and trend comprises the returns in the labor market (e.g. gender components of GDP, and use the results to argue about wage gaps), household technologies (availability of home the likely driving factors behind the observed decelera- appliances, electricity and internet connection), health tion in female LFP in the region. In particular, and unlike technologies (contraceptive methods), cultural factors Gasparini and Marchionni (2017) that carry out an analy- (religion, gender discrimination), and public policies sis based on observations at the country level, we estimate (taxes, cash transfers, children and elderly care services, fixed effects models of LFP based on a panel that follows maternity and parental leave). The variables that we are groups of individuals defined according to their level of most interested in are those related to macroeconomic schooling and age in each Latin American country over conditions, which are part of the second group of factors, the period 1987–2014. Working at a population-group such as the trend and cycle components of GDP. level allows us to incorporate a broad set of regressors, A strand of the literature studies the role of macroeco- combining variables linked to the macroeconomic con- nomic conditions by estimating the effect of recessions and text (cyclical and trend components of GDP) with others economic crises on the aggregate rate of female LFP. In par- related to demographic characteristics and public policies. ticular, these works try to assess the relevance of the added- We find that whereas female LFP is positively associated worker effect (AWE). The concept was originally conceived with the trend component of GDP, it has a countercyclical by Woytinsky (1940) and later developed by Ashenfelter behavior: large short-term expansions of GDP, beyond its (1980), Heckman and Macurdy (1980), Lundberg (1985) long-term increasing trend, are associated with a reduc- and Maloney (1987). The AWE refers to the entry into tion in female labor supply. The evidence is consistent the labor market of household secondary workers, usually with an inverse added-worker effect: better economic wives, after a transitory reduction in household income conditions for primary workers cause a delay in secondary (e.g. after the household head becomes unemployed). workers’ entrance into the labor market. This relationship u Th s, the relevance of the AWE with regard to female LFP is stronger for married women (either in formal or con- relies on a typical income effect that arises at the household sensual unions) with young children, who often act as sec- level in the context of a unitary model that assumes that ondary workers in their households, especially those with women are secondary workers and that leisure is a normal low educational attainment, living in rural areas, and from good. In the context of life cycle models with no liquidity low-income families. These results are consistent with the constraints, the AWE is expected to be negligible as long as hypothesis that the exceptionally high economic growth the loss in income due to unemployment is small compared in Latin America in the 2000s is a relevant factor behind to the primary worker’s lifetime earnings (Lundberg 1985). the deceleration in female LFP, which was particularly In addition to the income effect represented by the intense among vulnerable married women. AWE, unfavorable economic perspectives in the labor The rest of the paper is organized as follows. We market may lead to a substitution effect known in the lit - begin in Sect. 2 by briefly reviewing the literature on the erature as the discouraged worker effect, which operates in determinants of female LFP in Latin America. Section  3 the opposite direction. If the latter is outweighed by the describes the data sources that are used in the study. In Sect.  4 we present preliminary evidence on the associa- tion between the economic cycle and the recent decelera- Busso and Romero Fonseca (2015) apply a meta-analysis of the determinants tion of female LFP in the region. Section  5 describes the of women’s labor supply and estimate bivariate models with country fixed empirical strategy and Sect.  6 reports and discusses the effects to assess the role of each factor in explaining the evolution of female results. In Sect. 7 we conclude with some final remarks. LFP in Latin America. Their results suggest that the positive long-term trend is driven primarily by the expansion of health and household technologies and by the gradual change in cultural factors. Furthermore, traditional fac- 2 Background tors such as the increasing female educational attainment and the decreasing Disentangling all the factors that may account for the fertility rates have also contributed to the long-term expansion of female par- ticipation in the labor force and vice versa. Other works reach similar conclu- observed patterns in female labor supply is not an easy sions using aggregate decompositions, such as Peña et al. (2013) for Colombia task, since several potential driving forces are simultane- or Gasparini et al. (2015) for several countries in Latin America. ously at play. Busso and Romero Fonseca (2015) and Chi- In turn, the effect can be boosted if the increase in time away from the labor market of primary workers reduces the opportunity cost of carrying oda (2011) present a broad conceptual framework, which out market activities for secondary female workers, through the substitution divides the possible determinants of female LFP into of tasks within the household. Serrano et al. J Labour Market Res (2019) 53:13 Page 3 of 21 AWE, female labor force participation should exhibit a In contrast, in their study for Latin American coun- countercyclical behavior. tries between 1965 and 1987, Cox Edwards and Roberts In practice, the strength of these effects depends on (1994) show that the AWE is significant for low-income the relevance of other factors, such as the availability countries, although not for richer countries, such as of alternative strategies to cope with negative income Argentina or Chile. In a more recent work, Bhalotra and shocks (child labor, unemployment insurance, etc.) and Umaña-Aponte (2010) use a panel data set for 63 devel- the existence of imperfect credit markets together with oping countries for the period 1986–2006, and find that liquidity constraints (Mankart and Oikonomou 2017; the relationship between female employment and growth Garcia-Perez and Rendon 2016). Consequently, empirical is negative on average for Latin America and Asia, but studies for developed countries, such as the U.S. or the positive for Africa. We improve the empirical strategy United Kingdom, find small added-worker effects (Cul - of these earlier studies by using more disaggregated and len and Gruber 2000; Stephens 2002; Prieto-Rodriguez better-quality data (more Latin American countries, with and Rodriguez-Gutierrez 2003; Bredtmann et  al. 2017), longer and more comparable time series), by including or even no evidence of its existence (Layard et  al. 1980; more controls, and by focusing directly on female labor Maloney 1991). Instead, in developing economies such force participation instead of the employment rate. as those of Latin America, the AWE may be larger due In this paper we deepen the analysis of the main to the lack of unemployment insurance benefits, the fact hypothesis in Gasparini and Marchionni (2017): the that many households face financial restrictions, and the deceleration of the growth rate of female labor supply role of women as secondary workers, reinforced by solid could be related to the strong economic expansion expe- family structures with a strong attachment to traditional rienced by the Latin American countries in the 2000s. gender roles, and low levels of women’s skills and edu- The better economic scenario, which resulted in an cational attainment within some population groups. In improved labor situation especially for the less skilled addition, given the importance of the informal sector in male workers, could have encouraged an inverse added- the region, entry and exit barriers in the labor market are worker effect on their female partners. In other words, relatively low, which facilitates changes in female partici- women, especially the most vulnerable, may have decided pation (Basu et al. 2000; Maloney 2004). to postpone their entry into the labor market due to a Several empirical studies analyze the validity of the AWE lower pressure to search for a job. This paper overcomes hypothesis in the region. Martinoty (2015) uses the col- some methodological limitations and at the same time lapse of the Argentina’s convertibility regime as a natural deviates in some directions from the analysis in Gasparini experiment to evaluate the effect of changes in husbands’ and Marchionni (2017), besides extending the period of labor situation in the labor participation decision of their analysis. Their estimations are based on cross-country wives, finding evidence of a statistically significant AWE. panel data and include as unique regressors the cyclical Similar results are found by Cerrutti (2000) and Paz (2009) and trend components of GDP. Instead, in this paper we also for Argentina in the 1990s and 2000s, respectively, estimate multivariate models combining variables asso- Fernandes and Felicio (2005) for Brazil, and Parker and ciated with the macroeconomic context (the abovemen- Skoufias (2004) for Mexico. On the other hand, McKen - tioned trend and cycle components of GDP) with others zie (2004) and MacKenzie (2003) find no evidence on the related to demographic characteristics and public poli- presence of AWE when studying household strategies to cies. Among the regressors, we include the coverage of compensate the negative shocks from the financial crises conditional cash transfer programs, proxied by the pro- of 2002 in Argentina and 1995 in Mexico, respectively. portion of beneficiaries in the population. In addition, The literature that explores the dynamics of female we build a panel dataset disaggregating the adult popula- labor supply with aggregate level data provides more tion into groups defined by education and age for each mixed results, with some papers that even report a procy- country, which significantly increases the cross-section clical behavior for developed countries (Tachibanaki and variability. In this way, our models allow for a better iden- Sakurai 1991; Darby et al. 2001; Lee and Parasnis 2014). tification of the partial correlations between labor force participation and each of its covariates, as we control for fixed effects by country and by population group, among other variables. Regarding this issue, there is a variety of recent literature, both theoretical and empirical, which tries to reconcile the differences between estimates of elasticities of female labor supply with respect to wages, based on micro data 3 Data sources or aggregate data. For instance, Attanasio et  al. (2015) estimate a life-cycle Our analysis is mostly based on microdata from house- model to explain female labor supply in the United States, trying to bridge the discrepancies between micro and macro estimates. Among other results, they hold surveys, which are part of the Socioeconomic Data- find that the aggregate elasticities of labor supply vary throughout the eco - base for Latin America and the Caribbean (SEDLAC), nomic cycle, being stronger during recessions. Serrano et al. J Labour Market Res (2019) 53:13 Page 4 of 21 a project jointly developed by CEDLAS at Universidad restrict the sample to adults between 25 and 54 years old. Nacional de La Plata and the World Bank. Household Labor behavior of younger individuals is more related to surveys are not homogeneous across countries and in education decisions while the labor supply of older peo- some cases not even for the same country over time. ple mostly depends on the relevance and dynamics of the Given such heterogeneity, careful survey processing is pension system. necessary to ensure as much comparability as possible of estimates among countries and years. This is precisely 4 The economic cycle and female labor force one of the advantages of using the SEDLAC database, participation in Latin America where microdata are harmonized using similar defini - The deceleration since the early 2000s, after a strong tions of variables for each country/year and a consistent increase during the previous decades, has been a major and documented protocol (see SEDLAC 2014). In this change in the dynamics of female labor supply in Latin paper, we use SEDLAC microdata of 18 countries (the 17 America. Figure  1 shows that the deceleration is robust countries in continental Latin America and the Domini- to the group of countries chosen. While the growth of can Republic). Table  8 in the Appendix describes the female LFP was on average 0.91 percentage points a year corresponding surveys and the years they cover. between 1992 and 2002, it slowed down to 0.35 points a Our econometric estimations are based both on a non- year between 2002 and 2014. Even though the contrast balanced panel dataset of those 18 countries over the between the growth rates in both periods is not similar period 1987–2014, and on a more disaggregated panel in all Latin American countries, it is significant in most dataset that follows 9 groups of women defined accord - of them and sufficiently widespread to be visible in the ing to their age and education for each one of the 18 regional average. In some cases there are even signs countries. However, to compute the descriptive statistics of stagnation (Fig.  5 in the Appendix). Unlike women’s, that we show in Sect.  4, we use a smaller sample of only men’s labor supply is much higher and more stable over 15 countries for the shorter 1992–2014 period. With this time. Thus, the recent deceleration in the growth rate of restricted sample we build a balanced panel using lin- female LFP delays the closing of the gender gap in labor ear interpolations and extrapolations, for which we take participation. information from the most proximate surveys. Table 9 in Gasparini and Marchionni (2015, 2017) show that the the Appendix presents a schematic summary of the com- deceleration was stronger for women that are in more position of both the non-balanced and balanced panels. vulnerable conditions, especially those with low edu- All the statistics at the country or population-group level cation, living in rural areas, and who are married and are computed using the corresponding sample weights. with children. Vulnerable women usually have a weaker However, in order to describe the situation for the whole attachment to the labor market, and thus they are more region we use simple averages across countries instead prone to enter and exit the labor market depending on of population-weighted averages, to avoid that highly the economic situation inside and outside their house- populated countries, such as Brazil and Mexico, drive the holds. The fact that the deceleration in female LFP since results. The demographic, social and labor variables are the early 2000s was especially intense among vulnerable obtained from the SEDLAC microdata. The rest of the women suggests that changes in the macroeconomic variables, such as per capita GDP, some institutional and context could have played an important role. political variables, or the coverage of social programs are taken from alternative sources (e.g. the World Develop- ment Indicators from the World Bank, or CEPALSTAT). Table 10 lists the variables used throughout the study, their Female labor supply strongly and persistently expanded since the 1960s in definitions and the corresponding sources. Latin America (Chioda, 2011). An additional clarification before moving on to the 7 The slowdown is also evident when grouping the countries by sub-region next section: following the literature on labor supply, our (South and Central America) or by their initial levels of female LFP. analysis focuses on prime-age people. In our case, we Argentina, Bolivia, Brazil, Chile, Costa Rica, Ecuador, Honduras, México, Panama, Paraguay and Venezuela experience a deceleration in the growth rate in female LFP since the decade of the 2000s. We focus on labor force participation since most of the action seems to have taken place in that margin. Gasparini and Marchionni (2015) find that Most of the household surveys included in the sample are representative “Changes in hours of work for female workers were not large, not very dif- at the national level. The exceptions are Uruguay before 2006 and Argentina, ferent between decades, and not significantly different from those of males. where surveys cover only the urban population, which, however, represents Likewise, changes in unemployment seem to have been small and with no more than 85% of the total population. significant gender differences. These patterns reinforce the claim that the This group excludes Dominican Republic, Colombia and Guatemala, for dynamics of labor force participation are among the most noticeable labor which there are no comparable national household surveys previous to year phenomena with a clear gender dimension over the last decades.” Chioda 2000. (2011) and World Bank (2012), among others, share this conclusion. Serrano et al. J Labour Market Res (2019) 53:13 Page 5 of 21 GDP per capita Female LFP (left) 1994 1998 2002 2006 2010 2014 Fig. 1 Female labor force participation in Latin America. Source: own Fig. 2 Female labor force participation and economic growth. calculations based on microdata from national household surveys. Source: own calculations based on microdata from national Women aged 25–54. Unweighted means of Latin American countries. household surveys, per capita GDP (in PPP-adjusted constant US$) The series of 2 countries includes Argentina and Brazil. The series of from WDI. Women aged 25–54. Unweighted means of 15 Latin 9 countries adds to the previous Bolivia, Chile, Costa Rica, Honduras, American countries Mexico, Uruguay, and Venezuela. The series of 15 countries adds to the previous Ecuador, El Salvador, Nicaragua, Panama, Paraguay, and Peru. The series of 18 countries includes the previous 15 countries labor conditions for the primary workers improve during plus Colombia, Guatemala, and the Dominican Republic a period of strong economic expansion, female secondary workers could feel discouraged from participating in the labor market. The incentives for women do not necessar - In fact, the macroeconomic conditions in Latin Amer- ily imply an exit from the labor market, but a postpone- ica were significantly different in the 2000s compared to ment of the decision to enter the market, for instance, to the previous decade. While the average GDP grew at an allocate more time caring for their children or the elderly annual rate of 1.9% between 1990 and 1999, it grew 3.5% at home. Also, the better economic conditions inside and per year on average between 2000 and 2014, despite the outside the household can relieve the pressure on women effects of the 2008 global financial crisis. In addition to acting as secondary workers to take any kind of job, the higher growth rates, the 2000s were characterized by allowing them to wait until they find a job that fits their more macroeconomic stability than previous decades. preferences. Figure  2 shows that the strong increase in per capita Therefore, if the AWE outweighs the substitution GDP in the region over the 2000s occurred in coinci- effect, female LFP should exhibit a countercyclical behav - dence with the deceleration of female LFP. As discussed ior. In this sense, it is important to notice that the AWE in Sect.  2, the effect of economic growth on women’s should be much more relevant for women from more labor supply could be either positive or negative depend- vulnerable households. In fact, as discussed above, poor ing on whether it is the substitution or the income effect women, with low educational attainment and with young that prevails. On the one hand, an improved macro- children are more likely to act as secondary workers, as economic context can encourage female LFP through a they have a weaker attachment to the labor market and substitution effect, but on the other hand it can depress their labor decisions are more sensitive to the economic female labor supply because the better economic situa- situation in their households (Michalopoulos et al. 1992; tion in the household alleviates the pressure on women Kimmel 1998; Eissa and Hoynes 2004; Naz 2004; Tamm to look for a job outside the house, allowing them to 2009). It is precisely these women who have benefitted postpone their entry into the labor market. most from the economic expansion in the 2000s, through As above mentioned, the latter argument is a vari- improvements in the employment rate and earnings of ant of the hypothesis of the added-worker effect (AWE), men in their families. generally used to explain the increase in female LFP in Figure  3 illustrates this point by showing the evolu- response to unemployment shocks and the drop in fam- tion of some labor variables for high-skilled relative ily income during economic recessions. Conversely, as to low-skilled prime-age men. The hourly wage ratio between men with high and low education has substan- tially shrunk since the early 2000s. This fact suggests that Figure  6 in the Appendix shows the GDP per capita series for each one of in addition to the generalized increase in real wages, the the countries under analysis. 45 50 55 60 65 70 8000 9000 10000 11000 12000 13000 Serrano et al. J Labour Market Res (2019) 53:13 Page 6 of 21 Fig. 3 Unemployment and hourly wage ratios between men with high and low education. Men aged 25–54. Source: own calculations based on microdata from national household surveys. Note: the ratios are defined as high education/low education, where low education = less than secondary complete, high education = tertiary complete. Unweighted means of 15 Latin American countries economic improvement was greater in poor than in non- poor households. Also, there has been a pronounced decline in the unemployment rate of unskilled men, which fell from 6.1% in 2002 to 3.7% in 2014, in contrast with a more stable behavior of the unemployment rate for the skilled men, which fluctuated around 4%. Fig - ure  3 then highlights the existence of a potentially rele- vant added-worker effect especially for more vulnerable women who, given the positive assortative mating, are the ones likely married to low- education workers. Fig. 4 Evolution of public transfers in Latin America. Source: a own An additional factor that could explain the deceleration calculations based on non-contributory social protection programs database, Social Development Division, ECLAC; b own calculations in female LFP is the income effect due to an increase in based on microdata from national household surveys. a shows public transfers to families, especially from conditional unweighted means of 17 Latin American countries: Argentina, Bolivia, cash transfer programs (CCTs), which have strongly Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, expanded in Latin America since the early 2000s both El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama, in terms of number of beneficiaries and amounts trans - Paraguay, Peru, and Uruguay. b Women aged 25–54, unweighted means, monetary and non-monetary public transfers. Unweighted ferred. Figure  4a shows that the coverage of CCTs in the means of 15 Latin American countries region, measured as the percentage of the total popula- tion that are beneficiaries of these programs, grew from 2.7 to 18% between 2002 and 2010, and remained fairly CCTs in Mexico, Brazil and Colombia cover around 50% stable over the following years. CCTs consist of transfers of the poor, while coverage in Uruguay reaches 80%. to poor families with children. The transfers are usually The potential effects of CCTs on female LFP are ambig - monetary and conditional on households investing in uous. On the one hand, there may be a negative effect children’s human capital (education, health, and nutri- operating through three different channels. For women tion). Currently, almost all the countries in Latin America who had the urgency to get a job due to the difficult eco - have some kind of conditional cash transfer program, nomic situation in their homes, CCTs can provide the reaching a large fraction of the poor population. For instance, according to Stampini and Tornarolli (2012) CCTs programs are not the only policy tool for poverty alleviation. For instance, non-contributory pensions have strongly expanded in the region during the 2000s, adding another source of non-labor income for the more vulnerable households. Serrano et al. J Labour Market Res (2019) 53:13 Page 7 of 21 economic relief that allows them to delay their entrance on an unbalanced panel of 18 Latin American countries into the labor market. Moreover, since women are typi- over the period 1987–2014. cally the recipients of the cash transfer, they may perceive Y = α + βln(GDPpc) + η + μ ct c ct (1) ct the subsidy as earned income in exchange for their efforts to ensure compliance with the conditionalities associated Y = α + β cycle + β trend + η + μ ct 1 ct 2 ct c ct (2) with the program, which reduces their available time to where the left-hand-side variable Y is the LFP rate for ct engage in market activities while encouraging the tra- women aged 25 to 54 for country c at year t . In model ditional division of gender roles within the household (1) we include in the right-hand side the logarithm of the (Garganta et  al. 2017). Finally, the beneficiaries of CCTs real per capita GDP, while in model (2) we use its cyclical may believe that in order to continue to be eligible for the and trend components, estimated through the Hodrick- program, they should work less to remain poor. All these Prescott filter. The country fixed effects η capture both three channels would involve a negative effect of CCTs observable and unobservable factors that vary across on female LFP. On the other hand, CCTs may have posi- countries but that are fixed over time, avoiding potential tive effects on female LFP. If conditionalities require that sources of omitted-variable bias. children go to school, they may induce women to allocate As for other covariates of LFP, it is difficult to identify more time to market activities as they do not need to use statistically significant relationships given the small num - it for childcare anymore. This could also imply that chil - ber of countries and the limited variability that many social dren have less time to work, which could reduce house- and economic indicators exhibit over time. To gain more hold income and increase the demand for earned income sample variability we build a new panel with data at the inside the family (Busso and Romero Fonseca 2015). population-group level, where the groups are the result Figure 4b shows the evolution of labor participation of of combining three levels of educational attainment (low, women according to whether their households receive medium and high) and three age ranges (25–34, 35–44, or not public transfers, which include CCTs and other and 45–54). This latter panel then follows these nine popu - non-contributive pensions, either monetary or in kind, lation groups of women in each of the 18 countries over and exclude retirement pensions. While labor force par- the period 1987–2014. Based on the population-group ticipation rates are similar for the two groups of women level panel, we propose the specification in model (3): over the 1990s, since the early 2000s a gap develops due to a decline in labor supply of women from beneficiary Y = α + β cycle + β trend + β um gct 1 ct 2 ct 3 gct families. This fact is consistent with the hypothesis of the (3) + β cct + ϕX + θQ + η + ν + μ 4 ct ct gct c g gct negative income effect discussed above. According to the meta-analysis carried out in Busso and Romero Fonseca The left-hand-side variable Y is the LFP rate for women gct (2015), CCTs generally have a negative effect on female in group g for country c at year t . As in models (1) and LFP and hours worked, even though several studies for (2), the right-hand side includes the cyclical and trend countries in the region find the effect to be not statisti - components of per capita GDP for each country and year. cally significant. To assess the relevance of the added-worker effect we We have argued that there are several factors that could include the male unemployment rate of group g in coun- explain the deceleration in female LFP in Latin America, try c of year t um as an additional explanatory vari- gct and it seems that one of the most important is related to able. If the hypothesis of the inverse added-worker effect the economic cycle, even though the evidence presented is valid, we expect that as we add this factor, the absolute so far is merely descriptive. The relationships described value of the estimated coefficient of the cyclical compo - could be economically irrelevant or not statistically sig- nent of per capita GDP decreases. Also, in order to evalu- nificant. Therefore, in the following sections we deepen ate the relationship between the conditional cash transfer into the analysis of the role of the economic cycle on programs and the changes in female LFP, we include the women’s labor participation decisions to assess the extent coverage of such programs for each country and year to which macroeconomic conditions can account for the ( cct ). X are other controls that vary across countries ct ct recent slowdown in the rate of growth of women’s labor and through time, such as the share of the value added supply in the region. of the service sector in the GDP and the share of rural 5 Empirical strategy In order to study the dynamics of women’s LFP over the Our basic specification does not control for year fixed effects, but we per - economic cycle, we use a set of panel data models that form some robustness checks where we do, finding that the main results hold. See the discussion at the end of Sect.  6.2 and Table  6. Since GDP varies by allows us to control for some potential sources of bias. country and by year, it is not possible to control for year fixed effects sepa - We start with the simple models in Eqs. (1) and (2), based rately for each country. Serrano et al. J Labour Market Res (2019) 53:13 Page 8 of 21 6.1 Analysis at the country level population, and Q are covariates that also vary across gct We start with the simple models at the country level in the groups g , such as education, fertility, marital status, Eqs.  (1) and (2), as in Gasparini and Marchionni (2017). age of children, average wage of women, and gender wage Table 1 shows the OLS estimation results of country fixed gap. Tables 10 and 11 in the Appendix provide definitions effects models of labor force participation based on an and descriptive statistics of the variables. η and ν are c g unbalanced panel of 18 Latin American countries in the fixed effects by country and by group, respectively. 1987–2014 period. The dependent variables are female We estimate models (1), (2) and (3) for women, men, LFP (columns 1 and 2), male LFP (columns 3 and 4), and and the gender ratio (men/women), in order to explore the gender ratio in LFP (men to women, in percentage, differences in the effect of the economic cycle by gender. in columns 5 and 6). As explanatory variables we include The estimation results are presented and discussed in the logarithm of real per capita GDP, and alternatively its Sect. 6. cyclical and trend components. Although in model (3) we control for a quite large set The results in Table  1 suggest that changes in GDP are of potentially relevant factors, including the unobserved positively related to female LFP: a 10 percent increase in heterogeneity that varies across countries or across per capita GDP is associated with an increase in female groups but is fixed over time, it may still be insufficient LFP of 2.26 percentage points (pp.) on average. When and other potential sources of bias may persist. For decomposing per capita GDP, we find a statistically sig - instance, measurement error in the cyclical component nificant relationship both with the cyclical and the trend of GDP per capita could lead to attenuation bias of our components, but with opposite signs. Whereas the trend estimator of β . However, as we will see in the next sec- component is associated with an increase in female LFP tion, all our results indicate a strong and statistically sig- (2.56  pp.), the short-term movements are countercycli- nificant countercyclical behavior of female LFP. In any cal: a 10 percent increase in the cyclical component of case, we carried out robustness checks to other alter- the GDP is associated with a reduction in female LFP of native measures of the cyclical and trend components approximately 2.17  pp. This latter result is consistent and the results hold. Also, the reciprocal relationship with the hypothesis that the unusual strong economic between the business cycle and female LFP may bias our growth experienced by many Latin American coun- OLS estimates towards zero. This would be the case if, for tries during the 2000s contributed to the deceleration in instance, female LFP positively affects economic growth women’s labor supply. In fact, a back-to-the-envelope- (Tsani et  al. 2013; Elborgh-Woytek et  al. 2013). We calculation with the results from Table  1 suggests that address the concerns regarding this potential source of the cyclical component of the GDP growth of the Latin endogeneity by applying an instrumental variables strat- American economies over the 2000s may account for egy to identify the causal effect of the cyclical component 43% of the deceleration in female LFP. of GDP on female LFP. We propose to use exports prices In contrast with the case of women, changes in per as instruments for the cyclical component. The valid - capita GDP or its cyclical and trend components are not ity of this instrument is discussed in the next section. For associated with male labor supply, which shows very lit- the estimation, we use the two-stage least squares (2SLS) tle variation over the period under study. As a conse- procedure, considering the cyclical and trend compo- quence, the labor force participation gap between men nents as endogenous variables. and women is negatively correlated with the GDP trend, but positively correlated with the cyclical component. 6 Results In sum, the estimation results from models in Eqs.  (1) In this section we present the main results of our anal- and (2) indicate that female LFP increases with economic ysis. We start by discussing the models at the country level, then expand the analysis to a panel at the country- group level and we end by providing instrumental vari- ables estimates to reinforce the credibility of our results. 15 Our results are consistent with those of Bhalotra and Umaña-Aponte (2010), who find a countercyclical pattern of female LFP in Latin America and Asia. Using a similar regression framework, we find that the employment For the Hodrick-Prescott filter we use a smoothing parameter of 100 rate is also positively related to the GDP per capita growth for women but (Hodrick and Prescott 1997). We checked the sensitivity of our results to not for men. The effect of the trend component is positive and particularly changes in the smoothing parameter of the Hodrick-Prescott filter and to the strong for women, while the cycle is especially strong for males. In addition, use of three other filters: the Baxter-King band pass filter (Baxter and King unemployment is negatively related to GDP, and the effects of both cycli- 1999), the band pass filter (Christiano and Fitzgerald 2003) and a Butterworth cal and trend components are negative and statistically significant for both filter (Gómez 2001). In all cases, our results hold, i.e., we find a countercycli - genders. In turn, wages increase as GDP expands for both men and women. cal and statistically significant behavior of female labor participation. Results An increase in the trend component of per capita GDP is associated with available upon request. reductions in the gender wage gap, while short-term expansions widen it. We are thankful to one referee for suggesting the use of this instrument. These results are available upon request. Serrano et al. J Labour Market Res (2019) 53:13 Page 9 of 21 Table 1 Models of labor force participation Women Men Ratio men/women (1) (2) (3) (4) (5) (6) Log GDP per capita 22.6*** − 1.1 − 73.0*** (2.79) (0.87) (14.41) Cyclical component − 21.7*** 2.3 61.9*** (6.05) (1.58) (16.25) Trend component 25.6*** − 1.3 − 82.2*** (2.86) (0.93) (14.83) Observations 304 304 304 304 304 304 Countries 18 18 18 18 18 18 R-squared 0.56 0.66 0.04 0.06 0.49 0.57 Latin American countries, panel 1987–2014. Adults aged 25–54 Fixed effects (by country) OLS regressions. Unbalanced panel of 18 countries. Dependent variable: in columns 1 and 2 (3 and 4) is female (male) labor force participation as percentage of women (men) aged 25–54; in columns 5 and 6 is the LFP ratio (men/women) expressed in percentage. Log per capita GDP is the logarithm of real gross domestic product per capita. Cyclical and trend components of GDP are obtained by applying the Hodrick-Prescott filter to the log of per capita GDP. For detailed data definitions and sources, see Table 10 in the Appendix. Robust standard errors clustered by country in parentheses. ***Significant at 1% level, **5%, *10% development, reducing the gender gap in labor supply; families or for women living alone. Finally, the results in however, short-term economic expansions are associated columns 9 to 13 suggest that women with young children with a decline in the entry of women into the labor force, are more prone to react to improved economic condi- and thus with a widening of the gender gap. tions than those with older children or no children at all, We already suggested that an inverse added-worker a result again consistent with the inverse-AWE story. effect (AWE) could be the mechanism behind these As discussed in Sect. 4, another implication of the AWE results. Although we cannot provide causal evidence argument is that we should observe a stronger reaction with the data at hand, we can explore whether some pre- to the cyclical movements in economic activity for the dictions of this hypothesis are consistent with the evi- group of more socioeconomic vulnerable women. To this dence. In what follows we explore two implications of aim we estimate the female LFP models by education, the inverse AWE argument. First, since the mechanism household size, per capita family income, marital status, operates through adjustments within the household, the area of residence, and family types. We also build a vul- results should differ across women from different house - nerability index based on the principal components of hold arrangements. In particular, we expect the sensi- these variables. According to this index, vulnerable indi- tivity of LFP to cyclical fluctuations of GDP to be higher viduals are those with a very low educational attainment, for women living with her spouse and with young chil- living in rural areas, with many children, and with low dren, since, as argued above, they are more likely to act incomes. Table  3 shows the estimation results of model as secondary workers within the household. The results (2) by educational group, and for the vulnerable and non- in Table 2 are consistent with this expectation. The coef - vulnerable population, defined as quintiles 1 and 5 of the ficient for the cyclical component of GDP is higher (in vulnerability index, respectively. absolute value) for women living with a spouse (“mar- According to our results, the negative relationship ried”); in fact, the coefficient is negative but statistically between the cyclical component of per capita GDP and non-significant for the group of single women. Accord - female LFP is stronger for less educated, more vulner- ingly, the ratio in LFP between married and single women able women. Whereas the coefficient associated with falls (i.e. the gap in LFP between these groups becomes larger) as the cyclical component of the GDP increases (column 3). We also divide women into “head of household” and other members. We When classifying women by their family structure (col- use two alternative definitions for “head of household”: (i) status self-reported umns 4 to 8), the negative effect of the cyclical component in the survey, or (ii) family member with the highest earnings (economic defi - nition). In both cases the coefficient of the cyclical component is much higher of the GDP is significant only for women in two-parent (in absolute value) for group of women who are not heads of household. In families. The effect vanishes for women in single-parent fact, in the economic definition the coefficient is non-significant for female household heads. The results for other population groups are consistent with those pre- We are thankful to one referee for making this observation. sented in Table 3 and are available upon request. Serrano et al. J Labour Market Res (2019) 53:13 Page 10 of 21 Table 2 Models of female labor force participation by household structure Married or single Family structure Married (M) Single (S) Ratio M/S Two parents (T) Single‑parent (S) Single‑person (P) Ratio T/S Ratio T/P (1) (2) (3) (4) (5) (6) (7) (8) Cyclical component − 24.8*** − 8.8 − 23.3** − 24.0*** − 9.9 − 4.9 − 22.6*** − 23.6** (7.34) (8.01) (8.04) (6.63) (5.94) (7.70) (5.19) (9.07) Trend component 27.0*** 11.1*** 26.1*** 26.4*** 13.8*** 16.2*** 23.2*** 19.3*** (3.2) (3.15) (4.79) (2.80) (2.09) (3.86) (3.24) (2.89) Observations 268 268 268 304 304 304 304 304 R-squared 0.657 0.260 0.399 0.639 0.453 0.224 0.535 0.289 Youngest child is 0–5 (G1) 6–17 (G2) No children (NC) Ratio G1/G2 Ratio G1/NC (9) (10) (11) (12) (13) Cyclical component − 29.1*** − 24.0*** − 17.5** − 15.3* − 21.8* (8.07) (6.57) (7.62) (8.55) (12.16) Trend component 26.1*** 27.4*** 28.2*** 6.1 5.5 (3.19) (2.51) (3.26) (3.93) (3.78) Observations 303 303 303 303 303 R-squared 0.606 0.634 0.636 0.080 0.059 Latin American countries, panel 1987–2014. Women aged 25–54 Fixed effects (by country) OLS regressions. Unbalanced panel of 18 countries. Dependent variable: female labor force participation as percentage of women aged 25–54; and ratios of female LFP ratio between groups expressed in percentage. Cyclical and trend components of GDP are obtained by applying the Hodrick- Prescott filter to the log of real per capita GDP. For detailed data definitions and sources, see Table 10 in the Appendix. Robust standard errors clustered by country in parentheses. ***Significant at 1% level, **5%, *10% trend component, particularly for the more vulnerable/ the cyclical component is − 23.6 for the group of women less educated groups. with less than complete secondary schooling, it halves (− 11.3) for women with a degree from a tertiary institu- 6.2 Analysis at the country‑group level tion. In fact, that coefficient is negative and significant in In order to gain more sample variability, we estimate a regression for the ratio of LFP between low and high- model (3) based on a more disaggregated panel dataset educated women, suggesting a widening gap between where the units of observation are population groups these groups over the expansive phase of the business defined in terms of schooling and age for each country cycle. Columns 6 to 8 report the results for the vulner- and year. This also allows us to include other poten - ability groups defined above. Whereas the coefficient tial drivers of female LFP, besides the cyclical and trend associated with the cyclical component is negative (− 9.0) components of GDP, such as the male unemployment but not statistically significant for non-vulnerable women rate and the coverage of CCT programs. We also control (top quintile of the vulnerability index), the coefficient for other potential covariates of female LFP: factors that for the most vulnerable group is much larger in absolute are jointly determined with the labor supply decisions value (− 26.0) and highly significant. (education, marriage, fertility), and factors that are likely Unlike the results in Table  3, male labor supply does exogenous to the individuals (gender income gap, cost of not seem to move in line with changes in per capita GDP: childcare and eldercare services, women’s income, share the coefficients of the LFP regressions are much smaller 21,22 of the services sector in GDP, rural population). for men and non-statistically significant in most cases. Accordingly, the labor force participation ratio between men and women is positively related to the cyclical com- It should be noted that including women’s wages and a proxy of the impor- ponent of per capita GDP and negatively related to the tance of the services sector contributes to controlling for factors related to the labor demand, which in turn are correlated with the business cycle. There - fore, these controls would likely capture the fact that recessions or short-term economic expansions can generate compositional changes in the productive structure, thus affecting female LFP. The correlations of each of these additional controls with female LFP The results for men and for the ratio men/women (not reported) are avail - have the expected signs according to the empirical literature (see Busso and able upon request. Romero Fonseca 2015). Serrano et al. J Labour Market Res (2019) 53:13 Page 11 of 21 Table 3 Models of female labor force participation by educational attainment and vulnerability Education Vulnerability Low (L) Medium (M) High (H) Ratio L/M Ratio L/H Vulnerable (V) Non‑ vulnerable (N) Ratio V/N (1) (2) (3) (4) (5) (6) (7) (8) Cyclical component − 23.6*** − 16.9** − 11.3*** − 15.5*** − 19.9** − 26.0*** − 9.0 − 22.4* (7.12) (7.33) (3.30) (5.16) (7.29) (7.97) (6.04) (10.67) Trend component 22.1*** 10.3*** 10.1*** 20.8*** 18.9*** 29.1*** 13.5*** 29.8*** (2.63) (3.51) (2.94) (4.43) (2.73) (3.32) (3.14) (6.40) Observations 304 304 304 304 304 304 304 304 R-squared 0.557 0.219 0.260 0.351 0.397 0.520 0.235 0.256 Latin American countries, panel 1987–2014. Women aged 25–54 Fixed effects (by country) OLS regressions. Unbalanced panel of 18 countries. Dependent variable is female labor force participation as percentage of women aged 25–54. Columns show estimations of the models dividing individuals by different levels of educational attainment and quintiles of a vulnerability index; and ratios of female LFP ratio between groups expressed in percentage. Low education = less than secondary complete; medium education = secondary complete or tertiary incomplete; high education = tertiary complete. Vulnerable = individuals who are in quintile 1 of a vulnerability index based on the principal components of level of educational attainment, marital status and number of children in the household. Cyclical and trend components of GDP are obtained by applying the Hodrick- Prescott filter to the log of real per capita GDP. For detailed data definitions and sources, see Table 10 in the Appendix. Robust standard errors clustered by country in parentheses. ***Significant at 1% level, **5%, *10% Table  4 presents the estimation results of alternative Table  5 shows the results of estimating model (3) specifications of model (3) with fixed effects by coun - with the labor force participation ratio between men try and by group. As a general result, we find that the and women as the dependent variable. Cyclical expan- sign and statistical significance of the coefficients associ - sions are associated with a widening of the gender gap: a ated with the cyclical and trend components of GDP are 10-percent increase in per capita GDP is associated with robust across specifications. The absolute value decreases an increase in the gender gap that ranges between 3.2 pp. as we incorporate additional regressors into the model. and 5.8 pp. across specifications. CCTs coverage is statis - For instance, according to the basic model with controls tically significant only in specifications that include the (column 2) a 10-percent increase in the cyclical compo- full set of controls. A 10-pp. increase in CCTs coverage nent of per capita GDP is associated with a fall in female is related to a widening of the gender ratio in labor supply LFP of 2.17 pp., whereas the estimated fall is reduced to of about 1.8 pp. 1.72  pp. when we include male unemployment, CCTs Table 6 reports a range of specifications of models and coverage and the full set of controls (column 8). different estimation methods as robustness checks. For Our estimates suggest that men’s unemployment rate instance, we use pooled OLS, a first-difference estima - may explain part of the countercyclical behavior of female tor, the Arellano and Bond (1991) GMM estimator for a LFP. Indeed, when the model includes that variable, the dynamic model, and a random-effect estimator. We also coefficient associated with the cyclical component falls estimate model (3) adding year fixed effects to the coun - significantly. The partial correlation between male unem - try and group fixed effects. We add controls for educa - ployment and female LFP is always positive and statisti- tional group trends, and additional variables that refer cally significant in the regressions with controls. When to cultural or legal factors, although in these cases the men’s unemployment rate increases 1 pp., women’s labor number of observations falls. We also restrict the sam- supply rises around 0.21 pp. This result is consistent with ple to countries with available data for more than 10 and the added-worker effect and could explain part of the 20 years. In all cases, the main results hold: the counter- negative association between female LFP and the busi- cyclical behavior of female LFP seems a robust result. ness cycle. On the other hand, the coefficients for CCTs Of course, we are not controlling for all possible factors coverage are negative in the models including controls affecting female LFP and then our results could be biased (columns 4 and 8) although small in magnitude and never due to omitted variables. One of these variables could be 24 25 statistically significant. the expansion of pre-school and child-care coverage. Unfortunately we were not able to build such a variable for all countries spanning over the period analyzed in the The tables with the coefficients for all the variables in the regressions are available upon request. In part this result may be due to little variability in the data: we have data on CCTs coverage by country-year and not by group. We are thankful to one referee for this observation. Serrano et al. J Labour Market Res (2019) 53:13 Page 12 of 21 Table 4 Models of female labor force participation (1) (2) (3) (4) (5) (6) (7) (8) Cyclical component − 17.2*** − 21.7*** − 13.1*** − 20.3*** − 13.5** − 18.2*** − 10.1* − 17.2** (5.24) (5.08) (4.24) (5.28) (5.44) (5.99) (4.79) (6.39) Trend component 15.5*** 7.3** 15.8*** 10.0*** 15.3*** 8.6** 15.4*** 11.2*** (2.79) (3.25) (4.17) (2.67) (2.68) (3.27) (3.97) (2.64) CCTs coverage 1.3 − 2.0 1.2 − 1.6 (4.97) (3.22) (4.70) (3.30) Male unemployment 18.4 20.9** 15.5 21.0** (11.03) (9.12) (11.65) (9.72) Additional controls No Yes No Yes No Yes No Yes Observations 2736 2537 2511 2321 2669 2476 2445 2261 R-squared 0.837 0.854 0.831 0.851 0.845 0.858 0.839 0.856 Latin American countries, panel of education and age groups, 1987–2014. Women aged 25–54 Fixed effects (by country and by group) OLS regressions. Unbalanced panel of 9 groups in 18 countries. Dependent variable: female labor force participation as percentage of women aged 25–54. Cyclical and trend components of GDP are obtained by applying the Hodrick-Prescott filter to the log of real per capita GDP. CCTs program coverage as share of population who are beneficiaries (not available for Venezuela). Unemployment rate for men in each education and age group. Additional controls: average years of education, average number of children, share of married women, share of women in charge of old person, average age of children in household, female hourly wage, hourly wage ratio (men/women), service sector value added as share of GDP, rural population as share of total population. For detailed data definitions and sources, see Table 10 in the Appendix. Robust standard errors clustered by country in parentheses. ***Significant at 1% level, **5%, *10% paper. However, we do not expect the inclusion of that 2000s seems to have been largely attributed to improved variable to be a serious threat to our main results. For exports prices for the region. Since Latin American econ- countries for which data is available, the rate of growth omies are small, international prices can be considered in pre-school coverage was either constant or increasing exogenous and then largely unaffected by domestic vari - in the 2000s, and then it can hardly account for a decel- ables such as LFP. eration in female labor force participation. In any case, Table  7 shows the results using the exports price our results also hold for women without children and index for each Latin American country computed by with children older than 6, who were not affected by an the United Nations Conference on Trade and Develop- increase in pre-school and child-care coverage. ment (UNCTAD) as instrumental variable for GDP. In By all means, all the models estimated so far may all specifications, independently of whether we consider potentially suffer from endogeneity bias. Although we are the cycle or the trend as endogenous variables, the esti- aware that we cannot rule out all endogeneity concerns mated coefficient for the cyclical component of per capita with the data at hand, in the next section we carry out an GDP continues to be negative and statistically significant, additional robustness exercise to reinforce the credibility although larger than the fixed effects estimates from sub - of our results. section  6.2. This suggests that, if the instruments were valid, our previous results suffer from attenuation bias, 6.3 Instrumental variables so they could be interpreted as a lower bound of the true With the aim of alleviating the concern for endogene- effect of the economic cycle on female LFP. ity we use instrumental variables and estimate model (3) through two-stage least squares (2SLS). In order to 7 Concluding remarks instrument the cyclical and trend components of the In this paper we explore the relationship between female GDP we follow Besley and Pearson (2008), Brückner and labor force participation and the trend and cycle compo- Ciccone (2010) and McGuirk and Burke (2016), among nents of GDP based on fixed-effect models using panel others, and use exports prices. There is a large literature that discusses the close relationship between exports prices and the cyclical component of economic growth in Latin America (e.g. Erten and Ocampo 2013; Ocampo Prices are constructed using UNCTADstat Commodity Price Statistics, 2017). In particular, the large upswing in GDP over the international and national sources and UNCTAD secretariat estimates. Alternatively, we also used terms of trade and the international prices of the main commodities exported by each country as instruments for GDP In a sample of 6 countries with data from the early 1990s, the share of chil- (UNCTAD data). When using these instruments, we again find a negative dren aged 3 to 5 attending pre-school increased a rate of 1.1 pp per year in the and significant coefficient for the cyclical component of per capita GDP 1990s and 1.2 pp in the 2000s. (results available upon request). Serrano et al. J Labour Market Res (2019) 53:13 Page 13 of 21 Table 5 Models of relative labor force participation (male/female) (1) (2) (3) (4) (5) (6) (7) (8) Cyclical component 49.8*** 58.1*** 41.8** 56.4*** 32.0* 45.4** 26.4 45.0* (13.90) (16.71) (14.45) (18.86) (15.39) (19.01) (17.05) (21.81) Trend component − 53.2*** − 17.2 − 58.5*** − 28.0* − 56.7*** − 19.3 − 62.2*** − 30.6** (10.63) (14.81) (15.35) (14.27) (11.27) (14.59) (15.79) (12.90) CCTs coverage 9.9 17.3* 12.0 19.2* (19.36) (8.54) (19.53) (9.36) Male unemployment − 99.5** − 74.8** − 93.9* − 77.9** (42.65) (27.72) (44.98) (28.55) Additional controls No Yes No Yes No Yes No Yes Observations 2736 2537 2511 2321 2669 2476 2445 2261 R-squared 0.719 0.753 0.714 0.750 0.728 0.761 0.723 0.759 Latin American countries, panel of education and age groups, 1987–2014. Adults aged 25–54 Fixed effects (by country and by group) OLS regressions. Unbalanced panel of 9 groups in 18 countries. Dependent variable: labor force participation ratio (men/ women) expressed in percentage. See notes to Table 4 for more details. Robust standard errors clustered by country in parentheses. ***Significant at 1% level, **5%, *10% Table 6 Models of female labor force participation Dependent variable Obs. Female LFP (%) LFP ratio (men/women) Cyclical comp. Trend comp. Cyclical comp. Trend comp. (1) Base model − 17.2*** 15.5*** 49.8*** − 53.2*** 2736 (5.24) (2.79) (13.90) (10.63) (2) Pooled OLS − 14.1* − 0.6 35.3 2.0 2736 (7.46) (0.53) (22.12) (1.56) (3) Differences − 8.4** − 2.6 19.7* 12.4 2052 (3.70) (4.27) (10.92) (12.70) (4) Dynamic panel model/1 − 5.7*** 0.0 24.8*** − 0.0 1773 (2.16) (0.11) (6.61) (0.31) (5) Random Eec ff ts − 13.4*** 6.1*** 33.0*** − 11.8*** 2736 (2.60) (0.63) (7.73) (1.70) (6) Year fixed effects − 14.0** 8.2* 39.2** − 32.5* 2736 (6.35) (4.38) (17.86) (15.68) (7) Year fixed effects + interaction − 14.0** 8.2* 39.2** − 32.5* 2736 year and educational groups (6.42) (4.42) (18.04) (15.84) (8) Additional controls 1 − 21.7*** 7.3** 58.1*** − 17.2 2537 (5.08) (3.25) (16.71) (14.81) (9) Additional controls 2 − 19.9*** 8.7 42.5*** − 21.8 936 (4.81) (5.23) (11.21) (15.66) (10) Only country > 20 periods − 17.5** 19.2*** 55.7** − 64.9** 1224 (4.97) (4.50) (18.17) (18.23) (11) Only country > 10 periods − 17.7*** 15.9*** 50.7*** − 54.2*** 2646 (5.28) (2.82) (14.13) (10.80) Latin American countries, panel of education and age groups, 1987–2014. Adults aged 25–54 (1) Base model specification is the same presented in column 1 of Table 4 for female LFP and Table 5 for the LFP ratio (men/women). (2) Pooled OLS regression with standard errors clustered by country in parentheses. (3) First difference estimator with standard errors clustered by country in parentheses. (4) Arellano and Bond (1991) GMM estimator with robust standard errors; we instrument for cycle component using a double lag. (5) Random-effect estimator with robust standard errors in parentheses. (6) Base model adding year fixed effects with standard errors clustered by country in parentheses. (7) Specification (6) + interactions between education groups and year dummies. (8) Base model specification with control variables; these results are the same presented in column 2 of Table 4 for female LFP and Table 5 for the LFP ratio. (9) Specification (8) + additional controls such as percentage of married women using modern contraceptive methods, an indicator of legal abortion, percentage of women with a washing machine, percentage of non-practicing catholic and an index of labor market regulations. (9) Base model specification restricting the sample to countries with available data for 20 years or more. (11) Base model specification restricting the sample to countries with available data for 10 years or more. ***Significant at 1% level, **5%, *10% Serrano et al. J Labour Market Res (2019) 53:13 Page 14 of 21 Table 7 Models of female labor force participation FE 2SLS Cyclical component of GDP Cyclical and trend endogenous components of GDP endogenous (1) (2) (3) (4) (5) (6) Cyclical component of GDP per capita − 17.2*** − 21.71*** − 36.5*** − 28.7** − 34.1** − 32.4*** (5.24) (5.08) (11.36) (11.52) (12.67) (10.88) Trend component of GDP per capita 15.5*** 7.32** 15.9*** 7.9** 16.6*** 11.8* (2.79) (3.25) (2.79) (3.35) (2.55) (5.98) First stage for the cyclical component of GDP per capita Exports prices—cycle 0.170*** 0.178*** 0.168*** 0.180*** (0.037) (0.040) (0.038) (0.040) Exports prices—trend − 0.047*** − 0.022*** − 0.020*** − 0.010 (0.013) (0.007) (0.007) (0.008) F test 12.97 19.21 9.903 11.59 Overid. test (p value) 0.481 0.372 Additional controls No Yes No Yes No Yes Observations 2736 2537 2736 2537 2736 2537 Fixed effects OLS and instrumental variables 2SLS. Latin American countries, panel of education and age groups, 1987–2014. Women aged 25–54 Robust standard errors in parentheses ***p < 0.01, **p < 0.05, *p < 0.1 Columns 1 and 2 report the fixed effects (by country and by group) OLS estimates from columns 1 and 2 of Table 4. Columns 3 to 6 report the fixed effects (by country and by group) 2SLS regression results. Unbalanced panel of 9 groups in 18 countries. Dependent variable: female labor force participation as percentage of women aged 25–54. Instrument variables: cyclical and trend components of the UNCTAD´s exports price index. Over-identification test refers to the Hansen J statistic, where the joint null hypothesis is that the instruments are valid. See notes to Table 4 for more details. Robust standard errors clustered by country in parentheses. ***Significant at 1% level, **5%, *10% data from harmonized national household surveys for imply a loss of productivity, making women less likely to all Latin American countries. We find that female LFP work in the future, regardless of the macroeconomic con- is positively related to the trend component of per cap- ditions. Furthermore, it could mean a strengthening of ita GDP—long-term effect—and negatively related to the traditional gender roles in the household, negatively the cyclical component—mostly related to short-term affecting the perspectives of women to participate in the shocks. This latter link is stronger for vulnerable women, labor force in the long term. with low educational attainment, married, with young Acknowledgements children, and in low-income households, which is con- This paper is based on the research that the authors carried out within the sistent with an inverse added-worker effect. We believe project “Medición de las diferencias de género en las habilidades, las limita- ciones de la familia y las preferencias de carrera”, Gender and Diversity Division, these results may shed light on an intriguing fact: the Inter-American Development Bank. It is also a follow up of Joaquín Serrano’s significant deceleration in female LFP in Latin America dissertation at the Master’s Program in Economics at Universidad Nacional de in the 2000s, a decade of exceptionally high economic La Plata, in turn based on evidence from a recent book edited by two of the authors (Gasparini and Marchionni, 2015) with the support of IDRC-Canada. growth. We are grateful to Andrew Morrison, Monserrat Bustelo, Claudia Piras, Luana Our results have nuanced implications in terms of well- Ozemela, Guillermo Cruces, Matías Busso, Carlos Lamarche, Lorena Garegnani, being. On the one hand, the deceleration in female LFP Inés Berniell, seminar participants at AAEP (2016), Network of Inequality and Poverty (UNGS, 2014), IDB ( Washington-DC, 2015), LACEA (2017), and to the may reflect the fact that in a more favorable economic editor of the journal and two anonymous referees for valuable comments and context some women are no longer bound to enter the suggestions. labor market to take precarious low-quality jobs. How- ever, as suggested by Gasparini and Marchionni (2017), staying out of the labor market during some time could Serrano et al. J Labour Market Res (2019) 53:13 Page 15 of 21 Authors’ contributions Competing interests All authors actively participated in the design and implementation of the The authors declare that they have no competing interests. study. All authors conceived and designed the analysis; JS collected the data; MM, LG and PG contributed data and analysis tools; JS performed the Appendix statistical analysis. All authors wrote the final manuscript. All authors read and approved the final manuscript. See Tables 8, 9, 10 11 and Figs. 5, 6. Funding Not applicable. Availability of data and materials The raw data corresponds to national household surveys, whose microdata sets are publicly available. The specific processed datasets used in the study are available from the corresponding author upon request. Table 8 National household surveys used in this study Country Survey name Acronym Surveys used Argentina Encuesta Permanente de Hogares Puntual EPH 1992–2003 Encuesta Permanente de Hogares Contínua EPH-C 2003–2014 Bolivia Encuesta Integrada de Hogares EIH 1992, 1993 Encuesta Nacional de Empleo ENE 1997 Encuesta Contínua de Hogares ECH 1999, 2000 Encuesta de Hogares EH 2001, 2002, 2005, 2007–2009, 2011–2013 Brazil Pesquisa Nacional por Amostra de Domicilios PNAD 1988–1993, 1995–1999, 2001–2009, 2011–2014 Chile Encuesta de Caracterización Socioeconómica Nacional CASEN 1987, 1990, 1992, 1994, 1996, 1998, 2000, 2003, 2006, 2009, 2011, 2013 Colombia Encuesta Continua de Hogares ECH 2001–2005 Gran Encuesta Integrada de Hogares GEIH 2008–2014 Costa Rica Encuesta de Hogares de Propósitos Múltiples EHPM 1989–2009 Encuesta nacional de hogares ENAHO 2010, 2012–2014 Dominican Rep. Encuesta Nacional de Fuerza de Trabajo ENFT 2000–2014 Ecuador Encuesta de Condiciones de Vida ECV 1994, 1995, 1998, 1999, Encuesta Nacional de Empleo, Desempleo y Subempleo ENEMDU 2003–2014 El Salvador Encuesta de Hogares de Propósitos Múltiples EHPM 1991, 1995, 1996, 1998–2014 Guatemala Encuesta Nacional sobre Condiciones de Vida ENCOVI 2000, 2006, 2011 Honduras Encuesta Permanente de Hogares de Propósitos Múltiples EPHPM 1991–1999, 2001–2013 Mexico Encuesta Nacional de Ingresos y Gastos de los Hogares ENIGH 1989, 1992, 1994, 1996, 1998, 2000, 2002, 2004–2006, 2008, 2010, 2012, 2014 Nicaragua Encuesta Nacional de Hogares sobre EMNV 1993, 1998, 2001, 2005, 2009 Medición de Nivel de Vida Panama Encuesta de Hogares, Mano de Obra EMO 1989, 1991 Encuesta de Hogares EH 1995, 1997–2012 Paraguay Encuesta de Hogares (Mano de Obra) EH 1990 Encuesta Integrada de Hogares EIH 1997, 2001 Encuesta Permanente de Hogares EPH 1999, 2002–2014 Peru Encuesta Nacional de Hogares ENAHO 1997–2014 Uruguay Encuesta Continua de Hogares ECH 1989, 1992, 1995–1998, 2000–2014 Venezuela Encuesta de Hogares Por Muestreo EHM 1989, 1992, 1995, 1997–2011 Source: own elaboration Serrano et al. J Labour Market Res (2019) 53:13 Page 16 of 21 Table 9 Composition of the panel datasets used in this study argbol brachl colcri domecu slvgtm hndmex nicpan pryper uryven 1992 x~ x~ ~x 1993 ~x ~~ ~~ x~ ~ 1994 ~~ ~~ ~~ x~ ~ 1995 ~~ ~~ ~x 1996 ~~ ~~~x ~ 1997 ~~ ~~ ~ 1998 ~~ 1999 ~~ ~~ 2000 ~~ ~~ ~ 2001 ~~ ~ 2002 ~~ ~ 2003 ~~ ~~ 2004 ~~ ~ 2005 ~ 2006 ~ 2007 ~~ ~~ 2008 ~~ 2009 ~ 2010 ~~ ~~ 2011 ~~ 2012 ~~ x 2013 ~~ x 2014 x x The shaded cells correspond to the available surveys, which constitute the unbalanced panel that we use in the econometric estimates. The cells marked with the ~ and x signs are interpolated and extrapolated data, respectively, used to compute the descriptive statistics for the Latin American average in Sect. 4. Note that the Latin American average excludes Colombia, Dominican Republic, and Guatemala Serrano et al. J Labour Market Res (2019) 53:13 Page 17 of 21 Table 10 Description and sources of variables used in this study Variable Description Source Female labor force participation Female labor force participation as percentage of National household surveys women aged 25–54 Male labor force participation Male labor force participation as percentage of National household surveys men aged 25–54 Labor force participation ratio (men/women) LFP ratio (men/women) expressed in percentage National household surveys Log GDP per capita Real gross domestic product per capita (log) WDI Database, World Bank CCTs coverage Share of total population who are beneficiaries of CCTs beneficiaries: own elaboration based on data conditional income transfer programs from ECLAC. Population: WDI Database, World Bank Male unemployment Unemployment rate for men aged 25–54. National household surveys Years of education Average of years of education (log) National household surveys Number of children Average number of children (log) National household surveys Married women Share of married women aged 25–54 National household surveys Women in charge of old person Share of women who are in charge of old persons National household surveys (> 70 years old) Age of children Average age of children in household National household surveys Hourly wage gap (men/women) Gender ratio of average hourly labor income of the National household surveys main job, PPP adjusted. (men/women) Female hourly wage Hourly labor income of the main job of women, National household surveys PPP adjusted Service sector (value added) Value added of service sector (share of GDP) WDI Database, World Bank Rural population Rural population as share of total population WDI Database, World Bank Index of exports prices Value index of exports (f.o.b.) converted to U.S. UNCTAD dollars and expressed as a percentage of the average for the base period (2010) Table 11 Mean of main variables LFP Employment Unemployment GDP (log) CCTs coverage Service sector Rural pop. (value added) Women Men Women Men Women Men Argentina 67.1 94.1 62.7 89.6 6.6 4.8 16.0 27.4 62.9 8.4 Bolivia 73.8 97.0 71.0 94.9 3.7 2.2 4.8 26.8 50.2 31.9 Brazil 71.2 92.4 66.4 89.0 6.8 3.7 12.4 26.5 70.8 14.6 Chile 64.3 92.3 60.2 87.6 6.3 5.1 16.4 4.1 61.7 10.8 Colombia 72.9 96.1 66.2 91.1 9.2 5.2 9.5 10.1 58.0 23.8 Costa Rica 62.2 94.9 56.9 89.6 8.4 5.6 11.0 3.6 69.4 24.1 Dominican Rep. 57.8 90.3 55.1 88.4 4.6 2.1 8.7 23.4 66.9 21.9 Ecuador 63.3 96.7 60.8 94.5 4.0 2.3 8.5 17.4 51.8 36.5 El Salvador 61.3 92.8 59.8 89.3 2.5 3.8 6.7 2.8 61.9 33.7 Guatemala 50.1 96.5 49.4 95.3 1.4 1.3 6.2 24.7 59.6 48.9 Honduras 57.6 94.8 54.1 91.3 6.0 3.7 3.9 44.2 59.8 45.9 Mexico 58.5 96.2 57.0 91.8 2.5 4.5 14.7 22.5 62.2 21.0 Nicaragua 62.7 95.5 59.3 91.7 5.4 4.0 3.6 2.6 54.2 41.5 Panama 66.4 96.6 63.3 94.2 4.6 2.5 12.6 3.3 69.6 33.7 Paraguay 69.5 95.4 65.6 92.4 5.6 3.1 6.6 8.4 50.6 40.6 Peru 79.4 95.3 77.6 93.7 2.2 1.6 7.8 10.5 57.6 21.7 Uruguay 79.6 96.0 74.5 93.0 6.4 3.1 14.0 7.9 64.2 4.8 Venezuela 68.0 95.4 63.0 90.0 7.4 5.7 15.6 42.1 11.2 Latin America 65.9 94.9 62.4 91.5 5.2 3.6 9.9 15.6 59.6 26.4 The table shows averages over the period available for each country. Labor force participation (LFP) and employment as percentage of adults aged 25–54. Unemployment rate for adults aged 25–54. Real GDP in logs. CCTs coverage as the percentage of total population who are beneficiaries. Value added of Service Sector as percentage of GDP. Rural population as percentage of total population Serrano et al. J Labour Market Res (2019) 53:13 Page 18 of 21 Fig. 5 Female and male labor force participation. Latin American countries. Source: own calculations based on microdata from national household surveys. Adults aged 25–54 Serrano et al. J Labour Market Res (2019) 53:13 Page 19 of 21 Fig. 6 Per capita GDP. Latin American countries. Source: own calculations based on data from WDI Database. Note: per capita GDP in thousands of PPP-adjusted 2011 US$ Serrano et al. J Labour Market Res (2019) 53:13 Page 20 of 21 Author details Erten, B., Ocampo, J.A.: Super cycles of commodity prices since the mid-nine- Centro de Estudios Distributivos, Laborales y Sociales (CEDLAS), IIE-FCE, teenth century. World Dev. 44, 14–30 (2013) Universidad Nacional de La Plata, and CONICET, La Plata, Argentina. Depar- Fernandes, R., Felicio, F.: The entry of the wife into the labor force in response tamento de Economía, FCE-UNLP, Oficina 322, Calle 6 No 777, 1900 La Plata, to the husband’s unemployment: a study of the added worker effect Argentina. in Brazilian metropolitan areas. Econ. Dev. Cult. Change 53(4), 887–911 (2005) Received: 29 August 2018 Accepted: 21 September 2019 Garcia-Perez, J.I., Rendon, S.,: Family job search and wealth: the added worker effect revisited, working papers 16–34, Federal Reserve Bank of Philadel- phia (2016) Garganta, S., Gasparini, L., Marchionni, M.: Cash transfers and female labor force participation: the case of AUH in Argentina. IZA J. Labor Policy 6, 10 (2017). https ://doi.org/10.1186/s4017 3-017-0089-x References Gasparini, L., Marchionni, M.: Deceleration in female labor force participation in Arellano, M., Bond, S.: Some tests of specification for panel data: Monte Carlo Latin America. Econ.-LACEA J. 18(1), 197–224 (2017) evidence and an application to employment equations. Rev. Econ. Stud. Gasparini, L., Marchionni, M. (eds.): Bridging gender gaps? The rise and decel- 58, 277–297 (1991) eration of female labor force participation in Latin America. Center for Ashenfelter, O.: Unemployment as disequilibrium in a model of aggregate Distributive, Labor and Social Studies, La Plata, Buenos Aires, Argentina, 1 labor supply. Econometrica 48(3), 547–564 (1980) edition. https ://goo.gl/yeEQZ M (2015) Attanasio, O., Levell, P., Low, H., Sánchez-Marcos, V.: Aggregating elasticities: Gasparini, L., Marchionni, M., Badaracco, N., Serrano, J.: Characterizing female intensive and extensive margins of female labour supply. NBER Working participation changes. In: Gasparini, L., Marchionni, M., (eds.) Bridging Papers 21315, National Bureau of Economic Research, Inc (2015) gender gaps? The rise and deceleration of female labor force participa- Basu, K., Genicot, G., Stiglitz, J.E.: Unemployment and wage rigidity when labor tion in Latin America, vol. 1, chapter 4, pp. 199–260. Center for Distribu- supply is a household decision. Working papers 00–10, Cornell University, tive, Labor and Social Studies, 1st edition (2015) Center for Analytic Economics (2000) Gómez, V.: The use of butterworth filters for trend and cycle estimation in Baxter, M., King, R.G.: Measuring business cycles: approximate band-pass filters economic time series. J. Bus. Econ. Stat. 19(3), 365–373 (2001) for economic time series. Review of economics and statistics, MIT Press, Heckman, J.J., Macurdy, T.E.: A life cycle model of female labour supply. Rev. 81(4) (1999) Econ. Stud. 47(1), 47–74 (1980) Beccaria, L., Maurizio, R., Vázquez, G.: El estancamiento de la tasa de partici- Hodrick, R., Prescott, E.: Postwar U.S. business cycles: an empirical investigation. pación económica femenina en Argentina en los 2000 s. Mimeo (2015) J. Money Credit Bank. 29(1), 1–16 (1997) Besley, T.J., Persson, T.: The incidence of civil war: theory and evidence. Working Kimmel, J.: Childcare costs as a barrier to employment for single and married paper 14585, National Bureau of Economic Research (2008) mothers. Rev. Econ. Stat. 80(2), 287–299 (1998) Bhalotra, S.R., Umaña-Aponte, M.: The dynamics of women’s labour supply in Layard, R., Barton, M., Zabalza, A.: Married women’s participation and hours. developing countries. IZA discussion papers 4879, Institute for the Study Economica 47(185), 51–72 (1980) of Labor (IZA) (2010) Lee, G.H., Parasnis, J.: Discouraged workers in developed countries and added Bredtmann, J., Otten, S., Rulff, C.: Husband’s unemployment and wife’s labor workers in developing countries? Unemployment rate and labour force supply: the added worker effect across Europe. ILR Review, pp. 1–31 participation. Econ. Model. 41(C), 90–98 (2014) (2017) Lundberg, S.: The added worker effect. J. Labor Econ. 3(1), 11–37 (1985) Brückner, M., Ciccone, A.: International commodity prices, growth and the out- Maloney, T.: Employment constraints and the labor supply of married women: break of civil war in sub-saharan Africa. Econ. J. 120(544), 519–534 (2010) a reexamination of the added worker effect. J. Hum. Resour. 22(1), 51–61 Busso, M., Romero Fonseca, D.A.: Determinants of female labor force participa- (1987) tion. In: Gasparini, L., and Marchionni, M., editors, Bridging gender gaps? Maloney, T.: Unobserved variables and the elusive added worker effect. Eco - The rise and deceleration of female labor force participation in Latin nomica 58(230), 173–187 (1991) America, volume 1, chapter 6, pp. 199–260. Center for Distributive, Labor Maloney, W.F.: Informality revisited. World Dev. 32(7), 1159–1178 (2004) and Social Studies, 1st edition (2015) Mankart, J., Oikonomou, R.: Household search and the aggregate labour Cerrutti, M.: Economic reform, structural adjustment and female labor force market. Rev. Econ. Stud. 84(4), 1735–1788 (2017) participation in Buenos Aires, Argentina. World Dev. 28(5), 879–891 Martinoty, L.: Intra-household coping mechanisms in hard times: the added (2000) worker effect in the 2001 Argentine Economic Crisis. Working paper, Chioda, L.: Work and family: Latin American & Caribbean Women in Search of a working paper GATE 2015-05 (2015) New Balance. World Bank, Washington, DC (2011) McGuirk, E., Burke, M.: Economic shocks and varieties of conflict: global prices, Christiano, L.J., Fitzgerald, T.J.: The band pass filter. Int. Econ. Rev. 44(2), real income and local violence in Africa. Miemo. SSRN: https ://ssrn.com/ 435–465 (2003) abstr act=27762 63 (2016) Cox Edwards, A., Roberts, J.: The effects of structural adjustment on women McKenzie, D.J.: How do households cope with aggregate shocks? Evidence in Latin America. In: Horton, S., Kanbur, R., Mazumdar, D., (eds.) Labor from the Mexican Peso crisis. World Dev. 31(7), 1179–1199 (2003) markets in an era of adjustment, vol. 1. Washington, DC: World Bank, 1st McKenzie, D.J.: Aggregate shocks and urban labor market responses: evi- edition (1994) dence from Argentina’s financial crisis. Econ. Dev. Cult. Change 52(4), Cullen, J.B., Gruber, J.: Does unemployment insurance crowd out spousal labor 719–758 (2004) supply? J. Labor Econ. 18(3), 546–572 (2000) Michalopoulos, C., Robins, P., Garfinkel, I.: A structural model of labor supply Darby, J., Hart, R., Vecchi, M. (2001), Labour force participation and the business and childcare demand. J. Hum. Resour. 27(1), 166–203 (1992) cycle: a comparative analysis of France, Japan, Sweden and the United Naz, G.: The impact of cash-benefit reform on parents’ labour force partici- States, Japan and the World Economy, 13(2), pp. 113–133. Dell, M., pation. J. Popul. Econ. 17(2), 369–383 (2004) Jones, B. F., and Olken, B. A: Temperature Shocks and Economic Growth: Ocampo, J.A.: Commodity-led development in Latin America. In: Alternative evidence from the Last Half Century. American Economic Journal: Macro- pathways to sustainable development: lessons from Latin America, economics 4(3), 66–95 (2012) International Development Policy series No. 9 (Geneva, Boston: Gradu- Eissa, N., Hoynes, H.: Taxes and the labor market participation of married cou- ate Institute Publications, Brill-Nijhoff ), pp. 51–76 (2017) ples: the earned income tax credit. Journal of Public Economics, Elsevier Parada, C., Alejo, J., Gasparini, L.: Distributive implications of changing 88(9–10), 1931–1958 (2004) female employment. Revista Cuadernos de Economía, in press (2017) Elborgh-Woytek, K., Newiak, M., Kochhar, K., Fabrizio, S., Kpodar, K., Wingender, Parker, S., Skoufias, E.: The added worker effect over the business cycle: P., Clements, B.J., Schwartz, G.: Women, work, and the economy; macro- evidence from urban Mexico. Appl. Econ. Lett. 11(10), 625–630 (2004) economic gains from gender equity, IMF Staff Discussion Notes 13/10, Paz, J.A.: El Efecto del Trabajador Adicional: evidencia para Argentina International Monetary Fund (2013) (2003–2007). Cuadernos de economía 46, 225–241 (2009) Serrano et al. J Labour Market Res (2019) 53:13 Page 21 of 21 Peña, X., Bernal, R., Amador, D.: The rise in female participation in Colombia: Tsani, S., Paroussos, L., Fragiadakis, C., Charalambidis, I., Capros, P.: Female fertility, marital status or education? Ensayos sobre política económica labour force participation and economic growth in the South Mediter- 31(71), 54–63 (2013) ranean countries. Econ. Lett. 120(2), 323–328 (2013) Prieto-Rodriguez, J., Rodriguez-Gutierrez, C.: Participation of married World Bank.: The effects of women’s economic power in Latin America and the women in the European labor markets and the “added worker effect”. J. Caribbean. Technical Report 21315, Poverty and Labor Brief—LAC region Socio-Econ. 32(4), 429–446 (2003) (2012) SEDLAC.: Socioeconomic Database for Latin America and the Caribbean. Woytinsky, W.S.: Additional workers and the volume of unemployment in Technical report, CEDLAS-World Bank (2014) the depression. Pamphlet series 1, Washington: Social Science Research Stampini, M, Tornarolli, L. The growth of conditional cash transfers in Latin Council, Committee on Social Security (1940) America and the Caribbean: did they go too far? IZA Policy Papers 49, Institute for the Study of Labor (IZA) (2012) Publisher’s Note Stephens, M.: Worker displacement and the added worker effect. J. Labor Springer Nature remains neutral with regard to jurisdictional claims in pub- Econ. 20(3), 504–537 (2002) lished maps and institutional affiliations. Tachibanaki, T., Sakurai, K.: Labour supply and unemployment in Japan. Eur. Econ. Rev. 35(8), 1575–1587 (1991) Tamm, M.: Child benefit reform and labor market participation, p. 97. Ruhr Econ. Pap., No (2009)

Journal

Journal for Labour Market ResearchSpringer Journals

Published: Oct 16, 2019

There are no references for this article.