Introduction

Female Labor Force Participation (FLFP) is a large and complex phenomenon that still needs further investigation due to its key importance to social and economic development. As the literature concurs (e.g., ^{Chioda, 2016}; ^{Klasen, 2018}; ^{Novta & Wong, 2017}), in the last 25 years the Latin American region experienced the largest FLFP growth than any other region in the world, nevertheless, the studies on its determinants are still incipient with many opportunities for contribution.

In the last hundred years, the increase of female labor participation rate (FLFP) took place first in industrialized countries in the decades following World War II, rising up to 60% (^{Acemoglu, Autor, & Lyle, 2004}; ^{Fogli & Veldkamp, 2011}; ^{Goldin, 1991}). As for Latin-American & Caribbean (LAC) countries, FLFP experienced a strong increase starting at the end of the Twentieth Century, growing from a 36.6% in 1990 up to 51.4% in 2014, adding 63.3 million women to the labor market (^{The World Bank, 2018}). Novta (2017, p. 2) claims that the additions in FLFP in the region where the largest in the world in the period referred before, and “was driven by improvements beyond development levels” (2017, p. 9).

In the last two decades LAC started to shine in the world economic arena. As of 2014 LAC countries’ GDP represented 8.01%¹ of the World’s GDP; Mexico, Argentina, and Brazil were members of the G-20; and Mexico, Colombia, and Chile had become adherents of the OECD. This economic growth in the region had an impact on FLFP. Against the expected U-shape pattern (^{Goldin, 1994}; ^{Mammen & Paxson, 2000}; ^{Olivetti, 2013}; ^{Pampel & Tanaka, 1986}; ^{Psacharopoulos & Tzannatos, 1989}; ^{Tansel, 2002}) between FLFP and economic growth proxied using GDP per cápita, studies concur (^{Camou, 2015}; ^{Demirhan & Demirhan, 2017}; ^{Klasen, 2018}) that Latin American countries do not follow a U-shape.

Authors (^{Camou, 2015}; ^{Chioda, 2016}; ^{Gasparini & Marchionni, 2015}; ^{Klasen, 2018}; ^{Serrano et al., 2018}) agree that economic development, education, lower fertility rates, and economic policies such as childcare support, to be determinants of FLFP in LAC as U-shape theory studies found, but they do not consider them to be sufficient explanation for the countries in the region. Studies (^{Camou, 2015}; ^{Chioda, 2016}; ^{Gasparini, Marchionni, Badaracco, & Serrano, 2015}; ^{Klasen, 2018}) coincide in the need of further investigation of the determinants of FLFP in LAC, and point to women preferences and social factors as opportunities of research.

^{Serrano et al. (2018)} caution FLFP in LAC is reaching a ceiling that may result from women preferences and policy factors, as women delay their decision to participate in the labor market given the higher earnings of their spouses or the protection of new social programs.

Women’s role preferences and identity develop during their adolescence years (^{Burt & Scott, 2002}; ^{Vella, 1995}) influenced by their family structure (^{Kiecolt & Acock, 1988}; ^{Vincent, Peplau, & Hill, 1998}), the transmission of family values (^{Crespi, 2004}), and education and religious beliefs (^{Filler & Jennings, 2015}). Then, in adulthood, women´s labor outcomes are impacted by their choices according to the identity they recognize with: traditional - housewife, or modern working woman (^{Hayo & Caris, 2013}).

^{Gerson (1985)} claims that although women preferences may be forged during adolescent years, these can change during adulthood as result of new social forces and dynamics in their social milieu which may provide a new set of possibilities, resulting in different paths that two women may follow under similar circumstances. Also, ^{Cunningham (2008)} concluded that social and economic factors in adulthood may influence women’s decision to participate in the labor market, despite traditional values upbringing.

According to ^{Inglehart & Welzel (2005)}, socioeconomic development and the emergence of a knowledge society act as a force upon culture changing traditional values and beliefs, giving rise not only to more democratic institutions but also rising gender equality. As the authors assert, as result of shaping emancipative values, the orientation of gender roles changes, promoting “female empowerment and egalitarian democracy” (2005, p. 284).

In this sense, recent studies concluded that women’s identity (^{Fortin, 2009}; ^{Hayo & Caris, 2013}), emancipative values (^{Brieger, Francoeur, Welzel, & Ben-Amar, 2017}; ^{Inglehart & Welzel, 2005}) and gender egalitarianism (^{Cunningham, 2008}) affect FLFP.

^{Akerlof & Kranton’s (2000)} proposed the Identity Economic Theory (IET), where identity is defined as “a sense of self”, and claim that the choice of identity an individual makes is the most important economic decision. The IET incorporates a person´s identity, prescribed social norms, and social categories into an economic utility function that the individual seeks to maximize taking these factors into account (2000, p. 719). As a society eases the association of tasks to gender roles, we should expect an increase of FLFP, reducing the gap with men’s job rates (2000, p. 735).

Changes in prescribed social norms acting as constraints around expected gender roles, are captured through the Theory of Emancipative Values formulated by ^{Welzel (2013b)}. The theory states that as external constraints over the individual recede, emancipative values arise as a motivational source of human empowerment, emphasizing freedom of choice and equality of opportunities. A value is a belief that serves as a goal, orients behavior, and helps to evaluate situations, people, and events (^{Schwartz, 1994, p. 20}).

Since women preferences and social factors constitute opportunities of research for the Latin America region, this work seeks to contribute to the literature by analyzing the effect of identity, as a measure of women’s preferences, and emancipative values, as a reflection of social factors, upon FLFP. This investigation hypothesizes that these two explicative variables have been two strong drivers for women’s labor market participation over demographics and country economics.

Also, this study explores the effect of these explicative variables on OECD countries, as this group encompasses some of the most advanced and democratic economies in the world, and compares them with the LAC region, foreseeing a road of opportunity ahead in the region from which we may expect FLFP will continue to rise in the future, as result, mainly, from the augmentation of emancipative values.

This investigation is supported on the theoretical frameworks of the IET (G. ^{Akerlof & Kranton, 2000}) and the Emancipative Values Theory (^{Welzel, 2013b}). Employs pooled data from waves 3 to 6 (1994-1998 to 2010-2014) of the World Values Survey (WVS), applying a Probit regression; controlling for individual’s demographics, countries’ economic indicators proxied by GDP per cápita and male unemployment rate, country code, and survey wave.

This work is organized as follows. Section 2 presents an overview on the literature of the determinants of FLFP and the most recent studies of the subject in the LAC region. Section 3 describes the empirical strategy. Section 4 presents the empirical results and the analysis of the effect of the explicative variables on FLFP outcomes, contrasting the LAC region with OECD countries. Section 5 discusses the findings and states the final conclusion and suggestions for future researches.

Literature Review

Neoclassical theory states that supply and demand are forces at play in any given market (^{Marshall, 1890}). In the labor market, demand is in function of business/industry structure requiring skill dependent labor force at certain cost payable as wage; supply, is in function of labor force availability, where individuals, upon a utility curve, rationally decide to trade their leisure time in exchange of monetary pay (^{Boyer & Smith, 2000, pp. 200-201}). However, research has concluded that women’s job engagement is more complex than men’s because it is impacted by more factors as much at the micro as at the macro level.

For almost 60 years, since ^{Mincer’s (1962)} seminal work on FLFP and ^{Becker’s (1965)} Time Assignation Theory, studies on determinants on women’s’ job market participation from the supply perspective sprouted along three major categories: neoclassical labor economics; economic development; and social economics.

Theory states that unlike men who only have to choose their time between work and leisure, married women considers family chores as a third element of time usage, and their economic decision to trade their time is inelastic in function of family income, and elastic in respect to their child-caring duties (^{Mincer, 1962}). Families behave as a decision units, therefore women’s entrance into the labor market is pondered in the domestic nucleus-basically between wife and husband- choosing how to distribute the time of its members, so that those who are more efficient in market activities spend less time on consumer activities such as domestic tasks (^{Becker, 1965}).

^{Becker (1974)} conceptualizes marriage as an economic decision between a man and a woman to unite their lives as a couple, where the main gain is the breeding of their own children, and in order to maximize the outcome-raising healthy offspring-both cooperate allocating their time into market and non-market activities according to their best competencies. This result in women working less time, or not at all, in the labor market as long as the potential wage she could earn is less than their husband’s. Furthermore, ^{Becker (1991, pp. 30-51)} states that within the family, human capital investment is allocated according to sex biological advantages where market returns are more profitable; and since a woman is considered to be biologically best suited for household chores, a family rationally allocate more resources to boys schooling than girls, since men need to be better prepared for market production.

Likewise, since child rearing and domestic chores are much more energy demanding, women choose partial time and less effort intensive jobs, hence dedicating less time to human capital formation which results in a lesser income; while husbands dedicate their energy into full time jobs to maximize household income (^{Becker, 1985}).

Departing from these theoretical grounds, and from Western developed countries experience, scientists developed micro level economic models to understand the impact on the elasticity on women’s decisions, their family situations, fertility, education, technology advances, experience, wages, and public policies; having as central premise that domestic tasks are women’s role (^{Blundell & MaCurdy, 1998}; ^{Killingsworth & Heckman, 1986}).

From these models, researchers conceived empirical studies that have led to parsimonious explanations and understanding on the determinants that impact FLFP. For example, factors that have been linked as facilitators of women’s labor participation are: the contraceptive pill (^{Bailey, 2006}), home appliances (^{Cavalcanti & Tavares, 2008}; ^{Greenwood, Seshadri, & Yorukoglu, 2005}), and public policies such as family income taxes and childcare support (^{Attanasio, Low, & Sánchez-Marcos, 2008}; ^{Bick, 2011}; ^{Bishop, Heim, & Mihaly, 2009}; ^{Haan & Wrohlich, 2009}; ^{Jaumotte, 2003}). Life cycle and the number and age children point to be a strong factor (^{Bloom, Canning, Fink, & Finlay, 2009}; ^{Hotz & Miller, 1988}; ^{Mishra & Smyth, 2010}), although some studies don’t seem to agree to be a strong enough determinant (^{Agüero & Marks, 2008}; ^{Givord & Marbot, 2014}).

Taking a macro stand point of view, a stream of research grouped under economic development the structural and socio-economic changes that takes place in a country and impact FLFP, being ^{Esther Boserup’s (1970)} theory as the go-to reference in FLFP literature. ^{Boserup (1970)} claims that females’ job market involvement is impacted by a society’s² economic and technological advances. In an agricultural/pre-industrial low-income economy, women actively work to help on household income; in an industrial and technology advanced market, with better pays, men takes the bread-winner role and women assumes domestic duties; as economy further develops, women education expands, and re-enter the labor market that demands clerical and white-collar jobs as the service industry grows.

Based on ^{Boserup’s (1970)} theory, ^{Pampel & Tanaka (1986)} retake the U-shape theory, first postulated by ^{Sinha (1965)}, which states that income effect rules on low economic developed countries, and a substitution effect takes places on higher incomes. Studies followed to confirm the U-shape theory (^{Belke, 2016}; ^{Demirhan & Demirhan, 2017}; ^{Goldin, 1994}; ^{Mammen & Paxson, 2000}; ^{Psacharopoulos & Tzannatos, 1989}), although some have found that the theory doesn’t necessarily apply to low income (^{Lechman & Kaur, 2015}) and Latin American countries (^{Demirhan & Demirhan, 2017}), or even that the theory has little empirical support (^{Gaddis & Klasen, 2013}).

From the turn of the 21^st Century, social economics has become a venue of study in FLFP. From this approach of investigation, researchers concluded that the transmission of values and beliefs (^{Farré & Vella, 2013}; ^{Fernández, 2007}; ^{Fernández, Fogli, & Olivetti, 2004}), and attitudes (e.g., ^{Campos-Vazquez & Velez, 2013}; ^{Carriles, Beltrán, & Mata, 2019}; ^{Fortin, 2005}; ^{Giavazzi, Schiantarelli, & Serafinelli, 2013}) are strong and statistically significant determinants of women’s participation in the labor market.

Within this last stream of research, identity and emancipative values are two new lines of study. Pioneering the analysis of the effect of identity on FLFP, ^{Fortin (2009)} investigated its evolution in the United States from year 1977 to 2000. The author studied “Career Woman” vs. “Housewife” as competing identities of egalitarian and traditional gender roles, arguing that decisions such as fertility and labor market participation can be connected to them. ^{Fortin (2009, p. 31)} concluded that a preference of an egalitarian identity explained the increase of FLFP in the late 70’s and 80’s; and a rise of a preference of a “housewife” identity as result of societal and work environment factors, accounted for a decline in women’s labor participation by the end of the 20^th Century and the beginning of the 21^st.

^{Hayo & Caris (2013)} investigated the effect of identity in the Middle East and North Africa (MENA) region using WVS waves from years 2000 and 2005. From the analysis the authors concluded that women with a traditional role identity participate less in the labor market in the region, despite being Muslim or not, confirming that identity is a stronger determinant than religion.

Both investigations theoretical framework are supported on ^{Akerlof & Kranton’s (2000)} IET, focusing on two possible choices: traditional-housewife or egalitarian-working woman, where its choice is constrained by social expectations of the gender role.

^{Welzel (2013b)} claims that as social constraints decline in the human development process within a society, emancipative values rise as a motivational force of human empowerment to freely choose the path to develop her potential, and seek and demand equality of opportunities as any other human being. Supported on the Emancipative Values Theory (^{Welzel, 2013b}) as theoretical framework, ^{Brieger et. al. (2017)} investigated country-level on how emancipative forces motivated women to take leadership roles on corporate boards in 6390 firms in 30 countries. The authors determined in the cross-country comparison that emancipative forces to be positively and significantly linked to female taking leadership roles on company boards.

In the Latin American region, recent FLFP investigations have been approached from a neo-classical labor economics and economic growth perspectives. ^{Camou (2015)} investigated the incorporation of women to labor markets for eight Latin American countries (Argentina, Bolivia, Brazil, Chile, Colombia, Peru, México and Venezuela) in the 20^th Century, using a Gender Development Index and GDP as explicative variables. The author concluded that although there is a relationship between education, FLFP, and economic growth, there is no linear trend between the variables, and these are not enough to explain the variations between countries.

Novta (2017) analyzed FLFP in Latin American countries for the period of 1990-2014. The author asserts the expansion of FLFP was the largest in comparison to any other region of the world; nevertheless, gender gaps are still persistent and unequal across the countries of Latin America. Comparing periods from 1990 to 2014, Novta (2017, p. 8) illustrates that Latin American countries follow a U-shape, locating them at the bottom of the U. and claims that despite this situation, improvements in FLFP “was driven by improvements beyond development levels”, like women education and legal improvements, although policies on childcare services, promoting women higher education and laws against discrimination are still necessary for further improvements.

^{Gasparini et al. (2015)} studied FLFP growth in the region during the period of 1992-2012 using as explaining variables education, marriage, number of children, age, and area of residence, concluding that this factors favored women’s engagement in the labor market. However, the authors warn of a deceleration in FLFP due to better economic conditions in Latin America. Moreover, the authors suggest of further research opportunities on FLPF in the region as they considered their investigation as “imperfect and incomplete” (2015, p. 151).

^{Serrano et al. (2018)}, going deeper into ^{Gasparini et. al. (2015)} findings, studied the effect of GDP growth and conditional cash transfer (CCT) programs implemented through public policies on FLFP of the region for the period 1992-2014, in order to explain women’s labor participation deceleration after year 2000. The researchers concluded that better economic conditions and CCT had a decelerating effect on FLFP, mainly on women with lower education and from low income families or rural living, since better economic conditions lowers the pressure to seek low quality jobs.

^{Chioda (2016)} provides a complete assessment of the effect of family structure and dynamics with the household in Latin America. The author points to strong social norms still at play in the region, where woman attachment to household chores are associated to the beliefs that children may suffer if the mother works. However, the report also finds interesting that younger cohorts in the region display more attachment to the labor market in contrast to older women, which may reflect women’s preference for an identity of working-woman.

This study seeks to contribute to the literature in the following ways. First, by going beyond exogenous factors, as economic development or simply demographics, and incorporating women’s preferences in the model. Second, by integrating the social norms component through emancipative values as a reflection of social development and reduction of social norms constraints in the region. Third, by comparing these determinants with those of OECD countries to foresee the gap with advanced economies.

Data and Methodology

Methodology

This study considered the following regression lineal probabilistic model, where the dependent variable assumes a value of 1 if the individual in the survey participates in the labor market, and 0 otherwise.

Where

Y_ict: Is the labor participation of individual i in country c in period t, where 1 = participates, 0 = does not participates

X_1ict: Is individual’s i identity in country c in period t

X_2ict: Is individual’s i emancipative values in country c in period t

X_jict: Are individual’s i, j control variables of the country c in the period t

X_kct: Are the k control variables of country c in period t

ε: represents the non-observed economic variables

α: Are the constant factors not observed along time

t: period of time

c: country

Empirical results

To measure and compare the impact and significance of the independent variables upon female’s labor force participation decision, this study first analyzed the variables by types, and then grouping them up in a stepping up process until getting to the full model, as presented in Table 3.

Each of these analyses are referred as models 1 to 9, beginning with Model 1 trying the fixed effect country variable, and time effect variable identifying the World Values Survey wave number. Model 2, comprehends women’s demographic data, and incorporates Model 1 variables. Model 3, adds country’s exogenous variables GDP per cápita, and men’s unemployment rate to evaluate the effect of economic environment. Model 4, studies the effect of the Identity variable together with basic country and wave variables. Model 5 analyzes the effect of the Emancipative Values to contrast the size of its effect with Model 4. Model 6, experiments together with Identity and Emancipative Values which will serve to compare with Model 2. Model 7 and 8 evaluate Identity and Emancipative Values variables, respectively, in conjunction with variables considered in Model 3 to compare between these three models. Finally, Model 9 is the full model proposed by this research, to analyze the effect of all the determinants operating systemically upon the dependent variable.

Following, we first test the goodness of fit for each of the models. Then, we examine the effect of the explicative variables in the FLFP in Latin American countries. Third, we compare the results with those resulting from the analysis in OECD countries.

Conclusions

In the last 25 years the Latin American region experienced the fastest FLFP growth in the world. This study sought to contribute to the literature by analyzing the effects of Women’s Identity and Emancipative Values, as proxies to women’s preferences, and social factors, as determinants; and to identify the differences of this predictors with OECD countries, which represent the most advanced economies in the world. For this purpose, framed on the Identity Economic Theory and Emancipative Values Theory, we developed a probabilistic linear regression model, and exploited WVS and EVS data comprehending a 20-year period.

Women’s identity is formed during their youth; however, this may be adjusted later in life as result of personal experiences, social and economic shocks. As this study has shown, the role to which a woman identifies herself, housewife or working-woman, has economic ripple effects that goes from the person up to the aggregate economic level.

Emancipative values, as an empowerment force, is also at play on women’s labor participation. ^{Welzel (2013b)} suggested that as social constraints decline, emancipative values arise, serving as a motivation force of human empowerment to freely choose the path to full human potential. Within this perspective, this research analyzed the effect of Women’s Emancipative Values on FLFP, and as results have shown, it is effectively a powerful determinant for women’s engagement in the labor market.

Having analyzed the Latin American countries, the Probit regression marginal effects confirm that fertility, marriage, and income effect from country’s favorable economic situation, are disincentives to FLFP. However, as the analysis from the data of the last two decades show, Latin American women are starting to embrace the identity of working-woman that, together with the empowerment of emancipative values, and education, can counter the forces of traditional views.

Consequently, the results from this investigation lead to conclude that Women’s’ Identity and Women’s Emancipative Values are strong and significant determinants of FLFP. A self-view as a working-woman, and an ease of social constraints in the Latin American region have been key to motivate women into the labor market beyond simple mechanics from demographics and economic growth. Concurring with previous studies, higher education is also a key determinant to FLFP.

From comparison with OECD countries, Women’s’ Identity share similar views in both groups of countries, while Women’s Emancipative Values are three-fold greater in more advanced economies than in the Latin American region. Therefore, although there have been important social advances in the region, from the perspective of the Emancipative Values Theory is possible to infer that social constraining forces are still at play that are limiting gender equality. Future studies would need to identify and analyze the strength of the conservative values that are still at play in Latin America, that may be holding back FLFP growth in the region.

From a policies standpoint, actions are being fruitful and going into the right path, but as the comparison with OECD countries show, these are still incipient and efforts need to continue. To enroot gender equality societal values, these have to be promoted and taught at schools as part of the curricula, and egalitarian laws must be enforced in order to make permanent changes in the generations to come.