Abstract Twenty-five years after the fall of the communist regimes, the gender gap in employment varies widely across Central and Eastern Europe. This study examines the societal-level reasons for this variation and assesses the impact of different dimensions of neoliberally minded “economic development” strategies on gender inequality. We focus on Central and Eastern Europe, a segment of the world not typically addressed in the literature on gender and development. We rely on the 2008 and 2012 waves of the European Union Statistics on Income and Living Conditions survey as well as multiple macro-level data sources to analyze the association between development indicators, labor market context, social policy arrangements, and the gender employment gap. We find that typical growth indicators, global market integration, and social policy arrangements are not at all or only weakly associated with the gender employment gap in this region. Instead, the labor market context, specifically the degree of segregation and the size of the public and service sectors, are more important for shaping women’s labor market opportunities relative to men’s at both time points. Our findings contribute to the literature on the trade-offs between job segregation and aspects of gender inequality as well as to ongoing debates within the field of “gender and development” by pointing out important variations across regions. Starting in 1989, countries in Central and Eastern Europe (CEE) underwent a transition from centrally planned to market economies. Though countries in the region pursued widely varying strategies of privatization and liberalization, all of them experienced economic contraction and massive job loss as workers faced growing competition for—and heightened employer discretion over—entry into jobs in the emergent private sector. Women in particular were impacted by this transformation. Women’s employment rates, formerly among the highest in the world, declined, making CEE the only region where women’s labor market integration decreased rather than increased in recent decades. The financial crisis of 2008 and the resulting labor market contraction further exacerbated both women’s and men’s precarious positions. But the impact of these changes varied across post-socialist countries, and the gender employment gap—the difference in employment rates between men and women—has been highly variable. The gap has shrunk in some countries, widened in others, and remained stable in yet others. Twenty-five years after the collapse of state socialism, the gender gap in access to paid employment relative to the EU is among the smallest in Estonia and Lithuania and one of the largest in the Czech Republic and Romania. What explains this variation? What characteristics of post-socialist economic development, labor market structure, and/or welfare policy arrangements are associated with more or less gender inequality in access to paid work? This paper seeks to answer these questions by testing competing hypotheses about the correlates of gender inequality in employment within CEE, a region not yet systematically studied in a comparative light. By identifying the factors underlying variations in the gender gap in paid work, we contribute to the growing literature identifying the impact of economic development and welfare state policies on gender inequality. Our analysis is based on comparative survey data from ten formerly state socialist countries in CEE as well as societal-level information from a variety of published sources. In conversation with the literature on “gender and development,” we explore the relevance of neoliberal economic development indicators, including the level of economic growth and foreign investment as well as features of the labor market structure, for instance the size of the public and service sectors and job/sectoral segregation, as well as the impact of state austerity, such as the generosity of leave policies and childcare availability. We study these factors at two different time points: before and after the global economic crisis of 2008. We do so not because we intend to identify the impact of the economic crisis on women’s employment levels, as that would require a more systematic focus on other factors as well, but simply to identify possible changes and continuities in post-socialist economies. Our findings suggest that the main correlates of the gender gap in access to paid work are those related to the structure of the labor market: when jobs are available in female-dominated sectors and job segregation is high, the gender gap in paid work is smaller. This “silver lining of segregation” (Rubery 1988; Rubery and Rafferty 2013) operated in CEE countries before, during, and after the economic crisis as well. In contrast, the indicators closely monitored by international financial institutions, that is, the pace or level of economic growth or the depth of foreign capital penetration, do not predict gender differences in employment. Furthermore, contrary to the findings of comparative research on gender and welfare states focused mainly on Western European countries, we find that generous social protections, including paid parental leaves, do not necessarily shrink the overall gender gap in employment. In former socialist countries, labor market structure matters more than social policy arrangements for facilitating women’s access to paid work. First, we review the existing literature on the relationship between neoliberal economic development measures, social policy arrangements, and the gender employment gap. As we do so, we discuss the specificities of the CEE context and develop our hypotheses. Next, we describe our data and analytical strategy and present our results. We end with a discussion of our findings. Gender, Development & Labor Market Segmentation The relationship between economic development and social inequalities has been studied extensively (Lenski 1966; Kwon 2016). In this paper, we focus specifically on gender inequality and examine the impact of societal-level variables that are central to ongoing debates regarding the relationship between the gender employment gap and neoliberal economic development processes. This literature discusses the importance (or lack thereof) of the following factors: (1) the level and pace of economic growth; (2) the depth of foreign investment penetration; (3) job and sectoral segmentation; and (4) state austerity versus spending on policies targeting work-life balance. In an effort to make theories on gender and economic development “travel” to CEE countries (Burawoy 2015), we test the usefulness of existing explanations for understanding variation in the gender employment gap. Economic growth Scholars disagree on the impact of the level and pace of GDP growth on the gender employment gap. Economic growth may increase the demand for female labor as employment opportunities open for all (Elson 2009; Seguino 2000). Growth typically leads to the enhancement of infrastructural services, which could endow women with marketable skills as well as job opportunities, and free them from reproductive duties to be available for paid work outside the home (Kanbur, Lustig, and World Bank 2000). The faster the rate of growth, the quicker these changes are expected to take place. Such modernization theory–inspired explanations suggest a more or less linear association between both the level and pace of economic growth and women’s employment. Others present evidence for a positive correlation with a reversed causal arrow (Seguino, Berik, and Van der Meulen Rodgers 2010), for a U-shaped relationship (Boserup 1970; Elborgh-Woytek et al. 2013; Forsythe 2000) or one mediated by other factors Braunstein (2012). The relationship between negative economic growth (contraction) and the gender employment gap is even more complex. Within the European Union, a sharp negative change in GDP during 2008–2009 was associated with a narrowing gender gap in several countries, as men’s job loss exceeded that of women (Bettio et al. 2013; Smith and Villa 2014). Job segregation may have protected more women than men from the impact of market contractions, but gender differences in access to jobs were conditioned on the character of the existing demand and supply of labor and thus varied greatly from country to country (Karamessini and Rubery 2013; Rubery and Rafferty 2013; Smith and Villa 2014). CEE countries were first exposed to neoliberal reforms aimed at fast-paced GDP growth after the collapse of state socialism in 1990. The widely employed structural adjustment blueprint promoted privatization, global market integration, austerity in public spending, and strict monetary policies. CEE countries share a handful of features of what has been termed “post-socialist capitalism” (Bandelj 2016), including the historical legacy of state socialism and a geo-political position on the periphery of the European Union. However, CEE countries adopted market reform policies in different ways and with varying levels of intensity, resulting in wide variations in social inequality, economic vulnerability, and poverty (Bohle and Greskovits 2012; Hamm, King, and Stuckler 2012; Myant and Drahokoupil 2011). Studying the relationship between macro-economic policies and gender inequality in CEE countries can add important new insights to this field because women in this region entered the period of market integration from a significantly more emancipated position than women in other developing countries where structural adjustment had been adopted earlier. Prior to market reforms, women in CEE enjoyed high levels of education, low fertility rates, and significant labor market experience (Einhorn 1993; Gal and Kligman 2000). Researchers noted the deepening of gender inequalities in the early phases of the transition (Funk and Mueller 1993; Glass 2008), but women’s level of employment has been increasing steadily in each country since the late 1990s. This could indicate a decline in gender discrimination in the labor market and an increased demand for women’s skills, as economic development picked up after the initial shock. Given women’s high level of human capital, we thus expect that by 2008 rapid economic development will contribute to a shrinking of the gender employment gap, as women in CEE countries are well equipped to compete in increasingly competitive labor markets. The crisis of 2008 hit CEE countries to different degrees, leading to a sharp decline in GDP and employment in the Baltic region and more modest impacts elsewhere. Women were unevenly affected but we expect, for the reasons described above, that the gender employment gap will shrink as it did in other EU countries: women will be more likely than men to retain employment. In addition, we expect the gap to narrow more in regions hardest hit by the crisis, because it is here where men’s employment rates fell most sharply, which may have created a faster leveling-down process and may drive the narrowing of the gender gap. This would be in line with trends in several Western European countries during the crisis (Bettio et al. 2013). Overall, we expect that women in twenty-first-century CEE countries will benefit more than men from fast-paced economic development as employers recognize the usefulness of their human capital and their low cost. Global market integration Economic growth fueled specifically by investment liberalization and global market integration has been linked to the creation of expanded employment opportunities for women in developing countries, a process termed the “feminization of labor” (Standing 1989). These jobs are typically underpaid, dangerous, and/or exploitative in nature, and overall may lead to a decline in employment quality (Elson 2009; Mies 1999). Nevertheless, the proliferation of the “global assembly line” has offered employment opportunities for more women than men in several regions and has thus contributed to the closing of the gender gap in employment globally. Importantly, however, not all export-oriented investment leads to the feminization of labor. For example, in the case of CEE, women’s advantage declines or disappears as investments result in jobs that require better-educated and higher-skilled workers or when investments target sectors where men tend to work, such as construction or manufacturing (Braunstein 2012). Rates of foreign investment, a key indicator of global market integration, vary greatly in the CEE region (Bandelj 2007), with Estonia and Bulgaria experiencing the most investment in the years prior to 2008 and other Baltic countries, such as Lithuania and Latvia, receiving substantially less. CEE countries differ from most others within the European Union because their rate of inward FDI well exceeds their outward flow. Foreign investment is strongly related to fast-paced growth in certain sectors, including auto parts and textile manufacturing as well as finance, banking, business services, accounting, and tourism. Across these sectors, employers actively recruit and hire workers for multinational firms located throughout the region. For example in 2011, 27 percent of Slovenian and 38 percent of Estonian workers were employed in foreign-controlled enterprises. EU-wide, employment in foreign-owned firms averages 5 percent (Eurostat 2011). Women in CEE have high levels of human capital and significant experience in many of these sectors (e.g., textiles) and in white-collar professional employment (e.g., financial services) yet earn on average 5–25 percent less than men. Thus, we expect women to benefit disproportionately from job creation spurred by foreign direct investment: women in the region offer skilled yet cheap labor. On the other hand, unlike in other developing regions, women in CEE are not likely to be viewed by employers as docile or compliant. Instead, given the existence of lengthy maternity leaves and an unequal division of labor in the home, foreign employers may be more likely to consider women (especially mothers) unreliable and incapable of meeting intensive labor demands (e.g., Glass and Fodor 2011). This perception of women, discrimination based on the intersection of gender and reproductive duties, may counteract women’s attractiveness and prompt foreign employers to prefer men over women, even in traditionally female-dominated sectors. Although after 2008 foreign direct investment volume contracted radically, much of this contraction happened in heavy manufacturing sectors and thus we do not expect women to suffer a heavier toll than men. Labor market segmentation The differential impact on men and women of both the 1990 and the 2008 economic crises in the region highlighted the significance of labor market structure in mediating the impact of economic growth on employment. In particular, occupational and sectoral segregation have been linked to variations in the gender employment gap (Bettio et al. 2013; Karamessini and Rubery 2013; Milkman 1976; Rubery 1988). Although many scholars point to segregation as an impediment to gender equity in the United States and Western Europe (Esteves-Abe 2005), there is a silver lining to job segregation in countries undergoing rapid economic transformation. Under certain economic conditions—including massive economic restructuring—gender segregation may protect women more than men from unemployment and thus reduce the gender employment gap. First, jobs in the public (state) sector are typically more secure than those in private enterprises, which are more vulnerable to market fluctuations. Researchers have shown that women with small children often choose job security and family-friendly work hours available in public-sector, typically feminized (“mommy track”) jobs over higher wages in private sectors more vulnerable to economic volatility (Hakim 2006; Mandel and Semyonov 2005). Such “choices” may not be entirely voluntary, but due to prohibitively lengthy work hours or barriers to entry in private professional sectors or the lack of supportive family arrangements and inaccessible or insufficient childcare facilities. Nevertheless, gender segregation in female-typed public-sector jobs may serve to reduce the gender gap in employment during periods of economic contraction or crisis. As the availability of relatively secure, family-friendly public-sector jobs remains stable, women’s employment gains outpace those of men. The second way segregation may reduce the gender employment gap is through the construction of certain jobs or sectors as “feminine.” Women may enjoy more job opportunities and greater levels of job security relative to men when economic changes favor the type of work deemed appropriately feminine over those considered masculine. In these instances, strongly held convictions about the appropriateness of gender segregation prevent men from aspiring or even being considered for “female-typed” jobs even when no others may be available. There is evidence that men avoid fields where women comprise even a quarter of current incumbents (England and Li 2006; England et al. 2007), and that employer preferences are strongly shaped by gendered constructions of the nature of work (Kmec 2005; Reskin and McBrier 2000). Women have been drawn into paid work through the expansion of female-dominated jobs in the public sector and certain feminized professions like teaching and nursing (Goldin 1990). In addition, researchers have observed that during certain economic crises, entrenched sectoral and occupational segregation makes the replacement of women with men problematic and thus shelters more women than men from layoffs (Milkman 1976; Rubery 1988). This was the case during the recession of 2008 as well. The feminization of service and public-sector jobs and the slower rate of contraction in these areas protected women’s employment to a greater degree than men’s, resulting in a higher rate of job loss for men and a reduction in the gender employment gap (Bettio et al. 2013; Karamessini and Rubery 2013). While job loss was extensive in agriculture and manufacturing after the collapse of state socialism, employment in the service sectors multiplied across the region. Women have historically dominated in these sectors, which were considered “unproductive” and thus devalued under state socialism (Fodor 1997). Cross-country variation is significant, however, with well over 30 percent of people working in “market services” in Estonia, Hungary, and the Czech Republic but a significantly lower percentage in Bulgaria or Lithuania (Rutkowski 2006, p. 18). Women in CEE are also more likely than men to be working in public-sector employment, both in state administration and in jobs such as teaching, health, or social services (Burchell et al. 2014) There is evidence that occupational and sector-based segregation offered a degree of job protection, especially to young, childless, educated women after the collapse of state socialism (Fodor 1997; Ghodsee 2005; Glass 2008). Preliminary research also indicates similar trends for at least the first years of the 2008 recession (Fodor and Nagy 2014). At the same time, as more severe austerity measures were introduced after 2010, cuts in public employment may have led to disproportionate job loss among women, widening the gender gap in employment (Burchell et al. 2014). Nevertheless, we do not expect the direction of the correlation between segregation and the gender gap in employment to change between 2008 and 2012. For both years, we expect to find a positive relationship between the size of the service and public sectors as well as occupational and sector-based segregation and women’s labor market access in CEE. While much of the literature on the “inequality-inclusion trade off” (Pettit and Hook 2005) describes the relationship between women’s wages or career advancement and job segregation, we predict a similar outcome in terms of access to paid work. State austerity Social policy arrangements, especially those related to parenthood, no doubt impact women’s employment chances, especially in contexts where they are primarily responsible for raising children and tending to household duties. Paid leaves allow parents, typically mothers in CEE countries, to exit the labor market for a period of time following the birth of a child, and while limiting employment during the leave may facilitate greater employment continuity thereafter. The effect of such policies is dependent on the social class of workers and the length of leave. Leaves that are too short may dissuade women from high-income households from returning to work following the birth of a child, while leaves that are too long may increase employment discrimination against mothers (Glass and Fodor 2011; Misra, Budig, and Boeckmann 2011; Morgan and Zippel 2003), but they do not necessarily have a negative impact on all women’s long-term employment rates (Javornik 2010; Keck and Saraceno 2013). Publicly provided childcare is also associated with higher rates of female employment because it reduces the costs of employment for women and signals public support for their employment (Gornick and Meyers 2003; Pettit and Hook 2005). However, scholars have found important differences in the effects of childcare depending on the age children become eligible for public care. In countries where public childcare is provided only for children over age three, mother’s attachment to paid work is weakened. In contrast, when public childcare is available for children under age three, women are better able to sustain paid work following the birth of a child (Cukrowska-Torzewska 2015; Gornick and Meyers 2009). Overall, this scholarship suggests that the primary barriers to women’s employment are rooted in the gender division of labor in the family and household, which shapes the supply of women workers. To the extent that the state minimizes these family-based tensions, women are better able to enter paid work alongside men, thereby shrinking the gender employment gap. Post-socialist countries are rarely included in studies exploring the relationship between welfare institutions and the gender employment gap. Analyses that do consider these countries tend to limit country selection to a small sample, typically Hungary, the Czech Republic, and/or Poland (Budig, Misra, and Boekcmann 2012; Mandel and Semoyonov 2006; Pettit and Hook 2005). However, these countries are not representative of the former socialist region, and generalizations drawn from these cases tend to mask important regional variations (Javornik 2014). Our analysis focuses on CEE countries in order to identify important sources of variations within the region (see O’Conner, Orloff, and Shaver ). First, we analyze the effects of relatively lengthy parental leave policies on the gender gap in access to employment. While some form of parental leave was introduced in all socialist countries in the middle of the 1960s following the Soviet Union’s family policy measures, by the early 2000s the length, generosity, and take-up rate of parental leave varied widely across the region. In some countries, such as Hungary, the Czech and Slovak Republics, and Estonia, all new parents are eligible for at least three years of paid leave, while in others, such as Poland, leaves are means tested and thus only available to a small number of parents. In yet other countries, such as Slovenia and Romania, shorter and better-paid leaves are tied to social insurance schemes.1 As noted above, the evidence on the impact of very long family leaves is ambiguous and CEE countries are ideal for the examination of this question because several offer this option. We expect that employers will be less enthusiastic about hiring women (of childbearing age) if policies allow them to leave their jobs for extensive periods and employers are expected to take them back whenever they are ready to return. We thus hypothesize that the availability of lengthy parental leaves is detrimental to women’s labor market participation both before and after 2008, while shorter leaves may encourage employment continuity. Second, we analyze the impact of childcare for children below and over age three in order to specify the effect of public childcare on the employment gap. Former socialist countries offered generous childcare options for children over three years of age, but many of the facilities fell prey to the economic collapse following the fall of the communist regimes and the austerity measures imposed on local municipalities, making childcare more expensive and less accessible than it had been under communism (Saxonberg and Sirovatka 2006). Cutbacks in the quality and availability of public childcare were exacerbated by further reductions in state spending after 2008. We expect that the availability of public nurseries and kindergartens is associated with a smaller gender gap in employment, as it allows parents to reconcile work and family and reduces employers’ perception of women with children as unreliable workers.2 Finally, when fathers participate in raising children, women’s work burden is lowered. Paid paternity leaves have been shown to reflect societal approval of women’s role in paid labor and increase men’s childcare work in both the short and long run (Nepomnyaschy and Waldfogel 2007). In the CEE region paternity leave periods, where they are available, are quite short, so this is more an indication of how people generally feel about the division of labor in society than an actual reduction in women’s burden. Nevertheless, as an expression of attitudes it may be important, and we expect that countries with longer paternity leave days will have a smaller gender employment gap. Data and Methods We test our predictions using data from the 2008 and 2012 cross-sectional waves of the European Union Statistics on Income and Living Conditions. These are large representative samples of all EU member states, which are repeated each year. We use these two independent cross-sectional waves, as the longitudinal panel is too short (four years only) for our purposes and thus we could not cover the pre- and post-crisis years with a single longitudinal sample. We selected all of the post-socialist societies in the dataset, a total of ten, for this analysis. These countries include Bulgaria, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia, and Slovenia. We did not include other European Union member states because we wanted to control for theoretically relevant factors specific to formerly socialist societies in CEE. Each of the countries included in the analysis are significantly poorer than the EU average, have been exposed to strict structural adjustment requirements for the past twenty years, are relatively new locations for a growing volume of foreign investment, and are found at the peripheries of the European economy. There are other countries, for example Portugal, Greece, or Cyprus, which would qualify by these criteria as well. But CEE countries also share the state socialist legacy of women’s emancipation, an extensive welfare state, and full employment, which are particularly relevant for our purposes. Until 1990 most adult women were engaged in paid labor in CEE, dropping out only for short periods to raise children before returning to full-time employment. By the late 1980s, homemaking did not exist either as a concept or as actual practice in the region. And, as noted in the introduction, these countries are the only ones in the world that have witnessed a decline in women’s labor market participation since 1990. There are also important contemporary differences that distinguish these economies with regard to gender and employment. First, in contrast to many other European societies, women in CEE tend to work full-time. Nowhere in the CEE region does the percentage of women working part-time exceed 15 percent, while the EU 27 average is 32 percent. This is a very important distinction, as this means that the concept of “working” is quite different in post-socialist economies than it is in most other countries on the European continent. This is particularly relevant for our project, as we explain below, since our dependent variable is based on self-classification. Comparison with countries where “working” often means participation in part-time work is problematic. Second, in state socialist countries, women could retire early (at age 55) and significantly earlier than men. After the accession of these countries to the European Union, the low retirement age became untenable and governments began to gradually raise and equalize it. Nevertheless, the uniquely post-socialist practice of early retirement for women—which continued to be widely offered throughout the 1990s to alleviate pressures of high unemployment—means that older women’s labor force participation rates tend to be lower in many countries of this region than in other parts of the EU. We restrict our analysis to people between age 20 and 60 to minimize this problem, but we cannot completely eliminate it. The final difference between CEE countries and other EU member states relevant for this analysis concerns the length and popularity of parental leave policies that are supported by relatively conservative gender ideologies (Blaskó 2005; Takács 2008). Because of these unique cultural, political, and economic features, we limit our analysis to former state socialist countries in the EU. Univariate analysis: Individual-level dependent and independent variables We use a binary dependent variable describing whether the person considers himself/herself to be working full- or part-time as their main activity at the time of the interview (variable number PL031 in EU-SILC 2008 and 2012). EU-SILC uses direct self-classification to obtain this information. In other words, respondents were asked to identify their main activity on the basis of what they spend the most time on. Response options include working full- or part-time, being unemployed, a student, retired, disabled, in the military, or keeping house. We classified people as working if they identified their main activity as one of the first two categories: working full- or part-time. By this classification, workers include the self-employed, those working for benefits in kind, and those working in Central and East European countries’ extensive informal economy. These data yield somewhat different results from the Labor Force Surveys from which EU-wide employment information is typically available (see also Keck and Saraceno ). The Labor Force Survey (LFS) classifies respondents as “working” if they had spent at least one hour in the week prior to the survey at paid work; it is thus not surprising that in all cases LFS statistics are higher than those obtained from the self-reported SILC dataset. Only in a handful of cases is the difference larger than 5 percent, though. The most critical case is Slovenia, where SILC underestimates the proportion of women working by about 8 percent. Men’s rates are similarly lower, so this factor is unlikely to affect the gender gap. The difference is most likely due to the fact that sampling was carried out slightly differently in Slovenia than in the other nine countries, although still within the specifications of EU-SILC. We will interpret the results with caution. In 2008, only four of the ten post-socialist countries reached the average employment levels of the European Union for 15–64-year-olds (65.8 percent), and the employment rates in some of the countries, such as, for example, Hungary, Poland, or Romania, were among the lowest of the 27 countries (Eurostat online). The bars in figure 1 describe men’s and women’s level of employment in the countries we included in the analysis using our own dataset for the two time points studied. Women between 20 and 60 years of age are the least likely to be working in Hungary, Poland, and Romania (57–58 percent), and the most likely in the Baltic region (71–73 percent), with the Czech Republic, Slovakia, and Bulgaria in the middle (65–68 percent). In most countries, women’s level of employment declined between 2008, the year leading up to the crisis, and 2012, which is the year past the trough of 2011. The change, however, is relatively small, and especially so if compared to more radical changes in men’s employment levels. Figure 1. View largeDownload slide The percentage of men and women working in 2008 and 2012 in the population of 20–60-year-olds Source: EU-SILC, 2008, 2012. Figure 1. View largeDownload slide The percentage of men and women working in 2008 and 2012 in the population of 20–60-year-olds Source: EU-SILC, 2008, 2012. Figure 2 describes gender differences in access to employment, that is, the gender gap in paid work. In 2008, the gap was the largest in Romania and the Czech Republic, almost 20 percentage points, while it was quite small in Estonia and Lithuania. By 2012, the gap shrank everywhere: women started to catch up with men. In fact, the direction of the gender gap reversed in Lithuania and Latvia: women in these two countries are more likely to be working than men (at least by EU-SILC data), and the gender difference was reduced significantly in Bulgaria also. The changes are barely visible in some of the other countries, such as, for example, in Romania, Poland, and Slovenia. Note that in this paper we set out to explain the determinants of the variation among the countries at two points in time, before and after the main crisis period, rather than the change within each country, which would require an in-depth country analysis, rather than the macro-level comparison that is our ambition now. Figure 2. View largeDownload slide The gender gap (gender difference in percentage points) in employment rate among 20–60-year-olds in 10 post-socialist societies in 2008 and 2012 Source: EU-SILC, 2008, 2012. Figure 2. View largeDownload slide The gender gap (gender difference in percentage points) in employment rate among 20–60-year-olds in 10 post-socialist societies in 2008 and 2012 Source: EU-SILC, 2008, 2012. A number of individual-level factors are known to influence people’s probability of entering paid work, and we control for these in our models below. Specifically, we included controls for age, coded in years, as well as a squared term to account for non-linearity in the impact of age on work status. Table 1 shows the distribution/means of all of our independent variables. Table 1. Distribution of Individual-Level Variables 2008 2012 Men Women Men Women Age (average) 39.9 40.9 40.5 41.5 Secondary education 71.0 64.7 69.0 60.7 Tertiary education 14.3 19.8 17.0 25.2 Partner in household 63.1 65.6 60.6 63.4 Number of children under 6 0.18 0.17 0.19 0.18 Women 48.9 51.1 48.9 51.1 Working 76.4 63.5 71.1 61.6 N 55,918 58,943 59,735 62,517 2008 2012 Men Women Men Women Age (average) 39.9 40.9 40.5 41.5 Secondary education 71.0 64.7 69.0 60.7 Tertiary education 14.3 19.8 17.0 25.2 Partner in household 63.1 65.6 60.6 63.4 Number of children under 6 0.18 0.17 0.19 0.18 Women 48.9 51.1 48.9 51.1 Working 76.4 63.5 71.1 61.6 N 55,918 58,943 59,735 62,517 Source: EU-SILC 2008, 2012. Education is coded as a set of three dummy variables: elementary or less, secondary only, and college and above, with elementary or less as the omitted category. About 20–25 percent of women in the sample have postsecondary education, compared to 14–17 percent of men. While educational segregation is widespread in CEE, in terms of degrees women are better qualified than men. This phenomenon may be linked to the legacy of state socialism, where women started to enter college in large numbers in the 1970s, as well as to the widening of educational opportunities in the context of rising labor market gender discrimination and youth unemployment rates in the early 1990s. We also include an indicator of partnership status (if the respondent’s partner is living in the same household) and a measure to assess the number of the respondents’ children under six years of age who live in the same household. We expect education to have a positive association with work status, and having a partner or the presence of children to have little to no discernible impact for the whole of the population, as the positive impact for men is expected to cancel out the possibly negative effect for women. Societal-level independent variables Our main interest is the association between various societal-level factors described in the previous sections and the gender gap in employment. These data come from Eurostat, and a detailed description of the sources and the coding may be found in appendix A. Variables are lagged by one year, unless we note otherwise. The level of economic development is used as a control variable in each of the models. We experimented with several measures but found GDP per capita at PPP as a percentage of the EU 27 average as the most consistent over time and across countries. None of the ten countries reach the average “wealth” of EU countries; Slovenia comes closest at 84 percent in 2012, while Bulgaria’s GDP is 47 percent of the EU average only. This measure improved in every country between 2008 and 2012, except in Slovenia, where it declined, and the Czech Republic, where it stayed stable. Economic growth describes the three-year average of the country’s real GDP growth between 2006–2008 and 2009–2012. We also experimented with a measure of GDP growth in 2007 and 2011, but these are highly correlated with the three-year average rate, so we opted to use the former. It is worth noting, however, that when we discuss “economic growth” we mean medium-term growth and attempt to even out the annual fluctuation in the rather volatile years between 2006 and 2012. In the three-year period leading up to 2008, average growth rates ranged from a low of 1.6 percent in Hungary to a high of 8.2 percent in Slovakia. The Baltic countries also posted high rates of growth as well as a more radical decline starting in 2009. By 2012, economic growth rates recovered, and ranged from a negative 0.17 percent in Slovenia to a growth of 5.3 percent in Estonia. The volume and movement of foreign investment into a country is an indication of its degree of global market integration and the country’s dependence on international capital. In the years prior to 2008, foreign direct investors (FDI) favored Estonia and Bulgaria (10–24 percent), while foreign investment accounted for a much smaller (3–5 percent) fraction of the GDP of Slovenia, Lithuania, and Latvia. By 2012, Hungary and Estonia had the highest FDI flow relative to its GDP, but FDI stock amounted to almost 100 percent of Bulgaria’s GDP, while only to about 32 percent of that of Slovenia. We experimented with a measure of employment in foreign-owned companies as a percentage of all employment, but this indicator is highly correlated with the flow of FDI into the country and produced identical results. We also added measures that assess the concentration of women in specific occupations and sectors as well as the size of the sectors where women are more likely to be employed. There are several useful measures of occupational and sectoral segregation, including the Index of Dissimilarity (Duncan and Duncan 1955), the IP Index (Karmel and MacLachlan 1988), a tripartite classification offered by Hakim (2002), and model-based measures (Charles and Grusky 2004). We chose the IP index, because this is the one collected by EU countries in a harmonized way, thus comparison across countries and over time seemed the most reliable. The IP index, similarly to the index of dissimilarity, can be interpreted as the proportion of workers who would have to change occupations to achieve a perfectly equal distribution of men and women. Similarly to the index of dissimilarity, it is sensitive to the size of women’s share in the labor market. We note quite a great deal of variation across the ten CEE countries in this dimension but little change over time. Both types of segregations in both years are high in the Baltic countries and significantly lower and well below EU average in Romania and Poland. These differences are possibly due to both the legacy of the state socialist labor market and highly segregated educational institutions as well as the gender dynamics of international companies moving into the region. Given that women are more likely to cluster in the service sectors and in public employment, we introduced indicators, which measure the size of employment in this segment of the economy. As noted above, we expected these to correlate somewhat with measures of segregation and they indeed do. The Baltic countries, Hungary, and Slovenia have an especially large public sector, and in 2012 over 20 percent of workers were employed there. Fewer than 15 percent of workers work in this sector in Romania or Bulgaria. As table 1 shows, the overall proportion of jobs in the public sector did not decline by 2012 even in the context of austerity measures. The service sector is another part of the labor market where women are more likely to find employment than men. This is the sole sector that has been steadily growing since the early 1990s in most countries of the region. Employment and employment experience here are thus likely to lead to better chances of having a job, and when women dominate this sector, a smaller gender employment gap. In addition to measures assessing the labor market context, we included indicators of the generosity and character of state supports for working parents. In this regard, CEE countries are quite different from developed nations and researchers often find it hard to classify them within existing welfare state models (although see Ciccia and Verloo ; Javornik ). In four of the ten countries (Estonia, Hungary, Slovakia, and the Czech Republic), parents can choose to stay out of the labor force on paid, protected parental leaves for over 30 months. This option did not change in the five-year period of our study. While there are a number of ways in which the generosity of parental leave options may be measured (Budig, Misra, and Boeckmann 2015; Pettit and Hook 2009), such lengthy leaves are quite rare within the EU yet characteristic of countries with a state socialist past. We thus decided to take this opportunity and specifically focus on its impact on women’s labor market chances. In addition, this is where the most variation could be found across countries: all others provided at least one year of decently paid leave for working parents (with the possible exception of Poland, where payment was dependent on means tests). To examine the impact of lengthy leaves on the employment gap within the whole range of CEE countries, we included a binary variable, the value of which is 1 if the country allows a third (or even fourth) year of paid leave to new parents and 0 otherwise. We also use a measure of the proportion of children under three years of age in formal care institutions (“crèches”) and the percentage of three to six years old in formal full-time childcare (“kindergartens”). On average, according to Eurostat tables, 10–12 percent of young and 65–68 percent of older children are in daycare in the ten countries in our sample (see appendix A for further info). These numbers, especially the proportion of younger children, vary greatly across countries: 19 percent of the very young in Estonia and only 2 percent in Romania attend nursery schools. Not only do these numbers vary, but they are also far lower than the EU 27 average at 27–28 percent. The reason for this discrepancy is the presence of lengthy leave options, in addition to the dearth of childcare institutions, as well as a widespread conviction that children are better off at home with their mothers for the first three years of their lives (Saxonberg 2014). Given these specificities of the region, generalization about the impact of these variables must be handled carefully. Paternity leave represents the number of paid leave days reserved for fathers upon the birth of a child. These are typically rather short, except in Lithuania, where fathers may take up to 28 paid days of leave. In two countries, the Czech Republic and Estonia,3 no paternity leave is allowed, and the length ranges from 5 to 15 days in the rest. Strategies for the multivariate analyses Our survey data are clustered within countries, which require a multilevel approach to data analysis. But we only have ten cases at the second (macro) level and this raises serious questions about the appropriateness of this method. These ten countries are not a sample but represent the population of post-socialist countries that were members of the European Union in 2008 and 2012 when the data were collected. In our case, the small number of countries is a substantive as well as theoretical choice: as we noted above, Central and East European countries differ systematically from others in their labor market structures and gender regimes, which is the topic of the current study. Opinions differ on the usefulness of multilevel models, with a small number of Ns at the second level. Some statisticians claim that hierarchical models may be used even with a second-level N of 3 (Gelman and Hills 2007), while others argue that at least 20 or even 50 cases are necessary to gain a reasonably conservative estimate of the standard error of the coefficients (Bryan and Jenkins 2015; Maas and Hox 2005; Stegmueller 2013). The latter group argues that multilevel models with an N of 10, especially logistic regression models and especially ones with cross-level interaction terms, as is the case in this paper, may produce biased estimates and could underestimate the standard errors at the second-level parameters by as much as 30 percent. We have theoretical reasons to focus on the population of CEE countries within the European Union and do not want to abandon this restriction in our sample. We tend to agree with Gelman (ibid.) that while multilevel models may not be perfect in this case, the alternatives are worse. We thus ran and will interpret the results from a series of multilevel random coefficient logistic regression models.4 The dependent variable in all of our models is “working,” as defined above. In each model we included only one macro-level factor along with the GDP (as a control). We will primarily focus on the country level and interpret cross-level interaction terms with each of the macro-level coefficients and gender. We have at least two reasons to trust our results, notwithstanding concerns about the small N at the second level. First, our findings are primarily negative: we will show that a great number of factors typically considered important do not vary systematically with the gender employment gap in CEE countries. The standard errors of a handful of factors that do show significant association are quite small, and even if we increased them by 30 percent we would, in most cases, get statistically significant coefficients (we will indicate this when we discuss the results). Second, Gelman (2005) argues that when the N is large at the individual level, a two-stage analysis approximates very closely the results of a multilevel random slope regression. To obtain such a model, one simply runs logistic regression models for each country separately and as a second step plots the resulting coefficients for gender against each of the macro-level predictors of interest. This produces graphs, which can be easily interpreted visually, as well as linear regression equations, which predict the value of the logistic regression coefficient on each macro-level variable. We show two of these graphs below. We conducted a series of two-stage analyses along with the more complex multilevel equations for each time point and each macro variable, and our results from the two-stage models fully confirmed the findings of the multilevel regressions (table 2). Table 2. Distribution of Macro-Level Variables (10-country averages) 2008 2012 GDP as % EU average in 2007 and 2011 65 68 GDP growth 3-year average 5.6 1.8 FDI flow, 3-year average 6.2 3.03 FDI stock, 3-year average 45.8 55.3 Sectoral segreg (IP index) 20.1 21.3 Occupational segreg (IP) 27.8 27.7 % public sector empl. 18.0 19.3 % service sector employment 56.7 60.1 Kids lt 3 in care 10.3 11.1 Kids 3–6 in care 63.1 67.1 Has 3rd year parental leave paid 0.38 0.36 Paternity leave days 6.2 11.0 2008 2012 GDP as % EU average in 2007 and 2011 65 68 GDP growth 3-year average 5.6 1.8 FDI flow, 3-year average 6.2 3.03 FDI stock, 3-year average 45.8 55.3 Sectoral segreg (IP index) 20.1 21.3 Occupational segreg (IP) 27.8 27.7 % public sector empl. 18.0 19.3 % service sector employment 56.7 60.1 Kids lt 3 in care 10.3 11.1 Kids 3–6 in care 63.1 67.1 Has 3rd year parental leave paid 0.38 0.36 Paternity leave days 6.2 11.0 Source: See appendix A. Results from the multivariate models The first step is to examine the individual-level model without the cross-country difference predictors. Table 3 shows the output from two multilevel random coefficient logistic regression models predicting the probability of “working” using the individual-level variables only, for years 2008 and 2012 separately. Table 3. Coefficients from a Multilevel Random Coefficient Logistic Regression Model, Predicting the Odds of Working for Wages in 10 Post-Socialist Countries, 20–60-Year-Olds 2008 2012 Women −0.78** −0.59** (0.073) (0.104) Age 0.52** 0.49** (0.005) (0.005) Age squared −0.01** −0.01** (0.000) (0.000) Secondary education 0.35** 0.42** (0.010) (0.010) Tertiary education 1.72** 1.74** (0.029) (0.026) Parent of child 6 or younger −0.48** −0.42** (0.016) (0.016) Has partner in house 0.42** 0.47** (0.018) (0.018) Constant −8.94** −9.28** (0.367) (0.126) N 103,677 114,261 2008 2012 Women −0.78** −0.59** (0.073) (0.104) Age 0.52** 0.49** (0.005) (0.005) Age squared −0.01** −0.01** (0.000) (0.000) Secondary education 0.35** 0.42** (0.010) (0.010) Tertiary education 1.72** 1.74** (0.029) (0.026) Parent of child 6 or younger −0.48** −0.42** (0.016) (0.016) Has partner in house 0.42** 0.47** (0.018) (0.018) Constant −8.94** −9.28** (0.367) (0.126) N 103,677 114,261 Source: EU- SILC 2008, 2012. Note: ** indicates statistical significance at p < 0.05. The coefficients operate mostly as predicted: gender (being a woman) is negatively associated with work engagement for both years. It is notable that the impact of gender decreased considerably between 2008 and 2012: as we saw in the graphs above, gender inequality in employment declined during the crisis years in CEE (as in many other EU countries). We should repeat here that the decline is due to higher rates of men’s job loss relative to women’s rather than actual gains for women. Next we move to our main models and interpret the cross-level interaction terms associated with each macro-level variable and gender. Table 4 presents selected output from random coefficient logistic regression models predicting the probability of “working.” The independent variables include all the individual-level variables shown in table 3, GDP as a control, and one of the macro-level variables. We test these one by one because of the small number of N at the macro level. The table we show contains the main effect of gender, the macro-level variable in question, and the cross-level interaction term of these two. Table 4. Main Effects and Cross-Level Interaction Terms from a Series of Random Coefficients Logistic Regression Models Predicting Working for Wages on Individual-Level and Macro-Level Variables Main variables 2008 2012 Woman Main variable* Interaction Woman Main variable Interaction GDP growth, 3-year ave −0.72** 0.075 ** −0.011 −0.71** −0.030 0.057 (0.26) (0.04) (0.04) (0.158) (0.048) (0.057) FDI flow, 3-year ave −0.79** 0.024** 0.001 −0.78** −0.038 0.059 (0.12) (0.010) (0.01) (0.24) (0.058) (0.069) FDI stock, 3-year ave −0.81** 0.003 0.001 −0.77** 0.000 0.003 (0.21) (0.003) (0.004) (0.309) (0.004) (0.005) Sectoral segregation −1.95** 0.016 0.056** −2.46** −0.048 0.086** (0.38) (0.026) (0.018) (0.834) (0.036) (0.038) Occupational segregation −2.51** 0.023 0.061** −3.26** −0.047 0.095** (0.65) (0.031) (0.023) (0.965) (0.035) (0.034) Size of public sector −1.79** −0.026 0.056** −1.61** −0.052** 0.053** (0.349) (0.027) (0.019) (0.510) (0.020) (0.026) Size of service sector −1.94** −0.002 0.020** −2.33** −0.022** 0.029** (0.441) (0.012) (0.008) (0.639) (0.010) (0.010) Kids lt 3 years old in care −0.956** −0.009 0.016** −0.680** −0.013 0.008 (0.094) (0.009) (0.007) (0.149) (0.008) (0.010) Kids 3–6 years in care −1.06 ** 0.000 0.004 −1.13** −0.006 0.008 (0.285) (0.006) (0.004) (0.413) (0.005) (0.006) Has 3rd paid parental leave −0.776** 0.070 −0.012 −0.512** 0.196 −0.196 (0.095) (0.172) (0.150) (0.131) (0.167) (0.207) Paternity leave days −0.905** −0.006 0.016** −0.789** −0.020** 0.024** (0.079) (0.009) (0.007) (0.118) (0.008) (0.010) 103,677 114,261 Main variables 2008 2012 Woman Main variable* Interaction Woman Main variable Interaction GDP growth, 3-year ave −0.72** 0.075 ** −0.011 −0.71** −0.030 0.057 (0.26) (0.04) (0.04) (0.158) (0.048) (0.057) FDI flow, 3-year ave −0.79** 0.024** 0.001 −0.78** −0.038 0.059 (0.12) (0.010) (0.01) (0.24) (0.058) (0.069) FDI stock, 3-year ave −0.81** 0.003 0.001 −0.77** 0.000 0.003 (0.21) (0.003) (0.004) (0.309) (0.004) (0.005) Sectoral segregation −1.95** 0.016 0.056** −2.46** −0.048 0.086** (0.38) (0.026) (0.018) (0.834) (0.036) (0.038) Occupational segregation −2.51** 0.023 0.061** −3.26** −0.047 0.095** (0.65) (0.031) (0.023) (0.965) (0.035) (0.034) Size of public sector −1.79** −0.026 0.056** −1.61** −0.052** 0.053** (0.349) (0.027) (0.019) (0.510) (0.020) (0.026) Size of service sector −1.94** −0.002 0.020** −2.33** −0.022** 0.029** (0.441) (0.012) (0.008) (0.639) (0.010) (0.010) Kids lt 3 years old in care −0.956** −0.009 0.016** −0.680** −0.013 0.008 (0.094) (0.009) (0.007) (0.149) (0.008) (0.010) Kids 3–6 years in care −1.06 ** 0.000 0.004 −1.13** −0.006 0.008 (0.285) (0.006) (0.004) (0.413) (0.005) (0.006) Has 3rd paid parental leave −0.776** 0.070 −0.012 −0.512** 0.196 −0.196 (0.095) (0.172) (0.150) (0.131) (0.167) (0.207) Paternity leave days −0.905** −0.006 0.016** −0.789** −0.020** 0.024** (0.079) (0.009) (0.007) (0.118) (0.008) (0.010) 103,677 114,261 Source: EU-SILC 2008, 2012. Note: The models control for GDP level as well as all individual-level independent variables listed in table 1. In each model, only one macro level variable is included in addition to the level of GDP, thus each line is a separate model. The main variable always describes the variable in the row. ** indicates statistical significance at p < 0.05. Statistically significant interaction terms are highlighted by grey shading. Contrary to the findings of previous research, neither the level of economic development, nor the pace of economic growth, nor the flow or stock of foreign direct investment (or the proportion of employment in foreign-owned companies, not shown here) are associated with the gender gap in employment. In other words, while growth and the flow of foreign investment were positively associated with the overall level of employment in 2008, these factors did not impact gender difference in access to paid work.5 By 2012, not even the main effects are statistically significant for these variables.6 To illustrate this finding, see figure 3, which displays a scatter plot of the logistic regression coefficients associated with gender from individual country models against the average growth rate in each country in 2012. There is no linear association. Figure 3. View largeDownload slide Logistic regression coefficients for gender plotted against GDP growth rate, EU-SILC, 2012 Figure 3. View largeDownload slide Logistic regression coefficients for gender plotted against GDP growth rate, EU-SILC, 2012 We also find, however, that the availability of female-typed jobs, measured either through the IP index or through the size of the service or state sectors, are positively associated with gender equality in employment. Women are more likely to be employed in these sectors than men, and when these sectors are large and growing, women benefit from this advantage in terms of employment chances. Due to widely shared convictions about the gender appropriateness of jobs, as well as women’s work history and experience in these sectors and occupations, men are unlikely to replace women even when male-dominated work is unavailable. These findings are consistent over time. A visual example of the relationship between segregation and the employment gap is presented in figure 4, which, as above, plots logistic regression coefficients associated with gender against occupation segregation in each country. Note the visually clear and statistically significant relationship between the two. This representation is also useful to point out cases that do not fit neatly in the story. In this context, the Czech Republic is such a case, where relatively high levels of segregation are associated with a large gender employment gap.7 Figure 4. View largeDownload slide Logistic regression coefficients for gender plotted against occupational segregation index with linear regression line, EU-SILC, 2012 Figure 4. View largeDownload slide Logistic regression coefficients for gender plotted against occupational segregation index with linear regression line, EU-SILC, 2012 The association between state support for raising children and the gender employment gap is much less obvious than the impact of job segregation. According to table 4, the availability of lengthy leaves is not significant, that is, lengthy leaves are not associated with either more or less gender inequality compared to leaves that are less than three years, and neither is the proportion of children in kindergarten. The variable describing nursery school (“crèche”) places for small children (under three years of age) is positively associated with more gender equality in access to paid work only in 2008, but this is one of the cases where adding 30 percent to the SE may eliminate statistical significance. We must be cautious when interpreting this finding. Finally, the length of paternity leave, while positively correlated with gender equality in employment, is also only weakly significant in both years. Discussion and Conclusion This analysis considered societal-level mechanisms associated with the gender employment gap in CEE. The region had a long history of full employment for both genders before 1990, thus variations across countries deepening by the beginning of the twenty-first century describe the nature of the post-socialist transformation itself. We find great divergence among the countries and show that the difference in men’s and women’s chances of access to paid work is most strongly associated with the structure of the local labor market. A wide variety of “neoliberalisms” (Ong 2007) are emerging in CEE countries with correspondingly divergent patterns of gender inequality. The labor market factors that correlate with more gender equality in employment include the presence of large public and service sectors and higher levels of occupational and sector-based gender segregation. In other words, we identified a macro-level trade-off between segregated work opportunities and inclusion in paid work in the ten CEE countries we studied. Women are best able to find employment in economies where sectors and jobs deemed appropriate for women are plentiful. These conditions may not be surprising in countries where cultural expectations for both genders are restrictive (Bukodi 2006; Fodor and Balogh 2010) and employer discrimination is rampant and unsanctioned. In this context, when typically female-typed jobs are more widely available, women are more likely to find paid work. Importantly, macro-economic development variables or welfare provisions matter less for women’s relative employment rates than predicted: the level and pace of economic development, the depth of global market integration, the length of parental leave, or even the availability of nursery schools for small children are not (or are only weakly) associated with the overall gender employment gap. This negative finding is another important contribution of our research. The implications of our findings are threefold. First, unlike in developing countries in Latin America or Asia, economic growth and foreign direct investment in CEE did not lead to the feminization of labor. The gender gap in employment does not seem to be associated with either of these factors. In other words, economic growth did not “trickle down” to create more equality between men and women. In terms of foreign investment, this finding may suggest that patterns of investment are region specific. Employers seeking investment in certain regions in order to take advantage of a cheap and docile female labor force may view CEE as less profitable given the high rates of women’s education, cultural capital, and labor market experience, and many countries’ full membership in the EU, which provides at least rudimentary labor protections. Some foreign investors may also seek investments in this region due to its proximity to Western Europe. For instance, the surge of business service–related investment in this region may be due to the geographical and cultural proximity to Western markets. Unlike many labor-intensive and feminized sectors (e.g., textile and electronics manufacturing) in Latin America and Asia, the business services industry employs men and women in equal numbers and thus such investments are unlikely to significantly impact the gender employment gap. We interpret this to suggest that theories on the relationship between neoliberally oriented economic development and gender equality typically overgeneralize: the impact of macro-economic development policy is conditioned on the characteristics of the local labor market context. States, labor unions, and other labor market institutions thus have an important role in managing economic development to the benefit of all. Second, our results show that occupational and sectoral segregation has important implications for the gender employment gap; rigid gender segregation in the form of feminization of certain sectors or occupations creates greater employment opportunities for women relative to men and offers protection for women during economic crises or economic transformations. This is so because when job loss is uneven and threatens male-dominated sectors or occupations (such as the recession of 2008 or that of 1991 in the region), entrenched gender-based job segregation typically prevents the replacement of women workers with unemployed men, thus making women less vulnerable than men to unemployment or job loss. Note that the current analysis does not consider the quality of women’s jobs, merely their availability. Segregation typically disadvantages women in terms of wages, prestige, support, and mobility; however, in some cases it may also buffer women against job loss. Thus, while jobs in highly female-dominated sectors may not be the best jobs, they do allow women to sustain paid work at higher rates, leading to greater parity in employment between men and women. We note that this low-road trade-off is rarely considered by the policy-related literature, which is typically concerned with the elimination of gender-based job/occupational and sectoral segregation or the trade-off between segregation and the wage gap. Indeed, while the empirical literature on gender inequality often describes various aspects of the phenomenon (access to work, segregation, access to positions of authority, the wage gap, etc.), theoretical writings on gender and work rarely acknowledge the potential inconsistency among these dimensions. Our paper identified one such inconsistency on the macro level (segregation and access to paid work). This finding thus supports the conceptualization of the notion of gender inequality (or a gender regime) as a multilayered, multidimensional, and potentially internally inconsistent phenomenon. Such conceptualization may contribute to understanding the difficulty of fighting against gender inequality and the ongoing illegitimacy of such struggles in the eyes of much of the population of CEE countries. Finally, while we find that the social policy context is only weakly associated with the gender employment gap, our results do show that the availability of childcare is potentially important for shaping women’s employment rates relative to men’s. Although CEE countries are less likely to provide childcare before mandatory school age than other EU member states, the care is more likely to be close to full-time (over 30 hours) than in most other states, especially for children over three years old (Mills et al. 2014). Such childcare enables mothers of preschool-age children to be working for wages rather than simple providing a few hours of babysitting for toddlers. The role of grandparents, especially grandmothers, in providing regular childcare is also especially important in this region (Utrata 2015). Thus, institutional childcare is only a precondition to work and, as our results show, if acceptable jobs are not available, women’s labor force participation will remain low. It should be noted, however, that CEE countries are atypical of the EU in general, thus the relevance of these findings may be limited to this region. Interestingly, and contrary to previous research, the availability of very long parental leave does not significantly affect the size of the gender employment gap, at least there is no variation among CEE countries, where parental leave policies are typically quite generous overall. Again we hypothesize that the choice of taking lengthy parental leave is dependent on the labor market context, specifically the scarcity of available jobs. When, as in Hungary, for example, women with small children enjoy fewer employment options, they are likely to stay on parental leave for longer periods, occasionally using this time for informal, undeclared part-time work (Frey 2008). At the same time, for example in Estonia, although state policies guarantee similarly long leaves, women enjoy many more opportunities for work and are thus less likely to take advantage of its full length. This suggests that, as the impact of economic development and foreign capital penetration, the implications of social policies on employment outcomes must be analyzed within specific labor market contexts. Overall, our results address the literature on “gender and development” by emphasizing the importance of labor market context as an intervening factor between social welfare and macro-economic policies and actual inequality outcomes. This project identified one such intervening variable: labor market structure, specifically segregation by gender, which is associated with the gender gap in access to paid work in post-socialist countries. Careful analyses of other similar factors would greatly contribute to our understanding of how economic development impacts social structure and local and global inequalities. Notes 1 Some countries, such as Poland, Latvia, and Slovakia, have introduced austerity-driven policies after 2008, which encouraged mothers’ withdrawal from the labor market (Smith and Villa 2014). These, however, were not expected to have a measurable impact on women’s overall employment access by 2012. 2 The proportion of private daycare options is negligible in CEE countries. Grandparents, however, are more likely to pitch in than in developed Western European countries (Utrata 2015); indeed, co-habitation of generations is more significant in this region than in many others within the EU (authors’ calculation based on SILC datasets). 3 Paternity leave is unpaid in Estonia, so we only counted the days for which fathers get paid. 4 We used the program MLwiN developed in the Centre for Multilevel Modelling in Bristol, UK (see Raschbach et al., A User’s Guide to MLwiN. Version 2.1e (London: Centre for Multilevel Modeling, Institute of Education, University of London, 2004). Ample descriptions of the precise mathematical formula of the models are available in the reference above and elsewhere, so we decided not to include it here, to save space. 5 As we note earlier, the countries we analyze represent the whole population of CEE countries, which were members of the EU at the time of the data collection. Even though we are dealing with a population, we carried out and interpreted tests of significance throughout our analysis. We do this—following the argument of Blalock (1972), Rubin (1985), and Gelman (2011)—in an effort to account for variation from sources other than sampling from a finite population. 6 We tested two more variables that are expected to have an impact on the gender gap: the availability of part-time work and the wage gap. We assumed that when part-time options are more numerous, women are more likely to be working, and if the wage gap is smaller, the incentives for women to be employed are higher. Neither of these variables showed a significant association with the gender employment gap. 7 If we omit the Czech Republic, the R-squared of the linear regression model increases from 0.40 to 0.60. Appendix Appendix A. Description and Source of Macro-Level Variables in the Models Labor market/macro-economic characteristics GDP at purchasing power parity as a percentage of EU 27, in 2008 and in 2012 Source: Eurostat online, accessed April 23, 2015. Three-year average GDP growth rate between 2006–2008 and 2010–2012 (authors’ calculations) Source: Eurostat online, accessed April 23, 2015. FDI stock as % of GDP in 2006–2008 and 2010–2012 Source: Eurostat online, accessed April 23, 2015. Occupational/sectoral segregation indexes (IP index) Source for 2008: “Gender segregation in the labour market: Root causes, implications and policy responses in the EU” (pp. 33–34), http://www.lm.gov.lv/upload/dzimumu_lidztiesiba/situacija_latvija/gendersegregation.pdf. Source for 2012: Francesca Bettio, Marcella Corsi, Carlo D’Ippoliti, Antigone Lyberaki, Manuela Samek Lodovici, and Alina Verashchagina. 2013, “The Impact of the Economic Crisis on the situation of women and men and on gender equality policies,” Expert Group on Gender and Employment (EGGE) European Commission, p. 76. Service sector and public sector employment as % of all employment Source for both years: UNECE gender statistics, online, accessed November 21, 2013 (http://w3.unece.org/PXWeb2015/pxweb/en/STAT/STAT__30-GE__03-WorkAndeconomy/002_en_GEWEEmploySPN_r.px/?rxid=2e1ac0a5-3b4a-4d64-8b97-e57dbf62c369 and http://w3.unece.org/PXWeb2015/pxweb/en/STAT/STAT__30-GE__03-WorkAndeconomy/003_en_GEWEEmplSectSPN_r.px/?rxid=2e1ac0a5-3b4a-4d64-8b97-e57dbf62c369). Work-family balance-related provisions Percentage of children under three in formal childcare institution, 2007 and 2011 (of relevant age group) Source: Eurostat, accessed December 16, 2014 Percentage of children three to six in formal childcare in 2007 and 2011 (of relevant age group) Source: Eurostat, accessed December 16, 2014 Information on the length of parental and paternity leave was collected from a variety of online sources as well as through consultation with local experts during the period of 2012–2014. About the Authors Eva Fodor is Associate Professor of Gender Studies at the Central European University in Budapest. She received her PhD in sociology from the University of California–Los Angeles, where her research focused on gender, labor markets, and post-communist transformations. Currently she studies how state institutions are reengineered under illiberal-neoliberal regimes and the relevance of gender and other types of inequalities in these processes. Christy Glass is a Professor of Sociology at Utah State University. She earned her PhD in sociology from Yale University in 2005. Her research and teaching focus on gender inequality, social policy, and work. Her current research focuses on factors that contribute to gender and racial/ethnic inequality in work organizations in Central Europe and the United States. References Bandelj, N. 2016. “ On Postsocialist Capitalism.” Theory and Society 45( 1): 89– 106. https://doi.org/10.1007/s11186-016-9265-z. Google Scholar CrossRef Search ADS Bandelj, Nina. 2007. From Communists to Foreign Capitalists: The Social Foundation of Foreign Direct Investment in Postsocialist Europe . Princeton, NJ: Princeton University Press. Bettio, F., M. Corsi, C. D’Ippoliti, A. Lyberaki, M. S. Lodovici, and A. Verashchagina. 2013. “The Impact of the Economic Crisis on the Situation of Women and Men and on Gender Equality Policies.” European Commission, Brussels. Retrieved from http://www.eif.gov.cy/mlsi/dl/genderequality.nsf/All/F7B0CF0F70F22D33C2257A770040E7D0/$file/crisis_report__dec_2012_en.pdf. Blalock, H. M. 1972. Social Statistics . New York, NY: McGraw-Hill. Blaskó, Z. 2005. “ Dolgozzanak-e a nők? A magyar lakosság nemi szerepekkel kapcsolatos véleményének változásai 1988, 1994, 2002. [Should women work: changing in gender role opnions in Hungary].” Demografia 48( 2–3): 159– 86. Bohle, Dorothee, and Bela Greskovits. 2012. Capitalist Diversity on Europe’s Periphery . Ithaca, NY: Cornell University Press. Boserup, E. 1970. Women’s Role in Economic Development . London: Earthscan. Braunstein, E. 2012. Neoliberal Development Macroeconomics: A Consideration of Its Gendered Employment Effects. HRSG. v. United Nations Research Institute for Social Development, Geneva (Gender and Development Programme Paper, 14). Retrieved from http://www.unrisd.org/80256B3C005BCCF9/(httpAuxPages)/F95D244010CF453DC12579AD0049AB47/$file/Braunstein.pdf. Bryan, M. L., and S. P. Jenkins. 2015. “ Multilevel Modelling of Country Effects: A Cautionary Tale.” European Sociological Review 32( 1): 3– 22. https://doi.org/10.1093/esr/jcv059 Google Scholar CrossRef Search ADS Budig, Michelle J., Joya Misra, and Irene Boekcmann. 2012. “The Motherhood Penalty in Cross-National Perspective: The Importance of Work-Family Polciies and Cultural Attitudes.” Social Politics: International Studies in Gender, State and Society 19: 163– 93. Google Scholar CrossRef Search ADS Budig, M. J., J. Misra, and I. Boeckmann. 2015. “ Work-Family Policy Trade-Offs for Mothers? Unpacking the Cross-National Variation in Motherhood Earnings Penalties.” Work and Occupations 43( 2): 119– 177. https://doi.org/10.1177/0730888415615385 Google Scholar CrossRef Search ADS Bukodi, E. 2006. “Women’s Labor Market Participation and Use of Working Time.” In Changing Roles: Report on the Situation of Women and Men in Hungary 2005 , edited by I. Nagy, M. Pongracz, and I. G. Toth. Budapest: Tarki Social Research Institute. Burawoy, M. 2015. “Travelling Theory.” Presented at the Open Democracy/ISA RC-47: Open Movements. Retrieved from https://opendemocracy.net/michael-burawoy/travelling-theory. Burchell, B., V. Hardy, J. Rubery, and M. Smith. 2014. A New Method to Understand Occupational Gender Segregation in European Labour Markets. European Union DG Justice. Charles, M., and David B. Grusky. 2004. Occupational Ghettos:The Worldwide Segregation of Women and Men . Stanford, CA: Stanford University Press. Ciccia, R., and M. Verloo. 2012. “ Parental Leave Regulations and the Persistence of the Male Breadwinner Model: Using Fuzzy-Set Ideal Type Analysis to Assess Gender Equality in an Enlarged Europe.” Journal of European Policy Research 22: 507– 28. Google Scholar CrossRef Search ADS Cukrowska-Torzewska, E. 2015. “ She Cares and He Earns? The Family Gap in Poland.” Ekonomia 42: 43– 7. Retrieved from https://ideas.repec.org/a/eko/ekoeko/42_43.html. Duncan, O. D., and B. Duncan. 1955. “ A Methodological Analysis of Segregation Indexes.” American Sociological Review 20( 2): 210– 17. Google Scholar CrossRef Search ADS Einhorn, B. 1993. Cindarella Goes to Market: Citizenship, Gender and Women’s Movements in East Central Europe . London, New York: Verso. Elborgh-Woytek, M. K., M. M. Newiak, M. K. Kochhar, M. S. Fabrizio, K. Kpodar, M. P. Wingender, B. J. Clements, and M. G. Schwartz. 2013. Women, Work, and the Economy: Macroeconomic Gains from Gender Equity. International Monetary Fund. Elson, D. 2009. “ Gender Equality and Economic Growth in the World Bank World Development Report 2006.” Feminist Economics 15: 35– 59. Google Scholar CrossRef Search ADS England, P., P. Allison, L. Su, N. Mark, J. Thompson, M. J. Budig, and H. Sun. 2007. “ Why Are Some Academic Fields Tipping toward Female? The Sex Composition of U.S. Fields of Doctoral Degree Receipt, 1971–2002.” Sociology of Education 80( 1): 23– 42. Google Scholar CrossRef Search ADS England, P., and S. Li. 2006. “ Desegregation Stalled: The Changing Gender Composition of College Majors, 1971–2002.” Gender & Society 20( 5): 657– 77. https://doi.org/10.1177/0891243206290753 Google Scholar CrossRef Search ADS Esteves-Abe, M. 2005. “ Gender Bias in Skills and Social Policies: The Varieties of Capitalism Perspective on Sex Segregation.” Social Politics Summer: 180– 215. Google Scholar CrossRef Search ADS Eurostat. 2011. Employment in Foreign-Controlled Enterprises. http://ec.europa.eu/eurostat/tgm/table.do?tab=table&init=1&language=en&pcode=tps00004&plugin=1, last accessed November 24, 2017. Fodor, E. 1997. “ Gender in Transition: Unemployment in Hungary, Poland, and Slovakia.” East European Politics and Societies 11( 3): 470– 500. Google Scholar CrossRef Search ADS Fodor, E., and A. Balogh. 2010. “ Back to the Kitchen: Gender Role Attitudes in 13 East European Societies.” Journal of Family Research 22( 3): 289– 307. Fodor, É., and B. Nagy. 2014. “ An Ebbing Tide Lowers all Boats.” Revue de l’OFCE ( 2): 121– 51. Forsythe, N. 2000. “ Gender Inequalities and Economic Growth: A Longitudinal Evaluation.” Economic Development and Cultural Change 48( 3): 573– 81. Google Scholar CrossRef Search ADS Frey, Maria. 2008. “Hungary: Possibilities of and Barriers to Employment of Women on Childcare Leave.” In International Review of Leave Policies and Related Research 2008 , edited by Peter Moss and Marta Korintus, 30– 34. London, UK: Department of Business, Innovation and Skills. Funk, N., and M. Mueller. 1993. Gender Politics and Post-Communism: Reflections from Eastern Europe and the Former Soviet Union . New York: Routledge. Gal, S., and G. Kligman. 2000. The Politics of Gender After Socialism . Princeton, NJ: Princeton University Press. Google Scholar CrossRef Search ADS Glass, Christy, and Eva Fodor. 2011. “Public Maternalism Goes to Market: Recruitment, Hiring and Promotion in Postsocialist Hungary.” Gender&Society 25. Gelman, A. C. 2005. “ Two-Stage Regression and Multilevel Modeling: A Commentary.” Political Analysis 13: 459– 61. Google Scholar CrossRef Search ADS Gelman, A. C. 2011. “How Do You Interpret Standard Errors from a Regression Fit to the Entire Population?” October 25. Retrieved from http://andrewgelman.com/2011/10/25/how-do-you-interpret-standard-errors-from-a-regression-fit-to-the-entire-population/. Gelman, A., and Jennifer Hills. 2007. Data Ananlysis Using Regression and Multilevel/Hierarchical Models . Cambridge: Cambridge University Press. Ghodsee, K. 2005. The Red Riviera: Gender, Tourism, and Postsocialism on the Black Sea . Durham, NC: Duke University Press. Google Scholar CrossRef Search ADS Glass, C. 2008. “ Gender and Work during the Transition: Job Loss in Bulgaria, Hungary, Poland and Russia.” East European Politics and Societies 22: 757– 83. Google Scholar CrossRef Search ADS Goldin, Claudia. 1990. Understanding the Gender Gap: An Economic History of American Women . New York: Oxford University Press. Gornick, J. C., and M. K. Meyers. 2009. “Institutions That Support Gender Egalitarianism in Parenthood and Employment.” In Gender Equality: Transforming Family Divisions of Labor , edited by Janet C. Gornick and Marcia K. Meyers. London, New York: Verso. Gornick, Janet C., and Marcia K. Meyers. 2003. Families That Work: Policies for Reconciling Parenthood and Employment. New York: Russell Sage. Hakim, C. 2002. “ Lifestyle Preferences as Determinants of Women’s Differentiated Labor Market Careers.” Work and Occupations 29( 4): 428– 59. Google Scholar CrossRef Search ADS Hakim, C. 2006. “ Women, Careers, and Work-Life Preferences.” British Journal of Guidance & Counselling 34( 3): 279– 94. https://doi.org/10.1080/03069880600769118 Google Scholar CrossRef Search ADS Hamm, Patrick, Lawrence P. King, and David Stuckler. 2012. “Mass Privatization, State Capacity, and Economic Growth in Post-Communist Countries.” American Sociological Review 77: 295– 324. Google Scholar CrossRef Search ADS Javornik, J. S. 2010. “Exploring Maternal Employment in Post-Socialist Countries: Understanding the Implications of Childcare Policies.” Doctoral thesis, University of Southampton. Retrieved from http://eprints.soton.ac.uk/173805/. Javornik, J. 2014. “ Measuring State De-Familialism: Contesting Post-Socialist Exceptionalism.” Journal of European Social Policy 24( 3): 240– 57. https://doi.org/10.1177/0958928714525815 Google Scholar CrossRef Search ADS Kanbur, S. M. R., N. Lustig, and World Bank, eds. 2000. Attacking Poverty . New York: Oxford University Press. Karamessini, M., and J. Rubery. 2013. Women and Austerity: The Economic Crisis and the Future for Gender Equality . London and New York: Routledge. Karmel, T., and M. MacLachlan. 1988. “ Occupational Sex-Segregation: Increasing or Decreasing?” Economic Record 64( 3): 187– 95. Google Scholar CrossRef Search ADS Keck, W., and C. Saraceno. 2013. “ The Impact of Different Social-Policy Frameworks on Social Inequalities among Women in the European Union: The Labour-Market Participation of Mothers.” Social Politics 20: 297– 328. Google Scholar CrossRef Search ADS Kmec, J. A. 2005. “ Setting Occupational Sex Segregation in Motion: Demand-Side Explanations of Sex Traditional Employment.” Work and Occupations 32( 3): 322– 54. https://doi.org/10.1177/0730888405277703 Google Scholar CrossRef Search ADS Korpi, Walter, Tommy Ferrarini, and Stefan Englund. 2013. “Women’s Opportunities under Different Family Policy Constellations: Gender, Class and Inequality Tradeoffs in Western Countries Re-Examined.” Social Politics 20: 1– 40. Google Scholar CrossRef Search ADS Kwon, R. 2016. “ Can We Have Our Cake and Eat it Too? Liberalization, Economic Growth, and Income Inequality in Advanced Industrial Societies.” Social Forces 95( 2): 469– 502. https://doi.org/10.1093/sf/sow077 Google Scholar CrossRef Search ADS Lenski, G. E. 1966. Power and Privilege: A Theory of Social Stratification . New York, NY: McGraw-Hill. Maas, C. J., and J. J. Hox. 2005. “ Sufficient Sample Sizes for Multilevel Modeling.” Methodology 1( 3): 86– 92. Google Scholar CrossRef Search ADS Mandel, H., and M. Semyonov. 2005. “ Family Policies, Wage Structures, and Gender Gaps: Sources of Earnings Inequality in 20 Countries.” American Sociological Review 70( December): 949– 67. Google Scholar CrossRef Search ADS Mandel, Hadas, and Moshe Semoyonov. 2006. “A Welfare State Paradox: State Interventions and Women’s Employment Opportunities in 22 Countries.” American Journal of Sociology 111: 1910– 49. Google Scholar CrossRef Search ADS Mies, M. 1999. Patriarchy and Accumulation on a World Scale: Women in the International Division of Labor . London: Zed Books. Milkman, R. 1976. “ Women’s Work and the Economic Crisis: Some Lessons of the Great Depression.” Review of Radical Political Economics 8: 71– 97. Google Scholar CrossRef Search ADS Mills, M., P. Präg, RAND Europe, European Commission, and European Union (EU) (Belgium). 2014. “Gender Inequalities in the School-to-Work Transition in Europe. Short Statistical Report No. 4.” RAND Europe. Westbrook Centre, Milton Road, Cambridge CB4 1YG, United Kingdom. http://www.rand.org/randeurope/. Misra, Joya, Michelle J. Budig, and Irene Boeckmann. 2011. “Cross-National Patterns in Individual and Household Employment and Work Hours by Gender and Parenthood.” Research in the Sociology of Work 22. Morgan, Kimberly J., and Kathrin Zippel. 2003. “Paid to Care: The Origins and Effects of Care Leave Policies in Western Europe.” Social Politics 10: 49– 85. Google Scholar CrossRef Search ADS Myant, Martin, and Jan Drahokoupil. 2011. Transition Economies: Political Economy in Russia, Eastern Europe and Central Asia. Hobokne , NJ: John Wiley and Sons, Inc. Nepomnyaschy, L., and J. Waldfogel. 2007. “ Paternity Leave and Father’s Involvement with Their Young Children.” Community, Work & Family ( 10): 457– 72. O’Conner, Julia S., Ann Shola Orloff, and Sheila Shaver. 1999. States, Markets, Families: Gender, Liberalism and Social Policy in Australai, Canada, Great Britain and the United States . Cambridge, England: Cambridge University Press. Ong, A. 2007. “ Neoliberalism as a Mobile Technology.” Transactions of the Institute of British Geographers 32( 1): 3– 8. Google Scholar CrossRef Search ADS Pettit, B., and J. L. Hook. 2005. “ The Structure of Women’s Employment in Comparative Perspective.” Social Forces 84: 779– 801. Google Scholar CrossRef Search ADS Pettit, B., and J. L. Hook. 2009. Gendered Tradeoffs: Family, Social Policy, and Economic Inequality in Twenty-One Countries . New York: Russell Sage Foundation. Reskin, B. F., and D. B. McBrier. 2000. “ Why Not Ascription? Organizations’ Employment of Male and Female Managers.” American Sociological Review 65( 2): 210. https://doi.org/10.2307/2657438 Google Scholar CrossRef Search ADS Rubery, J. 1988. Women and Recession . New York: Routledge. Rubery, J., and A. Rafferty. 2013. “ Women and Recession Revisited.” Work, Emplyment, Society 27( 3): 414– 32. Google Scholar CrossRef Search ADS Rubin, A. 1985. “ Significance Testing with Population Data.” Social Service Review 59( 3): 518– 20. Google Scholar CrossRef Search ADS Rutkowski, J. J. 2006. Labor Market Developments during Economic Transition (Vol. 3894). World Bank Publications. Saxonberg, S. 2014. Gendering Family Policies in Post-Communist Europe: A Historical-Institutional Analysis . London: Palgrave Macmillan. Google Scholar CrossRef Search ADS Saxonberg, Steven, and Tomas Sirovatka. 2006. “Failing Family Policy in Post-Communist Europe.” Journal of Comparative Policy Analysis 8: 351– 79. Seguino, S. 2000. “ Gender Inequality and Economic Growth: A Cross-Country Analysis.” World Development 28: 1211– 30. Google Scholar CrossRef Search ADS Seguino, S., G. Berik, and Y. Van der Meulen Rodgers. 2010. An Investment That Pays Off: Promoting Gender Equality as a Means to Financial Development . Berlin: Frierdrich Ebert Stiftung. Smith, M., and P. Villa. 2014. “ The Long Tail of the Great Recession.” Revue de l’OFCE 133( 2): 85– 119. Standing, G. 1989. “ Global Feminization through Flexible Labor.” World Development 17: 1077– 95. Google Scholar CrossRef Search ADS Stegmueller, D. 2013. “ How Many Countries for Multilevel Modeling? A Comparison of Frequentist and Bayesian Approaches: How Many Countries.” American Journal of Political Science 57( 3): 748– 61. https://doi.org/10.1111/ajps.12001 Google Scholar CrossRef Search ADS Takács, J. 2008. “ Ha a mosogatógép nem lenne, már elvál tunk volna...”: Férfiak és nők otthoni munkamegosztása európai öszehasonlításban. [“‘If we didn’t have a dishwasher we would be divorced by now’:The gender division of labor in the household in a European comparative perspective.” Esély ( 6): 51– 73. Utrata, J. 2015. Women Without Men: Single Mothers and Family Change in the New Russia . Ithaca, NY: Cornell University Press. Author notes Direct all correspondence to Eva Fodor, Department of Gender Studies, Central European University, Budapest, Nador utca 9, 1051-Hungary; e-mail: firstname.lastname@example.org. © The Author(s) 2018. Published by Oxford University Press on behalf of the University of North Carolina at Chapel Hill. All rights reserved. For permissions, please e-mail: email@example.com.
Social Forces – Oxford University Press
Published: Mar 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera