Превосходство демократии: пол и материнство

Since the 1980s, the number of democracies across the world has approximately doubled. This rapid spread has brought substantial changes to people's life. Most studies have shown that democratic development leads to increased levels of literacy, education, industrialization, urbanization, and the overall well-being of citizens. Other positive outcomes of democratization have been demonstrated to include a middle class of educated professionals and intelligentsia, the principal carriers of democratic values who foster civic engagement and the development of a civil society. The sentiment of liberal democracy persists in today’s world. Countries are required to democratize in order to receive financial aid or be able to open up to foreign investment (Robinson 2004, Wolf 2001). Colonial empires imposed democracy and market economies on former colonies (Bollen and Jackman 1985; Crenshaw 1995) and economic problems are believed to be solved with the adoption of democracy (Lederer 1992; Tarrow 1991).

Scholarly interest in the processes of democratization has visibly intensified in the beginning of the Third Wave of world democratization that encompassed the democratization of Southern Europe in the 1970s, Latin America in the 1980s, former Soviet bloc in the early 1990s and the re-democratization of post-colonial Africa in the mid-1990s (Bermeo, 1992; Bratton & Mattes 2001; Higley & Gunther, 1992; Linz & Stepan, 1996; O’Donnell, Schmitterand, & Whitehead, 1996; Pridham, 1990; Stark & Bruszt, 1998; Spears, 2003). In addition, recent years have brought a renaissance of scholarly investigations coupling democratization jointly with globalization (understood as development and diffusion of a global, market economy). Among examples are studies on the economic dominance of democratic countries (Schwartzman, 1998); feasibility of democratic growth conditioned by a country's development (Neubauer, 1967); interaction of democracy and development (Przeworski et al., 2000); predictive power of diffusion vs. development on democratic growth (Wejnert, 2005); discourses on the impact of development on democratization (Przeworski & Limongi, 1997); emergence of a global democracy and unified global market economy (Kellner, 2002); an association of democratic growth with the transition to a market economy (Cardoso, 2005, Gills 2005, Przeworski 1991, Tenue 2002); simultaneous transitions to democracy and market economy across the former Soviet bloc (Wejnert, 2002); and the relations between people’s quality-of-life and global free market and democracy (Inglehart, 1997: 23). In the broadest term, democratic growth is considered a sign of modernity and democratic transitions are understood as a synonym for improvement of societal standard of living (Lipset, 1960, 1994). Hence, many studies have focused on a positive impact of democratization on development and societal well-being (Garrett, 2004; Shafer, 1994).

One might assume, then, that the growth of democracy would also improve the well-being of women as reflected by an increase in their representation in the workforce, equal pay for equal jobs, equal educational opportunities, and improvement in women’s health care and longer life expectancy. However, during the transition to democracy in former Soviet countries, the data show that women's employment declined substantially. So did their inclusion in politics. Unemployment rose faster among women than men. In some countries, a decline in the provision of women's health care, especially medical assistance at birth, led to an alarming increase in maternal mortality (Funk & Muller, 1993; Wejnert & Spencer, 1996; Wejnert, 2003). It is, therefore, uncertain whether democratic growth leads to expected increases in standard of living and in life opportunities both for men and women.

Understanding the relationship between democracy and gender equality in terms of cost and benefits of democratization, I turn to viewing democratic growth through a lens of countries' economic development. In this paper I posit that in addition to benefits, there are also costs associated with democratization. Especially when coinciding with a growth of capitalistic economy (Henisz, Witold, Bennet Zelner & Mauro Guillen, 2005), democratization leads to major transformations of socio-economic structures as expressed by a redistribution of available resources, restructuring of class systems and employment structure, and adjustments in life opportunities and culture (Wejnert 2002a). These changes empower some societal groups while disempowering others (Chua 1995) and women seem to be bearing the costs disproportionately.

The impact of democratic growth and global free market economy on women relative to men is rarely studied in a comparative way and on a large scale (one of a few exceptions is a book by Chua (2003) however author addresses outcomes of global democracy and free markets in terms of ethnic inequality)¹.Responding to this paucity and adding to the ongoing discourse on a synergy of democratic growth and the transition to a market economy², the paper tests whether the democratic growth, defined as a dynamic process of «democraticness,» or changing level of democracy (Gurr, Jagger & Moore, 1990, Wejnert, 2005), equally benefits men and women in the well-developed, semi-developed, underdeveloped, and post-communist countries.

Using hierarchical growth models (a type of hierarchical linear models–HLM) this study examines 149 independent countries from 1970 to 2000, a time of the Third Wave of democratization and a beginning of an expansion of global market (Huntington, 1991, Rosenau, 2003)³. The HLM analyses are complemented by the data collected using questionnaire interviews with men and women in six recently democratizing countries of Europe–Hungary, Poland and the Ukraine, Asia–Kyrgyzstan, and Africa–Mali and Senegal. In each country the interviews were conducted approximately ten years after democratic transition or re-democratization was initiated⁴.

Democracy, Development And Societal Well-Being

Although scholars acknowledge that democracy enhances development and societal well-being, they disagree as to what extent and in what types of societies the enhancement is sufficiently visible. The many positions can be roughly placed into four views.

Following a neo-Marxist approach, students representing the first view argue that the growth in the level of democratization mainly benefits societies in the well-developed, core countries, further enhancing these countries' position in the world system. Accordingly, they believe that development, articulated via urbanization, industrialization and an increase of standard of living is a function of a country’s historically determined position in the world system and that level of development alters the processes of democratization. Due to the structure of the world’s capitalistic economy, poorer countries (semi-peripheral and peripheral) are disadvantaged in the world trade and hence are unable to catch up with the modernized, well-developed world. This disadvantage alters the level and rate of each country’s democratization by delaying democratization processes (Hechter, 1984; Snyder & Kick, 1979; Wallerstein 1999, 1998, 2002)⁵ and by reducing sustainability of a democratic system during times of economic difficulties (Brown, 1992; Herspring, 2003).

The second view posits that the poorest and least developed states are benefiting from global market economy and frequently assisting it democratization, as much as the most developed states, but developmental improvement in the semi-developed «middle-level economies» is not sufficient. One of the reasons for such effect is the distribution of foreign direct investment (FDI). As shown, outcomes of foreign investment are determined by countries’ strength of economy and by the economic sector the investment is placed in. Since foreign investment is mainly in raw material and standardized manufacturing, the low manufacturing profit is sufficient to benefit poor peripheries where workers’ wages are low. Peripheries also harvest from manufacturing technology developed in richer nations that, as the result of FDI, is transferred to poor countries (Shafer, 1994). In sum, global growth is good for the poor (Dollar & Kraay 2000).

More technologically advanced middle economies contain a labor force that is trained in services where foreign investment is inhibited by limited world scale need for a service economy. To gain sufficient benefits the service oriented labor force of middle economies would need to compete with core countries for a knowledge economy, but is not skilled enough to win the competition. Thus, as argued by Garrett (2004), across semi-peripheries, the improvement in societal well-being and development is largely conditioned by foreign investors’ readiness to invest in knowledge technology, which rarely happens. Opinions regarding the benefits of foreign investments to the host economy vs. economy depending on foreign capital, however, are mixed⁶.

The third approach sees democratic growth as a symbol of progress, wealth, high standard of living, liberalization, freedom and happiness. It is a force advancing technological and cultural progression that operates across all types of countries. With the proliferation of modern technology, media, and the Internet, possibilities are opening for governmental political discourses further proliferating democratic growth, international collaboration and the improvement of market driven economy (Castels 2000; Hutton & Giddens 2000; Henisz, Zelner & Guillen 2005). Opened channels for political discourses form a base for creation of transnational alliances between groups demanding social and political justice, and democratic freedom (Kellner, 2002). Democracy is thus the inevitable final destiny of world history and the modern man (Fukuyama, 1992). Globalization and democracy are brought together in the general theoretical concept of development understood as the process of integration of social and economic systems into more encompassing ones, and as bringing individuals into common systems of collective action on the principles of equality and accountability (Kearny, 2001; Teune, 2002). The future danger in the globalization era could be however, a battle between democratic liberalism expressed as cosmopolitan tolerance vs. fundamentalism (Giddens, 1999).

The final approach understands that democratization diffuses facilitated by countries collaboration and closeness of communication within economic and political networks; influence of democratic neighbors on non-democracies (the «neighboring effect»); role modeling of non-democratic countries on well-established democracies and an imposition of democratic principles by economically powerful countries. The diffusion is further augmented by modern media communication that facilitates the spread of democratic ideas and promotes global economic ties (Wejnert, 2005). Democratic growth facilitated by diffusion, aims for modernity and a higher standard of living.

Democracy, Development And Gender Equality

As feminist scholars posit, democracy and modern development entails the enhancement of women’s social status and position (Beneria 2003; Beneria & Bisnath, 2004). The opening of economic and political opportunities to women, according to the UN Ambassador Muhammad Yunus, yields not only the bare improvement of women’s economic situation, but also empowers women’s involvement in family decision making, increases women’s political and legal awareness, reduces domestic violence and enhances social mobility of their children⁷. Hence, numerous studies demonstrate that a stronger position of women leads to a stronger civil society, growth of women’s political participation, and in turn, an increase of women’s entrepreneurship, higher female literacy, lower birth rate, lower infant mortality, less domestic violence, and more nutritious family consumption (Coleman, 2004).

Studies also show that the inclusion of women into the top decision making political structures redresses the imbalance in policy investments, because women policy-makers tend to invest in areas frequently overlooked by male policy-makers yet relevant to families, such as educational institutions, public health facilities, social and children welfare institutions (Bertrand, Duflo & Mullainathan, 2002). Investment in those infrastructures are intrinsically important, especially in lieu of recent findings showing (using simulation analyses) that the closing of the gender gap in education by increasing girls’ education by 3 years, would result in a lower birth rate by 1 child. In turn, it would lead to an increase of GNP per capita by 25-30% within one generation in any given country (Coleman, 2004).

Contrastingly, it has been shown that frequent in non-democracies large gender gaps in literacy of men vs. women are associated with skewed gender ratios of boys to girls (more boys than girls) that indicates discriminatory practices of inadequate nutrition and health care for female children, sex-selective abortions, female infanticide and insufficient health care for mothers (Murthy, 2001; Harriss-White, 1996). Coleman (2004) following other studies thus concluded that when substantial gender inequality exists democracies are rare because equal rights and freedoms are replaced by prevailing notion of male dominance and brotherhood. Male dominated culture is difficult to substitute by democracy but it can be modified to incorporate limited women’s rights (Coleman, 2005).

One may assume then that democratizing countries emphasize and promote gender equality. However, studies on the position of women relative to men in democratizing post-communist Eastern Europe and Russiademonstrate an increase of gender inequality in comparison to prior non-democratic periods (Funk &Mueller, 1993; Hauser, Heyns & Mansbridge, 1993; Wejnert & Spencer, 1996). At the time of democratic growth, women, as a result of dual roles as producers and mothers, and due to more tenuous employment status, faced a plethora of problems not experienced by men. Among these problems were a higher unemployment rate relative to men, unequal access to economic and political resources, and the feminization of poverty (Issraelyan, 1996; Lissyutkina, 1993; Wejnert, 1996 a, b). Specifically, by the late 1990s women’s unemployment skyrocketed to 60-70% in Russia and the Ukraine, while rates for men were also high but reached 40-50% (Kholodkovskioi, 1998; Zherebkina, 2000).

This contrasted sharply with the communist period when the female employment level was one of the highest in the world (e.g., as early as the 1970s the percentage of women 15 years of age and over in the labor force, on average 63 percent, substantially exceeded that of the USA--42 percent, and far exceeded that of most Western European countries, including Italy-28 percent, Spain-18 percent, and West Germany-39 percent) (First-Dilic, 1973; Bodrova & Anker, 1985). The disadvantaged position of women in the labor market was further amplified by gender discrepancies in earnings and by state propaganda encouraging women employees to return to domestic duties (Paradowska, 1992). Subsequently, in the first decade of the transition to democracy, women's status declined vis-à-vis that of men (Wejnert, 2003). For Eastern European and Russian women, political and economic gains of democratic and market economic transitions were overshadowed by negative consequences that women seemed to bear disproportionately (Drakulic & Slavenka, 1993; Reading 1992).

This study hypothesizes that democratic growth does not necessary lead to the improvement of gender equality, and for women, the outcomes of democratization vary depending on the level of a country’s socio-economic development. To test this hypothesize, the study presents analysis of the balance of costs and benefits of democratic growth on women relative to men and to society-at-large, using empirical data. First, the study investigates effects of democratization on women’s health and well-being across the world, and comparatively across the well-developed, underdeveloped, semi-developed and post-communist countries applying multilevel analyses. Second, the statistical analyses are supported by analyses of data collected during field studies in countries experiencing democratic transitions.

Methodology

Database

To conduct the study I constructed a database by merging variables derived from existing datasets including data on indicators of women’s well-being and of gender equality⁸. The database contains data on all sovereign countries, i.e., independent members of the international system in the world that had populations greater than 500,000, (Gurr, 1974; Gurr, Jagger & Moore, 1990; Jagger & Gurr, 1995a, 1995b). In order to examine the impact of democratization and market economy on the outcomes described above, three key sets of variables were depicted: (a) level of democratization, (b) socioeconomic variables, and (c) gender equality variables. To compare the effects of democratization on gender equality across all countries in the world with the effects across the poorest, semi-developed and wealthy countries, the worldwide analyses were followed by comparable analyses of groups of countries recorded according to their position in the world system of the core (well-developed), semi-peripheries (semi-developed) and peripheries (underdeveloped) countries (Wallerstein 1974). In addition, the effect of democratization was also assessed for countries of the former Soviet bloc to account for its distinct pathway of politico-economic development (Goldfarb, 1991; Nee & Stark, 1989; Stark & Bruszt, 1998)⁹. To record a country’s position in the world system, I used the Snyder & Kick (1979) classification supplemented by its more recent modification (Bollen & Appold, 1993; Smith & White, 1992)¹⁰.

Measurement of Democratic Growth

Countries can be viewed categorically as democratic or not, yielding a scaling of democracy as either «1» or «0» . This is not in keeping, however, with the dynamic nature of democracy, which represents a continuous variable, as many countries that have adopted democratic principles and are called «democratic» are only partially so (Dahl, 1998) and as countries’ level of democracy changes over time (Wejnert, 2005). Thus, from a measurement perspective, democracy can be viewed as a continuous, dynamic variable where one would refer to the changes in the level of democratization of a country (Gurr, 1974; Gurr, Jagger & Moore 1990; Jagger & Gurr, 1995a, 1995b). In addition to reflecting the actual nature of democracy, a continuous index of democracy increases statistical power of analyses because the outcome variables have enhanced variation relative to a categorical index. The continuous index of democracy used here is based on the Polity III and IV data (Jagger & Gurr 1995a, 1995b). In this system, a maximum score of 10 depicts a fully developed democracy, whereas a minimum score of 0 represents a lack of democracy.

Measurement of Socioeconomic Development and Gender Equality

Based on theoretical and empirical accounts of socioeconomic development, gender equality, as well as prior research on democratization (Bollen & Jackman 1989; Crenshaw, 1995; Lipset, 1994), several indicators (measures) of socioeconomic development and gender equality were depicted for this study¹¹. The yearly measures of each indicator for each country allowed us to observe changes in the indicators as a function of time as well as changes corresponding to democratic growth.

Grouped by their categories the indicators are:

Socioeconomic development (i) GNP per capita (GNP/c) measured in $1000 units¹²; (ii) non-agricultural labor force participation, and (iii) literacy. Although GNP per capita is the main measure of a country’s economic development, in order to incorporate poorly developed countries with a high GNP level (e.g., oil producing Saudi Arabia) additional measures are used that reflect level of development. In this study, literacy rate is measured as the percentage of each country’s population that is literate and non-agricultural labor force is measured as the percentage of the population who were part of the paid labor force in industry and services.

Gender equality (i) paid women’s labor force participation measuring women’s labor force as a percentage of the total labor force in each country; (ii) women’s literacy rate as the percentage of adult females 15 years and older who are literate; (iii) men’s literacy rate as the percentage of adult males 15 years and older who are literate (iv) ratio of women to men’s primary education as the percentage of girls among pupils in elementary schools; (v) ratio of women to men’s secondary education as the percentage of girls among pupils in secondary schools; (vi) maternal care measured as the percentage of births attended by professional health staff as the number of total births¹³; (vii) the fertility rate depicting the total number of births per women of a reproductive age; (viii) maternal mortality representing the ratio of maternal mortality per 100,000 live births; and (ix) women’s life expectancy as the duration (number of years) of a women’s life calculated at birth.

Statistical models

The strategy adopted in the analysis was to measure the impact of time and democracy on the various outcome variables (indicators) of socioeconomic development and gender equality by utilizing variations in democracy both within each country and across countries (to depict, for instance, the effects of a country’s historical legacy of democratic system or effect of countries regional location that might promote diffusion of gender equality independently of democratic growth). As prior studies attest variations within and between countries are significant contributors affecting the democratization processes and their outcomes (Wejnert, 2005; Przeworski, & et al., 2000)¹⁴. Neglecting this hierarchical structure would lead to underestimating the standard errors of the coefficients, which might lead to the interpretation that effects are significant when they are not. Multilevel modeling was thus used to account for the effects within and between countries.

Prior to performing these analyses, the time variable was rescaled to make the year 1980 = 0, when the mass democratic movements swept across Latin America and democratic movements started to emerge within the Soviet bloc (e.g. the Polish Solidarity movement, the pro-market economic reforms in Hungary, or the pro-democracy cultural movement in Yugoslavia)¹⁵. Each outcome variable was modeled as a function of time, democracy, and the interaction of time and democracy. In this way, one could interpret the coefficient on time and democracy as providing the correlation between the movement of time and of democratic growth¹⁶ on changes in the particular outcome variable. The interaction of democracy with time provides evidence on whether these correlations are strengthening or weakening over time.

The following equation summarizes the two-level hierarchical growth model employed for the world:

Y_ij=( β₀₀ + β₁₀Year_ij + β₁₀ Democ _ij + β₁₁(Democ_ij)(Year_ij)+ (e_oj + e_1jYear_ij + r_ij),

where e_oj ~ N(0, τ₀₀) e_1j ~ N(0, τ₁₀) r_ij ~ N(0, σ²)

Y_ij – the outcome variable (an indicator)

β₀₀ represents an average level of an indicator (intercept) in 1980

β₁₀Year_ij represents an average level of an indicator as a function of time

β₁₀Democ_ij represents an average level of an indicator as a function of democracy

β₁₁(Democ_ij)(Year_ij) represents an average level of an indicator as a function of time and democracy

τ₀₀ represents variation in an indicator level between countries (between countries’ intercepts)

τ₁₀ represents variation between countries’ temporal rate of an indicator change (between countries’ slopes)

σ² represents residual (the within country variance)

_ij – the subscript denotes the within country level (the level-1)

_{j –} the subscript denotes the between countries level (level-2)

The dependent variable Y is explained with an intercept β₀₀ and a slope β₁₀Year_ij(a function of time)_, as well as a function of democracy and a function of time and democracy. The fixed part of the model contains fixed effects for the intercept (β₀₀), for the effect of time (β₁₀Year_ij), for the effect of democracy (β₁₀ Democ_ij), and for the effect of democracy and time [β₁₁(Democ_ij)(Year_ij)]. The random part contains three estimates of variances: for the intercept (e_oj), representing variation in an indicator between countries in the world; for the slope of time (e_1j Year_ij), representing variation in the slope of the temporal rate of an indicator’s change between countries in the world; and for the within country residual (r_ij)_, representing variation in an indicator’s level within countries or the departure from the predicted score of the i^th country’s actual score of an indicator in 1980.

To allow the intercept and the slope to vary across countries, a structure of the variance-covariance was selected using the goodness-of-fit statistics, and the UN (unstructured) structure was indicated as best fitting the data (Singer and Willett 2003)¹⁷. Comparison of the results of the UN model with the simple model, which did not impose additional structure on the error covariance matrix (beyond the heteroscedastic structure of the intercept and slopes as outcome models) indicated that, once the covariance of the intercepts and slopes had been introduced, no additional autoregressive error structure needed to be added. Nonetheless, one more test was performed, with results leading to the same conclusion¹⁸.

The modeling was implemented in «SAS PROC MIXED,» a procedure that allows multilevel hierarchical modeling (Singer, 1998; Stinger and Willett 2003). It should be noted that in subsequent models, the fact that observations within the same country are more similar than observations among different countries creates dependence. This lack of independence was expressed as an intra-class correlation and was accounted for in the multilevel modeling.

Analysis of Perceived Effects of Democratic Growth on Women

To complement the results of multilevel regressions this paper includes study of perceived effects of democratization on quality of life as reported by women and men in recently democratizing countries. The data were collected inHungaryandPolandin 1998-1999,Ukrainein 2000-02,Kyrgyzstanin 2003-04 and re-democratizingSenegalandMaliin 2004; approximately ten years after the transition to democracy from a non-democratic system was initiated in each country.

The first two countries, Poland and Hungary, are similar in many ways: they are both located in East-Central Europe and hence are influenced by European culture and gender roles; before democratization were communist states with enforced gender equality and women's rights; had compatible average levels of development; had about the same ratio of men to women; a similar mean population age; and were about the same size.UkraineandKyrgyzstanused to be republics of the communist Soviet Union, but are located on different continents and are developed differently:Kyrgyzstanis peripheral country, and is located in more traditional Central Asia, while the more affluentUkraineis in the Eastern part ofEurope.Kyrgyzstan, being a predominantly Muslim country, is culturally similar to the African countries ofSenegalandMali, but is influenced by its communist past in terms of the rights for women.SenegalandMali, both prior French colonies, follow different developmental trajectories thanKyrgyzstan, but by the end of 1980s their average level of development was about the same.

To depict similarities and differences between countries, research in this study was conducted using the same interview questionnaire that was initially developed by Stycos, Wejnert and Tyszka (2001, 2002), and translated into four languages: Polish, Hungarian, Russian and French. The interviews were conducted with married women, age 18-49, and, to control for the reliability of assessment, with 25% of their husbands (every 4^th husband). In all six countries, the research was conducted in villages located near large cities that were capitals of geographical and administrative regions¹⁹. In the Polish study, the selected villages were located near Poznan and Konin (capitals of central and west districts) and in the second round, near Zielona Gora (capital of the south-west district); in the Ukraine near Kiev, Ukraine’s capital, and Kharkov–the industrial center and the second largest city in the Ukraine; in Hungary near the town of Gödöllo, North-East of Hungarian capital Budapest; in Kyrgyzstan, South and North of the country’s capitol Bishkek; in Senegal and Mali, near capitals’ Dakar and Bamako. Rural communities experienced greater economic hardships during the democratic transitions than urban communities (Brown & Bandlerova, 2000), but in villages located near cities, the urban influence was expected to enhance women’s economic opportunities, e.g., possibilities for selling farm products or crafts in city markets or commuting to jobs in the cities; as well as their well-being, e.g., closeness to medical facilities. In sum, it was assumed that in these villages, women would have a higher satisfaction with life domains than would be the case of women in remote villages.

In the study, respondents were asked to assess their own quality of life (also called subjective quality-of-life–QOL) prior and after democratic growth in each country. Based on studies carried out over the past forty years, perceived well-being or QOL, is best assessed by specific life concerns (also called life domains) such as jobs, housing, family, or income (Andrews & Inglehart, 1979; Andrews &Robinson, 1991). The predictive power of these life concerns probably derives from the fact that they appear to apply relatively equally to both rural and urban dwellers, men and women, young and old adults, renters and home owners, and homemakers and working women (Campbell, 1981; Haavio-Monnila, 1992; UNESCO 1983; Veenhoven, 1993).

The subjective QOL correlates with demographic and social classification variables (such as income, level of education, age and sex), and current living conditions, but when examined together, these factors typically do not explain more than 10% of the variance in people’s assessment of life domains and their satisfaction with life-as-a-whole (life in general) (Andrews & Robinson, 1991). Similarly, social psychological research has demonstrated that subjective perception of one’s QOL is associated with personality traits such as self-esteem, locus of control, depression or alienation (Costa & McCrae, 1980). Moreover, strong religious feelings are one of the determinants of QOL (Campbell, 1981; Campbell, Converse & Rogers 1976). Nonetheless, at the level of population studies, psychological traits and religiosity are only weakly correlated with subjective QOL.

To assess the factors that account for globally-rated perceived QOL, US scholars have developed and refined a special type of questionnaire method containing a list of 123 life concerns (domains) that could be integrated into 30 meaningful clusters of life activities (Andora 1980; Andrews & McKennel, 1980; Andrews & Robinson, 1991; Andrews & Withey, 1976; Campbell, Converse, & Rogers, 1976; Stycos, Wejnert & Tyszka, 2001). Because of possible differences in the most significant life concerns between other countries and theUS, all 30 concern clusters are often included in comparative international studies.

The questionnaire used in this study was built according to specifications designed by prior research on perceived well-being and contained standard clusters of life concerns. These clusters were grouped into fifteen main categories depicting women’s degree of satisfaction with (a) their family life, (b) their own housing, (c) the availability and quality of free time, (d) their health and the health care system, (e) political freedom, (f) cleanness of their environment, (g) the activities of religious institutions, (h) the performance of the economic system, (i) the activities of the government and the political system and (j) their own income.

Each category of concern was rated by each respondent on a 7-point word-anchored scale (where 1 means very dissatisfied and 7 very satisfied with a life domain) of Andrews and Withey (1976) which has been shown to provide high retest stability (Andrews & Robinson, 1991)²⁰. At the end of the questionnaire, respondents were also asked for their degree of general satisfaction with life-as-a-whole, which employed the same 7-point rating scale. In addition, respondents were asked for the depicted changes in their well-being since democratization had been introduced. In order to assess the reliability of the interviews, respondents were interviewed by trained interviewers and asked to respond twice to each question: a) how their life is currently, and how it was in the past using the 7-point word-anchored scale of Andrews and Withey (1976) and b) select a picture of «face measure»–emotional face measure of satisfaction (Andrews and Withey, 1976)²¹. To control potential bias of order effects, respondents were counterbalanced for response time-period order, half reporting on the past first, half on current time first. Accordingly, recall periods of up to 10 years have been used reliably in assessing QOL (Andrews 1991; Andrews & Withey 1976)²².

RESULTS

Descriptive statistics on the Effects of Democratization

As shown in Table 1, the comparative assessment of some of the most important indicators of women’s quality of life across countries in the world indicates that, on average, women’s quality of life, their economic position and gender equality is higher in democratic than in non-democratic states.

The Effects of Growth of Democracy on Women’s Well-being in Democratic vs. Non-democratic Countries in the world in 2000*

Notes: *Maternal death is measured as number of maternal deaths per 100,000 births given per year.

For illustrative purposes, a few countries representing each category and an average level of each indicator for well-established democracies and for non-democratic countries are presented in Table 1 (see Table 1). The demonstrated differences in women’s well-being are not longitudinal projections and attest only to differences between democratic vs. not democratic countries. The data do not accurately depict the position of women in countries that are in the process of transition to democracy or in countries where the democracy level is fluctuating. A strong example is Russia which had a democracy level of 2 (on a scale of 0-10) in the early 1990s and became more democratic with a score of4 in the mid-1990s and moved to a democracy level of 0 by early 2000s and where changes in democracy level correlated with varying social position of women.

A strong example is Russia which had a democracy level of 2 (on a scale of 0-10) in the early 1990s and became more democratic with a score of4 inthe mid-1990s and moved to a democracy level of 0 by early 2000s and where changes in democracy level correlated with varying social position of women.

The interaction between democratic growth and the well-being of women is illustrated on the comparative, world-wide scale in empirical analyses presented below.

Multi-Level Models

Tables 2, 3 and 4 show results from multi-level models for the world, as well as comparatively for the developed, semi-developed, underdeveloped and post-communist countries with each row of the table presenting the results for a different outcome measure. In order to provide information on the general time trends in the sample and compare them with the effect of democratization, each outcome measure is presented as an unconditional model that includes time effect and as a conditional model that includes democratization and time effects. The random parts of the models are presented in the Appendix (see Appendix, Table A [DR1] ). Cross-world models: The results of a broader world community support the common assumption of scholars, policy makers and the public alike that democratization improves societal well-being and thus is beneficial to countries’ modern development. When looking at the effects of time in comparison to the additional effect of democratic growth on indicators of societal development, it is evident that democratic growth further enhances the temporal trend of increase of societal development, i.e., correlates with an increase of GNP per capital (in the equation of democracy and time with GNP/c) and an increase in literacy rates (in the equation of democracy as well as democracy and time)²³.

Cross-world models: The results of a broader world community support the common assumption of scholars, policy makers and the public alike that democratization improves societal well-being and thus is beneficial to countries’ modern development. When looking at the effects of time in comparison to the additional effect of democratic growth on indicators of societal development, it is evident that democratic growth further enhances the temporal trend of increase of societal development, i.e., correlates with an increase of GNP per capital (in the equation of democracy and time with GNP/c) and an increase in literacy rates (in the equation of democracy as well as democracy and time)²³.

Table 2

Predicted Effects of Growth of Democracy on Women and Society-at-large Across the World: 1970-2000¹

In regards to women’s well-being, at first glance it seems that there are no discrepancies in the positive effects of democratic growth on women in comparison to the society-at-large, because indicators specific to women’s health are improving, i.e., fertility rate is reduced, maternal care is improving and maternal mortality is declining over time. However, a closer investigation of findings points to unexpected effects.

First, women’s labor force participation is negatively associated with the growth of democracy (negative covariate estimates in the equation of democracy with women’s labor force in Model 1, Table 2) where a yearly increase in the level of democracy by one reduces women’s labor force participation by -.11% while no effect on the labor force in society-at-large is depicted. In the equation of time with women’s labor force, however, women’s participation in the labor force is indicated to increase over time. Considering that from 1970 – 2000 in economically developed stable democracies the percent of women incorporated into the labor force either did not change or slightly increased²⁴ (Bodrova & Anker, 1985; United Nations, 2001; Wejnert, 1996a), the depicted decrease must reflect a change in either new or transitional democracies of less developed countries. I return to this hypothesis when discussing the results of countries categorized by the level of their development.

Second, a decline in primary and secondary schooling of girls, measured as a ratio of girls to boys in primary and secondary schools, is depicted with the growth of democracy over time (see Models 3 and 4, Table 2). It could be that the indicated decrease of women’s labor force participation caused a decline of the status of women in families and lowered women’s influence on decisions making within families, which eventuated in the low enrolment of female children in schools²⁵.

Third, unexpectedly and contradicting comparative statistical assessment of democratic and non-democratic countries (presented in Table 1 above), female life expectancy is indicated to decrease with an increase of democracy, i.e., the increase of democracy by a score of 1 per year correlates with the decrease of women’s life expectancy by –.05 year. Considering that according to literature, life expectancy in well-developed, strongly democratic countries is either stable or steadily increasing; the decrease must be influenced mainly by lower developed democratizing countries. I expect analyses of groups of countries to shed more light on these findings [DR2] .

Models of Groups of Countries: To assess whether the general findings are similar for all countries regardless of differences in their development level, the same models were assessed in cross regional analyses of the well-developed (core), semi-developed (semi-peripheral), underdeveloped (peripheral) and post-communist countries (see Tables 3 and 4).

Table 3

Predicted Effects of Growth of Democracy on Women and Society-at-large in Core, Semi-Peripheral, Peripheral and Post-Communist Countries: 1970 – 2000¹

Table 3. Continues….

Table 3. Continues….

Notes: *Coefficient at least twice its standard error. Values in parentheses depict standard errors. ** For clarity reported here are only conditional models that

predict the effect of democratic growth on outcome variables.

Total number of observations in core countries N=436, number of countries n=17; in semi-peripheries N=747, number of countries n=32; and in peripheries N=2480, number of countries n=100, post-communist countries N=426, number of countries n=20. ¹ In addition to being a separate

category, post-communist countries are also included in peripheries and semi-peripheries.

Table 4

Predicted Effects of Growth of Democracy on Women and Society-at-large by 1and 5 years of Democratic Growth in Core, Semi-Peripheral, Peripheral and Post-communist Countries: 1970-2000

Table 4. Continues….

*Coefficient at least twice its standard error. Values in parentheses depict standard errors. In the table are presented conditional models.

Total number of observations in core countries N=436, number of countries n= 17; in semi-peripheries N=747, number of countries n= 32; in peripheries N=2480, number of countries n=100; in post-communist N=426, number of countries n=20. In addition to being a separate category, post-communist countries are also included in peripheral and semi-peripheral group. ¹Dependent variables (indicators) are lagged by 1 and by 5 years.

Models of Groups of Countries: To assess whether the general findings are similar for all countries regardless of differences in their development level, the same models were assessed in cross regional analyses of the well-developed (core), semi-developed (semi-peripheral), underdeveloped (peripheral) and post-communist countries (see Tables 3 and 4).

Core countries: In core, well-developed countries, democratization is positively associated with most of the indicators of women’s well-being. The female labor force participation increases with an increase of democratization over time. Female health improves as indicated by the decline in fertility rate, improvement in maternal care, and maternal mortality decline, and an increase in life expectancy. The coefficient of democracy in the fertility equation is -.004; the coefficient of democracy in maternal mortality equation is -.13 and the coefficient in the interaction equation is -.029; the coefficient of democracy in the maternal care equation is +.05 and the coefficient in the interaction variable of the maternal care is +.005; and the coefficient of democracy in the female life expectancy equation is +.07. At the same time, women’s education increases as indicated by the positive coefficient of democracy in the elementary schooling equation (albeit not significant, perhaps due to limited variance in already achieved relatively high levels of education). In sum, in well-developed countries, democratization benefits women by increasing their job opportunities, schooling, health and expanding their life span.

The positive effect of democracy on indicators of societal well-being is also depicted. Of the tested indicators, the societal literacy level is expected to increase while the impact on already high GNP per capita is not significant. At the same time, the societal non-agricultural labor force was shown to decline suggesting an even stronger positive impact of democratic growth on expansion of women’s non-agricultural labor force (see Table 3).

Positive effects of democratic growth are also indicated in the analysis of democracy predicting indicators of women’s well-being by 1 and 5 years of democratic growth (see Table 4)²⁶ . Specifically, models with democracy effects lagged by 1 years conform initial decline of women’s labor force but the decline is reduced from -.11 to -.09 (p<.0001). In model with democracy effect lagged by 5 years, the negative effect is eliminated yielding positive coefficient of +.011 (p<.0001) (see Table 4). The maternal care is also indicated to improve with time; life expectancy is increasing (the indicator of life expectancy increases in variable of democracy and in the interaction variable of time and democracy); fertility is continuously decreasing in models predicting fertility by 1 and 5 years of democratic growth; maternal mortality is indicated to decrease; and there are no negative effects on girls’ schooling. Similar positive effects on indicators of societal well-being are depicted in lagged analyses.

Semi-peripheries: In contrast, negative effects of democratic growth on women’s well-being in semi-peripheries are depicted. Women’s participation in the labor force is predicted to decline with increasing democratization in the interaction equation, contrasting depicted increase of women’s labor force over time (coefficient of the effect of time on women’s labor force is +.17). Women’s education is also indicated to decline. In an average country, the percentage of females among students in elementary and secondary schools is indicated to decline by -.05% and -.1% respectively. Moreover, these indicators also decline in the interaction variable of time and democracy, with coefficients of -.007% in elementary and -.015% in the secondary education equation. The negative effects are even stronger in light of absence of influence of democratic growth on men’s literacy (neither in the equation of democratic growth or democracy with time).

The depicted increase in maternal mortality is also disturbing especially that it coincides with a decline in female life expectancy. With an increase of democracy level by a score of +1, female life expectancy is expected to decline on average by -.009 per year and maternal mortality to increase by +1.4 maternal deaths per 100,000 live births. The negative outcomes on women’s well-being are expected to continue over time as indicated by the negative coefficients in the equations of democracy and time with labor force, and life expectancy indicators.

The only positive impact of democratic growth is the decrease in fertility rate and the increase in maternal care. As literature indicates, during democratic growth modernization of medical facilities and better training of medical personnel takes place in part due to the increase of foreign aid and professional contacts with medical personnel in well-developed countries. An increase of maternal care attests to the possibility to reverse the negative trend in maternal mortality and life expectancy, under the condition of sustainability of democracy. However, the sustainability of democratic growth in not well-developed countries is uncertain, as new fragile democracies often revert to totalitarianism or became dangerous democracies (Herspring, 2003; Owen, 2005; Przeworski et al., 2000).

At the societal level, results of the impact of democratic growth on societal developmental are mixed in semi-peripheries. As shown, democratization does not inevitably lead to an overall higher societal standard of living; the coefficient of the interaction of democracy with GNP per capita (GNP/c) is negative and significant (-.09) but positive in the equation of democracy and time. The initial negative trend is consistent with economic decline in transitional democracies described in literature (Herspring, 2003), and with prior investigations on the disadvantaged position of semi-peripheral, middle economies in the processes of the global diffusion of democracy and market economy (Garrett, 2004). The impacts on other indicators of societal well-being are not significant except for the indicated decline of societal literacy overtime, which in part accounts for the shown decline in female enrolment in elementary and secondary schooling.

The analysis of the democracy predicting women’s and societal well-being by one and five years of democratic growth indicate a further reduction of women’s well-being after one year of democratic growth and only in some indicators a limitation of the negative effects after the fifth year. Specifically, female life expectancy is predicted to decline by the one year of democratic growth with no improvement indicated by the fifth year. The ratio of girls to boys in primary and secondary schooling is predicted to decline across five years. The negative effects on women’s labor force are eliminated but not reversed to a positive effect. Contrastingly, the societal non-agricultural labor force is predicted to increase by the fifth year of democratic growth but the decline of the GNP/c is not eliminated. In addition, the negative effect on female schooling is correlated with a decline of societal literacy level across the five years (see Table 4).

Peripheries: It seems that women in peripheral countries do not benefit from democratic growth either. However, in contrast to semi-peripheries, some indicators of women’s well-being become positive in the interaction models of democratic growth and time.

First, the observed decrease in women’s labor force participation weakens over time; the coefficient on democracy is negative (-.14), while the interaction of democracy with time is positive (+.005). Second, in contrast to semi-peripheries, a negative effect on maternal mortality rate is not detected. Third, the regression in the ratio of girls to boys students over time is indicated (the coefficient on the interaction variable of democracy with time is -.01 and -.13, respectively), but the coefficients are not significant. Fourth, democratization positively influences fertility levels (with an increase of democracy level by one, women’s fertility rates decreased by -.0013% over time). Fifth, in the life expectancy equation, the coefficient of the interaction variable is still negative (-.007) and significant.

Particularly alarming is the negative effect of democratic growth on female life expectancy in semi-peripheries and peripheries. Although more research needs to be conducted to explain the phenomena, a plausible explanation could be a correlation between traditionalism and diffusing models of a small family. At the time of democratization and modernization of economy, practices of family planning diffuse (e. g., indicated by the decline in fertility rate) (also see Bongaarts, 1994; Bongaarts & Watkins, 1996; Rosero-Bixby & Casterline, 1993). In poorer countries, small family models coincide with the traditional cultural preference for male offspring and the new expectations often lead to sex-selective abortion, female infanticide and inferior baby-girls’ nutrition and health care (Murthy 1996). Thus, on one hand, declining fertility indicates improvement of future women’s health; on the other, in traditional, poorer countries, declining fertility could signal the deliberate sex-selective practices indicating a latent negative function of democratization and diffusing models of small families (Coale & Banister, 1996; Sabu, 1997; Shin, Lee & Young, 1981)²⁷.

In respect to societal development, no negative relations between democratic growth or democracy in interaction with time and socioeconomic indicators were found. The positive effect of democratic growth over time on women’s labor force participation; no negative effects on maternal mortality; positive albeit not significant impact on schooling; and no negative effects for a society–at–large, attest to the validity of Garrett's (2004) hypothesis that peripheries are better off during democratic growth than semi-peripheries and are consistent with arguments that the outcomes of globalization are determined by the local developmental conditions of each country (Rosenau, 2003).

Similar to semi-peripheries, the analysis of the democracy predicting women’s and societal well-being by one and five years of democratic growth indicate a persistent decline of women’s well-being. Women’s labor force is indicated to decrease after 1 and 5 years of democracy growth; female life expectancy is declining after the first year and no improvement is demonstrated after the fifth year of democratic growth; ratio of girls relative to boys in elementary and secondary schools decreases; maternal care is indicated to decrease by the first year with no improvement indicated by the fifth year. However, the positive effect on the fertility rate sustains. The worsening women’s well-being contrasts no negative impact of democracy growth on societal quality-of-life (see Table 4).

Post-communist countries: Being an aggregate of mixed low to middle level economies with high literacy rate and a high provision for women’s health, post-communist countries follow their own path when reacting to global democratic growth and free market economy (see the last part of the Table 3).

First, in sharp contrast to the results for countries worldwide and across prior groups of countries, a number of women’s health indicators are negatively associated with an increasing level of democracy. Female life expectancy is negatively associated with democratization (the coefficient on democracy in this model is negative and statistically significant) and the decline is observed with the increase of democracy level rather than in the equation on democracy and time suggesting an immediate negative effect of democratic growth that fades with time (albeit insignificantly).

Second, unexpectedly, the fertility level increases where the coefficient on democracy in the fertility equation is positive and statistically significant, but this effect is reduced over time (the coefficient on the interaction of democracy and time is negative). According to the literature, the fading with time negative effect may depict substantial differences in women’s attitude about fertility before and after mid-1990s. There is some evidence that newly democratic governments promote larger families in order to solve problems associated with growing unemployment, but years after the initiation of a democratic transition, women facing economic difficulties decided to move back into the labor market and forego larger families (Albanese, 1996, David & Titkov, 1994; Kligman 1992; Kurczewski, 1990).

Third, similar to peripheries and semi-peripheries the coefficient of women’s labor force participation and democracy equals -.25 and is statistically significant. This indicates that an increase in the level of democracy by a score of +1 (on scale 0-10) is associated with a decline of -.25% in women’s labor force participation. However, the positive coefficient in the interaction of time and democracy suggests that with time, the negative association between women’s labor force participation and democracy becomes less strong. In contrast, the relationship between democracy and measures of the societal labor force was not significant, but the coefficient was negative. Considering a strong negative effect on women’s labor force participation, we may expect that the non-significant, negative correlation in the equation of democracy and societal labor force was driven largely by females.

Fourth, former communist countries used to provide equal educational opportunities to men and women, which led to gender equality in education (Stycos, Wejnert & Tyszka, 2002; Wejnert, 1996a). As indicated by a declining ratio of girls to boys in elementary and secondary schools overtime, this equality may be lost during democratic transition. It is plausible that the re-birth of cultural traditionalism is one of the causes of the declining school attendance of girls in these countries. Democratizing former communist countries often express their newly gained freedom by returning to traditional gender roles that were banned during communism. Consequently, the early marriages of girls are being promoted, motherhood is prioritized over women’s employment, and women’s domestic duties are glorified, which reduces parental interest in schooling of girls (Wejnert & Djumabaeva, 2004).

On the societal level (see last rows of Table 3), the only significant effect is depicted in the interaction of democratic growth with the GNP. The coefficient in the equation of democracy and time with the GNP per capita was negative (-.02) and significant, suggesting a decline in countries' economic development as the processes of democratic and market economic transitions unfold. These relationships are consistent with literature discussions of the economic decline experienced by transitional democracies of the former Soviet bloc (Herspring, 2003; Kolodko, 2001, 2002), which I assume, in part caused the depicted increasing gender inequality²⁸ .

In post-communist countries, many of the negative effects of democratic growth on women’s life are constant by the fifth year of democratic growth. Women’s labor force decreases in democracy and democracy with time equation. A regression is indicated in the ratio of girl to boy students in primary education over time (despite a depicted increase of the coefficient in democracy equation). By the fifth year of democratic growth, female life expectancy increases (the coefficient of democracy in female life expectancy equation is +.02), but the coefficient in the interaction equation is -.002.

The impact of democratic growth on societal well-being fluctuates between the positive and negative coefficients but the negative outcomes prevail. The negative coefficients are depicted in the interaction of democracy with GNP per capita (GNP/c) and societal non-agricultural labor force, as well as the interaction of democracy and time with GNP/c. The positive coefficient, however, is indicated in the equation of democracy and time with literacy (see Table 5).

Table 5

Summary of Outcomes of the Predicted Effects of Growth of Democracy on Women and Society-at-large in the World, as Compare to Core, Semi-Peripheral, Peripheral and Post-Communist Countries: 1970-2000

In all models reported in Tables 2, 3 and 4, the low p values for the effect of democratic growth and a significant difference in the obtained values of -2LL between models with and without democracy variables confirms the statistically significant relationship between indicators of societal and women’s well-being and countries’ democratization. In the final part of the study, I test if the findings are supported by subjective, perceived changes in women’s quality of life in democratizing semi-peripheral and peripheral countries.

Perceived Quality of Life Models

Table 6 provides results from the questionnaire interviews conducted on the perceived (or satisfaction with) quality of life in the recently democratizingPoland,Hungary,Ukraine,Kyrgyzstan,SenegalandMali. During the interviews, women and men subjectively assessed changes in social, economic and political changes as well as changes in women’s quality of life since the initiation of democratic reforms in their countries.

Although reported assessments vary across life domains and countries, few conclusions can be drawn in light of depicted similarities of responses.

First, the cross-countries comparison reveals a clear pattern of pair-like clustering of responses: Poland and Hungary versus Kyrgyzstan and Ukraine versus Senegal and Mali. This pattern overlaps paths of countries' economic and political developments suggesting that a country’s economic and political situation determines satisfaction with result of transition (Rueschmeyer, 1998; Saleci 1992) more than, for instance, shared historical and cultural roots²⁹.

Second, satisfaction with life in general varies across countries, but the level of satisfaction seems to be most strongly determined by countries’ economic performance.Kyrgyzstan andUkraine have experienced a downward trend in economic development since the time of democratic growth (Kolodko, 2002), where respondents reported the lowest satisfaction with life in general (3.57 and 3.71, respectively, on scale 1-7 with 7 meaning «very satisfied»). In the economically progressing countries ofSenegal andMali, democratic growth coincided with the highest life satisfaction (4.75 and 4.08, respectively). Furthermore, only inMali, the poorest of selected countries, respondents believed that their life prior to democratic growth was worse than during the years following the democratic transition. An argument that the democratic growth is less costly for low developed states and the least developed countries gain more from the transition to a global market economy and democracy than semi-developed was supported (Dabrowski & Antczak, 1996; Garrett, 2004; Kearny, 2001).

This finding was complemented by reports on satisfaction with the national economy and the activity of new democratic government where low satisfaction was reported in economically declining post-communist countries, while satisfaction was much higher in economically progressingMaliandSenegal. Accordingly, the depicted mean in satisfaction with state economy ranged from 2.11 to2.15 inpost-communistUkraineandKyrgyzstan, was slightly higher in

Table 6

Level of Satisfaction and Dissatisfaction with Selected Life Domains as Expressed by Women in Poland, Hungary, Ukraine, Kyrgyzstan, Mali and Senegal at the end of the 1990s and early 2000*

Table 6. Continues……

Note: *Subjective assessment of Quality of Life of women. Percent of women who are strongly satisfied are reported as score 6-7 (on 1-7 satisfaction scale with 7 depicting very strong satisfaction), and strongly dissatisfied as 1-2 score (on the same 1-7 satisfaction scale). **Change in satisfaction with life domains is reported on scale 1-3; where 1 depicts better, 2=same, 3=worse satisfaction). ¹ Due to freshly gained religious freedom and religious activity being still sensitive issue, the data on satisfaction with activity of religious institution were not collected in Kyrgyzstan. ² To avoid bias of an assessment, the change in women’s position and women’s employment was assessed by women and men combined.

PolandandHungary(2.57 and 2.88, respectively), but in postcolonial African states ranged from 3.91 to 3.39. Satisfaction with the activity of democratic government ranged from 2.08 to 2.99 across post-communist states, while inSenegalandMali, was 3.75 and 3.70; satisfaction with political freedom ranged from 2.15 to3.35 inpost-communist countries while inSenegalandMaliwas 4.08 and 3.64, respectively. The satisfaction with macro economic and political changes was similar to satisfaction with personal income that ranged from 2.75 to3.33 inpost-communist states while was 3.90 and4.17 inSenegalandMali.

Third, in addition to the division of countries into two groups of respondents of more versus less satisfied with democratic changes, the less satisfied post-communist countries were further divided into subgroups: Poland and Hungary, where nearly 40% of respondents claimed to be either very dissatisfied or dissatisfied with economic and political activity of democratic government and Kyrgyzstan and Ukraine, where almost 70% respondents were dissatisfied (see Table 6).

In particular, while in Polandand Hungaryrespondents were particularly satisfied with gains in political freedom, this was not the case in Ukraine and Kyrgyzstan where reported satisfaction was near or at the level 2.00. One explanation could be that in the former republics of the Soviet Union, the concept of political freedom was frequently misinterpreted as the complete elimination of governmental accountability to its citizens shown by, for example, the falsification of elections, as well as limited crime control. The lack of control led to excessive corruption, unaccountability of regimes, drug trade, the spread of Mafia and other crimes, all of which according to public opinion, resulted from the extreme leniency of the democratic government (Wejnert, 2003).

The reported satisfaction with gains in political freedom was similar to the perception of gains in civil rights. As expressed by the early 2000 grass-root protests, the impingement of Presidents and re-run elections, the depicted societal dissatisfaction was a sign of germinating societal revolt against the ruling elites inUkraine andKyrgyzstan that, in the public eye, were not democratic.

The environmental conditions were one of the major areas of concern in Ukraineand in Poland, but in Hungary, where the democratic government focused on the protection of the natural environment, 54.3% of respondents reported to be very satisfied with environmental conditions. Furthermore, relative to Hungary, Ukraineand Kyrgyzstan, Polish respondents reported low satisfaction with religious life and the activity of the church. During the communist period, Poland, unlike any other Soviet bloc country, placed fewer restrictions on religious practices and the Polish church played a strong role in the initiation of democracy. Therefore gains in freedom of religion were not as visible as in other countries, and hence, not much appreciated.

Fourth, regardless of the limited rights for women in predominantly Muslim African states and their relatively low income, African women were much less critical of newly democratic states’ governing, and state policies. The limited criticism observed inSenegal andMali sharply contrasted reports in post-communist countries. Specifically, only 10-20% of women inSenegal andMali were dissatisfied with the economy, government and environment, a striking contrast to the 40-70% of women in post-communist countries. Less than 20% of African women were dissatisfied with their own income, compared to more than twice as many in the former communist state. Comparison of these results with economic progress further supports hypothesis of a greater gain of democratization experienced by peripheral than semi-developed countries.

Fifth: across all countries, the most satisfying elements were the domains of private life, such as one's family, life accomplishments, personal health, relationships with others, the amount and quality of free time, community life and housing. These domains were assessed at equally high (4.00–4.5 on scale 1-7) levels of satisfaction across all countries. Any assumption of differences in psychological or cultural predisposition's of the more optimistic African respondents versus the more pessimistic European and Central Asian respondents should be therefore excluded (Czapinski, 1994). Next part of the analyses contains discourse on plausible causal explanations of reports.

Reported changes in women’s well-being following the democratic transition: As Table 6 shows, opinions regarding changes in women’s societal position after democratic transition significantly differed between African and the former communist countries. Specifically, in Senegal and Mali respondents reported enhancement of women’s social position evaluating their social position before the democratization as a score of 2.7 and 2.3 on scale 1-3 (where 1 means better and 3 worse). Women’s employment opportunities, income and own life were reported to improve (on the same scale 1-3 were reported as being score of 1.8, 1.7 and 2.2 prior to democratization) (although according to literature the employment opportunities are still limited in these countries). In contrast, respondents in post-communist states reported decline in women’s social position indicating past position as 1.2-1.3 score (only Polish respondents were less critical). In particular, women’s income and employment opportunities were perceived to decline (before democratization reported as score of 0.5) and past life opportunities were reported to be better (score of 1.3-1.8). The exception were Hungarian women reporting that they life opportunity did not significantly changed.

What are the potential causal explanations of the differences?

First, it is plausible that the difference in perceived outcomes of democratic growth on women could be a function of culture. Accordingly, respondents inSenegal andMali subdued and accustom to male dominance may be more compliant with changes introduced by predominantly male political authorities. Contrastingly, the more critical attitude towards democratic growth in post-communist democracies could be explained by the ideology of gender equality introduced by the communism system, which could raise nostalgia for the communist ruling while increasing criticism of post-communist democratization.

Second, it is also plausible that the differences are cause by political factors. In African countries, due to limited women’s right prior to democratization, gains made toward women’s basic rights to education, inheritance or custody of children in case of divorce (Coleman 2005) could be perceived as significant improvements brought about by democratic change. At the same time, in formerly communist states gains made during democratization in freedom of speech, allowed a previously voiceless society to openly criticize democratic ruling elites (Binyon 1983).

Third, not disregarding the cultural and political implications, in lieu of this study’s findings the most compelling seems to be socio-economic explanation according to which the progress of economic development in countries would influence women’s perception of quality of life. Accordingly, substantive opinion of democratic changes seems to be formed in relation to the fluctuation of women’s social position which correlated with economic development, where the social position would be strengthening or weakening by women’s gained (or loss) economic opportunities, services and rights.

Thus,SenegalandMalihave experienced rapid modernization of specific to women medical and other facilities and services relative to the prior democratization period. The modernization in part was supported by foreign aid (conditioned, for instance, by required girls’ schooling and other rights for women), which probably affected more positive opinions about impact of democracy on women’s life. Contrastingly, negative responses in democratizing post-communist countries correlated with experienced economic difficulties, limitation of women’s employment and deterioration of medical maternal services (Romaniuk 2002; Solcova 1984; Stephenson 1998; Titkov 1993).

Overall, the results on the subjective perception of quality of life, support findings of the multilevel regressions showing that democratic transitions and the globalization of free market economy are costly for all developing countries, they are more costly to women than men and are positively influenced by an economic progress. By the tenth year of democratic growth, as results of perceived changes in women’s position indicate, in economically progressing countries a reversal of the trend in indicators of women’s social position and life opportunities is observed. This reversal occurs independently of the observed average level of development, i.e., it is reported in relatively low developedMaliandSenegal, and to some extent inPolandandHungary, but not in the economically decliningUkraineandKyrgyzstan.

CONCLUDING COMMENTS

The study adds to the scholarly debates on democracy in three ways. First, conceptually, it points to the importance of the extremely broad perspective that Mills is famous for and that allowed this study to address the complexities of efforts to extend democracy along with its unanticipated consequences. Second, methodologically, the study provides a foundation for scholarly work on the development of methodologies that carefully detect complexities of good intentions in conjunction with unforeseen consequences. Third, considering a feminist perspective, the study offers a new conceptualization of the outcomes of democracy on women showing that societal gains from democratization and global economic development are not gender neutral.

Specifically, in contrast to most studies indicating that the growth of democracy improves societal development and people's well-being, and that it is gender neutral, this study demonstrate that when democratic growth is assisted by a country's inclusion to a global market economy, in all except for highly developed countries in addition to positive effects, there also are costs associated with the transitions, which are disproportionately paid by women. The negative effects on women's well-being are in terms of decline in women’s representation in the workforce, unequal educational opportunities, decline in women’s health care and shortening of life expectancy. These effects become even stronger after one year of democratic growth and are reduced but not eliminated by the fifth year; it is only after ten years, as the reported subjective perception of QOL suggests, a reversal of the trend in some indicators and mainly in poorer, peripheral countries is observed.

In lieu of literature (Barendt & Musiolek 2005; Wejnert & Djumabaeva 2004), it seems plausible that the disadvantaged treatment of women is partly the result of newly democratic governments’ issuance of unintentionally gender discriminatory policies early on during the democratization period. Specifically when co-opting with the economic instability, transitional leadership, and limited resources, ruling elites limit provisions specific to women. For instance, in the early 1990s in democratizing post-communist states, women's employment declined substantially as well as their inclusion in politics. Unemployment rose faster among women than men, while in an attempt to artificially reduce unemployment rate, the state sponsored propaganda to encourage women to enter domestic spheres (Wejnert 1996). In new democracies, birth clinics and maternity units in hospitals were closed and resources were relocated to provide for other impoverishing hospital units (Wejnert & Djumabaeva 2004). To encourage foreign investment, governments in democratizing countries pass policies allowing foreign industry to employ workers at low wages and without benefits. Most frequently, those low wage workers are women (Barendt & Musiolek 2005).

A closer scholarly and policy look at the outcomes of these two major global trends, democracy and global free market, provides material to generate protective, pro-women policies during the initial few years of democratic growth. Such policies could mark an era of building greater gender equality across the world by sheltering domains of women’s well-being that are shown to decline. Examples of such policies could be tax breaks and institutional incentives provided to industry for hiring women or keeping employed women laborers; micro-financing and provision of state funds to support the development of women’s entrepreneurship; tax breaks for women-run enterprises; increased age limits of marriage for women; family allowances for each school-age child that attends primary or secondary school; tax exemption for organizations that provide health services and maternity benefits for women; tax exemption and the provision of state funds to women’s health clinics and health units in hospitals; and lower tuition costs for students specializing in gynecology and specialized women’s medical services.

Gender equality could reverse the trends reported by the United Nations Commission on the Status of Women (Morgan 1984) that while women represent half of the global population and one-third of the labor force, they receive only one-tenth of the world's income and own less than 1 percent of world property. Women perform 60-80% of all agriculture work in the word and in the developing countries they produce more than 50% of food (in African countries over 90%). Nonetheless, of 22 million people that die yearly from starvation, the majority are women and children under 5 years old.

The need of a special attention paid to pro-women’s policies and services can be best summarized by the following statement of Kofi Annan, secretary general of the United Nations «study after study has shown that there is no effective development strategy in which women do not play a central role. When women are fully involved, the benefits can be seen immediately: families are healthier; they are better fed; their income, savings and reinvestment go up. And what is true of families is true of communities and, eventually, of whole countries.» (Annan 2002).

Gender equality is desperately needed and would greatly honor women’s place in the world, but it is not accomplished by democratization alone.

¹ Chua addresses unexpected outcomes of global democracy and free markets in terms of ethnic inequality and argues that free markets have concentrated wealth in the hands of resented ethnic minorities, including theUS as a "market-dominant minority". "Market-dominant minorities" include Chinese in Southeast Asia, Croatians in formerYugoslavia, Whites in Latin America, Lebanese in West Africa, and Jews in post-communistRussia and theMiddle East

² Herein, synergy is understood as the combined action of democratic growth and market economy that is greater than the sum of their effects individually.

³ In this paper I focus only on two aspects of global changes—the spread of democracy and its assistance in the globalization of a market economy, understanding however, that the complete process of globalization incorporates also social and cultural dimensions. I examine the social and cultural dimensions only insofar as they interact with the discourse on global democratization.

⁴ Data were collected inPoland in1999, in theUkraine in 2000-2002, inKyrgyzstan in 2003-04, and inSenegal andMali in 2004.

⁵ The evidence regarding the world system approach is mixed (e.g., Bollen & Appold, 1993).

⁶ While one group of scholars argue that dependence on foreign investment is detrimental to countries’ economy (Chase-Dunn 1975; Dixon and Boswell 1996), another forcefully argues the opposite (Firebaugh 1999). To solve the problem, Kentor & Boswell (2003) show that in the middle-economies insufficient foreign investment initially negatively effects countries’ economic development, however, it also stimulates within-country economic initiatives and innovativeness that contribute to reversal of the trend within 15 years.

⁷ Professor Muhammad Yunus, an economist fromBangladesh, set up “Village Bank” (micro-credit) and promoted the formation of local government based on the participation of rural people. He was appointed by the UN Secretary General to the International Advisory Group for the Fourth World Conference on Women in Beijing from 1993 to 1995, served on the Global Commission of Women's Health (1993-1995) and in 2002 was appointed UNAIDS Ambassador.

⁸ The database was pre-tested in prior studies (Wejnert, 2003; Wejnert, 2005). The data was derived from the database Nations, Democracy and Development (Wejnert, 2001) particularly its subset Gender, Democracy and Development created by merging data from Bank's (1993) Cross-National Time Series, 1815-1973; the World Bank’s (1999) World Development Indicators 1960-1998; The World’s Women 2000: Trends and Statistics (UN, 2001); Polity III data (Jagger & Gurr 1995b) and coding variables from descriptions of political, economic, and social institutions in volumes of the Statesman's Yearbook (2000); the World Handbook of Political and Social Indicators (Taylor & Jodice, 1983); Osmanczyk's (1982) Encyclopedia of the United Nations and International Relations, and the journal Freedom Review published by Freedom House. In the database all indicators were recorded annually from 1970-2000 for 149 independent members of the international system. For the analyses, the years 1970-2000 were selected to represent the last three decades of the twentieth century, the so-called “third wave” of worldwide democratization, a period which saw a substantial increase in transitions to liberal democracy and market economy (Huntington, 1991). The database is being prepared for commercial use (Wejnert, 2001).

⁹ The following 20 countries were selected as former communist countries: East Germany, Poland, Hungary, Czechoslovakia (Czech Republic), Rumania, Bulgaria, Russia, Moldavia, Georgia, Azerbaijan, Turkmenistan, Kyrgyzstan, Uzbekistan, Albania, Estonia, Latvia, Lithuania, Ukraine, Belarus, Armenia.

¹⁰ As core countries were recorded: Australia, Austria, Belgium, Canada, Denmark, France, Germany (West Germany), Greece, Italy, Japan, Luxemburg, Netherlands, Norway, Sweden, Switzerland, United Kingdom, United States, Yugoslavia; the semi-peripheries are: Argentina, Bosnia-Herzegovina, Burma, Columbia, Croatia, Cuba, Cyprus, East Germany, Finland, Hungary, India, Iran, Ireland, Israel, Jordan, Kenya, South Korea, Kuwait, Lebanon, Malaysia, Macedonia, Pakistan, Peru, Philippines, Portugal, Russia, Singapore, South Africa, Spain, Sri Lanka, Turkey, United Arab Emirates, Uruguay, Venezuela; and peripheries are: Algeria, Bangladesh, Benin, Bolivia, Brazil, Burundi, Cameroon, Central African Repub., Chad, Chile, Colombia, Congo, Costa Rica, Dominican Repub., Ecuador, Egypt, El Salvador, Ethiopia, Fiji, Gabon, Gambia, Ghana, Guyana, Honduras, Indonesia, Iraq, Ivory Cost, Jamaica, Liberia, Libyan Arab Rep., Madagascar, Malawi, Mali, Mauritania, Mauritius, Mexico, Morocco, Nepal, New Zealand, Nicaragua, Niger, Nigeria, Oman, Panama, Papua New Guinea, Paraguay, Poland, Rwanda, Saudi Arabia, Senegal, Sierra Leone, Somalia, Syrian Arab.Rep.,Tanzania,Thailand,Togo,Trinidad&Tobago,Tunisia,Uganda,Upper Volta,Yemen Arab. Rep., Zaire, Zambia, Zimbabwe, Sudan, Bahrain, Guinea-Bissau, Morocco, Afghanistan, Angola, Mozambique, Namibia, Lesotho, Swaziland, Bhutan, Moldavia, Bulgaria, Rumania, Iceland, Georgia, Azerbaijan, Turkmenistan, Tajikistan, Kyrgyzstan, Uzbekistan, Kazakhstan, China, Mongolia, North Korea, South Korea, Cambodia, Laos, South Vietnam, North Vietnam, Haiti, Guatemala, Czechoslovakia (Czech Rep.) Slovakia, Albania, Estonia, Latvia, Lithuania, Ukraine, Belarus, Armenia. Due to missing data on indicators of societal and women’s well-being nine countries are not included in the study:Bosnia-Herzegovina,Croatia,Macedonia,North Vietnam,Tajikistan,Kazakhstan,Mongolia, Yemen Arab.Rep.,Zaire,Denmark.

The world system measure depicts the world system position of countries as fixed at its 1950 level, which, considering arguments about the dynamic nature of countries’ position in the international market (e.g., Smith & White 1992), presents a limitation to the analysis. Nonetheless, to test my finding further, I conducted the same analysis with the modified variable of the world system position that reflected dynamic changes in measurements during last decades as indicated by Smith & White (1992). The control analysis yields similar results (the coefficient of the variable “world position” in analyses of democratic growth changed from -.8 (s.e.=.3) in this study to -.9 (s.e.=.4) in the control study. o my knowledge, there is no available dynamic measure of countries’ positions in the world system for all sovereign countries covering the duration of my study. Such a study needs to be conducted in the future.

¹¹ Most researchers studying democratization are focusing on its impact on society-at-large; therefore, to demonstrate the distinctiveness of effects of democratization on women, this study compares effects on society-at-large with impact on women.

¹² GNP per capita is the gross national product, converted to US dollars using the World Bank Atlas methods, divided by the midyear population. GNP is the sum of gross value added by all resident producers plus any taxes (less subsidies) that are receipts of primary income (employee compensation and property income) from nonresident sources. Data are in current US dollars. For more information see World Bank (1999).

¹³ The category of professional health staff accounts for all medically trained professionals including doctors, nurses and midwives.

¹⁴ For example findings showing interdependence of the effect of a country’s GNP/c and regional economic affluence on democratic growth (Wejnert, 2005) or arguments that stable democracies, primarily clustered in region of Europe and Americas, have the highest GNP/c (Przeworski, & et al., 2000).

¹⁵ The movements are examined by, for example, Banac (1992), Bermeo (1992), Gredelj (2002), Nee and Stark (1989), and Wejnert (2002).

¹⁶ The democratic growth depicts increases and decreases in level of each country democratization.

¹⁷ The UN model was selected out of the tested CS, CSH, HF, ARH, AR, and UN models, using goodness-of-fit statistics from Akaike's Information Criterion (AIC); Akaike's Information Criterion corrected for sample size (AICC); Bayesian Information Criterion (BIC); and a subsequent Likelihood ratio test (–2 RLL). The UN structure indicates that the model does not place any structure on the variance for intercepts and variance for slopes, nor is any structure imposed on the covariance between these two.

¹⁸ One of the strengths of PROC MIXED is that it allows adding complexity to test for autoregressive error structure once the covariate has been taken into account, and to compare different structures for the error covariance matrix. In this study, following Singer (1998), the extra complexity was added to the assessment of the covariate estimate for the autoregressive parameter, and fit statistics pointed toward the conclusion that no additional complexity of the autoregressive error structure needed to be added to the final model.

¹⁹ The data were collected by Wachowiak and Wejnert in Poland in 1999 (306 respondents); in Hungary by Lakocsa, Stycos and Wejnert in 1998 (369 respondents); in the Ukraine by Wejnert, Plisovskaya-Muller and Borisovna in 2000-2002 (200 respondents), in Kyrgyzstan by Wejnert and Djumabaeva in 2003-2004 (400 respondents), and in Senegal and Mali by Wejnert and Lo in 2004 (200 respondents).

²⁰ Life satisfaction is reported on the 1-7 scale with 1 as completely unsatisfied.

²¹ Respondents were asked to select a picture of face that comes closest to expressing how they feel about a particular life domain.

²² In the case of the longer period of recall, interviewers can follow the Brown (1979) recall method by asking respondents to first give concrete examples for their responses prior to providing their ratings. This method increases test-retest reliability significantly for recall periods of 3-7 years (r = .77 with method vs. r = .56 without method, p<.01). Retest stability is often assessed in the study: 40% of each cell of the sample (total n=40) is randomly selected and re-interviewed by a different interviewer within two weeks of the first interview. Retest stability is indexed by interclass correlation, which reflects both rank order of respondents and magnitude of ratings. The retest reliability was tested in the pre-test of the study showing high stability of responses. The pre-test was conducted inPoland by Stycos, Wejnert and Tyszka in 1996 (50 respondents).

²³ Notably however, these findings contradict earlier arguments that a high correlation of literacy and democracy refers to an increase of literature, either causing countries' democratization or spurring democratic growth (Almond & Verba, 1989; Lipset, 1960). Rather, the an increasing literacy rates seems to support recent investigations showing that the increase of literacy is a function of democratic growth (Wejnert, 2005) and rapid socioeconomic development succeeds democratic growth (Wejnert, 2002b).

²⁴ In1970, in most highly democratic countries the percentage of women 15 years of age and over in the labor force did not exceed 40% (US 43%, Germany 39%, Italy 28%), by the 2000 with the exception of the United States, whose rate had increased to 50 percent, other democratic countries showed little improvement.

²⁵ Such assumption is concordant with Coleman’s (2004) and other studies.

²⁶ The long term effects were measured in models with dependent variables (indicators) lagged by 1 and by 5 years.

²⁷ E.g., inChina, evidence suggests that the practice of female infanticide was the cause of the high female mortality rate in the 1930s & 1940s. This practice declined subsequently, and the ratio of females to males in successive cohorts increased. Enhanced female mortality in the 1950s was due to the great famine in which female children suffered more than males. The introduction of birth control in the 1970s did not affect the sex ratio at birth but led to the voluntary cessation of childbearing after the birth of a male child. The escalation in the ratio of male-to-female births in the late 1980s when ultrasound testing became widely available was largely due to the increased number of sex-selective abortions of female fetus (Coale, Ansley J. & Judith Banister, 1996).

²⁸ Considering the limited foreign investment in many of the post-communist countries it is expected that the decline changes into the positive growth in the future (Kentor and Boswell, 2003).

²⁹ Grouping these countries according to shared historical and cultural roots, e.g., as being old statehoods of Poland, Ukraine and Hungary vs. newly created post-colonial/post-tribal states of Mali, Senegal and Kyrgyzstan, or as Christian vs. Muslim countries, would create distinctly different patern.

Notes: *Coefficient at least twice its standard error. Values in parentheses depict standard errors. Total number of observations in core countries N=436, number of countries n= 17;

in semi-peripheries N=747, number of countries n= 32; in peripheries N=2480, number of countries n=100; in post-communist N=426, number of countries n=20.

¹ In addition to being a separate category, post-communist countries are also included in peripheries and semi-peripheries.

[DR1]Would it be possible to insert the table 2 below the first paragraph of p.15? It is much better if the table is preceded by a small text that introduces is. However, I understand if editorially such change is impossible

[DR2]Again the table 3 and 4 should come after one small paragraph of text. The first paragraph below the table (top of p. 21) would be perfect here.

It is :

Models of Groups of Countries: To assess whether the general findings are similar for all countries regardless of differences in their development level, the same models were assessed in cross regional analyses of the well-developed (core), semi-developed (semi-peripheral), underdeveloped (peripheral) and post-communist countries (see Tables 3 and 4).