Over the last hundred years, political, legal, and social rights for women have significantly increased. In most countries women are enfranchised (Banaszak Reference Banaszak1996; Finnemore and Sikkink Reference Finnemore and Sikkink1998; Markoff Reference Markoff2003; Ramirez, Soysal, and Shanahan Reference Ramirez, Soysal and Shanahan1997) and the inclusion of women in politics has grown enormously, though at dramatically different rates in different areas of the world (Caraway Reference Caraway2004; Coleman Reference Coleman2004; Inglehart and Norris Reference Inglehart and Norris2003a; Inglehart, Norris, and Welzel Reference Inglehart, Norris and Welzel2002; Kenworthy and Malami Reference Kenworthy and Malami1999). Nonetheless, except in a very few countries, political and economic equality lag far behind their potential (Poe, Wendel-Blunt, and Ho Reference Poe, Wendel-Blunt and Ho1997). We focus on one such right that has wide-ranging implications—a woman's right to abortion. The right to an abortion is opposed transnationally in ways that other women's political and economic rights are not (Buss and Herman Reference Buss and Herman2003). Indeed in 2000, less than one-third of countries allowed women access to abortion on demand.
Given the normative importance of the issue for both supporters and opponents and the level of contention related to the issue in many states, as well as the fact that data are readily available, it is surprising that while there are quite a number of single-country (Abramowitz Reference Abramowitz1995; Adams Reference Adams1997; Alvarez and Brehm Reference Alvarez and Brehm1995; Banaszak Reference Banaszak1998), comparative-country (Htun Reference Htun2003; Yishai Reference Yishai1993), and regional studies (Boland Reference Boland1993; David and Skilogianis Reference David and Skilogianis.1999; Miller Reference Miller2001; Pillai and Wang Reference Pillai and Wang1999; Sirokova and Buresova Reference Sirokova and Buresova1993;), there are very few global studies (two exceptions are Pillai and Wang Reference Pillai and Wang1999 and Ramirez and McEneaney Reference Ramirez and McEneaney1997) that look at the factors that impact the legalization and liberalization of a woman's right to abortion. This is surprising for two central reasons. First, abortion is perhaps one of the most divisive political issues today in the United States, and though there are many scholars in American politics, sociology, women and gender studies, and other fields who explore this issue, very few attempt global analyses. The realization that some rights are gendered and that the path to these rights is different from that of nongendered rights makes analyses of issues like this one necessary for understanding how gendered rights advance and what are the different obstacles that social movements face when tackling gendered issues. Second, a person's control of his or her body, regardless of gender and application, is perhaps the sine qua non of rights generally.
We examine the economic, cultural, and political opportunity structures that facilitate and constrain women's right to abortion globally and pose two related questions. First, which factors explain the liberalization of different types of abortion laws? Second, which factor or set of factors is the most important in explaining abortion laws? We focus our analysis in this article on abortion rights contemporarily (that is, we use cross-national data for the year 2000). We examine different levels of legally acceptable reasons for abortion, as the complete prohibition of abortion is relatively rare. While we recognize the limitations inherent in a global analysis (most notably the availability of data for certain important explanatory variables and the sacrifice of depth for breadth), it is nonetheless important to understand how and when rights that are particular for women—and not just extensions of previously existing rights that men have exercised—are extended, and not just in the West.
We begin by developing a three-pronged approach that includes a look at the impact of politics, economics, and culture to answer these questions. We then parse these out into the structural and empowerment components of each, and posit a theory of rights in which empowerment factors are more important for explaining change than are structural factors. Of the three competing explanations for liberal abortion policies, there is little agreement about what facilitating and constraining factors are globally operative. The reason that there is no agreement in the field about which of these factors is most important in explaining abortion rights is that they each are operationalized in different ways, and it is these different ways that best explain the evolution of rights. That is, each of these factors may take a structural or an empowerment form. For example, when we talk about the importance (or lack thereof) of politics in influencing rights, using a global or comparative approach, we mean the political structure of a given state. But what may be more important is the practical politics in that state. In the instance of abortion rights, this may be the importance of women's suffrage or the presence and impact of women in legislative bodies. In terms of economic factors, a state's gross domestic product or level of urbanization may be more important than empowerment aspects of the economy, such as women's presence in the labor force. Culturally, factors such as the religious structure of a state may explain rights, or the impact of a world culture on a specific state may be more important. We contend that there are important differences between a structural and an agency analysis of the emergence of rights.
To test this approach, we examine their impact on the liberalization of abortion rights globally using a distributed lag model with ordered logistic regression. We find that an empowerment approach explains liberalization of abortion better than a structural approach in terms of politics, but that a structural approach is a better predictor of liberalization of this policy in terms of culture, and that both empowerment and structural factors are important predictors when economic factors are taken into account. We conclude with a discussion of the implications of these findings for understanding policy change and human rights on a global scale.
THE LEGALIZATION OF ABORTION CROSS-NATIONALLY
Women's access to medical abortion can be viewed in different ways. Many in the U.S. evangelical movement view it as a crime and as antifamily (Buss and Herman Reference Buss and Herman2003). American feminists have argued that access to abortions is a right and “that restricting the right of abortion would constitute effectively an attack on the sexual equality of all women and not merely a limitation of individual interests in procreative autonomy” (Yamin Reference Yamin1996, 425). There are those who criticize both of these approaches for perpetuating a Western ethnocentrism (Renteln Reference Renteln1990). While we recognize this possibility, we take the view that control over one's own body is a human right, and as such, we situate our inquiry within debates about human and civil rights generally and women's rights specifically. See, for example, Lynn Freedman and Stephen Isaacs (Reference Freedman and Isaacs1993) and Rebecca Cook and Bernard Dickens (Reference Cook and Dickens2003) for a full explication of some of the arguments supporting this view.
In exploring the factors that lead to women's access to abortion, we start with two sets of questions. First, which factors are most important for advancing or limiting women's rights: political, economic, or cultural? Second, is there a difference between structural and empowerment components of these factors? If there is, which is more important in explaining the liberalization of abortion rights?
Political Factors
One of the key factors posited to advance rights is the system of contestation: a democratic system will bring with it democratic inclusion (Davenport Reference Davenport1999; Gurr Reference Gurr2000; Wilensky Reference Wilensky2002). The idea that the development of democracy per se will lead to abortion rights for women is not frequently expressed. A clue as to why this might be so comes from Francisco Ramirez and Elizabeth McEneaney (1997) in their comparison of the extension of suffrage with the expansion of reproduction rights. They find that “the liberalization of abortion laws differ from franchise rights in that the former uniquely apply to women and are not derivative from rights earlier extended to men” (Ramirez and McEneaney Reference Ramirez and McEneaney1997, 7).
Regime type, argues Yael Yishai, takes a backseat to ideology. Yishai suggests that interest group activity has had only marginal impact on policy output in Sweden, the United States, Ireland, and Israel. Rather, policy output is a function of predominant political ideologies and public ideas—“a set of durable beliefs having a broad scope that pertain to personal matters and involve deep emotions” (Yishai Reference Yishai1993, 207). Attitudes toward women's social role is an example of a public idea that predicts the legality of abortion in the four democracies studied. She concludes that “women have failed to shape abortion policy, not because they were short of organizational resources, but because their road to influence was strewn with public ideas in which they, too, equivocally believed” (Yishai Reference Yishai1993, 209). She writes further:
The ideology of the welfare state in Sweden, religious tenets consolidated by a powerful Catholic Church in Ireland, norms espousing both individual liberty and minimal state interference rooted in the American creed, and demographic considerations sustained by a system favoring a strong and intrusive state in Israel, these determinants help explain abortion policies in the four countries. (Yishai Reference Yishai1993, 224)
Despite the sense that democratic structures produce greater human rights, abortion rights are also positively associated with nondemocratic regimes. In 1920, the USSR became the first country in the world to legalize abortion, and it has been hypothesized that
[t]here is nothing in Marxism that stands in the way of legalized abortion. Marxist values of materialism and atheism offer no reason to pause, and the doctrines of class struggle and building a classless society require the full commitment of all who can work; concern with children and family distracts from these tasks and reflects attachment to bourgeois values that are to be eradicated. Unrestricted abortion facilitates faster progress toward utopia. (Flood Reference Flood2002, 191)
All of this points us to the null, but nonetheless still interesting, hypothesis that political structures are not associated with abortion rights. The only exception to this may be if the structure of and support for the state could somehow be positively impacted by the legalization of abortion.
Finally, in analyses to date, suffrage does not appear to matter much, though economics does. The introduction of female suffrage has not been found to have a significant impact on the probability of a state's having liberalized abortion law (Ramirez and McEneaney Reference Ramirez and McEneaney1997). Vijayan Pillai and Guang-Zhen Wang (1999) also do not find that political-legal equality or socioeconomic equality correlate to abortion rights in their study of the developing world.
Economic Factors
If democracy is not a direct causal factor in the rise of abortion rights, modernization might be so through its relationship to women's equality. The economic modernization of society extends both democratic contestation and democratic inclusion (Doorenspleet Reference Doorenspleet2004; Huntington Reference Huntington1991; Lipset Reference Lipset1960; Przeworski and Limongi Reference Przeworski and Limongi1997; Ramirez, Soysal, and Shanahan Reference Ramirez, Soysal and Shanahan1997), and several researchers have found a relationship between modernization and a change in values supporting women's inclusion in general (Inglehart and Norris Reference Inglehart and Norris2003b; Steel Reference Steel, Warner, Stieber and Lovrich1992). At the same time, cross-national studies of abortion rights worldwide (Ramirez and McEneaney Reference Ramirez and McEneaney1997) and in developing countries (Pillai and Wang Reference Pillai and Wang1999) have not found a relationship between economic development and abortion rights. Some argue that since the liberalization of abortion laws is not affected by the level of economic development or political democracy, there is little support for standard modernization arguments (Pillai and Wang Reference Pillai and Wang1999; Ramirez and McEneaney Reference Ramirez and McEneaney1997).
Both democracy and modernization, though, are large structural and institutional phenomena that do not necessarily empower women. The political mobilization of groups, as well as their economic empowerment, increase the rights of group members (Chinn and Truex Reference Chinn and Truex1996; Dahl Reference Dahl1989; Lipset Reference Lipset1960; Marx Reference Marx1996), and allow for more independence to demand a variety of different rights. As Janeen Baxter and Emily Kane (1995, 194) argue, the “ties of social, economic, and interpersonal dependence of women on men” at both the societal and individual level “affect both men's and women's interests and their interpretations of gender inequality.” They suggest that women's dependence on men discourages egalitarian gender attitudes that suppress challenges to gender inequality, thus legitimatizing an unequal status quo. They conclude that longer hours in paid employment, location in a middle-class position, and higher education are associated with more egalitarian gender attitudes (Baxter and Kane Reference Baxter and Kane1995). Focusing on the United States, Kristin Luker (Reference Luker1984) also finds that women's labor force participation affects individuals' views on abortion. As more women enter the labor force, their traditional role primarily as wife and mother is deemphasized, while the desire of women to have more control over their fertility is emphasized. A relationship between women's labor force participation and abortion rights has been demonstrated, lending support to the idea that economic empowerment is related to abortion rights (Ramirez and McEneaney Reference Ramirez and McEneaney1997). These contrary findings suggest the need to further explore the possible relationship between women's economic inclusion and their abortion rights.
Cultural Factors
In regional and case studies, cultural factors have been found to be important predictors of abortion policy, especially religion. Lee Ann Banaszak (Reference Banaszak1998) examines the differences in abortion opinion among the German mass public. At the time of reunification, East German law, which allowed abortion on demand for the first 12 weeks of pregnancy, was more liberal than the West German law that required a medical-rational or demonstrable social hardship for a legal abortion. According to Banaszak, the differences in the aggregate abortion opinion between East and West Germans can be explained by both individual factors and social environmental effects. While this can be explained through political analysis (e.g., communist countries are more likely to have the most liberal abortion policies), she attributes the differences in East–West opinion to differences in religion, women's participation in the workforce, educational attainment, and the social environment of the two regions (Banaszak Reference Banaszak1998).
Ted Jelen, John O'Donnell, and Christian Wilcox (1993), in the context of Western Europe, argue that there is a correlation between the proportion of a country's population that is Catholic and the strictness of its abortion law. According to their research, “all of the countries with the most restrictive laws have large Catholic majorities, while none of the most permissive nations do” (Jelen, O'Donnell, and Wilcox Reference Jelen, O'Donnell and Wilcox1993, 376). In addition, the visible presence of a large number of Roman Catholics causes a pro-choice countermobilization among non-Catholics. Indeed, much of the cross-national research has shown that individuals' religious denominations and the strength of religious beliefs have a strong impact on their stance on abortion (Banaszak Reference Banaszak1998; Jelen, O'Donnell, and Wilcox Reference Jelen, O'Donnell and Wilcox1993; Legge Reference Legge1983a; Legge Reference Legge1983b; Pillai and Wang Reference Pillai and Wang1999).
Finally we take into close consideration the argument about the impact of world culture as a decisive factor in the expansion of abortion rights. Using a world culture model, Ramirez and McEneaney (Reference Ramirez and McEneaney1997, 8) argue that the political incorporation of women
is a worldwide process deeply influenced by world models of progress and justice and strongly associated with becoming a legitimate nation-state. These models are articulated and transmitted through international organizations, social movements, and certified expertise. Through these carriers world models shape the constitution and activity of nation-states and inform policies regarding the appropriate status of women within them.
The push for abortion rights, they claim, initially began at the nation-state level. However, abortion policy specifically, and reproductive rights more generally, were moved to the international realm when population growth became a worldwide concern. Intricately tied to both suffrage and abortion rights is the idea of “autonomous and responsible adulthood” for both men and women (Ramirez and McEneaney 1997, 9). They also suggest that membership in international nongovernmental organizations exposes member states to world models that favor liberalization of abortion and suffrage laws. Their analysis confirms the world culture model prediction that membership in international nongovernmental organizations has a positive effect on the likelihood of acquiring the franchise and the liberalization of abortion laws (Ramirez and McEneaney Reference Ramirez and McEneaney1997).
Related to the world culture model is the date of independence, which, we believe, has an effect on the extension of both female suffrage and liberalized abortion rights. Recent nation-states have enacted female suffrage and abortion liberalization at a much faster rate than have states that gained their independence prior to 1900 (Ramirez and McEneaney Reference Ramirez and McEneaney1997). This finding suggests that these recently independent nation-states are influenced by world norms of justice and progress.
ANALYTICAL FRAMEWORK
We analyze the determinants of abortion laws using an ordered range of grounds on which abortion is permitted in order to examine and discuss the degrees of liberalization of abortion policies. Our abortion rights scale was developed by coding information in country reports in the 2002 Abortion Policies: A Global Review of the United Nations Population Division of the Department of Economic and Social Affairs (The Population Division of the United Nations Secretariat 2002). We developed a six-point scale that ranges from 0, countries with no abortion rights, to 5, countries where abortion on demand is available. The categories of our abortion scale include no abortion rights (0); abortion allowed in cases of rape, of incest, or when the life of the woman is in danger (1); abortion allowed because of physical health considerations (2); abortion allowed because of mental health considerations (3); abortion allowed because of socioeconomic considerations (4); and abortion allowed on demand (5).
We grouped together laws permitting abortion for reasons of rape, incest, and existential health threats. We then grouped together countries that allowed abortion for these reasons, as well as for non-life threatening physical health problems. We made another category that permitted abortions for the previous reasons and also for mental health reasons. Our rationale for the ordinal ranking of mental health reasons as a more permissive reason for abortion allowances than physical health reasons is that physical health considerations can empower caretakers more than women themselves, and thus should be considered more restrictive. Ranking them together, we feel, will needlessly combine different levels of tolerance for women's rights to abortion. We then grouped together all countries that permit women access to abortion due to social or economic considerations, in addition to other rationales. While some countries that permit women access to abortion due to social or economic considerations allow a great deal of leeway for them to make this claim, we distinguish this from abortion on demand because it still puts a woman in the position of having to justify her choices to a state-controlled body. To account for the nature of the scale, we use ordered logistic regression for the multivariate analyses with a distributed lag model.
Countries with any form of forced abortion policies were excluded from the analysis, notably China and Vietnam. In addition, eight other countries were excluded because their policies did not fit the logic of our ordered scale. These countries include Australia, Brazil, Fiji, Guatemala, Japan, Mexico, Panama, and Qatar. These exclusions do not change the significance of any of the empirical analyses. After these exclusions and the dropped observations due to missing data, we are left with 112 countries in our analysis.
We pulled together data from seven sources to examine the political, economic, and cultural factors that are important to the legalization of abortion: Women in Parliament: Archived Data. (Inter-Parliamentary Union 2005); Women in Parliaments: 1945–1995 (Inter-Parliamentary Union 1995); Abortion Policies: A Global Review (the Population Division of the United Nations Secretariat 2002); POLITY IV (Marshall and Jaggers 2003); Convention on the Elimination of All Forms of Discrimination Against Women (The Division for the Advancement of Women 2005); Women's Suffrage: A World Chronology of the Recognition of Women's Rights to Vote and to Stand for Election, Inter-Parliamentary Union (Inter-Parliamentary Union 2004); and World Development Indicators (World Bank 2003).
The political factors we examine include Polity scores to measure political structure and the percentage of women in lower houses of parliament to measure female political empowerment. The Polity score is ordinal and ranges from −10 to 10, measuring the level of institutionalization of a regime and its level of democracy by adding up component scores of the nature of executive recruitment, competition, constraints, and the regulation and competitiveness of political competition (Marshall and Jaggers 2003). This is a one-year lagged variable, measuring political structure for 1999. The causal logic we employ here is that changes in abortion policy occur after changes in political structure, as democracy has been found to increase a variety of rights for all sorts of groups. Because of the nature of regime change in the 1990s, the creation and implementation of new policies as a result of regime change happened fairly quickly, and thus only a one-year lag is necessary in this case (Asal and Pate Reference Asal, Pate, Gurr and Marshall2005; Gurr Reference Gurr2000). In initial analyses, we also included the year suffrage was first granted for women, but this variable was highly collinear with several other explanatory variables and so was dropped from the final analysis. The number of women in parliament is used as the measure of women's empowerment. This is a 10-year lagged variable, measuring women's empowerment for 1990. Our rationale for the 10-year lag in this case is that in stable regimes, major policy change may take significant time, and the empowerment of women through their inclusion in parliamentary bodies will not have an immediate effect on changes in the passage of legislation, but rather will likely take a number of terms to occur.
The economic factors we examine include urbanization as the structural measure and percentage of females in the labor force as the empowerment measure. In initial analyses, we also included gross domestic product (GDP) as a structural economic component, but it was highly collinear with all of the other political and economic variables in the model and was dropped. We use urbanization as a measure of economic modernization with a 10-year lag. Our rationale for the 10-year lag is similar to that given for the political empowerment variable. The effects of modernization are not immediate and take time to translate into policy change. We measure women in the labor force also for 1990, and do so because the effects of women's inclusion in the labor force are often not immediate, particularly with respect to policy changes.
The cultural factors we examine include the percentage of Catholics as the structural variable, using religion as a proxy for culture, and ratification of Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW) as a measure of empowerment. Of the 112 countries in our analysis, 51 have populations of which 50% or more are Catholic. Though the percentage of other religious groups in a particular country was marginally statistically significant, particularly for Muslims, these groups were dropped because of problems stemming from collinearity. We also developed a measure of fundamentalism by combining the percentage of Catholics and Muslims, but the measure had to be dropped because of collinearity.Footnote 1 While world culture can be operationalized in a variety of ways, one key component is contact with international organizations and entering into multilateral treaties. Within the context of abortion rights, we operationalize the possible impact of world culture by a state's ratification of CEDAW. The convention,
adopted in 1979 by the UN General Assembly, is often described as an international bill of rights for women. Consisting of a preamble and 30 articles, it defines what constitutes discrimination against women and sets up an agenda for national action to end such discrimination. (The Division for the Advancement of Women, 2005)
CEDAW calls for the complete equality of women in all fields. While the convention does not mention abortion, it does call for women to have complete control of their lives. If international agreements and organizations are creating a world culture with certain expectations of inclusion for women, CEDAW seems to be a clear manifestation of it. We use ratification rather than a state's signature to the convention because it signifies a deeper level of commitment to the agreement and its goals that ought to be reflected in higher status of women within those societies.
Though there are many other interesting political, economic, and cultural components that may in part help explain the emergence of rights, this set of variables presents the most parsimonious approach that still enables us to model the key theoretical arguments of this type of approach. Using these variables we test the following hypotheses:
H1: Political Hypotheses
- H1a:
The higher the level of democracy, the more likely women will have the full range of abortion rights.
- H1b:
The higher the number of women in (the lower house or only house of) parliament, the more likely women will have the full range of abortion rights.
H2: Economic Hypotheses
- H2a:
The higher the level of urbanization, the more likely women will have the full range of abortion rights.
- H2b:
The greater participation of women in the workforce, the more likely women will have the full range of abortion rights.
H3: Cultural Hypotheses
- H3a:
The higher the percentage of the Catholic population in a country, the more likely women will not have the full range of abortion rights.
- H3b:
Countries that have ratified CEDAW will be more likely to allow the full range of abortion rights.
ANALYSIS
In simple bivariate analyses of each of these factors, both their structural and empowerment forms are important for explaining the liberalization of abortion policies and in the directions expected. As countries become more democratic, their abortion policies become more liberal; as the percentage of women in parliament grows, so do the different forms of abortion allowances; as the percentage of women in the labor force grows, so do the different allowances for legalized abortion; the greater the urbanization of a country, the more liberal their abortion policies; and if countries ratify CEDAW, they are more likely to have liberal abortion policies. Religiosity, particularly Catholicism, works in the direction we expect as well: As the percentage of Catholics in a country grows, the legalized allowances for abortion decreases. A more sophisticated approach tells a slightly different story.
Table 1 lays out the results of an ordered logistic regression analysis of these factors and their components on the legalization and liberalization of abortion policy. We find support for all our hypotheses except for the impact of democracy and the ratification of CEDAW. In this analysis, political structure is no longer important, but the political empowerment of women is. This is not to suggest that political structure is irrelevant, but rather that its impact is indirect. In a separate path model, levels of democratization predict increases in the percentage of women in national legislative bodies (beta coefficient is .426 and is significant at p < .000). The results show that as the percentage of women in legislative leadership positions increases, so does the allowance of abortion policies. The level of urbanization and the percentage of females in the labor force have the same positive and significant effect, while the percentage of Catholics in a country has a negative and significant effect.
Table 1. Ordered logistic regression of factors that liberalize abortion policies cross-nationally, 2000
* p < .05 using a one-tailed test
** p < .01 using a one-tailed test
*** p < .001 using a one-tailed test
Table 2 lays out the changes in probabilities for each of the types of abortion policies. In the case of the political empowerment of women, the probability of legalized abortion on demand jumps from less than 16% to almost 70% as the percentage of women in the parliament goes from none to close to 40%. Empowered women in decision-making positions matter to abortion rights, and they matter a lot.
Table 2. Probability tables, selected variables
- Polity 2 set at:
mean
- Women in parliament set at:
mean, except for estimation row
- Urbanization set at:
mean, except for estimation row
- Women in labor force set at:
mean, except in estimation row
- Percentage of Catholics set at:
mean, except in estimation row
- CEDAW set at:
ratified
Both the structural and the empowerment components of economics are positive and significant predictors of increases in abortion allowances in this model. Again looking at Table 2, we see that the probability of abortion on demand changes from 4.4% to 78.6% as urbanization increases from 5.15% to 100%. At the same time, the probability of legalized abortion on demand increases from 2.8% to 62% as the percentage of women in the labor force increases from 9.7% to 53.5%. When the impact of economics is considered, both the structural and the empowerment components are important for explaining the liberalization of abortion. However, the marginal change for the empowerment component is less than that for the structural component (.016 and .166, respectively). Structure matters slightly more at the lowest ends of the measures and much more at the higher ends.
Finally, in the case of culture and norms, religion is a good predictor of these policies, but the world culture variable has no explanatory power. As expected, as the percentage of Catholics in a country goes from its lowest point (none) to its highest point (99.8%), the probability of legalized abortion on demand drops from 39.3% to 8.6%. The cultural opposition to abortion entrenched in Catholicism prevails in more Catholic countries, with religious positions greatly influencing their politics and policies, at least in the case of this policy area.
CONCLUSION
We began by posing two central and related questions: First, which factors explain the liberalization of different types of abortion laws? Second, which factor or set of factors is the most important in explaining abortion laws? We find that there is a dynamic relationship between structure and agency that is played out on the world stage. In some cases agency predominates, in other cases structure predominates, but in all there is an interactive mechanism between the two that sets up the causal dynamics of change. Our analysis of abortion rights in the year 2000 supports our belief that when it comes to abortion, the political empowerment of the people who are directly affected by such a right is more important than political structure, but political structure is indirectly important in that it helps predict political empowerment. With respect to the economy, modernization as measured here by urbanization has a larger marginal impact influencing the liberalization of this policy, while the percentage of women in the workforce is important but matters slightly less. Rights are determined not only by a society's rules and economic and political structure but also, and in this case especially, by the people who have the right to make those rules. The broader culture in which these decisions are made is an important, but not the only, predictor of liberalization.
While much of the framing of abortion politics is within the context of religious values and culture, we find that the empowerment of women economically and politically has the largest impact. We believe that the reason is that, unlike other human and political rights (e.g., the outlawing of torture), abortion does not involve the transfer or expansion of a right previously granted only to males. Abortion is a procedure that is inherently gendered. In this instance, we believe that these rights emerge not because of large structural forces but when women themselves, as individuals and as a group, are empowered. Thus, the most important factors that explain the liberalization of abortion policies are not found in the larger cultural frame but, rather, in the economic and political frames that change how and where women work and hold positions of power.
One of our negative findings, that political structure does not directly matter when other factors are taken into account, is intriguing. First, it shows that, like much of the state feminism literature, in politics and economics particularly there is a great deal of give and take between structure and agency. Second, while much evidence supports the relationship between democracy and other rights, the fact that it is not directly correlated with abortion raises the point that, as Ramirez and McEneaney (Reference Ramirez and McEneaney1997) suggest, not all rights are created equally. Certain rights are, by their nature, gendered, and will thus find different and perhaps harder routes to being achieved. This is the heart of gender politics and has important implications for our understanding of how rights advance and about the different obstacles that social movements face when pushing for gendered rights in particular. Perhaps most importantly, the results show that women themselves must first be empowered in order to realize such rights. Because specifically gendered issues, and by extension gendered rights, have systematically been marginalized for so long, the presence of females in positions of power is necessary for these issues to be recognized. Women cannot depend upon men to do it for them.
The story of the liberalization of abortion rights tells us that women's political incorporation matters, especially for abortion on demand. Our findings also suggest the central role of women leaving the home and entering the workforce—especially within the context of modernizing or modernized societies. Exploring the right to abortion cross-nationally leads to interesting theoretical developments about how “women's rights” and rights in general develop. Comparing these results with the results of other human rights studies shows that the mechanism for achieving “human rights” is not the same one for achieving “women's rights”—they are overlapping but different issues, and in the particular case of abortion, a change in rights is linked to the empowerment specifically of those people who are directly impacted by the right.
