Poverty, Ethnicity, and Gender in Eastern Europe During the Market Transition Edited by Rebecca Jean Emigh and Iván Szelényi 1 The Racialization and Feminization of Poverty? Rebecca Jean Emigh, Eva Fodor, and Ivan Szelenyi Quite understandably—and quite correctly—poverty is usually studied as a persistent, unchanging social problem that, hopefully, can be ameliorated through specific social policies. Indeed, the biblical saying that, "the poor are always with us" has withstood the test of time. The present volume, however, tackles poverty from a different angle. We ask how poverty changes during an epochal transformation, in this case, the transition from economies based on socialist redistribution to those based on capitalist markets. We use this major transformation as an epistemological lever to provide insight into the causes and nature of poverty, in the same spirit of drawing on difference as an analytic tool, our approach is also explicitly comparative. We compare and contrast poverty— and what we argue are associated social processes of racialization and feminization—in different countries during this transition. We hope that this approach not only provides sociological insight but also illuminates policy debates. During the past decade, there have been dramatic and sweeping changes in the countries of the former East European communist bloc; all have moved toward a market economy. In some places, marketization has been very rapid. At the same time, however, poverty has increased in all the countries. In this chapter, we offer some hypotheses about the relationship between poverty, markets, and ethnicity in this region and suggest how the evidence from these countries, as reported in the later chapters, addresses these hypotheses. POVERTY, MARKETS, AND REDISTRIBUTION To define our theoretical puzzle, we draw on the distinction between neoclassical economists and traditional institutionalists. These two positions provide opposing views concerning the origins of poverty. They also recommend strikingly different policies to eliminate poverty. For decades—during the ideological hegemony of traditional institutionalism—the received wisdom among social scientists was that unregulated markets were the primary sources of poverty. Welfare redistribution by governments was necessary to reduce poverty (Polanyi 1944; Townsend 1970) whether it was defined in either relative or absolute terms. Recent evidence suggests that welfare states (in particular, the encompassing "social-democratic" welfare states, in contrast to their more minimalist "liberal" counterparts) have been remarkably successful in alleviating poverty—especially long-term poverty—and minimizing inequality while maintaining growth (Headey, Goodin, Muffels, and Dirven 1997; Kenwortby 1999; Korpi and Palme 1998). Neo-classical economists, on the other hand, have a dismal view of the welfare state. They argue that unconstrained markets generate rapid economic growth that eventually benefits the bottom of the social hierarchy. Neo-classical economists believe that welfare redistribution offers no cure for poverty; it only suppresses the symptoms and prolongs the disease. Governmental welfare redistribution locks the poor into welfare dependence and usually supports the middle class because it does not reach those in extreme poverty anyway. Thus, poverty will be reproduced indefinitely unless the poor take responsibility for their own welfare (Friedman and Friedman 1980; Hayek 1944; 1960; Stigler 1970). Over time, the social-democratic model of the welfare state has also lost ground ideologically to the neoliberal model of the limited welfare state (especially in terms of its poverty-reducing effect), and the more limited model is being instituted in Central and Eastern Europe (Esping-Andersen 1996; Standing 1996). At the same time, however, the ability of welfare states to reduce poverty has been weakened by capital flight that has undermined some of states' key economic-planning institutions (Stephens 1996) and by an increasingly globalized labor and skill market, which has undermined labor-market policies (Esping-Andersen 1996). Kuznets (1955) offered a theory that was consistent with this neo-classical position but posed the relationship between economic growth and social inequality in a more complex way. He analyzed income distribution data over a long period of time using countries with different levels of economic development. He concluded that income inequality increased during an economic takeoff but that inequality declined after economic growth reached a certain level. Thus, the relationship between economic growth and social inequality can be well described by an inverted U-shaped curve. Assuming that free markets produce the fastest possible economic growth, especially in highly developed economies, Kuznets' theory suggests that the expansion of markets may initially increase inequality. By implication, poverty, at least in relative terms, will also increase. Even an increase in absolute poverty is not unimaginable from this theoretical perspective. Eventually, however, free markets will correct the socially undesirable effects of economic growth. Markets will reduce inequality, thereby reducing not only absolute, but also relative, poverty, In a recent comparative study of inequality and poverty, Lai and Myint (1996) formulated the question of the relationship between equality, poverty, and economic growth in a particularly clear and provocative way. They found that income distributions in the core capitalist countries of the world during the past quarter century did not follow the pattern observed by Kuznets. In particular, inequalities in Western Europe and North America increased during the last decades of the twentieth, century. Similarly, Li, Squire, and Zou (1998) found little relationship between development and inequality in their crosscountry time-series analysis. Nevertheless, though Lai and Myint (1996) find little evidence that economic growth reduced inequality (relative poverty), they did find evidence that economic growth reduced absolute poverty. Thus, free markets and economic growth may produce greater inequality, but they may reduce absolute poverty. POVERTY AND THE UNDERCLASS Whether economic growth and marketization eliminate absolute poverty is our central analytic question. We realize that social inequality is a serious problem, but here we focus on underclass formation and. therefore, on absolute poverty. Thus, we arrive at the problem of the tricky and often debated definition of the underclass. Unfortunately, this term has been recently adopted by ideologues of the political Right, who combined it with the "culture of poverty" tradition (Lewis 1966). They define the underclass in behavioral terms, as the "welfare dependent." the subset of poor mired in crime and blamed for its own condition (Levy 1977; Russell 1977), However, we prefer to return to the concept as it was originally used in the 1960s and early 1970s. As far as we can tell, the term "underclass" was coined by Myrdal (1963; 1964), who used it to designate the unique position of the long-term poor who were not benefiting from the posrwar economic boom. These individuals were locked into long-term structural unemployment because they did not have, and could not acquire, the education and skills demanded in a diversified economy. We believe that William Julius Wilson (1978) intended to use the term this way (Aponte 1990), though he may not have been entirely clear, as he sometimes shifted between an economic and behavioral definition of the term "underclass.'* Here, we simply highlight that Wilson expanded Myrdal's notion of underclass in useful ways, by moving away from an economically deterministic understanding of the underclass and by bringing the issue of racism into his analysis. Thus, although the concept of the underclass has been surrounded by numerous debates, we argue that much of the conceptual unclarity can be resolved. We distance ourselves from attempts to define the term in behavioral terms (Ricketts and Sawhill I98S). which unfortunately only point out the social problems resulting from persistent poverty, and return to its original, structural definition. While recognizing that structural definitions may also have slightly different variants (Aponte 1990), we make use of a three-part definition of the underclass, drawing from Myrdal (1963:14, 3S-39), Wilson (1978:156-157; 1987:10-12, 60), and Massey and Denton (1993:3-9): (1) extreme poverty (2) that is persistent and thus tends to be lifelong and intergenerationally transmitted, and (3) spatial segregation. We also follow Myrdal (1963:14-16) and Wilson (1978:1-23; 1987:20-62) in hypothesizing that an underclass forms during economic transformations that adversely affect some segment of the population. In the last 50 years, this transformation has been the decline in highly paid, low-skill work brought about by deindustrialization, leaving an entire segment of the population unemployed and unemployable, thus creating persistent poverty. Thus, we define the underclass in terms of the objective life chances of its members. These individuals are separated from the rest of society by a castelike cleavage; members of the underclass become the "untouchables," the "undeserving poor," or the "no-hopers." In the next section, we argue that an underclass is likely to form in conjunction with the presence of certain social characteristics, and in particular, ethnicity or gender, UNDERCLASS FORMATION AROUND ETHNICITY AND GENDER? All evidence at our disposal indicates that during the transition from socialist redistribution to capitalist markets, both absolute and relative poverty increased. Within this empirical context, we formulate our hypotheses by drawing on the neo-classical and institutionalist perspectives. As market institutions become consolidated in former postcommunist countries, will absolute poverty decline, following a "Lai and Myint" curve, as neo-classical theory predicts, or will the poor reproduce themselves unless the government intervenes, as institutionalists predict? Of course, many factors impinge on this relationship. For example. Blank (1997a; 1997b) argues that economic growth in and of itself does not bring an end to poverty. Instead, the effects of growth on poverty are mediated by other factors, primarily the labor market and wage rates (see also Ayres and Clark 1998; Cornia and Danziger 1997). Here, we consider another factor. In particular, we hypothesize that whether poverty will be reproduced under the conditions of well-functioning markets depends on the social characteristics of the cleavages that separate the poor from the not-so-poor and on the social characteristics of markets. We focus on two sets of socially achieved distinctions, ethnicity and gender, which can be used to mark differences in economic status. Although Myrdal (1963:45) originally explicitly included other groups, including whites, as candidates for the underclass, we note that this process of underclass formation has been generally associated with urban blacks and, to a lesser extent, with Puerto Ricans (Tienda 1989). This empirical focus on racial and ethnic minorities, we argue, is linked to a fourth component of the underclass that suggests quasi-behavioral aspects to the underclass. Myrdal (1963:37) argued that members of the underclass are viewed as "outcasts." as socially unnecessary individuals who cannot participate in social life. Underclass formation is accompanied by an act of exclusion by the rest of society. Thus, although neither race nor ethnicity is by definition required for underclass formation, these minorities are more likely to be its members. Because underclass formation has this element of social exclusion, we draw on theories of social classification that explain this process. As Weber (1978:387-393) noted, ethnicity is one way to mark and exclude others. Thus, we argue, following Bourdieu (1991:221), that ethnicity is an outcome of classificatory struggles. Ethnic groups are categories of ascription and identification constructed by social actors (Barth 1969:10). This link between poverty and ethnic and racial categories is a classic social science problem. Weber (I97S) described ethnicity and class as two separate bases of social stratification. Yet, the two systems of classification not only overlap and reinforce each other (Jenkins 1994:206; McAll 1990:222; Williams 1989:410) but also tend to become each other. Those who are poor become "■ethnic," while members of this ethnicity are more likely to be called "poor." We make a stronger claim, however, that underclass formation is linked to the process of racialization. We use the term "racialization" in contrast to "ethnicization." While there are multiple definitions of these terms, we use "race" to refer to social actors" cultural distinctions based on physical or biological differences and "ethnicity" to their cultural distinctions based on social differences. Thus, while there may be a link between poverty and ethnicity (Caplan 1996; McAll 1990:173, 221-222; Ward 1989), we also link the process of underclass with race. We argue that the process of racialization, the process of turning cultural distinctions based on social differences into cultural distinctions based on physical differences, may be the exclusionary classificatory process linked to underclass formation. Thus, the process of racialization, the presentation of a social phenomenon such as poverty as if it were biologically determined, may lock certain groups into underclass positions. When an ethnic minority is concentrated in poverty, there may be a tendency toward the formation of an underclass if poverty is racialized. Ethnic labels, such as Roma, create social boundaries (Barth 1969:9-10; Jenkins 1994:202; McAll 1990:66-67; Weber 1978:388}. Once the ethnic difference is created, it may be racialized: that is, social differences may be labeled as biological. For example, British and Irish ethnic differences became increasingly racialized between the fifteenth and the nineteenth centuries (Smedley 1993:52-61; Williams 1989:431). This process of racialization may create a Roma underclass. Thus, we argue that the racialization of poverty is the outcome of a classificatory struggle. If this classificatory struggle is successful, a social phenomenon—poverty—is presented as if it were racial, not social, and therefore, as if it were genetically and biologically determined. From this perspective, racialization arises from a strategy used by the positively privileged ethnic group to keep themselves out of the category "poor" and/or to blame the poor for their own poverty at the same time. ("I cannot be poor, since I am not African American or Gypsy" or "They are poor since they are African Americans or Gypsies. It is their biological heritage or culture that makes them poor.") Our conceptualization of racialization is similar to Miles' position (1982). He argues that social actors in structurally similar social, economic, or geographic positions to ethnic minorities are likely to use race to separate themselves from these minorities. Understood this way, racialization implies a classificatory struggle: the social actors who are involved in struggles to define poverty as a racial issue pursue a specific agenda. As actors pursue this agenda, of course, they construct reality. Nevertheless, they do so in reference to reality. In the absence of an ethnicity that is amenable to racialization, attempts to racialize a cleavage will fail. There are, indeed, some dispositions that increase the likelihood of successful racialization. These dispositions are historically variable: at one time they may be sufficient to create a sharp cleavage between the poor and the not-so-poor; at another time, they may not serve this purpose. For example, being Jewish may have been an ethnicity around which to racialize poverty 100 years ago in Eastern Europe, bat it may no longer be possible to do so. Today, being Jewish may be an ethnicity from which a privileged status can be constructed. Thus, in the social construction of reality there must be some objective point of reference that can be used, though how this objective reality is used is a product of struggles between classifiers and classified. In many ways, the Jews and Roma of Europe are examples of racialized ethnicities. There are, or were, sizable groups of ethnic minorities, and, thus, they could be targets of exclusionary classificatory practices. Historically, though Jews were a religious group, they were often separated out as a separate race (Rogger 1986:37). Even highly assimilated Jews were often distinguished from their neighbors by virtue of their complexions, facial features, and other physical characteristics (Pinchuk 1990:17), Similarly, Roma are distinguished from ethnic majorities on the basis of skin color or facial features and assumed to be descendants of some biologically inferior. non-European people from the East (Crowe 1991:151; Fraser 1992:249). Roma were slaves in Romania for nearly 500 years (Beck 1989; Crowe 1991:151). In fact, there is little evidence for a common Roma ancestry and no genetic basis for Roma ethnicity. The Roma are highly heterogeneous, viewed as a singular group, as "Gypsies," only by outsiders. An ethnic category is thus turned into a racial category by a hostile, outside world. Who is Roma is a result of an interaction between both the classifiers and the classified. In addition, we argue that those who are poor are more likely to be classified as Roma. In this volume, we focus on Roma ethnicity. It is the largest cross-national group in Central, Eastern, and Southern Europe that is concentrated in poverty and, therefore, likely to be candidates for the underclass. Of course, there are other poor ethnic groups in these countries, but the Roma allow us to examine the link between poverty and ethnicity, racialization and underclass formation, using a cross-national, comparative perspective. We believe that race is the product of classificatory struggles, thus, (1) no assumptions should be made about common ancestry; (2) empirically, it is untenable that race (or ethnicity) can be determined with certainty by anthropological characteristics (like darkness of the skin, facial characteristics, color of the hair, or eyes), and (3) therefore, the assumption made even by many nonracist researchers (Kemeny 1976), namely, that all members of the society can agree with absolute certainty about who is a member of any particular racial or ethnic group (e.g.. the Roma) is untenable. Who is Roma, Jewish, Hungarian, or Romanian is a result of classificatory struggles that involve both those who are classified this way and those who try to classify (hem. Thus, we argue that the object of our study should be the varieties of classificatory systems, not the "correct" way to determine who belongs to which group. This understanding of Roma ethnicity is closest to Gkely's work (1983) and also similar to the positions of Stewart (1997), Lucassen (1991), and Pronai (1995). Empirical research conducted in Central, Eastern, and Southern Europe makes it very clear that estimates of the size of the Roma population depend on how the classifiers define ethnicity. Different systems of classifications have been used in different countries and in different studies: (1) respondents were classified as Roma if they spoke the Romany language (this is what most censuses did before 1980); (2) respondents were asked to identify themselves as belonging to Gypsy "nationality" (most censuses asked this question after the fall of communism); (3) respondents were asked to identify the ethnic background of their ancestors (Szelenyi and Treiman 1992); (4) experts or jurors (schoolteachers, headmasters, social workers, doctors, the police) were asked to identify Gypsies in particular communities (Kemeny 1976; Kemeny. Havas. and Kertesi 1995; Tomova 1995; Zamfir and Zamfir 1995); (5) interviewers were asked whether they believed that respondents were definitely or possibly Roma (Hungarian Central Statistical Office 1993; Szelenyi and Treiman 1992; Szonda Ipsos 1997). The size and composition of the Roma population in Bulgaria, Hungary, Romania, and Slovakia vary depending on how the Roma population is estimated. If individuals who self-identify as Roma are used as the benchmark, other estimates of the Roma population can be compared to the estimate obtained by using self-identification in the following way (these are rough figures, since estimates of the size and proportion of Roma population vary across countries). If the number of individuals who self-identify as Roma equals 1, then the number of individuals whose native language is Romany is 0.5, the number of individuals reporting Roma ancestors is 3, the number of Roma reported by expert judges is 5, the number that interviewers identify as definitely Roma is 6 or 7, and the number that interviewers identify as definitely or probably Roma is 10. Thus, estimates of the Roma population vary between 0.5 and 10, or about 20-fold (Kemeny, Havas, and Kertesi 1995; Szelenyi and Treiman 1992; Szonda Ipsos 1997), depending on the classificatory scheme used to define the population. Drawing on this understanding of the racialization of poverty, we develop an analogous concept of the feminization of poverty, that is, a classificatory struggle around gender. Thus, the feminization of poverty occurs when women are concentrated in poverty and when biological, not social, causes are proposed as the explanation of this concentration. When poverty is feminized, it becomes viewed as inherent: young women with low intelligence and high fertility have many children, thereby creating a cycle of poverty. This image is often linked to what is assumed to be women's natural role in child-raising activities. Becker (1991:37-48), for example, argues that child-raising activities are biologically easier for women than for men. As a consequence, he argues that it is rational for women to specialize in such activities. Indeed, in popular parlance, poor women with many children are often viewed as victims of biological processes of fertility and childbearing. Thus, the "welfare" mother is treated as a quasi-biological construction. In a process analogous to racialization, when poverty is feminized, it is presented as if it were a naturally occurring attribute of sex, that is, women's biological capacities, instead of as a social construction. While there are some objective, biological bases for women's role in reproduction, much of women's concentration in poverty is related to women's inferior market position with respect to men and to women's greater social responsibilities for child raising. There is some indication that in the United States, some women, in particular, some female ethnic minorities, form an underclass (Casper, McLanahan, and Garfinkel 1994; Rodgers 1996; Stack 8 Poverty. Ethnicity, and Gender in Eastern Europe The Raciaiization and Feminization of Poverty? 9 1974). Such women tend to become single mothers at a young age, and for them poverty is not merely a life-cycle phenomenon. They do not cease to be poor when their children reach adulthood, and their children are also more likely to be poor and remain poor. For our purposes, Central and Southern Europe is again a strategic research site—this time because poverty appears not to be feminized, at least not in the same way as in the United States. During the transition from socialist redistribution to capitalist markets, the raciaiization and feminization of poverty followed different trajectories. Under socialism, Roma experienced discrimination. Still, however, they did not constitute an underclass—though they were the most underprivileged group in these societies—because the socialist economy operated with full employment. Kornai (1980), with great accuracy, called the socialist re distributive economy an "economy of shortage." While economies based on market capitalism suffer from cyclical crises of overproduction—they have a chronic problem of excessive supply—socialist, redistributive economies represent their mirror images. They suffer from chronic shortages, even with respect to the allocation of labor. Threatened with chronic labor shortages, socialist societies institute laws that require individuals to be employed. Under these circumstances, there was sufficient demand for labor to draw Roma into the labor market, in addition to the legal compulsion requiring Roma to be employed. Thus, under socialism, Roma had access to a steady cash inflow for the first time in their history. Gypsies were employed where labor was in the most demand—precisely in the most privileged sectors of the economy, steel, mining, and construction, though racial prejudice slotted them into the most underprivileged positions. When socialism crumbled, these privileged bastions of the socialist economy were the hardest hit: what used to be the most privileged sectors of the economy became the least privileged ones. Thus, with the transition to capitalist markets, the Gypsies became the poorest of the poor. With the collapse of the mining, steel, and construction industries, Roma were the first to lose their jobs and had no hope for finding new employment (Barany 1994; Gheorghe 1991). This process is similar to the effect of deindustrialtzation in creating poor, black ghettos in major American cities. Deindustrialization in former socialist countries, however, has a much greater impact on the creation of a Roma underclass, because socialist ideology idealized those sectors in which the genuine proletariat, "the workers with hammers in hand," labored! Socialist planners were obsessed with "department one" and with the "production of the means of production" in a way that was unprecedented in market economies. As a result, when socialism collapsed, the Roma population moved out of the underprivileged stratum of the privileged economic sectors and into the underclass of postcommunist society. The Roma have become geographically isolated as well. One unusual pattern of geographic isolation in Central and Southern Europe is rural, ethnic ghettos. Although African American ghettos in the United States are urban phenomena, Roma ghettos are found in villages. Not only are rural ghettos unique in comparison to the United States, but they are also novel historical developments in Central and Southern Europe. Traditionally, the Roma minority—if settled, and in the successor states of the Austro-Hungarian empire it tended to be settled—lived in Gypsy settlements next to villages inhabited by ethnic majorities. During the past several decades, there was a dramatic ecological change in rural social spaces. The ethnic majority population began to escape small, isolated villages and moved into centers of rural growth or cities. The vacated spaces were filled by Roma, who were able to move out of the Roma settlements and into the villages. The housing conditions of Roma improved dramatically when they relocated, as they often moved out of mud huts in the settlements. Nevertheless, the flight of the peasant population front these isolated villages was disastrous for the Roma. Peasants in the villages mediated between the rural Roma poor and the rest of society. Peasants consumed the services and products of the Roma. Rural Gypsies were hired as entertainers, as musicians at weddings and funerals, and as seasonal laborers. Peasants also bought goods that Gypsies produced. The out-migration of the peasants was, in many ways, analogous to the out-migration of the black middle class from urban black neighborhoods in the United States. Historically, rural Roma lived in an asymmetrical, but symbiotic, relationship with the peasants who belonged to the ethnic majority. When they lost these social ties, they were turned into a ghetto poor. Though the ethnic dimension is different, this process of the creation of a ghetto is similar to the way in which poor African Americans were turned into an underclass because of the loss of their middle class. Nevertheless, the overlap between rural residence, poverty, and ethnicity is not complete. Gypsy ghettos are also visible in the cities of Central and Southern Europe. These ghettos of urban Roma are created by a similar process, when those who are doing better economically escape the Roma neighborhoods. In addition, as Tarkowska (Chapter 4 in this volume) indicates, poverty is prevalent in rural regions among the population of ethnic majorities in postcommunist societies, because these regions experience little economic growth and suffer from the collapse of socialist state farms. Tarkowska argues that an underclass is formed if its members are socially, economically, and geographically isolated from the rest of the society. By her criteria, the Roma qualify as an underclass. Like the social condition of the Roma, women's position has been changing during the transition to capitalist markets. Their trajectories, however, differ dramatically. Women were also discriminated against under socialism. Unlike Gypsies, who ended up at the bottom of the privileged economic sectors, however, women were locked into the underprivileged sectors of the socialist economy, though they were allowed to climb the social ladder in these sectors. Women were therefore substantially overrepresented in the tertiary sector, the orphan of the socialist redistributive economy. They were also excluded from the royal path to power and privilege: Communist Party membership. Thus, they frequently compensated for their underrepresentation in the party, for their shortage of political capital, with higher levels of education. Indeed, by the end of the socialist epoch, women tended to be better educated than men. Though well-educated women often held relatively poorly paid and low-prestige jobs in the tertiary sector, they were also employed in reasonably high positions in banking, insurance, advertising, tourism, and similar industries. In this respect, the market transition affected women in very different ways than it affected Roma. When the transition to market capitalism occurred, women were concentrated in the sector that experienced the most growth. Furthermore, 10 Poverty, Ethnicity, and Gender in Eastern Europe The Raeialization and Feminization of Poverty? 11 Communist Party membership lost its prestige, so this form of political capital in which men were privileged lost its value, while education became more important. Well-educated women were not threatened by unemployment, and often they found themselves in good jobs with highly desired skills (Fodor 1997). As a result, at least in the early stages of market transition, there is little evidence of the feminization of poverty. In fact, the transition from socialist redistribution to capitalist markets benefits women, at least temporarily. As marketization proceeds, however, women's relative advantage may be eroded. Young men are rapidly entering the tertiary sector. It may take less than a generation for women to lose this relative advantage that accrued to them virtually by accident as a result of market transition. While in the advanced. Western capitalist economies, raeialization and feminization are two reinforcing processes that accentuate the boundaries between the poor and the not-so-poor, during the epochal transition from socialism to capitalism, we hypothesize that raeialization, but not feminization, creates an underclass. Thus, while raeialization may prevent a long-term reduction in absolute poverty and may be apparent in the early stages of transition, feminization may not have the same effect or may fake place at a much later stage in the transition. This difference between the dynamics of feminization and raeialization highlights two crucial issues. First, the feminization of poverty is not inevitable and is a social, not a biological, construction. Second, it illustrates that both (he institutional is t and the neo-classical theories suggest (hat markets have a singular effect, though the theories differ in the way that they characterize this effect. The institutionalist theorists argue that markets create a permanent underclass, while neo-classical theorists argue that they eliminate this type of poverty. Neither position, however, incorporates the historical trajectories of markets, and, therefore, neither position captures the relative advantages of the socialist legacy for women that are specific to a transition from socialist redistribution to capitalist markets. Thus, drawing on Myrdai and Wilson's use of the term, "underclass," our own perspective, inspired by Bourdieu, combines a focus on the objective problems of structural unemployment with an analysis of the subjective classificatory struggles of ethnicity and gender to predict when sharp distinctions are drawn between the "hopelessly poor" and the "deserving poor." Thus, we focus on two mechanisms—raeialization and feminization—that can reproduce poverty. Poverty may become racialized or feminized when ethnic minorities or women are overrepresented among the poor. This over-representation is a necessary, but insufficient, condition. Raeialization and feminization occur when a social phenomenon—here, poverty—is attributed to supposedly ascribed characteristics of actors. If the raeialization or feminization of poverty is successful, the cleavage between the poor and the not-so-poor becomes rigid, and social actors perceive this cleavage as natural, created by the "laws of nature." The outcome is that the resultant poverty is durable and long-term. It is likely to be transmitted to the next generation. In drawing attention to these processes, we try to navigate between the Scvlla of "objectivism" and the Charybdis of "subjectivism." Objectivism in this context would mean that a particular ethnicity or gender is locked into poverty because of certain objective, unalterable attributes of that group. Subjectivism, on the other hand, would imply that ethnicity and gender have nothing to do with poverty; the fact that the poor are ethnic minorities or women is purely a subjective assessment by social actors. Instead, our conceptualization lies between these two positions, because we call attention to the ways in which actors draw on differences to make social classifications. TRANSITION TO MARKET ECONOMY: THE "LABORATORY" TO EXAMINE THE EFFECTS OF MARKETIZATION ON POVERTY Societies in transition from socialist redistribution to capitalist markets provide an ideal site to consider hypotheses about the relationship between markets, inequality, and poverty. In these postcommunist countries of Central and Southern Europe, there is an ideal, almost "laboratory"-like situation. Comparisons between these countries are a powerful tool to examine our hypotheses concerning the interactive effects of marketization, ethnicity, and poverty, because these countries vary along these dimensions. Table !. I presents some of the major characteristics of these countries. In terms of economic development and degree of market penetration, Hungary, Poland, and Slovakia are more advanced than Romania and Bulgaria. Poland is a relatively homogeneous society with respect to ethnicity, while Hungary is a country with a substantial Roma minority. Thus, the comparison of these two countries will show, whether the presence of an ethnic minority that is concentrated in poverty creates an underclass. The comparison of Hungary with Slovakia, Romania, and Bulgaria (all of which have at feast one ethnic minority) will suggest how much impact marketization has on the creation of an underclass. If neo-classical theory is correct, there should be no underclass in Hungary. If Roma poverty is similar in these four countries, our "revisionist" institutionalist theory will be supported. There is another interesting comparison between the ethnically heterogeneous countries. Hungary has a single ethnic minority, while Romania and Slovakia have two minorities: Roma, who are concentrated in poverty, and Hungarians, who are not. Thus, a comparison between Slovakia. Romania, and Hungary will help to explain how the presence of a second minority, which apparently has no elective affinity with poverty, affects the growth of a Roma underclass. Bulgaria falls into a unique third category. It is a multiethnic society, in which more than one ethnicity is overrepresented among the poor (Roma. Pomaks [they were not presented as a separate ethnicity in the data used for Table I.I], and Turks are concentrated in poverty). With respect to our hypothesis about raeialization, we expect to find little evidence of an underclass in Poland because the Roma population is small. Thus, we hypothesize that there will be differences in the degree of underclass formation depending on how many ethnicities exist in a country and how many ethnicities have an elective affinity with poverty (see Table 1.2). We hypothesize that (I) if there is one ethnic group that is overrepresented among the poor, underclass development will be strong; (2) if there are multiple groups overrepresented among the poor, underclass formation may be weaker, since 12 Poverty, Ethnicity, and Gender in Eastern Europe classificatory struggles using race to separate the poor and (he not-so-poor cannot focus on a single group; (3) if there are two ethnic groups, one that is concentrated in poverty and one that is not, the racialization of poverty may be especially strong. The ethnic group that is concentrated in poverty will be blamed, with particular force, for its own poverty, since the nonpoor ethnic group will be cited as an example that ethnic minorities are not necessarily poor. If our hypotheses are correct, the racialization of poverty will be especially strong in Hungary, Slovakia, and Romania. The comparisons between these countries also address our hypotheses about the feminization of poverty (see Table 1.2). Feminization may be more advanced when (1) a racially defined underclass is less developed, because class ificatory struggles can use only gender, not race, to mark poverty; and (2) if marketization is more advanced, it may overcome the temporary advantage that women had in the early stages of market transition. Thus, the feminization of poverty in these countries may increase with marketization. Consequently, the feminization of poverty will be strongest in Poland, which is relatively homogeneous ethnically and where marketization is the furthest under way, followed by Hungary, where marketization is also well under way, but which has a sizable Roma minority. THE EVIDENCE FROM CENTRAL AND SOUTHERN EUROPE Poverty The comparison of poverty and inequality under socialism End post-communism is a complex task. Examining poverty was a taboo in communist societies (Chapters 3. 4 in this volume; Warzywoda-Kruszynska and Grolowka-Leder 1996:68). In the re form-oriented countries, such as Hungary and Poland, data were collected on low-income groups during the socialist epoch, but this evidence is fragmentary (Bokor 1987; Ferge 1969; Frackiewiez 1983; Jarosz 1984; Kemeny 1972; Tulli 1978; Tymowski 1973). Despite these incomplete data, scholars agree that both absolute and relative poverty increased substantially after the breakdown of state socialism (Cornia 1994; Kornai 1996; Milanovic 1994) The causes of the increase in poverty and prospects for its amelioration are less obvious. Is an increase in poverty the price of transition? Is it the price that societies pay for radical, but delayed, economic transformation (Chapter 2 in this volume)? Or. stated more theoretically, is there an inverted U-shaped curve describing the relationship between poverty and marketization during the postcommunist transition? If so, then absolute poverty may increase during the early stages of marketization but may decline as the economy grows and as fully, developed market institutions restructure the economy. The existence of an inverted U-shaped curve of poverty during postcommunist market transition would be consistent with Lai and Myint's theory of the relationship between economic growth and poverty. So far, there is no evidence of a Kuznets' curve of income distribution in postcommunist transitional economies. As the economies of Hungary and Poland grow, inequality remains high and may even c « s O 05 ■o c a a 3 <° ■D a a- DO « 3 X rf I s a. £ B a o -"8 "8 « & S.£ 1 Q 3 C b w O u tl w 8 c o £ 8. c o II 2 "8 c 'E E 8. a "5 8 n « u P ^ is & ° in p as? — —* ro g. >- E 00 f*l ^ ^ VI iri CO fS 3 ~ o 00 a - -s 2 S t~. ^ Si> o Poverty, Ethniciry, and Gender in Eastern Europe poverty line may seem less dramatic, but the actual proportion of poor is higher. His estimates are given in Table 1.4. Table I-J Percent or Population Below the Poverty Line (Poverty Measured by Absolute Income {$4.00 per day]) Country 1987-88 1993-94 Bulgaria 2 33 Hungary <1 3 Poland 6 19 Romania 6 39 Slovakia 0 <1 Source Komai 1996. Table 1.4 Estimates of Poverty (% of the Population) 19S9 1990 1991 1992 1993 Bulgaria [45% of 1989 average wage) 13.8 52.1 53.6 57.0 Hungary l40°/o of 19S9 average wage) 14.5 19.4 Poland (40%qi1989 average wage) 24.7 43.1 41.2 43.7 Romania (45% or 1989 average wage) 33.9 21.4 29.7 51.5 Slovakia (40°/oof 1989 average wage) 5.8 9.4 27.7 34.1 Source: ComialW. As far as we know, the only survey with data for most of the Central and Southern European countries that used the same methodology on a large, random sample of the population is the study "Social Stratification in Eastern Europe after 1989," conducted by Szelenyi and Treiman (1992). Because of the methodology, it provides the most comparable data for these countries. Following the conventions of the literature and for reasons of comparability, The Racializaiion and Feminization of Poverty? 17 they defined "poor" as those whose per capita household income was below 50% of the nations) median. This poveny line is a relative measure, because its monetary value differs in each country depending on the national income level. We present detailed data on poverty in 1992 in Table [.5. The results are consistent with Comia's findings: the differences among the countries lie, by and large, in the same direction in Tables 1.4 and 1.5, although the degree of difference among the countries is smaller in Table 1.5, Thus, Bulgaria and Poland have the highest rates of poverty, followed by Slovakia and Hungary. Since the base here is 1992, not 1989, income, the poverty rate is lower. (In all countries, real incomes declined substantially between 1989 and 1992, and the decline may have been the sharpest in Bulgaria and Romania. This may account for the astonishingly high poverty rate found by Cornia in these two countries.) Ethnicity In this volume, we focus on the Roma, because they are the single largest ethnic group found in the countries of Central and Southern Europe that is likely to be concentrated in poverty. According to Davidova (1995), the Roma population of Europe during the mid-1990s was about 8 million; perhaps over 5.5 million, lived in the Czech Republic, Slovakia, Hungary, Poland, Romania. Serbia, Macedonia, and Bulgaria. An additional 700,000 Roma lived in the other countries of the Central, Southern, and East European region (Albania, other successor states of Yugoslavia, Russia, and the Ukraine). Thus, almost 80% of Europe's Roma population can be found in the former socialist states of Europe. Aside from these countries, Spain and France are the only European countries with a substantial Roma population (750,000 and 340,000. respectively.) Davidova's estimates are rather conservative, so her figures may actually understate the size of the Roma population in the region (Table 1.6). While the estimates for the 1950s and 1960s for Yugoslavia and Romania are highly dubious, it is reasonable to argue that the Roma population increased about threefold over the past three to four decades, This population growth can be attributed solely to high levels of fertility and declining rates of mortality, because international migration was not a major factor in the growth of the Roma population. Although it is likely that Roma are concentrated in poverty, there is very little statistical evidence to support this assertion. Szelenyi and Treiman's study (1992) again provides one of the few sources. Table 1.7 displays their findings for Bulgaria, Hungary, and Slovakia for individuals reporting Roma ancestors (about 2.5-3.0%of the respondents). These data show that the Roma tend to have larger families, less education, and higher rates of unemployment than the general population and the poor. Since Szelenyi and Treiman used a nonstratified, proportional random sample of the population, the N for the Gypsy minorities is small, which makes the relatively consistent results across these countries even more remarkable. 1S Poverty, Ethnicity, and Gender in Eastern Europe Table 1.5 Poverty Rates for Women and Men, Bulgaria, Hungary, Poland, and Slovakia, 1992 (Population Between 20 and 69 Years Old [Percentages, with Ns in Parentheses)) The extent of poverty within: Bulgaria Hungary Poland Slovakia Whole Population 12.6% 9.0% 13.5% 10.9% (567) (422) (421) (500) Men 11.0 7.8 12.9 9.2 (237) (150) (205) (202) Women 14.1 10.1 14.1 12.6 (330) (208) (216) (298) Economically 18.9 7.5 12.7 9.8 Active* (361) (20S) (309) (343) Men 9.5 7.0 12.6 9.2 (162) (106) (168) (175) Women 12.3 8.0 12.8 10.3 (199) (99) (141) (168) Retired** 15.6 8.2 13.8 8.7 (137) (76) (62) (63) Men 35.5 10.2 14.6 7.6 (57) (38) (28) (19) Women 15.6 6.8 13.6 9,4 (80) (38) (34) (44) On Maternity 25.5 27.1 26.2 32.3 Stave/ at home (49) (67) (39) (81) (Women only) Ratio - Female to Male poverty 1.28 1.29 1.09 1.37 (Whole papulation) Notes: N for the whole sample population of Bulgaria = 4,485; Hungary = 3,987; Poland = 3,120; Slovakia = 4.577. The "poor" are defined as those whose per capita household income does not exceed 50% of ihe median national per capita income. * Economically active includes those working and looking for work, ** The reared also includes those on long-term sick leave. Source: Sielenyi and Treiman 3992. The Racialization and Feminization of Poverty? 19 Table 1.6 Estimates of the Size of the Roma Population in Central and Southern Europe Country 1950s- 1960s Minimum, 1990s Maximum 1990s Czech Republic (Czechoslovakia) 200-224.000 250,000 300,000 Slovakia 480,000 520,000 Bulgaria 230,000 600,000 700,000 Hungary 190,000 550,000 600,000 Poland 30,000 50,000 60,000 Macedonia (all Yugoslavia) 85,000 220,000 260,000 Serbia 400,000 450,000 Romania 104,216 1,800,000 2,500,000 Source: Davidova 1995. Gender We can use Szelenyi and Treiman's survey to consider the extent to which women are concentrated in poverty as well. As discussed earlier, their measure of poverty is a relative one. They defined the poor as those whose per capita household income did not exceed 50% of the median national per capita household income in 1992. Since they use per capita household income, as opposed to individual earnings and benefits, the gender gap reflects differential poverty rates among single people. This not only is the conventional definition but meshes well with social reality in Central and Southern Europe, because approximately 40% of the population over 15 years of age is single at any one time (Central Statistical Office 1991). Table 1.5 presents the gender difference in poverty, using this measure. Table 1.8 presents another measure of poverty, "potential poverty rates," which we define as less than 50% of the national median individual income. This indicator shows how poverty would be gendered if all adults had to make ends meet on their own income—a real threat for many people, given the high rates of divorce. The two measures together, we believe provide a more differentiated picture of gendered poverty in the region than either measure alone. The last lines in Tables 1.5 and 1.8 present the ratio of female to male poverty. u ■a a □ c si Ä "3 03 T3 § a. § a. <7\ i*l 3 O r*l "1 oi r~ Ö -3" d "1 M — 0- c .2 n y c u ■a ■3 '<š 's (i U 5 1 i— 12 oo VI V) JjJ ■i i r of c u E 5 '1 E « 'S. E oj g u c E u o 5 'as g E E U z Z, Nui u a. B 0-, — VI q d 3 CTI OS O a I Ži «1 li a sä II i i II ■o ^ iS 01 i ~ 8.1 Jí.i a B I i lil 1 i-i OD g -O i á & a í.šJŠ ch E S 5.5 li spěí 1 1 E a s 3 £Jl Ol "1 1 a w Ě 1 E I ■s S y u 3 E o ctí li The Racialization and Feminization of Poverty? 21 Tables 1.5 and 1.8 illustrate that both the level of poverty and the gender gap differ across countries. In Poland there is a high poverty rate but tittle gender difference. In contrast, in Slovakia, the poverty rate is smaller, but the gender gap is larger. The poverty rate is higher in Bulgaria than in Hungary, but they have about the same-size gender gap. A poverty ratio of 1.25 would rank around the middle of the distribution in Casper, McLanahan, and Garfinkel's (1994:594) eight-country study, about the same as in Germany and Canada, lower than in the United States and Australia but higher than in Sweden, the Netherlands, or Italy. Table 1.8 Potential Poverty Rates for Men and Woraen>in Five Countries, 1992 (Population Between 20-69 Years of Age) Bulgaria Hungary Poland Slovakia M W M W M W M W Whole Population 6.1 8.1 3.3 9.4 6.7 14.8 1.6 10.8 Gender Ratio 1.32 2.S 2.2 6.75 Notes: M=men and W=women; N for whole populations of Bulgaria = "1,48s; Czech Republic = 5,054; Hungary = 3,987; Poland - 3,121; Slovakia = 4.577. Source: Szelényi andTreiman 1992. As Table 1.5 shows, the gender gap in poverty is not limited to either the economically active or the retired: rates are quite similar in both categories. Poverty, even feminized poverty, does not have an "elderly" face in Central and Southern Europe; the retired are generally not worse off than those who are active in the labor force. Rather, the typical poor family consists of a wage-earner father and a mother on maternity leave (or, less often, keeping house), as indicated by the high rates of homemakers living in poor households. Women's plight is particularly apparent when we examine "potential poverty" rates in Table 1.8, which reflect differences in earning capacities in the labor market and illustrate women's dependence on a male breadwinner. The potential poverty rates in Table 1.8 are all higher than the actual poverty rates in Table 1.5, dramatically so in Slovakia. In sum, then, these figures show that the gender gap in poverty exists in Central and Southern Europe, even though levels of state welfare spending as well as women's labor force participation rates are particularly high. This suggests that to understand gendered poverty, we must go beyond mere labor force participation rates and must examine, as Smith (1984) suggested, women's integration into the changing economy. The sizable gender gap in poverty also suggests that women living alone or with children, even when they are working and receiving social benefits, are perhaps the most vulnerable group in postcommunist societies. In general, women's poverty in Central and Southern Europe has few of the "welfare mother" or "state-dependence" associations; instead it is related to the social organization of a system of production where 22 Poverty, Elhnicity, and Gender in Eastern Europe The Socialization and Feminization of Poverty? 23 women's wages are insufficient to support themselves and/or a family. Thus, as we suggested, the concentration of women in poverty may not be as noticeable in some of these countries in Central and Southern Europe as elsewhere, because of the socialist legacy that provided women with some comparative advantages, though poverty does seem to be gendered. Our more specific hypothesis, that women are more likely to be concentrated in poverty in the more marketized countries, seems to hold, at least in general. Taken together, Tables 1.5 and 1.8 suggest that women may be less concentrated in poverty in Bulgaria, as we predicted. The small gender difference in poverty in Poland, however, is somewhat anomalous in terms of our hypotheses. The Interaction between Ethnicity and Gender Furthermore, not all women benefit from the legacies of communist "exploitative" emancipation. We argue that a significant bifurcation is emerging not only between the genders but also among women themselves. We expect this dual opportunity structure to be strongly associated with ethnicity, especially among the Roma. The interaction between gender and race creates a double disadvantage for Roma women. Table 1.9 presents our evidence. Almost half of the Roma, while only around 10% of the non-Roma population of the three countries for which data on ethnicity are available, live in poverty. But Roma women are even more likely than Roma men to be poor—by less than only 10% in Hungary but by almost 40% in Bulgaria. Thus, we argue that one of the characteristics of the newly emerging poverty in Central and Southern Europe is that important differences are created among women. This bifurcation, we propose, is structured along racial lines: women at the very bottom of the social hierarchy are minority women, in particular, of Roma ethnicity. Table 1.9 Poverty Rates among the Roma Population in Three Countries, 1992 (Population 20-57 Years Old) Bulgaria Hungary Slovakia Men Women Men Women Men Women ROMA 40.4 55.3 41.4 44.3 32.1 35.8 NON-ROMA 10.0 12.5 7.0 10.6 9.4 13.3 RATIO FOR ROMA 1.37 1.07 1.11 N (ROMA) 6! 1 46 42 1 49 57 1 48 Source: Szclenyi and Treiman 1992. Even before the collapse of the communist regimes, state authorities attempted to push Roma women out of the labor force. As the demand for labor tightened, Roma women were considered to be the most easily dispensable workers, particularly because their fertility rates were higher than those of the average population. In 1979 at a Politburo meeting discussing the lack of work opportunities available for the Gypsy, retired prime minister Jeno Fock argued, "We should seek employment for the [Gypsy] youth, not for the women, particularly not women with a lot of children. The fact that fewer Gypsy than non-gypsy women work is essentially related to the particular situation of the Gypsy population. ... At times when employment opportunities are tight, we should give preference to the youth over women. . . . Gypsy women's problem can be best treated as a welfare question" (From the minutes of the meeting of the Politburo of the Hungarian Communist Party 288.5.770. April 18. 1979). This quote demonstrates that starting very early, the state made no efforts to provide Roma women with the means to fend for themselves in a non-state-centered economy. They avoided utter destitution because of state paternalism, but once the safety net crumbled, they were left to their own devices and sank into deep poverty. Thus, we argue that while most women collected significant resources during the slate socialist era, minority women could not do this, because they had been treated as a "welfare problem" even during the communist regime. To summarize our propositions, we argued that (I) feminized poverty in Central and Southern Europe is not primarily the result of the growth in the rate of single mothers or the insufficiency of welfare checks: rather it is due to women's inferior position in the labor markets. However. (2) we added that women's poverty, on average, is not the kind of hopeless destitution experienced by women on welfare in the United States, because Central and Southern European women had gained some resources in the communist era, which turned out to be marketable during the transition to a market economy. But not all women can do this: (3) minority women, devoid of these resources, are the most vulnerable social group: they are thrust to the fringes of society and experience downward mobility as does no other segment of the society. CHAPTERS IN THIS VOLUME The contributors to this volume also addressed the hypotheses about the formation of an underclass through the processes of racialization and feminization of poverty. These chapters address our hypotheses in two ways. First, we draw on these chapters to provide evidence for our argument. Second, more importantly—and more interestingly—is that the authors themselves consider these hypotheses. They often disagree with us in fascinating and fruitful ways. We provide here some hints about the outlines of the debate and hope that readers will draw their own conclusions about the veracity of our hypotheses. The transition to a market economy in Poland has been more rapid than in the other countries we examine, as the popular term for the Polish transformation, "shock therapy," implies. Tarkowska in Chapter 4 examines the changing nature of poverty in this context. In doing so, however, she stresses the multidimensionality of poverty; deprivation is not just an economic matter of few resources but also a personal, social, and cultural phenomenon. Poverty, according to Tarkowska, is a many-faceted disadvantage that excludes individuals from full participation in society. Not surprisingly, Tarkowska 24 Poverty, Ethnicity, and Gender in Eastern Europe The Racialization and Feminization of Poverty? 25 advocates using the term '"underclass" and adopts a broad definition of it that includes social, economic, and cultural correlates of poverty; spatial and social isolation; and the transmission of these attributes to the next generation. She also points to the changing nature of this multidimensional poverty. During socialism (the study of poverty was taboo, so accurate accounts are scarce), poverty was generally temporary, linked to a phase in an individual's life cycle. It was also dispersed throughout many groups in society. It tended to be, as Tarkowska points out, "private, invisible, unofficial, and uninstitutionalized." In contrast, Tarkowska argues that poverty during the market transition is much more likely to be a long-term condition that is concentrated regionally, not dispersed throughout society. It stems from unemployment and low levels of education, not from the stages of an individual's life. As is already clear from the preceding tables, poverty is widespread in Poland. Furthermore, Tarkowska's chapter clearly illustrates that poverty began to increase during the 1980s, before the fall of communism, and continued to increase until 1994. Despite the decrease in poverty after 1994, however, Tarkowska provides some evidence that long-term poverty is increasing. She also argues that women are overrepresented among the poor and points to several indicators of the feminization of poverty. Most telling, perhaps is the evidence that long-term unemployment is more common among women than men. Tarkowska's assessment about the feminization of poverty is consistent with our hypothesis that this process should be relatively advanced in Poland, because of the degree of raarketization and the absence of an ethnic minority. In fact, Tarkowska makes the strong claim that in Poland the formation of an underclass is in progress, though it is not being created around classificatory struggles over ethnicity. In fact, Poland is a relatively homogeneous country with respect to ethnicity. Tarkowska cites workers from former state farms as an example of an emergent underclass or "new poverty" in Poland. This is an interesting argument and has some parallels in Hungarian historiography. During the interwar years. Hungarian ethnographers (in particular, Erdei [(1943) 1983] and Illyes [(1936) 1955] defined the laborers working on latifundia (majorsdgi cseledek) for an annual contract as a "society below society." According to Erdei and Illyes, these laborers tended to be ethnic Hungarians, but they were separated from the rest of the society by a eastelike cleavage. They were likely to reproduce their living conditions of extreme poverty across several generations. Typically, state farms were created from land previously cultivated as latifundia. Thus, Tarkowska may see the resurgence of an underclass comprising the ethnic majority, which was partially integrated into society under state socialism, just to be shunned again with the collapse of communism. Tarkowska argues, therefore, that underclass formation is possible without racialization. Her analysis provides the strongest evidence against neoclassical theory: markets do not erase absolute poverty even in ethnically homogeneous countries. Like Tarkowska, Radicova and Vasecka in Chapter 6 point to the changing nature of poverty during the market transition and also note the shift in poverty from being a life-cycle phenomenon under socialism to a long-term condition under capitalism. However, the context is quite different because Slovakia is a highly heterogeneous country, and the transition to a market economy has been slower than in Poland. They argue that the transition has been hindered by the retention of socialist-style redistribution and that the payment of welfare benefits has even increased over the past few years. This large welfare state stifles the growth of the market economy because it reduces individuals' incentives to work. Because of the strength of redistribution, which provides resources to ail segments of society, they argue that poverty is not linked either to gender or to ethnicity. They note that the Roma are exceptional in this respect, as they tend to be concentrated in poverty. They also point out that other Slovaks tend to discriminate against the Roma and that they are socially isolated, though they also suggest that some of the difficulties of the Roma lie in their poor educational attainment. In Slovakia real wages decreased between 1989 and 1994, after poverty had declined between 1958 and 1988. Unlike the other countries where poverty seemed to have decreased after the mid-1990s, in Slovakia, poverty continued to increase, according to Radicova and Vasecka's figures from 1992 and 1996. There is also some indication that poverty is deepening as well: Radicova and Vasecka argue that long-term unemployment rose between 1993 and 1995. They also point out that some segments of the Roma population, in particular, those who live in camps, are desperately poor and highly isolated from the rest of society. Yet, despite some indicators of increased and deepening poverty, Radicova and Vasecka hesitate to label any particular portion of Slovak society as an underciass. instead, they offer a different interpretation by pointing to the continued role of the state in redistributing income, which maintains a minimum standard of living for all Slovaks. According to them, no underclass is being formed in Slovakia around Roma ethnicity, because redistribution from the middle class to the poor by the Slovak welfare state prevents the creation of extreme poverty. Similarly, they argue that poverty is not feminized, because it is not linked to intrinsic biological characteristics of women. Although they note women's disadvantages in the labor market and the greater likelihood that they, instead of men, head single-parent households, they downplay the link between gender and poverty. Radicova and Vasecka argue that in Slovakia the patriotic, centra] right, populist government maintains an overgrown welfare system. Since they adopt a neoliberal perspective, they view the welfare state negatively. Analytically, however, their position suggests that the dismantling of the welfare state, along with increased marketization, may lead to the creation of an underclass. Their position may be consistent with the argument that the Roma did not constitute an underclass during socialism and that the development of the underclass would result from the expansion of the market and the decay of redistributive mechanisms. Such an argument parallels what we have called the traditional institutional perspective. We hypothesized that there would be strong underclass formation and some feminization of poverty in Slovakia, given the large Roma minority population and the overall level of marketization. Some of Radicova and Vasecka's evidence supports this conclusion, though they would disagree about our use of the term "underclass." Nevertheless, it is quite plausible that income ruvcuy, cuuiiuEy, ana uenuer in eastern Europe redistribution by the Slovak state prevents the formation of an underclass, contrary to our hypotheses. Mitev, Tomova, and Konstantinova in Chapter 2 address the role of the state in a different way. They argue that the market transition itself did not increase poverty. Instead, they argue that poverty increased because of the delay in implementing market reforms and because of the way in which privatization and restitution became divisive political issues. They argue that maintaining unprofitable, state-owned enterprises leads to economic decline. Privatization, however, must not concentrate ownership among the incompetent. Thus, their ' argument suggests that economically sound and timely market reforms could alleviate poverty, which has increased dramatically since 1989. Their chapter shows that Roma are more likely to be chronically poor than other ethnic groups. They also argue that the social distance between Roma and other Bulgarians and the stigmatization of Roma have been increasing recently, supporting our argument that an underclass is forming around Roma ethnicity in Bulgaria. Mitev. Tomova, and Konstantinova's argument also points to another important element of Roma ethnicity that has obvious implications for arguments about raciaiization. Like the other authors in this volume, they note that Roma ethnicity is highly heterogeneous and is perceived as an ethnicity only by the non-Roma, Subgroups of Roma differ by religion, language, and lifestyle. Thus. Mitev, Tomova. and Konstantinova argue that the only element common to all Roma is the experience of discrimination. Rephrased in the language of our hypotheses, we suggest that in Bulgaria, classificatory struggles successfully racialize Roma ethnicity. Similarly, as our hypothesis that there would be relatively little feminization of poverty in Bulgaria would predict, the poverty rates for men and women are nearly the same. Mitev, Tomova, and Konstantinova. however, would add a precautionary note: there are indicators that women's social and economic position is considerably worse than men's and that it is declining. The unemployment rate is higher for women than for men. They also note that women suffer most during the market transition, as the growth of the informal sector and of production for own use places a larger burden on women than men. At the same time, the overall loss of income tends to reduce women's share of the family budget more than that of the men. Ladanyi's argument in Chapter 3 neatly complements the chapters of Mitev, Tomova. and Konstantinova and Radicova and VaSecka. While Mitev, Tomova, and Konstantinova argue that the state should implement more market reforms to eliminate poverty, and Radicova and Vasecka argue that underclass formation has been prevented by the continuation of redistribute policies, Ladanyi argues that the rapid market reforms and the rise of the neoliberal state in Hungary, combined with social processes that classify individuals by ethnicity, are creating a highly racialized underclass He argues that poverty became a much more serious and chronic problem after the fall of communism, especially among the Roma. While state socialist programs to combat poverty often reduced inequality, they were also counterproductive because special programs for the "disadvantaged" were disguised forms of discrimination. With the decline in welfare state spending, the programs are too expensive to maintain. His chapter also provides insighl into these social processes of classification by The Racielization and h'emmization ot Koveny.' showing that estimates and social characteristics of Roma vary depending on the method used to identify them. As he aptly notes, raciaiization occurs through classificatory struggles: those who are poor are more likely to be called Roma and are more likely to be treated as Roma by those around them. As in Poland, the economic crisis in Romania started in the 1980s, before the fall of communism. Poverty grew dramatically after the market transition. As Magyari, Magyari-Vincze, Popescu, and Rotariu note in Chapter 5, the transition to market capitalism in Romania has been relatively slow, and the state sector continues to be a large employer and owner. There is still a relatively large agricultural sector, and production for own use is important. Magyari et al. argue that poverty is both racialized and feminized because of classificatory struggles around ethnicity and gender. They point to the numerous ways that Roma and women are socially and culturally marginalized and disadvantaged in Romanian society. Though they distance themselves from the underclass terminology m some ways, they make a strong argument about the depth and breadth of Roma poverty, which in their view is linked to a "culture of poverty" that is intergenerationally transmitted through socialization. They also emphasize that the status of Roma as ethnic minorities is not isolated from the more general issue of multicultural ism in Romania and. in particular, is linked to the status of Hungarians. While our hypotheses also suggest that poverty will be racialized around Roma ethnicity in Romania, Magyari et al. take issue with our suggestion that women may be advantaged during the transition to a market economy because of their advantageous position in the labor market. They point out that poverty entails more than labor market position and that what we consider advantages may be, in fact, the double burdens of femininity. Thus, by arguing that poverty is feminized in Romania when our perspective predicts little feminization of poverty, Magyari et al. challenge one of the central hypotheses of this proposal. They also point to ways in which our conceptualization of poverty must be reexamined through empirical research. We suggested that the feminization of poverty occurs only if women are overrepresented among the poor. At the household level of analysis, therefore, poverty can be feminized only if female-headed households are overrepresented among the poor. Magyari etal. challenge this premise because they argue that poverty may be feminized within households if there are systematic asymmetries in access to scarce resources within the family or household unit. For example, females may be undernourished within patriarchal families or may be poor in nonpoor households. Finally, following in this spirit of inquiry into the possibilities of underclass formation, we put Michael Stewart's Chapter 7, which questions the very use of the term "underclass" as the conclusion. In so doing, we hope that his chapter, far from providing the final, concluding word, serves instead to renew the debate about our hypotheses and our terminology. Stewart suggests that although there is little doubt that poverty has increased dramatically in Southern and Central Europe, we should think twice before applying the term "underclass." He notes that the supposed social exclusion that sometimes seems to accompany poverty is perceived as such only by the nonpoor. The poor themselves are, all the while, engaged in a variety of social activities and strategies, without which they would certainly not be able to survive. Stewart reminds us of the important point 23 Poverty, Ethnicity, and Gender in Eastern Europe running throughout this volume that acts of social classification have a directionality inherent in them. A negative group label is often used to maintain privilege and hierarchy, not to describe social groups neutrally. Acts of social science classification are not necessarily exceptional in this respect. CONCLUSIONS We use our theoretical perspective and the available evidence from Central Europe and Southern Europe to make these hypotheses: 1. If the neo-classical position is correct, the relationship between economic growth and poverty will follow an inverted U-curve. As marketization proceeds, poverty will increase for a while, but eventually the benefits of economic growth generated by the market economy will trickle down lo the poor, decreasing absolute poverty. 2. By incorporating the perspectives of Mitev, Tomova, and Konstantinova (Chapter 2) and Radicova and Vasecka (Chapter 6) into the neo-classical framework, it is possible to argue that a gradualist strategy of market transition may initially delay the growth of poverty, but when poverty begins to increase, it may actually reach higher levels than under conditions of "shock therapy." Thus, we can imagine two inverted U-curves with slightly different shapes and heights, depending on whether the transition was gradual or sudden. In the case of shock therapy, we expect rapid growth of poverty early in transition. The height of the U curve, however, may be relatively low, and the curve may decline rapidly after peaking. Such a curve may describe well the dynamics of poverty in Poland and Hungary: it increased rapidly right after 1989 but peaked as early as 1993 and declined gradually in the past five years. In countries that followed gradualist policies (Romania and Bulgaria), the growth of poverty may have been smaller and slower during the early years of the transition (in Romania poverty probably declined right after 1989), but poverty continued to increase after it began to decline in Poland and Hungary. Furthermore, in Romania and Bulgaria absolute poverty may have been higher than where shock therapy was implemented. 3. We also hypothesize that the presence of an ethnic minority affects the shape of the U-curve. Ethnicity may be used to launch successful struggles to racialize poverty and, therefore, to create an underclass. We argue that the raeialization of poverty is likely to occur as the U-curve begins to decline. Without raeialization, poverty may decline monotonicatly. With raeialization, this decline may stop if an underclass emerges that reproduces itself and consequently perpetuates absolute poverty. While we do not expect that feminization will have strong effects at this stage of marketization, it is possible that feminization and raeialization may have analogous effects on the shape of the U curve describing the relationship between marketization and poverty. We can summarize the different ways that poverty is constructed under the conditions of the market transition in Central Europe, How do raeialization and feminization alter the theoretical controversy between traditional institutionalist and neo-classical economists? If there is an ethnic minority in a society that has an elective affinity with poverty, there wilt be a tendency toward the formation of an underclass, depending on the outcome of the classificatory struggle. Where such an underclass develops, it is less likely that absolute poverty will be reduced as market institutions develop and as economic growth occurs. When The Raeialization and Feminization of Poverty? i" poverty is feminized, an underclass may also develop that inhibits the elimination of absolute poverty. Raeialization and feminization of poverty, however, are strongly affected by access to labor markets. In the transition from socialist redistribution to capitalist markets, the historical legacy of socialism may provide some advantages for women in the labor market at the same time that it disadvantages the Roma. These historically specific attributes of markets are missed by both the institutionalist and the neo-classical economists. The evidence presented in this introduction begins to address these issues, as do the other chapters in this volume. ACKNOWLEDGMENTS We would like to thank Rachel Cohen and Jánoš Ladányi for their assistance with this chapter. REFERENCES Aponte, Robert. 1990. "Definitions of the Underclass: A Critical Analysis." Pp. 117-137 in Sociology in America, edited by H. Gans. Newbury Park, CA: Sage. Ayres, Ron, and David A. Clark. 1998. "Capitalism. Industrialization and Development in Latin America: The Dependency Paradigm Revisited." Capital and Class 64:89-U8. Barany, Zolton D. 1994. "Living on the Edge: The East European Roma in Postcommunist Politics and Societies." Slavic Review 53:321-344. Barm. Fredrik. 1969. "Introduction." Pp. 9-38 in Ethnic Croups and Boundaries: The Social Organization of Cultural Difference, edited by F, Barth. Boston: Little, Brown. Beck, Sam. 1989. "The Origins of Gypsy Slavery in Romania." Dialectical Anthropology 14:53-61. Becker, Gary S. 1991. A Treatise on the Family. Cambridge: Harvard University Press. Beskid, L. 1997. "Ubostvo w Polsce" (Poverty in Poland). Raport Rády Strategii Spoleczno-Gospodarczej przy Raczic Ministrow RP. Blank, Rebecca M. 1997a. // Takes a Nation: A Neiv Agenda for Fighting Poverty. Princeton, NJ: Princeton University Press. Blank, Rebecca M. 1997b. "Why Has Economic Growth Been Such an Ineffective Tool Against Poverty in Recent Years?" Pp, 27-42 in Poverty and Inequality: The Political Economy of Redistribution, edited by J. Neill. New York: W. E. Upjohn Institute for Employment Research. Bokor, Agnes. 1987. Stegěnység a mai Magyarorszagon (Poverty in Hungary Today). Budapest; Magvetd Kiado. Bourdieu, Pierre. 1991. Language and Symbolic Power. Cambridge: Harvard University Press. Caplan, Lionel. 1996. "Dimensions of Urban Poverty: Anglo-Indian Poor and Their Guardians in Madras." Urban Anthropology 25:311-349, Casper, Lynne M., Sara S. McLanahan, and Irwin Garfinkel. 1994. "The Gender-Poverty Gap: What We Can Learn from Other Countries." American Sociological Review 59:594-605. Central and Eastern Europe Business Information Center. 1998/1999. "Country Commercial Guides 1998/1999," American Embassy Staff. (http://www.mac.doc.gov/eebic/ceebic.html). 30 Poverty, Ethnicity, and Gender in Eastern Europe The Racialization and Feminization of Poverty? 31 Central Statistical Office. 1991. Statistical Yearbook of Hungary, 1990, Budapest: Central Statistical Office. Cornia, Giovanni Andrea. 1994. "Income Distribution, Poverty, and Welfare in Transitional Economies—A Comparison between Eastern Europe and China." Journal of International Development 6:569-607. Comia, Giovanni Andrea, and Sheldon Danziger. 1997, Child Poverty and Deprivation in the Industrialized Countries, 1945-1995. Oxford: Clarendon Press. Crowe, David. 1991, "Conclusion." Pp. 151-157 in The Gypsies, edited by D. Crowe and J. Kolsti. London: M. E. Sharpe. Davidova, Eva. 1995, Romano drom—Cesty romu: 1945-1990. Olomouc: Univerzita Palackeho v Olomouci. Erdei, Ferenc. [1943] 1983. "A magyar tarsadalom a ket vilaghaboru kozfttt" (Hungarian society during the inrerwar years). Pp. 291-346 in A magyar tarsadahmral (On Hungarian Society), edited by F. Erdei. Budapest: Akademiai Kiado. Esping-Andersen, Gosta. 1996. "After the Golden Age? Welfare State Dilemmas in a Global Economy." Pp. 1-31 in Welfare States in Transition: National Adaptations in Global Economies, edited by G. Esping-Andersen. London, England: Sage. Ferge, Zsuzsa. 1969. Tarsadalmunk retegzodese (Social Stratification in Hungary). Budapest: KOzgazdasagi es Jogi Konyvldado. Fodor, Eva. 1997. "Power, Patriarchy, and Paternalism: An Examination of the Gendered Nature of State Socialist Authority." Ph.D. thesis. Department of Sociology, University of California, Los Angeles, Frackiewicz, L. 1983. Sfery niedostatht (Spheres of Deficiency). Warsaw: IWZZ. Fraser, Angus M. 1992. The Gypsies. Oxford: Blackwell, Friedman, Milton, and Rose Friedman, 1980. Free to Choose. New York: Avon Books. Gheorghe, Nicolae. 1991. "Roma-Gyspy Ethnicity in Eastern Europe," Social Research 58:829-844. Hayek, Friedrich August von. 1944. The Road to Serfdom. Chicago: University of Chicago Press. Hayek, Friedrich August von. 1960. The Constitution of Liberty. London: Routledge and Kegan Paul. Headey, Bruce, Robert £. Goodin, Ruad Muffels, and Henk-Jan Dirven. 1997. "Welfare over Time: Three Worlds of Welfare Capitalism in Panel Perspective," Journal of Public Policy 17:329-359. Hungarian Central Statistical Office. 1993. Labor Force Survey. Budapest: Hungarian Central Statistical Office. Illyes, Gyula. [1936] 1955. A pusztdk nepe ("The People of the Pusta"). Budapest: Szepirodalmi Konyvkiado. Jarosz, Maria. 1984. Nierownosci spoleczne (Social Inequalities). Warsaw: KiW. Jenkins, Richard. 1994. "Rethinking Ethnicity: Identity, Categorization, and Power." Ethnic and Racial Studies 17:198-227. Kemeny, Istvan. 1972, "Az atacsony jovedelmu csaladok eletkorfilmenyei Magyarorszagon" (Living Conditions of Low Income Families in Hungaiy), Manuscript, Budapest. Keminy, Istvan, 1976. Beszamolo a magyarorszdgi cigdnyok helyzetevel foglalkozd 1971-ben vegzett kutatdsrol (Report on the 1971 Survey of Hungarian Gypsies). Budapest: MTA Szociclogiai Intezete. Kemeny, Istvan, Gabor Havas, and Gabor Kertesi. 1995. "Beszamolo a magyarorszagi roma (cigany) nepesseg helyzetevel foglalkczo 1993 oktobere es 1994 februarja kozott vegzett kulatasrol" (A Report on the Survey of the Hungarian Roma Population, 1993-94). Manuscript. Kenworthy, Lane. 1999. "Do Social-Welfare Policies Reduce Poverty? A Cross-National Assessment." Social Forces 77:1119-1139. Koniai, János. 1980. The Economics of Shortage. Amsterdam: North Holland. Komai, János. 1996. "Reform of the Welfare Sector of Post-Socialist Countries—a Normative Approach." Paper presented at the National Research Council Meetings, November 1996, Washington, DC. Korpi, Walter, and Joakim Patme. 1998, "The Paradox of Redistribution and Strategies of Equality: Welfare State Institutions, Inequality, and Poverty in the Western Countries." American Sociological Review 63:661-687. Kuznets, Simon. 1955. "Economic Growth and Income Inequality." The American Economic Review 45:1-28. Lai, Deepak, and H. Myint. 1996. The Political Economy of Poverty, Equity, and Grotvth: A Comparative Study. Ox ford:-Clarendon Press. Levy, Frank. 1977. "How Big Is the Underclass?" Urban Institute, Washington, DC. Lewis, Oscar. 1966. "The Culture of Poverty." Scientific American 215:19-25. Li, Hong Yi, Lyn Squire, and Heng-fu Zou. 1998. "Explaining International and Intertemporal Variations in Income Inequality." ZTie Economic Journal 108:26-43, Lucassen, Leo, 1991. "The Power of Definition, Stigmatisation, Minoritisation. and Ethnicity Illustrated by the History of the Gypsies in the Netherlands." The Netherlands Journal of Social Science 27:80-91. Massey, Douglas S., and Nancy C. Denton. 1993. American Apartheid. Cambridge: Harvard University Press. McAll, Christopher. 1990. Class. Ethnicity and Social inequality, London: McGill- Queen's. Milanovic, Branko. 1994. "A Cost of Transition—50 Million New Poor and Growing Inequality." Transition 5:1—4. Miles, Robert, 1982. Racism and Migrant Labour. London: Routledge and Kegan Paul. Myrdal, Gunnar. 1963. Challenge to Affluence. New York: Pantheon Books. Myrdal, Gunnar. 1964. 'The War on Poverty." The New Republic 150:14-16. Okely, Judith. 1983, The Traveller-Gypsies. Cambridge: Cambridge University Press. Pinchuk, Ben-Cion. 1990. Shtetl Jews under Soviet Rule. Oxford: Basil Blackwell. Poíanyi, Karl. 1944. The Great Transformation. Boston; Beacon Press. Pronai, Csaba. 1995. Cigänykutalás és kulturális antropológia (Gypsy Research and Cultural Anthropology). Budapest-Kaposvár: N.p. Ricketts, Erol R., and Isabel Sawhill. 1988. "Defining and Measuring the Underclass." Journal of Policy Analysis andManagment 7:316-325. Rodgers, Harrell R., Jr. 1996. Poor Women, Poor Children: American Poverty in the 1990s. Armorik, NY: M. E. Sharpe. Rogger, Hans. 1986. Jewish Policies and Right-Wing Politics in Imperial Russia. London: Macmillan Press. Russell, George. 1977. "The American Underclass." Time. August 28, pp. 14-27. Smedley, Audrey. 1993. Race in North America: Origin and Evolution of a World View. Boulder, CO: Wesrview. Smith, Joan. 1984. "The Paradox of Women's Poverty: Wage-Earning Women and Economic Transformation." Signs: Journal of Women in Culture and Society 10:291-310. Stack, Carol B. 1974. All Our Kin: Strategies for Survival in a Black Community. New York; Harper and Row. Standing, Guy. 1996. "Social Protection in Central and Eastern Europe; A Tale of Slipping Anchors and Tom Safety Nets." Pp. 225-255 in Welfare States in Transition: National Adaptations in Global Economies, edited by G. Esping-Andersen. London, England: Sage. 32 Poverty, Ethnicity, and Gender in Eastern Europe Stephens. John D. !996. "The Scandinavian Welfare States: Achievements, Crisis, and Prospects." Pp. 32-65 in Welfare Stales in Transition: Naiionai Adaptations in Globul Economies, edited by G. Esping-Andersen. London. England: Sage. Stewart, Michael, 1997. The Time of the Gypsies, Boulder, CO: Westview. Stigler, George. 1970. "Director's Law of Public Income Distribution." Journal of Law and Economics 13:1-10. Szelenyi, Ivan, and Donald Treiman. 1992. "Social Stratification in Eastern Europe after 1989." NSF Grant Proposal. Szonda Ipsos. 1997. "Survey 'TF3' Machine Readable Data Set." Szonda Ipsos, Budapest. Tienda, Maria. 1989. "Puerto Ricans and the Underclass Debate." The Annals of the American Academy of Political and Social Sciences 501:105-119. Tomova, Ilona. 1995. The Gypsies in the Transitional Period. Sofia: International Center for Minority Studies and Intercultural Relations. Towns end, Peter. 1970. The Concept ofPoverly. London: Heinemann. Tuili, R, 1978. Jednoosobowegospodarstwa domowe (Single Households). Warsaw: lFiS PAN. Tymowski, A. 1973. Minimum socjalne. Metodyka I proba okreslenia. (Social Minimum, Method and Measurement). Warsaw: PWN. Ward, David. 1989. Poverty. Ethnicity and the American City 1840-1925. Cambridge: Cambridge University Press. Warzywoda-Kntszynska, Wielislawa, and J, Grotowka-Leder. 1996. Wielkomejska bieda ic okresie iransformaeji (Poverty of a Great City in the Period of Transition). Lodz: Instyhit Socjologi Uniwersytetu Lodzkiego. Weber, Max. 197S, "The Belief in Common Ethnicity." Pp 387-393 in Economy and Society, vol. 1, edited by G. Roth and C, Wittich. Berkeley and Los Angeles: University of California Press. Williams, Brackette. 1989, "A Class Act: Anthropology and the Race to Nation across Ethnic Terrain." Annual Review of Anthropology 18:401-^44. Wilson, William Julius. 1978. The Declining Significance of Race: Blacks and Changing American Institutions. Chicago: University of Chicago Press, Wilson, William Julius. 1987, The Truly Disadvantaged. Chicago: University of Chicago Press. Zamfir, E. and C. Zamfir. 1995. Dimensiunile Saraciei (Dimensions of Poverty), Bucharest: Editura Expert.