Globalization,Political Institutions,and Redistribution in Central and Eastern Europe

Introduction

The last few decades have put governments under unprecedented budgetary strain. The Great Recession of the late 2000s and the European sovereign debt crisis of the early 2010s exacerbated the considerable fiscal constraints imposed by demographic change and intensifying globalization. At the same time, they vastly increased the number of citizens dependent on state-sponsored social programs. Policymakers have thus faced difficult choices as they have sought to address emerging needs, alleviate rising economic insecurity, and adapt existing social policies to a context of “permanent austerity” (Pierson, 1998).

The sustainability of existing welfare states in an era of deepening international integration has spurred a rich literature that puts forth two competing hypotheses. While the “race-to-the-bottom” or “efficiency” perspective predicts that governments will reduce redistribution to prevent capital flight, the “compensation” argument expects social safety nets to expand to shield citizens from heightened risk (Cameron, 1978; Ohmae, 1995; Rudra, 2008; Rodrik, 1998). Although multiple studies initially discarded the former, claiming that democratic cabinets respond to public demands for economic redistribution, recent work casts doubt on the resilience of generous welfare states (Heimberger, 2021). Indeed, as Busemeyer (2009) notes, the newer “phase[s] of globalization fundamentally transformed the economic and political basis of the compensatory welfare state” as remaining closed to international trade and capital was no longer a policy choice political elites were willing to consider. Because global integration has been a gradual process with long-reaching consequences, its long-term effects might have begun to unravel only recently. The impact of globalization could thus have changed or crystallized over time, leading to different effects in different periods and countries (Busemeyer, 2009; Ruggie, 1997).

Has the redistributive capacity of modern democracies eroded in the contemporary period? We set out to explore this question in the context of the new democracies of Central and Eastern Europe (CEE). The enormity of external pressures and the relative weakness of representative institutions render this region a least-likely case of welfare state endurance in a globalized world. In the aftermath of the collapse of communism in 1989, the changing economic, ideational, and demographic environment made the communist-era welfare state unsustainable. Influential international financial institutions (IFIs) that postcommunist countries depended on for financing and policy advice promoted tighter eligibility criteria and decreased government responsibility (Deacon, 2000). Influenced by these IFIs, transition states went beyond any other in terms of liberalization (Appel & Orenstein, 2018; Vachudova, 2005), slashing corporate taxation rates, introducing flat tax systems, privatizing pension schemes, deregulating housing and labor markets, and overhauling healthcare and educational systems. Declining labor unions and delegitimized left-wing parties were unable to counteract this policy course. Faced with aging populations, rising debt burdens, and scarce domestic capital, Central and Eastern European governments were expected to cut social spending and benefit entitlements. The deteriorating economic conditions of the 2010s and the illiberal turn that the region took after its accession to the European Union should have accelerated the trend toward lower redistribution.

Has this been the case? Drawing on existing work, we argue that national governments in Central and Eastern Europe have in fact sought to alleviate the rising uncertainty and widening income differentials brought about by economic openness. Given the region’s legacy of comprehensive social provision and high level of support for state-sponsored redistribution, we expect policymakers to have deeply entrenched incentives to maintain existing welfare programs even in the presence of strong, mainly external, pressures for welfare state retrenchment. We believe that this effect is stronger in institutional environments which raise the costs of unpopular reforms. Specifically, we focus on accountability and electoral disproportionality. While the latter reflects the connection between representatives and their constituents by measuring the discrepancy between the share of votes cast and the share of legislative seats gained by political parties, the former captures voters’ ability to punish officeholders for implementing unwanted policies or for failing to deliver on their promises. Low accountability and high disproportionality decrease the probability that public discontent will translate into electoral and reputational losses for incumbents. Globalization is thus associated with redistributive efforts in political contexts that motivate elected legislators to cater to the electorate’s demands.

We test these hypotheses with data from 11 Central and Eastern European countries between 2004 and 2018. Contrary to the race-to-the-bottom hypothesis and consistent with the compensation perspective, we find that economic openness is on average related to higher redistribution. This effect is particularly pronounced when accountability is low and disproportionality is high. In line with recent work (Ibenskas & Polk, 2021), this suggests that policymakers in the region are not isolated from public pressures and remain sensitive to the electoral costs of welfare retrenchment. In fact, our results point to a persistent effort to shield citizens from the disruptions induced by globalization despite the adoption of comprehensive deregulatory reforms meant to attract foreign capital (Appel & Orenstein, 2016). The presence of political institutions that promote accountability and participation is thus strongly linked to post-tax-and-transfer inequality in Central and Eastern Europe. The recent undermining of representative structures in the region thus threatens to bring higher inequality and to leave vulnerable social groups unprotected from risk.

Our results thus qualify existing research on the postcommunist transition which sees CEE political elites as strongly oriented toward the demands of international markets in disregard of domestic concerns and grievances. They show that Central and Eastern European policymakers might have embraced a differential strategy to economic reform which combines rapid deregulation, decisive opening to international trade and investment, and higher levels of economic redistribution. Although a rich literature has highlighted the challenges to democratic consolidation and the disappearance of traditional left-wing parties in the region (Berman and Snegovaya, 2019; Haughton & Deegan-Krause, 2015; Holmes, 2006; Kornai and Rose-Ackerman, 2004; Sikk, 2012; Snegovaya, 2021; Stanley, 2019; Tavits & Letki, 2009), our work suggests that political elites in young democracies shy away from unpopular reforms and continue to invest in robust welfare states. Consistent with Ibenskas and Polk (2021), we find indications that incumbents respond to popular demands for a solid social safety net.

More generally, this is one of very few quantitative studies which systematically examine the determinants of inequality reduction in the postcommunist world (Petrova, 2021). Contrary to existing work, which has tended to focus on social spending, we look into economic redistribution, which reflects broader welfare state dynamics. Furthermore, in contrast to earlier research, which has analyzed welfare state reforms during the first two decades of the transition, we zoom in on the 2000s and the 2010s. This period—especially the years after the Great Recession—has received relatively little scholarly attention. Exploring the drivers of redistribution in the aftermath of the economic crises that shook Europe and the world can thus improve our understanding of the interaction between economic globalization and political institutions at a time of heightened risk, rising discontent, and pressing fiscal constraints.

The paper is structured as follows. We begin by briefly summarizing existing scholarship on redistribution. We then highlight how globalization and domestic politics interact to shape welfare state development in postcommunist Europe. We empirically explore this interaction with cross-sectional time-series analysis of 11 states between 2004 and 2018. In a first step, we document divergent trends in redistribution drawing on individual-level data from the European Union Statistics on Income and Living Conditions (EU-SILC) database. Because of its unprecedented detail, the EU-SILC database is the highest-quality data source to measure income inequality and redistribution dynamics. We proceed to show that economic openness is associated with higher levels of redistribution, especially in contexts characterized by low disproportionality and high vertical and diagonal accountability. The fifth section presents additional robustness checks. We conclude by discussing possible implications for future research.

Literature Review

The question of how globalization affects governments’ ability to provide welfare ¹ has generated a heated debate between proponents of the “race-to-the bottom” hypothesis and adherents of the compensation perspective (Busemeyer, 2009; Heimberger, 2021). The former posits that economic integration puts pressure on political elites to cut social expenditure, lower tax rates, reduce employer contributions, and limit workers’ rights in an effort to attract foreign direct investment (Garrett, 1998; Rudra, 2008). Coupled with small states’ reliance on external capital for growth and development, capital mobility constrains policymakers and induces them to adopt business-friendly policies (Katzenstein, 1985; Ohmae, 1995). The race-to-the-bottom theory thus expects the competition for mobile factors of production to unleash welfare state retrenchment, with policies converging on the “least common denominator.”

The compensation perspective, in contrast, predicts that redistribution increases in response to the rising economic volatility brought about by globalization (Cameron, 1978; Garrett, 1998; Rodrik, 1998). International integration exposes the domestic economy to market instability. Fearing the electoral repercussions of policy inaction, political elites have the incentive to step in and shield citizens from heightened risk. Economic openness thus propels policymakers to expand the social safety net to compensate those hurt by intensifying economic competition.

The debate about the impact of globalization on redistribution has often taken place in the context of the advanced industrialized world. While these countries have undoubtedly been affected by economic openness, they have also been able to influence the speed and rules of economic integration. Postcommunist states, on the other hand, re-inserted themselves in the global economic system from the position of rule-takers who faced tremendous external pressures to deeply restructure their economies (Myant & Drahokoupil, 2012). Their welfare institutions confronted particularly difficult questions as they proved crucial for maintaining social peace but often fell short of adequately addressing the challenges of the market economy.

Welfare state reform was ultimately “difficult, politically challenging, and economically burdensome” (Inglot, 2008). In the words of Gans-Morse and Orenstein (2007, p.14), “CEE politicians and policymakers were caught in a double crossfire: first between the exigencies of market transition and the social legacies of the communist past, and second, between domestic political demands and pressures from international institutions.” On the one hand, the international financial institutions (IFIs) on which the region depended for financing and policy advice promoted a liberal reform agenda in all domains of social security (Kuitto, 2016). The World Bank (WB) and the International Monetary Fund (IMF), in particular, recommended that governments cut and streamline public expenditure, privatize pension systems, introduce means testing, and decrease the generosity of benefits (Deacon & Hulse, 1997; Kuitto, 2016, Careja & Emmenegger, 2009, Orenstein, 2008, 2009). While other prominent organizations, such as the European Union, the International Labor Organization (ILO), and the European Bank for Reconstruction and Development (EBRD), promulgated a Bismarckian model based on universal benefits (Deacon & Hulse, 1997), high debt levels and pronounced economic vulnerability often forced CEE policymakers to follow the IMF and the WB’s programs (Casey, 2004; Careja & Emmenegger, 2009; Muller, 2001).

The implementation of neoliberal reforms was often embraced by domestic political and economic elites (Madariaga, 2020). Consistent with the “race-to-the-bottom” hypothesis, reformers in Central and Eastern Europe believed that ballooning deficits, generous social benefits, and the high taxes necessary to finance them would threaten CEE countries’ ability to attract foreign capital at a time when they desperately needed funds to restructure and grow. The desire to create a business-friendly environment therefore led national governments to cut taxes, reduce social spending, liberalize labor markets, and engage in competitive deregulation (Appel & Orenstein, 2018; Cerami & Vanhuysse, 2009). Economic redistribution was consequently expected to decrease. These pressures were particularly strong in South-Eastern Europe and the Baltics, where democratic institutions were weaker (Madariaga, 2020) and domestic coalitions proved incapable of counteracting the ideologically homogeneous neoliberally inclined political elites (Aidukaitė, 2010; Bohle & Greskovits, 2012).

Nevertheless, existing work on CEE welfare states has documented considerable path dependence, viewing reforms as partial and heavily constrained (Haggard & Kaufman, 2008; Inglot, 2008). Postcommunist social protection systems were characterized by compromise solutions and hybrid arrangements rather than by radical breakthroughs (Brusis, 1999; Cook, 2010). Recognizing the social costs of the transition and their electoral consequences, policymakers sought to aid vulnerable groups, alleviate rising insecurity, and avoid political conflicts (Brusis, 1999). ² This was especially true in more democratic societies, where representative institutions enabled societal interests to influence welfare reforms (Cook, 2007; Orenstein & Haas, 2005). Democracy generally favors the formulation of demands for higher redistribution and induces political elites to compensate citizens for rising insecurity (Boix, 2003; Meltzer & Richard, 1981; Huber & Stephens, 2012). In contrast, retrenchment was much easier in environments where executive power was concentrated and representation was limited (Orenstein, 2008; Cook, 2007). Consistent with Swank (2002), political institutions thus mediated the impact of globalization on economic redistribution in Eastern Europe.

Interestingly, this conditioning effect has not always extended to political parties and other social partners. An influential literature views left parties as representing the interests of low-income constituencies and pursuing a more egalitarian agenda (Hicks, 1999; Huber & Stephens, 2001). While this perspective has found empirical support in Latin America and the advanced industrialized world, scholarship on Central and Eastern Europe has been less conclusive. Although Lipsmeyer (2002) and Cook et al. (1999) indicate that communist successor parties in the region have resisted the tightening of financing and eligibility criteria, Tavits and Letki (2009) have shown that leftist cabinets have not raised general government, health, and education spending. In contrast, Careja and Emmenegger (2009) find that left incumbency is associated with a significant increase in total public and social expenditure. More recent research points to a partisan convergence whereby ideological differences among parties on the left and the right of the political spectrum have disappeared or, at least, considerably declined, during the transition (Appel & Orenstein, 2018; Coman, 2019).

Existing work has thus discussed a variety of factors that affected the development of CEE welfare states during the postcommunist transition. While informative, these studies have reached different conclusions, often producing conflicting narratives about the effect of globalization and political institutions on redistribution in the European periphery. Such disagreement hinders our understanding of the nature and intensity of external pressures as well as the capacity of democratic institutions to withstand these pressures. This problem is especially acute during the later stages of the postcommunist transition, when governments’ need to attract foreign capital might have lessened, existing institutional arrangements constraining globalization might have collapsed, the role of international organizations might have changed following accession into the European Union, and democratic institutions might have eroded. How has international integration shaped redistribution dynamics in the region during this period? Have representative institutions retained their ability to counteract retrenchment pressures or have they hollowed out? Our work seeks to answer these questions by focusing on the second half of the postcommunist transition.

Theory

The combination of hardening budget constraints, exacerbating income insecurity, and high support for a generous social safety net has made welfare reform difficult. On the one hand, consistent with the race-to-the-bottom argument, governments might be under pressure to cut social expenditure, lower tax rates, and reduce employer contributions in an effort to stimulate economic activity and attract foreign direct investment (Rudra, 2008). On the other, as the compensation hypothesis has posited, rising economic volatility and intensifying precariousness motivate incumbents to step in and shield citizens from heightened risk (Cameron, 1978; Garrett, 1998; Rodrik, 1998). We expect the incentive to increase redistribution in response to worsening economic conditions to have been particularly strong after the Great Recession.

A rapidly growing literature has shown that voters oppose austerity, particularly when it entails welfare state retrenchment (Bremer & Bürgisser, 2020; Talving, 2017). Raising taxes and cutting public spending and services are highly visible political decisions that deeply affect citizens. They deprive beneficiaries of (often much needed) resources at times of intensifying vulnerability. The material losses that these policies bring are likely to spur discontent, especially during crises, when government choices are under greater public scrutiny (Bansak et al., 2021; Hübscher et al., 2020; Talving, 2017). Indeed, dissatisfaction with austerity can lead people to mobilize against proposed reforms, resulting in mass protests (Genovese et al., 2016) and social unrest (Ponticelli & Voth, 2020).

The expectation of social backlash is starkly evident in Central and Eastern Europe, where citizens have historically been actively antagonistic to welfare state retrenchment. Attempts to slash benefits and eliminate entitlements provoked mass demonstrations throughout the 1990s and the early 2000s (Inglot, 2008). In fact, widespread protests forced governments across the Visegrad countries to promptly abandon their plans to restructure pension and unemployment systems (Inglot, 2008; Orenstein, 2008). In line with Pierson (1994), who expects social policies to generate self-sustaining constituencies, locking welfare state reform, powerful coalitions of program beneficiaries and state bureaucrats successfully blocked or strongly moderated liberalizing reforms during the first years of the transition (Cook, 2007). Existing work indicates that postcommunist citizens exhibit higher support for state-sponsored redistribution, lower acceptance of inequality, higher propensity to view socio-economic differences as too high, and higher inclination to believe that it is the responsibility of the state to assist vulnerable groups and provide social services (Alesina and Fuchs-Schundeln, 2007; Kluegel and Mason, 2004; Loveless and Whitefield, 2011; Mason, Kluegel and Khakhulina, 2000; Pop-Eleches and Tucker, 2017). Transition states thus largely sought to maintain a strong welfare state commitment and compensate their citizens for the trauma of economic opening and system restructuring (Orenstein & Haas, 2005), opting for “the preservation of social order via protest avoidance” (Vanhuysse, 2009).

Beyond social mobilization, retrenchment decisions carry direct electoral costs. Although not always consistently (Arias & Stasavage, 2019; Bansak et al., 2021; Passarelli & Tabellini, 2017), existing work has shown that voters punish policymakers who adopt unpopular social reforms. Implementing fiscal consolidation hurts incumbents’ re-election chances and seriously undermines governments’ approval and popularity (Hübscher & Sattler, 2017; Hübscher et al., 2020; Jacques & Haffert, 2021; Talving, 2017). Political elites thus have few incentives to engage in retrenchment. They recognize that moves to reduce the generosity of social provision could jeopardize their political future. In fact, vulnerable governments strategically time or altogether avoid austerity reforms to avoid negative electoral consequences (Hübscher & Sattler, 2017).

Consistent with previous research, we maintain that CEE’s insertion into the global economy was accompanied by measures to compensate the losers from globalization. The Great Recession and the European sovereign debt crisis exposed the volatility associated with economic openness. As trade and foreign direct investment slowed down globally, the region, which has historically depended on external capital and demand (Bohle & Greskovits, 2012; Myant & Drahokoupil, 2012), witnessed negative growth and rising precariousness, which exacerbated inequality (Brzezinski, 2018). As foreign financing and economic activity declined, entire industries stumbled, household income fell, and unemployment reached an average of 12.2% (Brzezinski, 2018). We propose that, confronted with deepening economic hardship in a globalized world, national governments opted to continue shielding their populations from intensifying risk in an attempt to protect themselves from the electoral fallout likely to accompany decreases in social spending and services. Thus,

Hypothesis 1

Globalization is associated with an increase in economic redistribution.

Is this effect constant or is it conditional on institutional configurations? Existing work has shown that institutions structure the policy-making environment and shape interactions among different actors (Huber et al., 2020; Immergut, 1992). Crucially, they determine the incentives and constraints that political elites face while in office (North, 1991). Consequently, they can make different policies more or less costly. By involving multiple actors, for example, institutions can disperse responsibility for policy decisions and obstruct straightforward blame attribution. Furthermore, by forcing consensus seeking, they can result in reforms that do not radically depart from the status quo or appeal to different constituencies (Enns et al., 2014). The impact of globalization on redistribution can thus vary across different institutional contexts (Cook, 2007; Swank, 2002).

We expect that governments are more likely to compensate the losers from globalization in institutional environments that make them more sensitive to the electoral and reputational repercussions of inaction or unpopular reforms. The fear of punishment incentivizes policymakers to step in and increase redistribution to correct for the impact of economic openness. The prospects of such punishment, however, are a function of a country’s institutional framework. Indeed, voters’ capacity to penalize elected officials varies across different institutional settings. While some institutions shield politicians from popular discontent, others make it relatively easy for citizens to credibly threaten electoral losses. The latter should motivate incumbents to avoid welfare cuts.

We find two such institutions particularly important. First, electoral disproportionality, or the discrepancy between the share of votes and the share of legislative seats gained by political parties, hinders voters’ ability to effectively discipline their political representatives. High disproportionality enables political parties to attain higher/lower representation in the national legislature than their vote share warrants. This undermines the connection between legislators and their constituency. If a significant loss of votes does not translate into an equally significant loss of seats, political parties might feel less inclined to pursue reforms in line with the preferences of their electorate. Deviating from what citizens want would not necessarily incur losses for incumbents. Similarly, listening to popular demands would not always bring rewards. In contrast, if lower popularity at the voting urns results in lower representation in the national assembly, policymakers will be under greater pressure to avoid unpopular policy choices. Disproportionality thus affects political elites’ behavior in office by shaping their perceptions of risk.

Thus,

Hypothesis 2

Governing elites are more likely to compensate the losers from globalization in political systems characterized by lower levels of disproportionality.

Similarly, accountability captures voters’ ability to hold policymakers responsible for their actions. Higher accountability makes it easier to punish incumbents for going against citizens’ wishes. Doing so threatens to provoke public discontent, which undermines officeholders’ chances of securing re-election and raises the costs of unpopular reforms. When accountability is low, on the other hand, governing elites can get away with policy proposals that do not enjoy widespread popular support. The perception of threat is much lower given that constituents cannot effectively punish policymakers. Accountability therefore strengthens the link between citizens and their representatives by introducing incentives for legislators to adopt reforms closer to their electorate’s preferences. In light of the region’s seeming embrace of state-sponsored redistribution, it increases the payoffs of shielding people from potential economic losses.

Hypothesis 3

Governing elites are more likely to compensate the losers from globalization in political systems characterized by higher accountability.

A brief clarification is in order. We conceptualize accountability as taking place both at election time and between electoral contests. For the latter, we make a distinction between “performance accountability” and “policy-making accountability” (Rose-Ackerman, 2005). Performance accountability entails transparency and free media which allow the public to exercise some oversight over elected officials and to “hold the government to account for both overall policies and the day-to-day implementation of programs” (Rose-Ackerman, 2005, p. 5). Policy-making accountability, on the other hand, requires that policies are a “reflection of the interests and needs of the population” (p. 5). This is achieved through institutions that “channel and manage public participation by individuals and groups in policy making” (p. 6). While multiple such institutions exist, we see active civil society as particularly important, especially in light of the weakness of labor unions and the modest reliance on tripartite groups and social partners in the region (Ost, 2009). Indeed, Rose-Ackerman highlights that postcommunist states placed little emphasis on popular control and accountability outside of the electoral realm during the initial years of the transition, increasing the risk of disillusionment, distrust, and popular disengagement from political life (p. 2). We therefore highlight the mutually reinforcing nature of performance and policy-making accountability between electoral contests, underlining the role of a vibrant civil society and an independent media, which permits the free flow of information, exposes government irregularities, and enables mobilization by constituencies with a strong preference for maintaining welfare state arrangements.

Our argument thus views electoral arrangements, media freedom, and civil society deliberation as crucial for redistribution dynamics in Central and Eastern Europe. By determining how votes translate into seats, how easily information about policy flows, how engaged citizens are in politics, and how easily they mobilize in defense of their interests, these institutions shape voters’ capacity to punish incumbents for unpopular policy choices. Ultimately, this affects how policymakers see their incentives while in office. We thus expect that disproportionality and accountability will mediate the effect of globalization on economic redistribution in the region.

Empirical Strategy

Results

Table 1 presents our findings. Model 1 focuses on the effect of disproportionality and globalization on redistribution. Models 2 and 3 replace disproportionality with vertical and diagonal accountability, which more closely capture the mechanisms we are interested in. ¹⁰ All regressions include the three measures of economic openness.

OPEN IN VIEWER

Table 1.

Effects of Globalization and Politics on Redistribution.

	Model 1	Model 2	Model 3
	b/se	b/se	b/se
Capital openness	8.530*(4.01)	6.277* (2.72)	9.214** (3.10)
FDI flows	.025 (0.02)	.045* (0.02)	.029 (0.02)
Trade	.094** (0.03)	.099*** (0.03)	.124*** (0.03)
Disproportionality	−.293* (0.12)
Vertical accountability		10.383*** (2.47)
Diagonal accountability index			4.055* (1.58)
Electoral democracy	7.286 (8.66)
Left-wing parties	.002 (0.01)	−.013 (0.01)	−.008 (0.01)
Veto points	−1.546 (0.89)	−1.550 (0.87)	−.683 (0.87)
Voter turnout	−.022 (0.04)	−.012 (0.04)	−.028 (0.04)
Market inequality	−.217 (0.15)	−.436** (0.14)	−.307* (0.14)
LN(GDP per capita)	−3.406 (6.13)	−2.552 (4.55)	.294 (4.70)
GDP growth	−.118 (0.08)	−.018 (0.07)	−.015 (0.07)
Unemployment	.347** (0.12)	.441*** (0.11)	.497*** (0.12)
Age dependency ratio	.487* (0.21)	.710*** (0.19)	.578** (0.21)
Budget deficit	.030 (0.10)	−.134 (0.09)	−.140 (0.10)
EMU membership	−1.771* (0.70)	−1.897** (0.69)	−1.909* (0.77)
Crisis	−1.987 (5.00)	−4.464 (4.69)	−5.504 (4.96)
Constant	−463.162 (849.53)	−584.394 (794.73)	−629.469 (851.09)
R-squared	.558	.599	.559
N	131	143	143

Globalization emerges as a meaningful predictor of economic redistribution. Although FDI inflows fail to reach statistical significance in two out of three models, trade and capital account openness return positively signed and statistically significant coefficients. In line with hypothesis 1, higher economic openness—both in terms of capital and goods and services—is associated with a more active role for the state in socio-economic affairs. This effect is not negligible in size—a one-standard-deviation change in capital openness and trade is linked to a 0.43 and a 0.44 standard-deviation change in redistribution, respectively. ¹¹ Integration into the world economy thus appears to motivate governments to alleviate income differentials. This finding goes against the expectation that welfare states will shrink in response to intensifying globalization.

FDI inflows are statistically significant in one out of three models. ¹² This implies that the amount of foreign capital that countries receive does not matter for redistribution as much as the policy framework that governments adopt to attract investment does. In other words, the general legislation regulating the unobstructed entry and exit of capital, not the actual entry per se, is what ultimately affects inequality alleviation. Fluctuations in the amount of investment do not have the capacity to influence welfare state outcomes to the same extent as broader restrictions on cross-border financial transactions do. In this sense, our results suggest that the latent pressures that globalization places on states meaningfully predict policy changes.

Politics also shape redistribution. Consistent with our expectations, accountability promotes redistribution while disproportionality suppresses it. As models 2 and 3 indicate, governments exposed to public pressures seek to reduce income inequality more than those that are shielded from criticism and mobilization. The electorate’s ability to garner information, participate in public deliberations, organize against unpopular reforms, express disagreement with policymakers, and support other political parties matters for efforts to narrow down income differences. In contrast, higher disproportionality is associated with lower redistribution. The larger the incongruence between electoral results and legislative seats, the lower the incentives policymakers have to alleviate income differentials. This might be because disproportionality makes elected officials less wary of punishment at the voting booth or lowers the perceived electoral gains linked to efforts to redistribute income. The effect of these variables is slightly weaker than the impact of globalization: as Figure 2, which presents standardized coefficient plots based on models 1, 2, and 3, illustrates, the effect sizes of disproportionality, vertical and diagonal accountability range between −0.15 and 0.30-standard-deviations.OPEN IN VIEWER

The rest of our political variables warrant further discussion. Partisanship, turnout, democracy, and veto points do not reach statistical significance. While unusual, these findings are not surprising. Existing work on CEE has shown that political parties in the region do not follow the same logic as their Western European counterparts. They were less prone to interfere with the market or increase spending on general government, education, or health care during the 1990s (Coman, 2019; Tavits & Letki, 2009). Our results suggest that this trend persisted throughout the 2000s and the 2010s. This lack of differentiation between left- and right-wing parties corroborates Appel and Orenstein’s argument (2016) that programmatic differences in the socio-economic stances of political parties disappeared in the course of the transition (Berman & Snegovaya, 2019). It also follows the path-dependent nature of welfare state reform, whereby political parties are reluctant to abolish existing welfare schemes that enjoy support (Pierson, 1994).

The lack of statistical significance of the electoral democracy index is more intriguing. It implies that the mere presence of the institutions traditionally associated with democratic governance is not sufficient to induce higher redistribution. This is particularly true in political systems characterized by high degrees of disproportionality and lack of programmatic differences among political formations. Furthermore, as Huber and Stephens (2012) have shown in the context of Latin America, democracy sometimes requires time in order to yield results. As to veto points, it might be that our sample does not exhibit sufficient variation on this indicator during the 2000s and the 2010s. Institutional arrangements of this nature vary little over time, which makes their effect likely to disappear in fixed effects models. It is, therefore, not surprising that the electoral democracy and the veto points indices come out insignificant.

Several other findings stand out. Higher unemployment is correlated with higher redistribution. This is because demand for benefits is greater during bad economic times and automatic stabilizers lead to a greater reduction in income differences. Similarly, more vulnerable people—young or elderly—lead to higher redistribution. These two groups generally increase claims on the social protection system. In contrast, higher market income inequality and membership in the Economic and Monetary Union are associated with lower redistribution. In line with Appel and Orenstein (2016), this suggests that the fiscal constraints driven by the expectation of—and explicit preparation for—EMU membership limited the redistributive potential of the welfare state.

The discussion so far has revealed that globalization and politics shape economic redistribution. How do these two forces interact? Our argument expects incumbents to engage in greater redistribution in response to globalization pressures when political systems are more accountable to voters. To test this expectation, we create interaction terms between our measures of disproportionality and accountability and the three variables that capture international economic integration. The interaction terms with trade and capital account openness return statistically significant coefficients, while those with FDI inflows fail to reach statistical significance (Table 2). Following Brambor et al. (2006), we examine the resulting margins plots to identify the conditional impact of globalization over the range of disproportionality and accountability.

OPEN IN VIEWER

Table 2.

Effects of Globalization and Politics on Redistribution.

	Model 4	Model 5	Model 6	Model 7	Model 8	Model 9	Model 10	Model 11	Model 12
	b/se	b/se	b/se	b/se	b/se	b/se	b/se	b/se	b/se
Disproportionality	.16*** (0.00)	−.003* (0.00)	.011*** (0.00)
Disproportionality × KAO	−.022*** (0)
Disproportionality × FDI		−0 (0)
Disproportionality × Trade			−0*** (0)
Vertical accountability				−.459** (0.15)	.095*** (0.03)	−.060 (0.06)
Vertical accountability × KAO				.581*** (0.15)
Vertical accountability × FDI					.001 (0)
Vertical accountability × Trade						.001** (0)
Diagonal accountability							−.245*** (0.05)	.036* (0.02)	−.108** (0.03)
Diagonal accountability × KAO							.322*** (0.05)
Diagonal accountability × FDI								.001 (0)
Diagonal accountability × Trade									.001*** (0)
Capital openness	.156*** (0.04)	.080* (0.04)	.037 (0.03)	−.814*** (0.22)	.061* (0.03)	.039 (0.03)	−.424*** (0.09)	.088** (0.03)	.070* (0.03)
Foreign direct investment	0 (0)	.001+ (0)	0 (0)	0* (0)	−.001 (0)	0.001* (0)	0+ (0)	−0 (0)	0 (0)
Trade	.001** (0)	.001* (0)	.002*** (0)	.001*** (0)	.001*** (0)	−.001 (0)	.001*** (0)	.001*** (0)	−0 (0)
R-squared	.642	.569	.701	.652	.605	.632	.679	.565	.646
N	131	131	131	143	143	143	143	143	143

The bottom six panels on Figure 3 show that trade and capital account openness are associated with lower redistribution at lower levels of vertical and diagonal accountability. In fact, the effect of these two variables on redistribution is negative (and particularly pronounced in the case of capital account openness). ¹³ In other words, when governments are less accountable to the electorate, incumbents seem to lack incentives to use social services and benefits to reduce income differentials. In contrast, when accountability exceeds a threshold of 1 (for trade) and 1.5 (for capital openness), the effect of globalization becomes positive: officeholders use social spending to alleviate inequality. This suggests that, by itself, economic openness is not enough to induce a more active role for the state in socio-economic affairs. It is only when voters have the capacity to effectively punish governing parties that the latter engage in redistribution in response to external shocks.OPEN IN VIEWER

The upper panels on Figure 3 confirm this. The impact of capital account openness and trade is positive at low and medium levels of disproportionality, but turns negative once votes are decoupled from parliamentary seats. In this sense, when the electoral fortunes of political parties determine their influence within the legislature, the political elite is more likely to attempt to address rising inequalities brought about by globalization through social transfers. In contrast, when the link between votes and seats is weakened, incumbents are less likely to redistribute income. Institutions clearly moderate the effect of economic openness.OPEN IN VIEWER

What explains this? We posit that national governments alleviate income inequality because they are concerned about the electoral and reputational costs associated with welfare state retrenchment or inaction in the face of rising income differentials. This concern is likely to be stronger in countries with a more dynamic civil society which is interested in politics, informed about policy, and prepared to mobilize in defense of its interests. Implementing unpopular reforms in such an environment is riskier than when citizens can’t or don’t actively engage in politics. Unwanted policies are likely to be more visible and public backlash against retrenchment stronger—under such circumstances.

To look into this possibility, we examine the average marginal impact of trade and capital account openness over the range of engaged society (V-Dem’s v2dlengage index), one of the principal dimensions of diagonal accountability. Ranging from 0 (“public deliberation is never allowed”) to 5 (“large numbers of non-elite groups as well as ordinary people tend to discuss major policies among themselves, in the media, in associations or neighborhoods, or in the streets”, p. 151), the indicator captures the extent of public deliberations when important policy changes are considered. The interaction plots on Figure 4 show that economic globalization puts downward pressure on redistribution in societies where public deliberation is non-existent or very limited. This suggests that policymakers feel free to reduce redistribution when civil society is weak and cannot credibly threaten to punish incumbents. Where citizens actively participate in debates and discussions, on the other hand, trade and capital account openness have a positive effect on redistribution. Indeed, the marginal average effect of capital account openness and trade is positive and statistically significant when the index exceeds 1.5 ¹⁴ and 0.5, ¹⁵ respectively. This lends support to the idea that politicians might be concerned about public mobilization against unpopular reforms and might thus base their policy decisions on the electoral costs they expect these reforms to entail.

Do all governments behave in a similar fashion? Our theory posits that incumbents will maintain or increase redistribution in an attempt to protect their popularity. By corollary, fears of a potential backlash should be particularly acute among incumbents who feel vulnerable. If a government enjoys solid political support, it should be less concerned about the effect potential retrenchment might have on its electoral performance. In contrast, a cabinet conscious of its precarious hold on power might be more hesitant to cut benefits because it should be especially sensitive to public discontent.

To test this logic, we look at different government types. We focus on two scenarios: one where a single party occupies all government seats and enjoys a parliamentary majority and another where multiple parties are necessary to form a majority government. ¹⁶ The interaction term plots below (Figure 5) reveal an interesting pattern: trade and capital account openness are linked to higher redistribution under minimal winning coalition governments. Multi-party cabinets that are barely keeping their grasp on power are reluctant to let economic integration raise income differentials. In contrast, single-party majority governments appear less concerned with taking action. Under such cabinets, the impact of trade on the welfare state is not statistically different from zero. The effect of capital account openness is in fact negative. This yields support for the idea that parties whose position in power is more precarious have greater incentives to cushion the disruptions induced by economic openness.OPEN IN VIEWER

Lastly, what policy instruments do governments use to reduce income differentials? Central and Eastern European countries cut tax rates and adopted flat taxation systems during the 2000s (Appel, 2018; Appel & Orenstein, 2018). Policymakers must thus have relied exclusively on social spending to redistribute income. To check whether our argument explains variation in spending, we replace our dependent variable with social protection expenditures. The results, presented in Figure 6 and Table A3 in the Appendix, are largely consistent with our findings. They suggest that higher disproportionality and limited accountability depress the amount governments spend on social benefits. From the globalization variables, only capital account openness has a meaningful effect on social protection spending. The interaction terms between trade, capital account openness and the political variables that we highlight reveal that politicians in more accountable political systems increase the resources they dedicate to social needs in response to globalization. Indeed, trade and capital account openness have a negative effect on social protection spending at low levels of accountability. Once the diagonal and vertical accountability reach the value of 1.5, however, globalization induces positive redistribution.OPEN IN VIEWER

Robustness

Our analysis so far suggests that deepening globalization, higher vertical and diagonal accountability, and lower disproportionality are associated with higher redistribution in Central and Eastern Europe. To test the robustness of these findings, we include additional controls, experiment with different model specifications, use an alternative measure of globalization, de-trend our variables, and run Prais Winsten regressions with country dummies. The results, presented in the Appendix, remain largely similar.

Furthermore, in an attempt to check whether our conclusions can be extended to the universe of postcommunist states during the entire transition, we broaden our sample to include 12 countries from the Commonwealth of Independent States. This requires us to rely on the Standardized World Income Inequality Database (SWIID) for our dependent variable (Solt, 2020). SWIID provides a less precise instrument than the redistribution measure we calculated using the EU-SILC database. Nevertheless, it allows us to add the 1990s and the early 2000s to our analysis. We thus re-run our models against 23 countries from Central and Eastern Europe and Central Asia between 1990 and 2004. While trade and capital account openness fail to reach statistical significance, the interaction terms with FDI inflows reveal a logic reminiscent of our main findings. Foreign capital generally has a negative effect on economic redistribution at low levels of accountability. This implies that, seeking to attract foreign investment, governments in the region typically avoid intervening to reduce inequality when they do not feel directly threatened by voters. Nevertheless, FDI inflows’ impact becomes positive once accountability surpasses the threshold of 1.5. In other words, in settings where citizens have the capacity to hold incumbents responsible for their actions, globalization does indeed lead to more active efforts to decrease income differentials.

Conclusion

This paper examines the impact of globalization and politics on economic redistribution in Central and Eastern Europe between 2004 and 2018. Following a turbulent transition from communism, countries in the region faced the imperative to redesign their welfare state and adapt to a new economic reality dominated by higher economic openness, stronger demographic pressures, and tighter fiscal constraints. The Great Recession of the late 2000s and the European sovereign debt crisis of the early 2010s highlighted the risks of economic integration and further exacerbated these constraints. Policymakers thus confronted the unenviable task of reconciling rising economic insecurity with urgent budgetary considerations.

How did this reconciliation play out? Did governments prioritize attracting foreign capital by retrenching the welfare state? Or did they seek to protect their citizens from intensifying precariousness? Our findings suggest that globalization is associated with active attempts to shield people from risk. Partly because of the legacy of communism, Central and Eastern Europeans tend to dislike high economic inequality and support an active role for the state in socio-economic affairs. As a result, substantial majorities expect the state to be involved in the provision of social benefits and services. Incumbents thus have the incentive to increase or maintain redistribution to meet popular demands. Contrary to the “race-to-the-bottom” hypothesis, office holders strive to reduce income differentials and limit the disruptions induced by economic integration.

This is especially true in political systems characterized by higher accountability and lower disproportionality. Welfare state retrenchment is costlier in environments where citizens can effectively hold the government responsible for its policy choices. Higher accountability forces political elites to heed public demands. Deviations from these demands could potentially result in backlash against elected representatives, especially where civil society is informed, organized, and engaged. Similarly, lower disproportionality strengthens the link between voters and policymakers. By tying legislative seats to electoral performance, higher proportionality makes punishing incumbents for unpopular reforms easier. Institutions therefore mediate—and reinforce—the impact of economic openness on economic redistribution. They increase politicians’ incentives to ameliorate the economic disruptions caused by globalization.

Our analysis further indicates that the governments that engage in compensation tend to be more vulnerable than those that do not. Minimal government coalitions comprising multiple governing parties are more likely to shield citizens from the economic risks brought about by deepening globalization. In contrast, supermajorities that comfortably command more than 50% of legislative seats feel less threatened by the potential discontent that retrenchment might provoke. These results suggests that political elites are aware of the risks associated with cutting welfare provisions and undertake unpopular reforms only when they feel secure enough in their hold on power.

Our analysis thus problematizes the idea that peripheral countries have lost the ability to maintain their welfare states and meaningfully engage in redistribution in an increasingly globalized economic order. The economic crises of the late 2000s and the early 2010s put tremendous pressure on European policymakers, forcing incumbents throughout the region to implement painful economic reforms. The budget constraints imposed by these financial crises cast doubt on the sustainability of postcommunist welfare states. Nevertheless, our work shows that cutting spending and entitlements was not the modal strategy of Eastern European governments. This implies that political calculations and institutional factors continue to shape the effect of external pressures on existing welfare states, and that peripheral governments still have room to maneuver despite the perceived constraints imposed by globalization.

Supplemental Material