Silent Non-Exit and Broken Voice. Early Postcommunist Social Policies as Protest-Preempting Strategies

Introduction. The 1990s as a Critical Juncture in Shaping Policy Pathways

Numerous observers of the postcommunist transitions in Central and Eastern Europe (CEE) have pointed out that these historically unique instances of large-scale social change were also accompanied by often severe social costs. ^[1]

Beyond the short honeymoon periods after the fall of the Berlin Wall, the region’s reform-minded governments generally could not prevent protracted costs associated with such reforms, often resulting in the ousting of these governments at the polls. One important communist inheritance involved jobs, social rights, and social provisions, all of which had been comparatively secure and virtually universal. ^[2] Thus postcommunist voters, to a much greater extent than those in the liberalizing parts of Latin America during the 1980s and 1990s, expected extensive interventions from their elected governments to ensure the presence of social safety nets. ^[3] Indeed, high reliance on state social support survived into the twenty-first century among voters who had been raised and socialized during the communist era. ^[4] Not surprisingly, governments in the first postcommunist years found it extremely difficult to retrench welfare state programs. For instance, Christine S. Lipsmeyer shows that in five postcommunist countries during the mid-1990s, between 75% and 88% of respondents believed that their government should definitely or probably be responsible for ensuring a job for every working-age individual. Between 57% and 91% thought that their government should spend ‘much more’ or ‘more’ on retirement provisions, and between 95% and 98% expressed the view that government should definitely or probably be responsible for a decent standard of living for the elderly. ^[5] Especially in the case of old-age pensions, support for state involvement was patterned relatively straightforwardly according to self-interest along lines of age and class: support systematically increased with age, and decreased with income. ^[6] Postcommunist support levels for state protection against social risks were not only high. Consistent with findings on Western welfare states, those most at risk typically recorded the highest levels of support.

Beyond expressing their dismay about the social costs of transition via elections, vulnerable workers could actively resist reforms by mounting collective protests of the kind that have frequently disrupted reforms in the developing world. ^[7] Compared to the situation under communism, the postcommunist transitions’ losers possessed greater freedom and had more political resources to draw from in organizing protests against reforms. Consider the plight of workers whose jobs were threatened. Across postcommunist Europe (with the exception of the Czech lands), unemployment shot up rapidly after it had been at near-zero official levels for many decades. Crucially, the threat of job loss was not equally distributed but was strongly stratified along geographical (urban-rural) as well as educational lines. For instance, in 1995, 1998, and 2003, the unemployment rate in Poland ranged between 1.5% and 4.4% among university graduates, compared to between 35.5% and 39% among those with a basic vocational education. ^[8] In all four Visegrád countries, for instance, variation in regional unemployment levels tended to be significantly larger than what one encountered in Western economies. ^[9] Vulnerable workers were closely linked in terms of both social-economic and professional status and geographic location. Workplaces, being sites of active social networks, provided an ideal environment for the sharing of grievances and the coordination and mobilization of workers for protests against government reforms. The greater the specificity of the attribution of blame for reform costs, the greater the probability of protest was likely to be. ^[10]

And yet, the early postcommunist years have seen surprising little disruption through strikes and collective protests. ^[11] For instance, workdays lost through strikes and lockouts in postcommunist Europe were neither dramatically nor consistently higher than in the advanced democracies in Western Europe between 1990 and 1995, even though economic hardship in the later countries was far less pronounced. ^[12] Acknowledging such political quiescence, I propose in this essay that we should account for it by looking at how policymakers exercised political agency in order to reduce the potential for social mobilization in the aggregate. I argue that early postcommunist governments that were strongly committed to economic reforms made strategic use of state power in their efforts to ensure political order and to safeguard their initial reforms.

Given that this early phase of transition represented a historic juncture, the scope for such proactive political agency by those in power was especially large. Under communism, special interest groups had either been formally incorporated into the state apparatus or were repressed and marginalized. When the political and social spheres were ‘opened up’ after 1989–1990, this period of early transition represented an archetypal stage of ‘unsettled times,' of the sort that comparative historical institutionalists and students of institutional change and path dependency have regarded as being (temporarily) highly conducive to the exercise of agency and the implementation of strategic measures. ^[13] Policymakers faced the complex task of directing the simultaneous transformations of multiple policy domains. Existing administrative capacities were doubtless strained and overloaded while new policies were being put into place. Yet this context also provided critical opportunities for postcommunist politicians to shape policies for their own benefit. Laws, institutions, even constitutions—the very rules of the political game—were being shaped and reshaped. Obviously constrained by the institutions and capabilities at hand, key policymakers nonetheless found themselves with unusually numerous opportunities to ‘rework’ this existing institutional material. ^[14] Recent empirical studies of the drivers of institutional and policy outcomes in CEE corroborate such an agency thesis. For instance, much of the large diversity among postcommunist regimes (despite the monolithic legacy of communist rule) can be traced to actions taken within a relatively short window during the early 1990s. ^[15] And the influence on living conditions achieved by CEE governments, far from resulting mainly from transitional effects or (non-) responses to international pressures, was exerted through specific policy decisions and policy reforms. ^[16]

Theoretically, this essay contributes to the development of an analytical political sociology analysis of postcommunist policy pathways, and applies such an analysis to the reinterpretation of existing empirical studies on Hungary and Poland. Specifically, using microcollective action theory and building on the theory put forward in my book Divide and Pacify, ^[17] I argue that, as a strategic alternative geared toward the avoidance of political blame, governments in both these democracies during the critical juncture of the 1990s proactively created (rather than compensated or accommodated post hoc) transition losers and/or labor market outsiders in an effort to avoid political protests and to stifle disruptive expressions of political voice. Analytically, this distinction between blame avoidance and protest avoidance mirrors a differentiation between ‘routine’ politics (aimed at reducing electoral losses) and ‘contentious’ politics (aimed at reducing political disruption). ^[18] Microcollective action theory in the tradition of Mancur Olson, Russel Hardin, and Gerald Marwell and Pamela Oliver ^[19] in turn helps to elucidate how the break-up of hitherto relatively homogeneous clusters of threatened workers into newly competing interest groups has shaped the nature of distributive conflict in the formative first decade of these new democracies.

How Social Policies Generated Competing Interest Groups and Reduced the Potential for Collective Mobilization

The key thesis of the 'divide and pacify' theory is that soon after 1989–1990, both the Hungarian and the Polish governments fragmented and thereby ‘immobilized’ the losers of the postcommunist reforms by breaking up groups of well-networked and formally organized at-risk workers into various subgroups marked by fewer common social ties and newly conflicting material interests. ^[20] In the seven years after 1989 alone, hundreds of thousands of workers were transferred out of the labor force and onto early pensions and disability pensions. Whereas the number of 60-plussers remained stable in Hungary and grew by ten percentage points in Poland between 1989 and 1996, the number of old-age pensioners increased by, respectively, one-fifth and 46%. In the same period, the number of disability pensioners also increased by one-half in Hungary and by one-fifth in Poland. ^[21] By establishing these new categories of working-age labor market outsiders, governments could defuse the potential for protests by exposed workers simultaneously threatened by job loss. To achieve this, ‘divide and pacify’ strategies revolved around two, essentially Bismarckian, social mechanisms. First, material benefits were selectively provided to particular target groups whose opposition to reforms would have been most effective in disrupting economic reform progress. Second, these policies ‘created’ distributional conflict between groups that, until that moment, had shared similar objective interests, thereby newly subdividing them into different social and administrative status categories. ^[22]

For my purposes here, these strategies meant that at-risk workers were newly separated, de facto and administratively, into four categories. Precarious citizens of labor-market age were internally divided into (1) regular jobholders, (2) unemployed workers, and (3) irregular or premature (that is, early or disability) pensioners. There were also (4) regular (old age) pensioners. This essay analyzes in greater detail the consequences of this four-way split for political mobilization and collective action as they were first identified in my book Divide and Pacify. The key point of departure here is that seemingly similar individuals—say, newly unemployed miners or steelworkers (group 2) and newly retired early pensioners who until recently had been miners or steelworkers (group 3)—may nevertheless have had a very different and jointly lower capacity for collective mobilization because they had become members of analytically distinct social groups on the basis of traits such as their physical input resources, prior extent of organization, social-network ties, and individual agency. As an exercise in analytical political sociology, I aim to further flesh out these traits and identify the sociological reasons why early ‘divide and pacify’ policies could produce a marked reduction in collective mobilization. Table 1 analyzes these four groups along eight different collective-action dimensions identified by Gerald Marwell and Pamela Oliver, and Chris Hasselmann. ^[23]

The rise in unemployment post-1989 was always bound to create a new category of unemployed workers (group 2). Similarly, offloading workers at risk of job loss onto public pension systems (group 3) had long been a strategy for labor-force reduction in the advanced welfare states of continental Western and Southern Europe. But there it evolved gradually, over several decades rather than just a few years. ^[24] As I have pointed out elsewhere, what made the early exit template in postcommunist Hungary and Poland so remarkable was the unprecedented speed and scale with which this template began to produce its effects. Hundreds of thousands of working-age Hungarians and Poles newly retired in the scope of less than one decade. ^[25] Moreover, the political consequences of the breaking up of clusters of exposed workers into the four interest groups displayed in Table 1 have been particularly far-reaching in the postcommunist case.

Collective action microtheories help reveal certain social mechanisms that have contributed to the stifling of political protest as a result of these social policies. Ever since Mancur Olson’s seminal Logic of Collective Action, such theories start from the observation that to achieve their goals, notional ‘interest groups’—sets of individuals possessing similar interests—typically need to organize collectively and exert pressure on governments. Collective action, however, possesses strong public-good characteristics. ^[26] On the positive side, putting pressure on governments to implement favorable policies tends to display a significant degree of jointness of supply. That is, the cost of realizing this goal rarely increases and conversely the benefit derived from it rarely decreases, in exact proportion to the number of group members. All else equal, large groups therefore enjoy an advantage in organizing, as more members may lead to lower per capita costs of contributing to collective action. ^[27] On the down side, however, collective protests provide nonexcludable benefits. If successful, even group members who did not participate cannot be prevented from consuming the benefits that were obtained. Free-riding thus becomes a core problem for collective action.

Deciding independently from one another, individuals will prefer to spend their limited resources on an (excludable) private good while hoping to freeride on those (non-excludable) public goods financed by others. If, as a worker or pensioner, I choose to use my working day to earn informal wages in the black market while my peers use theirs to protest our firm’s imminent closure or our pension’s real value erosion, I may earn extra income while keeping my regular job or pension revenue. Yet if everyone reasoned this way, the firm would close, pensions would be cut. In line with Mancur Olson’s seminal argument, providing participants with selective benefits—private rewards or penalties—may then become the best hope for successful collective action. However, the social and political-economic characteristics of interest groups themselves form an important further determinant of the likely success of collective action mobilization. Crucially, these later characteristics are to some degree malleable by public policy—and postcommunist governments have used this opportunity to break political voice against reforms.

Broken Voice

Like regular (elderly) pensioners (group 4), all three groups of vulnerable workers had high interests at stake (dimension 1). These were the transition’s losers: materially disadvantaged actors for whom the outcome of state intervention, whether in the form of higher welfare benefits or better job protection and working conditions, had high marginal utility. Dimension 2 captures the group-size effect.

Silent Non-Exit

Post-1989, pensioners constituted a classic example of a single-issue voting constituency, for whom the real value of their public pensions formed a common interest that was both easily identifiable and of high importance (since it was by far their main source of income). ^[54] With elderly pensioners (group 4) now joined by the early and disability pensioners (group 3), pensioners’ electoral clout left them with mainly the weapon of peaceful—i.e. regular ‘democratic’ and non-disruptive—political voice in the polling booth. Postcommunist turnout levels were generally as low as, or lower than, those in the first four general elections in new Latin American democracies. ^[55] Within postcommunist Eastern Europe, nowhere were average postcommunist turnout rates lower during those elections than in the prototypical ‘divide and pacify’ cases, Hungary (66%) and Poland (46%). And crucially, these already low general levels were not likely to be randomly distributed across the total voting population. Rather, non-voters generally tend to be overrepresented at the bottom of the distribution of socioeconomic status, skill/training, and, importantly, age.

Frances Fox Piven and Richard Cloward famously showed how low turnout levels reduce the political influence of the poor. ^[56] The use of voting as an expression of political protest could be expected to be diminished by the well-known tendency of economically weak actors to record low levels of actual involvement in politics. For instance, in a study of over 25,000 citizens across twelve European nations, the unemployed, compared to employed citizens, recorded significantly lower levels of democracy satisfaction in all twelve cases, and significantly lower levels of political interest and of voter turnout in eleven cases. ^[57] In his 1996 Presidential Address to the American Political Science Association, Arendt Lijphart argued that ‘low voter turnout means unequal and socioeconomically biased turnout’. ^[58] Decreasing voting turnout levels of the kind observed in postcommunist democracies are thus likely to have disproportionately reduced the policy influence of the socioeconomically most disadvantaged groups—with one exception: elderly voters tend to record high turnout rates. ^[59]

These arguments indicate a further policy feedback mechanism affecting the various groups benefiting specifically from postcommunist welfare programs. To the extent that any particular welfare-state interest group recorded comparatively lower turnout levels, it was also likely to be increasingly excluded from policy influence over a key policy variable that is systematically affected by turnout levels—welfare spending. ^[60] For instance, stronger voter turnout in OECD democracies raised spending ‘on every kind of social program, as one would expect if one assumed that the social programs cater to the lower income groups whose voter turnout differs most over time and across countries’. ^[61] If lower levels of postcommunist voter turnout and democratic engagement affected especially younger voters such as (low-skilled) families with children and workers, rational politicians could be expected subsequently to target these younger age groups whenever benefit cuts or other retrenchment measures were necessary. Such age-targeted retrenchment became increasingly evident as the 1990s went by. For instance, despite fast-rising numbers, pensioners in Hungary and Poland benefited from generous benefit replacement rates and better-than-average inflation protection throughout the 1990s, in stark contrast with younger generations of welfare state dependants such as the unemployed and families with children. ^[62] By 2004, Hungary and Poland ranked first and second among a group of six new European Union (EU) member states in terms of the adequacy of pension benefits. Pensions represented respectively 75% and 77% of average consumption by pensioners in the poorest quintile of the income distribution, as compared to 42% and 52% in Latvia and Lithuania. But these same two countries ranked second and third lowest among the new EU member states in terms of spending on child allowances, with respectively 0.2% and 0.3% of their GDP spent on this program, as compared to 0.7% and 0.9% in Estonia and Slovenia. ^[63]

The unemployed fared no better over time. Immediately after 1989-1990, Hungary and Poland offered respectively the second and third most generous unemployment replacement rates (70% and 65% in the first six months) within a sample of ten postcommunist states. But by the late 1990s, Hungary had cut those rates marginally (to 55%), whereas Poland had enacted the most drastic cuts of the entire sample (by 30 percentage points). ^[64] Similarly, while these two countries offered by far the most generous maximum duration periods for payment of unemployment benefits in postcommunist Europe in the early 1990s (24 months), they had severely cut these payment periods (by respectively 12 and 6 months) by the end of the decade. ^[65] Other indicators of eligibility and generosity similarly show that after an initially generous starting point in early transition, unemployment benefits were significantly retrenched subsequently. ^[66]

The divide and pacify theory can account for these developments. It assumes that the overriding objective for governments was to split up threatened workers early on into groups 1, 2, and 3, in order to prevent mobilization for collective protests. Interestingly, in regions where unemployment was a larger sociopolitical problem, governments often made special provisions allowing greater generosity. Thus, in Poland, a law enacted in 1996-1997 stipulated that if the local unemployment rate was more than double the national average, and the unemployed had more than 20 years of service, an unemployed spouse, and supports a child of less than 15 years of age, then s/he was entitled to a maximum duration of 18 months. ^[67] However, once the threat of disruptive collective action was reduced by the adverse sociability, group size, and organizational effects discussed above, politicians could safely retrench the benefits going to the electorally less important group (group 2), while safeguarding benefits for the larger electoral constituency of pensioners (groups 3 and 4).

To add insult to injury for younger welfare generations, the positive general age-voting turnout effect could, over time, be increasingly expected to interact with a separate population aging effect, as postcommunist societies faced increasing old age dependency rates into the present century. ^[68] In other words, whereas early and disability pensioners (group 3) were now largely stripped of the ability to mobilize for collective protests, they could still join normal (elderly) pensioners (group 4) in a concomitantly enlarged electoral constituency, free-riding on the later group’s higher levels of voter turnout to boot. Pensioners’ influence on policy as a welfare state pressure group was thus boosted, and their power was further increased vis-à-vis younger generations of welfare state dependents. Unlike pensioners, transition losers stuck in threatened jobs (group 1) or on unemployment programs (group 2) found that their ability to engage in peaceful voice was reduced rather than enhanced. As a collective action category, these groups are likely to remain the true losers of the postcommunist transition: the equivalent of what Mancur Olson famously referred to as ‘“forgotten groups”—those who suffer in silence’. ^[69]

Conclusions. Pathways toward Elderly Policy Bias in Backsliding CEE Democracies

This essay in analytical political sociology has argued that Hungarian and Polish governments have successfully resorted to Bismarckian social policy strategies, aimed at creating, then crystallizing, new lines of conflict between groups of exposed workers in order to stabilize political order and reduce disruptive voice by means of disruptive protest repertoires. At the same time, these strategies are likely to have reduced the use of peaceful voice—voting— among most of the actors targeted, as postcommunist electoral turnout rates have been low and decreasing especially among disadvantaged groups. The one group likely to have been better able to substitute electoral voice for disruptive voice are pensioners, a group that was numerically enlarged through the addition of hundreds of thousands of working-age pensioners in early transition, and which recorded higher turnout rates than younger groups of welfare dependents.

The analysis presented here has implications for wider scholarship. Research on labor market economics ^[70] and political science ^[71] has indicated that, even in coordinated (as opposed to liberal) market economies, and even under leftwing (as opposed to right-wing or Christian-conservative) coalitions, the material interests of labor market outsiders and insiders clash along a large number of social policy and taxation dimensions; such conflict is most often resolved at the expense of the former group. Yet these literatures on advanced welfare states exhibit a degree of status-quo bias as they almost invariably study the institutions, incentives and strategies that govern distributive conflict between existing insiders and outsiders. The present analysis has, in addition, pointed to generalizable reasons why, and ways in which, social policies can be used to proactively create new outsiders. The Western literature can in turn inform another enduring stylized observation about postcommunist polities—the unexpected weakness of trade unions wide across the postcommunist region. ^[72] Trade unions have even stronger incentives than leftwing parties to defend the interests of labor market insiders given that unions do not face the electoral constraint of attracting the support of upscale voters, and insiders tend to be more unionized and more influential than outsiders. ^[73] The same basic constellation of distributive conflict was applicable to, and possibly enhanced in, the postcommunist context, where transition losers such as the jobless and the elderly were among the very first constituencies to be shaken of the radar of union elites.

More speculatively, the successful silencing of economically weak actors may have contributed to the now well-known dramatic ‘illiberal backsliding’ in this region, spearheaded, perhaps not by coincidence, by Poland and Hungary, and accelerating after the 2008 crisis. This regression has been evident in decreasing levels of support for liberal politics (as measured by class-based economic protests, electoral turnout levels, and support for liberal and nonextremist parties) coupled with rising levels of dissatisfaction, disengagement and disillusion regarding the new political and economic order. ^[74] These liberal-democratic disillusions, though generally high, have been especially pronounced among the economic losers of transition, and they were almost invariably stratified along educational and regional lines. For instance, almost one in four Polish university graduates and almost one half of those with basic vocational education believed that ‘the establishment of a market economy is not beneficial’ to them personally in 2000. Respectively 19% and 49% of respondents felt that ‘Poland before 1989 was a better country to live in’ for people of their kind; 11% and 44% agreed that ‘it doesn’t really mater whether a government is democratic or not’ for people of their kind; and 66% and 81% assented to the proposition that ‘people like them had no influence on national issues’. Similar cleavages revealed themselves between residents of villages and those living in the largest urban areas. Responses to the last four questions came out to, respectively, 47% vs. 35%, 47% vs. 24%, 39% vs. 25%, and 85% vs. 78%. ^[75] Social policies, I submit, have contributed to this tale of silent non-exit, broken voice, and political anomie in postcommunist democracies.

Into the present century, the early-transition social policies analyzed in this essay have had significant path-dependent consequences for these postcommunist welfare regimes. If critical junctures such as 1989-1990 are marked by the multiplicity of institutional and policy alternatives, they are necessarily followed by the contraction of this feasible set. In the Hungarian and Polish cases, the creation of working-age pensioners subsequently led to soaring pension system dependency rates and necessitated radical systemic reforms in later years. However, their increased electoral clout could now preemptively influence the policy platforms of vote-seeking politicians, making it harder than before to cut benefits for current pensioners and shifting welfare cutbacks the costs of and systemic pension reforms onto younger workers and future taxpayers. Throughout the late 1990s and into the early 2000s, public expenditures for pensioners have been made increasingly at the expense of younger generations. ^[76] Between 1999 and 2003, total public pension spending in Hungary and Poland, at respectively 11% and 8% of GDP, was well above the OECD average (7.5%), even though both countries still boasted younger populations than most OECD countries. ^[77] In fact, pension spending in Poland by the early 2000s was as high as in the single highest pension spender in the entire OECD – Italy. The key difference was this: Italy at that time was a demographically old society (with 31 65-plussers for every 100 working-age Italians), whereas Poland was still a demographically young nation (with only 20 65-plussers for every 100 working-age Poles). Pensioners appear to have been the most politically favored among the different sociological categories of transition losers (low educated citizens in rural areas). ^[78] The relative incomes of pensioners were not just higher in 2002 than they had been in 1991 in Hungary and Poland (as well as in the Czech lands). Pensioners’ relative incomes were also higher than those of the unemployed (which were in turn higher than those of workers) in all four Visegrád countries in every single year between 1991 and 2002.

More generally, by the mid-2000s, Poland was ahead of even Japan and the three ‘usual Southern suspects’ (Greece, Italy, and Portugal) in recording the single strongest overall welfare state bias towards elderly people. Other CEE countries such as Slovakia, the Czech Republic, Hungary, and Slovenia, also then still demographically much younger societies than the Southern European ones, similarly recorded generally the most heavily pro-elderly biased welfare states in the entire OECD. ^[79] The welfare state in ‘middle-aged’ Hungary (with 3.9 working-aged people per 65-plusser) spent 4.8 times more on every elderly as on every non-elderly citizen. But in slightly older Estonia (with only 3.6 working-aged people per 65-plusser), the welfare state spent only 2.9 times more. And Poland occupied pole position on this elderly bias in social spending measure. In this then still ‘young-to-middle-aged’ society (4.8 working-aged people per 65-plusser), the state spent 8.6 times as much on every elderly Pole as on every nonelderly Pole. Yet in the equally young New Zealand, the state spent only 2.7 times as much. A recalculation of the same elderly bias in social spending (EBiSS) indicator for the years 2010-2011 shows strong path dependency in the elderly orientations of welfare states. Poland remained the most pro-elderly-biased welfare state in the entire OECD sample. The Polish welfare state spent on average 8.7 times as much on each elderly Pole as it spent on each nonelderly Pole. ^[80] Following at some distance, Greece and Italy (EBiSS values around 7), Slovakia (around 6.5), then Czech Republic, Slovenia and Japan (around 5.5), were all positioned on the high-EBiSS side of the spectrum as well.

These observations indicate a special source of policy trouble with Central European welfare states. ^[81] Unlike in the case of elderly societies such as Japan, Greece or Italy, the high levels of pro-elderly welfare state orientation in CEE were the policy result of a politically induced early ‘push’ (great abnormal pensioner booms) and its path-dependent dynamics, rather than of demographically necessitated ‘pull’ (population aging). In the first thirty years of postcommunist transition, CEE’s demographic window of opportunity— the ability to prepare for population aging while still benefiting from relatively younger societies—was largely spurned, not to say smashed. ^[82] That window has now closed. CEE societies, Poland foremost among them, now face fast population aging in the decades ahead. Looking towards future developments, fast-accelerating demographic aging in Hungary and Poland, as in most of CEE, is likely to shift the distributional logic of these welfare regimes still further along the pathway of old age bias in social policy.