Echoes from the past: the relationship between satisfaction with economic reforms and voting behavior in Poland.
The return of many reformed communist parties to power implies, at least on the surface, that this logic has played out in the voting booth. Popular press accounts of the 1992 Lithuanian elections, the Polish elections of 1993, and the 1994 Hungarian elections have suggested that people, weary of economic hardship, are casting their lot with postcommunist parties. Cross-national analyses of aggregate voting data in democratizing societies confirm the relationship between macroeconomic deterioration and the rejection of incumbents (Pacek 1994, Remmer 1991). It is not clear, however, that economic dissatisfaction is the primary motivation behind casting a vote for a former communist party.
Negative economic evaluations, which are widespread in postcommunist countries (see Evans and Whitefield 1995), should be connected to some notion of who is to blame for the situation, and how people decide to attribute blame must be understood within the context of a country's transition history. Briefly stated, we argue that transitions from authoritarian rule are centrally constitutive events, the varied interpretations of which become an important part of a country's political culture. Subjective perceptions of the transition from authoritarian rule encompass not simply what happened but also what people believe could or should have happened - and who deserves the blame when times get tough. These assignments of blame are "prior understandings" that fundamentally structure attitudes toward reforms, though subjective perceptions of one's economic situation have the power to reshape these understandings to some degree. With regard to voting, we argue that a vote for a reformed communist party is more a matter of how one understands the transition history - how blame for bad economic circumstances is attributed - than of simple dissatisfaction with reforms or evaluations of one's own personal situation.
We are not the first to suggest that the transition process itself has a lasting effect (O'Donnell and Schmitter 1986, Stepan 1986). Some have gone so far as to argue that different "modes" of transition (e.g., reform, revolution, imposition, pact) are critical determinants of democratic stability and regime consolidation (Karl 1990, Karl and Schmitter 1991). Critics of this approach have pointed to its deterministic tendencies, the inherent difficulties in categorizing countries according to this framework, and the fact that such a narrow conceptualization of transitions does not account for different outcomes of similar transition processes and similar outcomes of disparate processes (Cavarozzi 1992, Hagopian 1993, Hartlyn n.d.). Instead, they believe, we need to pay attention to both long-run historical processes and more proximate factors, such as newly formed institutions, economic interests, or culture.
Our purpose is neither to pursue the attack on modes of transition nor to defend the approach from critics. Insights from the debate, however, can be applied to the study of political behavior in societies undergoing transitions from authoritarian rule. Above all, we argue that transitions do matter, or at least can provide a contextualizing framework for political attitudes and behaviors. At the same time, we agree with critics of the modes-of-transition approach that transitions must be grasped more broadly in the context of long-run historical processes and economic dynamics. For our purpose, the influence of transitions must be viewed not only in the terms describing the nature of elite power transfers but also in the meanings that they have for ordinary people. Simply put, for the average citizen in the street, what looks to one like a "pact" may appear to another as an "imposition."
Based on our experience with the Polish case, we have developed a model that links understandings of the past, economic circumstances, and satisfaction with reforms. We believe the model has applicability to other postcommunist and, perhaps, postauthoritarian societies more generally. Poland is a good candidate for such analysis because the political and economic transitions had been under way long enough by 1993, the date of our survey, to give people a clear understanding of the economic effects of the transition and the contours of the political system. To test for hypothesized relationships between support for reforms and voting behavior, we used data from a national random sample survey of the Polish population immediately after the September 1993 parliamentary elections.(2)
Thus far, debates over the extent to which transitions "matter" have focused on patterns of elite coalitions and institutions without addressing mass politics directly. Likewise, studies of mass political behavior in transitional societies tend not to make explicit links to theories of regime transition.(3) Our study represents a step forward in delineating a micrologic of political behavior for transitional societies that simultaneously are contending with economic reform and the incorporation of the vestiges of an old elite into a democratic order.
TRANSITIONS FROM COMMUNISM AS FOUNDING EVENTS
Founding events are particularly decisive moments for a society, and understandings of them have tremendous salience for the political culture by shaping patterns of discourse (Eliade 1963, Shils 1975, Weber 1968). These streams of discourse, in turn, provide a contextualizing framework for individuals' political thinking (Eckstein 1988, Feldman 1988, Hartz 1955, McClosky and Zaller 1984, Zaller 1992). The East European transitions from communism are classic examples of just such a founding event; they constitute a contested past in which people struggle over what should be remembered and forgotten, and who should be blamed, honored, excused, or praised (Middleton and Edwards 1990, 9). Interpretations of the transition from communism are "prior understandings" that invoke one's cumulative experiences and interpretations of history and help structure political attitudes and behaviors.
Understandings of founding events are not simply factual accounts uniformly held by everyone; they are subjective interpretations filtered through variations in contemporary circumstances (Coser 1992, Halbwachs 1951, Kinder and Sears 1985). Yet, these interpretations are not pure constructions of the present that are refashioned with each change in ephemeral circumstances; understandings of the past possess a certain element of temporal consistency (Middleton and Edwards 1990, Schwartz 1982) in the same way that ideologies and core values are believed to operate (Feldman 1988).
The Polish transition from communism, as is often the case with founding events, was actually a lengthy and complex struggle around which many competing historical accounts have emerged.(4) A prominent discourse reflecting the variation in understandings of the transition history encompasses notions of who is to blame for current economic problems. There was general consensus in 1993 that the economic situation was undesirable in some way, irrespective of the exact problem definition.(5) Just as in the United States in the later phases of the Vietnam War, variations in social discourse centered more around how to improve a bad situation than whether the situation was bad or good. In both cases, debates on how to extricate the country from its predicament presupposed the notion that someone was to blame for it (Schuman and Scott 1989).
The general discourse of blame in Poland may be divided roughly into two currents based upon how the transition from state socialism is understood, The first reveals a deep antipathy toward the communist past, not only as a theoretical system but also as it was actually experienced. It contends that state socialism (1) limited Poland's potential and damaged society and individuals through its inherent flaws or debased form of realization and (2) still bears the responsibility for contemporary problems.(6) The second current of discourse embodies the notion that society was betrayed by opposition leaders who later implemented the early reforms: Either the transition was hijacked by a small elite, or society's goals from the early 1980s were not (or only minimally) realized through the enacted reforms.(7) Affect for the original Solidarity movement may vary in this latter understanding of blame, but its core sense is that something went awry in the transition from state socialism, and the first-wave reformers are to blame.(8)
Blame of Communist System and Blame of First-wave Reformers are measured by the extent to which respondents agree with statements that attribute responsibility for Poland's declining standard of living to 45 years of state socialism and to the first-wave reformers (see Appendix A for questionnaire items). The overall assignment of blame to the early reformers is higher than that attributed to the communist system; Table 1 shows mean values of 3.6 and 3.1, respectively, on a scale of 5. These understandings of blame are inversely related (-.34 Pearson correlation), but more than one-fourth of the sample expressed a high degree of blame for both systems, as revealed in Table 2.
TABLE 1. Mean Values of Blame, by Selected Groupings Blame Blame Communist First-Wave System Reformers Grouping (mean) (mean) Whole sample 3.10 3.60 Men 3.15 3.61 Women 3.06 3.59 Past Communist Party members 2.65 3.80 Past Solidarity members 3.39 3.53 Primary school unfinished 2.99 3.70 Primary school completed 3.05 3.77 Trade school completed 3.10 3.79 High school completed 3.15 3.39 College completed 3.19 2.94 Age 29 and under 3.21 3.46 Age 30-39 3.17 3.58 Age 40-49 3.02 3.68 Age 50-59 2.93 3.70 Age older than 59 3.12 3.59 Income bottom quintile 3.01 3.79 Income second quintile 3.16 3.58 Income third quintile 3.02 3.74 Income fourth quintile 3.13 3.51 Income top quintile 3.24 3.23 Place of residence rural 2.97 3.78 Place [less than] 20,000 people 3.05 3.67 Place 20,000-49,999 3.21 3.65 Place 50,000-100,000 3.08 3.47 Place [greater than] 100,000 3.25 3.35 Church once a year/never 2.85 3.68 Church few times a year 2.94 3.65 Church at least once a month 3.02 3.61 Church once a week or more 3.32 3.54
With few exceptions, most categories of people blame the first-wave reformers more than the communist system, but a trend in one type of blame across a given variable does not imply the opposite trend for the other type of blame. For example, members of the pre-1989 Communist Party are much less inclined to blame Poland's state socialist past for contemporary problems than are former Solidarity members. In contrast, former Solidarity members blame the first-wave reformers only slightly less than do former Communist Party members. There is a dramatic drop in blame attributed to reformers among those who finished at least high school. Some of this may be due to an economic effect, but those who completed high school are also less inclined to feel betrayed by elite reformers; they have a similar social status and, perhaps, had congruent transition goals from the outset.
TABLE 2. Relationship between Blame of the Communist System and Blame of First-Wave Reformers First-Wave Blame Low Medium High Communist system blame Low 6.0% 1.5% 34.3% Medium 1.0% 6.9% 3.0% High 18.5% 2.1% 26.8% Note: 1,581 respondents (85.6% of the sample) fall into one of the four extreme categories. High blame = responses of 4 or 5; low blame = responses of 1 or 2. N = 1,847.
A clear cohort effect is also present. As age rises, blame of the communist system decreases except for those 60 years and older, and the opposite pattern is evident in blame of first-wave reformers. We argue that this cohort effect exists because people over age 60 came to political consciousness before the consolidation of state socialism. Those who survived World War II and remember the hardships and political struggles in its aftermath have a fundamentally different perspective from those born just a few years later. This cohort was socialized during the most concentrated period of state socialism (Sulek 1990), which accounts for the sharp difference between the two oldest age groups. There is no overall trend across income levels, although people in the highest income bracket attribute more blame, on average, to the communist system and less to the first-wave reformers. Finally, as frequency of church attendance rises, average levels of communist system blame markedly increase, but there is no analogous inverse pattern in the blame assigned to the first-wave reformers.
A CAUSAL MODEL OF ATTITUDES TOWARD ECONOMIC REFORMS
We argue that prior understandings of the transition, manifested here as blame for Poland's current circumstances, have a direct effect on attitudes toward the economic reforms implemented since 1990. The more one blames the communist system, the greater is the satisfaction with reforms; the more one blames the first-wave reformers, the lower is the satisfaction with reforms, all else being equal. Although people assign each blame separately, these attributions are not entirely independent of each other. Accordingly, we further hypothesize that the extent to which one blames the communist system will have a direct effect on the level of blame attributed to the first-wave reformers; the greater one's blame of the communist system, the more likely one is to interpret the actions of the first-wave reformers positively by "forgiving" any of the reformers' perceived transgressions or missteps. We argue that communist system blame is causally prior to first-wave reformer blame because most of the experiences integral to the formation of communist blame have a temporally prior status. People's memories of the communist system have existed for much longer and are presumably more stable than their understandings of the first-wave reformers, the memories of whom have a more recent genesis.
Although attributions of blame are rooted in understandings of political history, they are not immune to influence from contemporary personal circumstances. Assessments of how one's life has changed will have some effect on the way the past is interpreted and, thus, on how blame is attributed. Therefore, evaluations of change in living situation not only will positively and directly affect evaluations of reforms but also will dampen or intensify each type of blame. Poles who think they have done well in the last two years will blame first-wave reformers less, and their prosperity may intensify the degree to which they blame communism. Poles who have done poorly may remember communism as "not all that bad" and lower the degree of blame they assign to it, while increasing their antipathy toward first-wave reformers.
Several exogenous variables also are incorporated into the model. First, we include objective socioeconomic measures of Income and Occupation. We expect that the higher one's income, the more likely one is to evaluate change in living situation positively and to have a favorable attitude toward economic reforms. Of all possible occupational categories, we believe that farmers, owners, blue-collar workers in state-owned industrial enterprises, and white-collar workers in the state sector are the most likely to have class-type interests and identities.(9) We hypothesize that owners will be more likely to be satisfied with reforms, whereas farmers will be less likely to be satisfied. Income and occupation are included as controls in all other equations.
We also include reported Membership in Solidarity until 1981 and the Communist Party at any time until 1989. Since actual membership is underreported, these variables are probably best understood as distinguishing a core of people that still identify with these groups.(10) Those who report past membership in the party are less likely to blame the communist system and more likely to blame the first-wave reformers for Poland's problems. Those who claim past Solidarity membership are more likely to blame the communist system, but they are not any more or less likely to blame the first-wave reformers, all else being equal. Although some former members have a continued identification with the Solidarity movement, many of them are among those most disappointed with the former opposition leaders, who are seen as having come to power at the cost of workers' influence.
Finally, we control for Age, Gender, Education, Religion (frequency of church attendance), Size of Place of Residence, and National Particularism, which reflects the extent to which one feels Poland should be cautious toward outside influences and "follow its own road." Since plots of age against the endogenous variables revealed curvilinear relationships, we included a squared age term to capture this effect. To account for an observed step function in education relative to the dependent variables, we include a dummy variable for completion of high school and an interaction term between years of schooling and completion of high school. We expect that those who finished high school will blame the first-wave reformers less than those without a high school diploma, all else being equal. People who are more religious are more likely to blame the communist system and less likely to blame the first-wave reformers, since the fall of state socialism has opened the door to increased Church influence in public and political life. National particularism will have a positive effect on first-wave reformer blame and a negative effect on satisfaction with reforms, which are often considered "foreign" influences (Miller, Reisinger, and Hesli 1996).
The hypothesized relationships are specified by the following structural equation model:
CHANGLIV (Y1) = [[Beta].sub.0] + [[Beta].sub.1] INCOME
+ [[Beta].sub.2] FARMER + [[Beta].sub.3] OWNER + [[Beta].sub.4] BLUESTATE
+ [[Beta].sub.5] WHITESTATE + [[Beta].sub.6] AGE + [[Beta].sub.7] SQAGE
+ [[Beta].sub.8] SEX + [[Beta].sub.9] SIZEPLACE + [[Beta].sub.10] EDUC
+ [[Beta].sub.11] FINHS + [[Beta].sub.12] ED*HS + [E.sub.Y1];
BLAMECOMM (Y2) = [[Beta].sub.0] + [[Beta].sub.1] CHANGLIV
+ [[Beta].sub.2] INCOME + [[Beta].sub.3] FARMER + [[Beta].sub.4] OWNER
+ [[Beta].sub.5] BLUESTATE + [[Beta].sub.6] WHITESTATE
+ [[Beta].sub.7] FORMSOL + [[Beta].sub.8] FORMCOMM+ [[Beta].sub.9] AGE
+ [[Beta].sub.10] SQAGE + [[Beta].sub.11] SEX + [[Beta].sub.12] SIZEPLACE
+ [[Beta].sub.13] EDUC+ [[Beta].sub.14] FINHS + [[Beta].sub.15] ED*HS
+ [[Beta].sub.16] RELIG + [E.sub.Y2];
BLAMEFIRST (Y3) = [[Beta].sub.0] + [[Beta].sub.1] CHANGLIV
+ [[Beta].sub.2] BLAMECOMM + [[Beta].sub.3] INCOME
+ [[Beta].sub.4] FARMER + [[Beta].sub.15] OWNER + [[Beta].sub.6] BLUESTATE
+ [[Beta].sub.7] WHITESTATE + [[Beta].sub.8] FORMCOMM
+ [[Beta].sub.9] AGE + [[Beta].sub.10] SQAGE + [[Beta].sub.11] SEX
+ [[Beta].sub.12] SIZEPLACE + [[Beta].sub.13] EDUC + [[Beta].sub.14] FINHS
+ [[Beta].sub.15] ED*HS + [[Beta].sub.16] RELIG + [[Beta].sub.17] NATION + [E.sub.Y3];
ECONSAT (Y4) = [[Beta].sub.0] + [[Beta].sub.1] CHANGLIV
+ [[Beta].sub.2] BLAMECOMM + [[Beta].sub.3] BLAMEFIRST
+ [[Beta].sub.4] INCOME + [[Beta].sub.5] FARMER + [[Beta].sub.6] OWNER
+ [[Beta].sub.7] BLUESTATE + [[Beta].sub.8] WHITESTATE
+ [[Beta].sub.9] AGE + [[Beta].sub.10] SQAGE + [[Beta].sub.11] SEX
+ [[Beta].sub.12] SIZEPLACE + [[Beta].sub.13] EDUC + [[Beta].sub.14] FINHS
+ [[Beta].sub.15] ED*HS + [[Beta].sub.16] NATION + [E.sub.y4],
BLAMECOMM = blame communist system;
BLAMEFIRST = blame first-wave reformers;
CHANGLIV = evaluation of change in personal living situation;
ECONSAT = satisfaction with economic reforms;
INCOME = monthly income in hundreds of thousands of Polish zlotys;
FORMCOMM = reported membership in Communist Party until 1989;
FORMSOL = reported membership in Solidarity until December 13, 1981;
BLUESTATE = blue-collar workers in state-owned enterprises;
WHITESTATE = white-collar workers in health, education, and social services;
SQAGE = age squared;
EDUC - years of schooling;
FINHS - completion of high school;
ED*HS = interaction term of education and high school completion;
RELIG = frequency of church attendance; and
NATION = national particularism.
To analyze the structural equation model we use the ordinary least-squares method. Figure 1 shows the standardized coefficients of the model of satisfaction with economic reforms.(11) For clarity, we have shown only the paths among the endogenous variables; full model results are given in Table 3.
The results depicted in Figure 1 support the hypothesized relationships. As blame attributed to the communist system increases, satisfaction with economic reforms rises, whereas increasing levels of first-wave reformer blame decrease satisfaction with reforms. Blame of first-wave reformers has a stronger influence, however. This difference in relative effect is not surprising, since memories of the first-wave reformers are more recent and more closely linked with the reforms themselves. A comparison of the total effects (shown in brackets in the figure) indicates that the added effect of both types of blame has a greater influence on satisfaction with economic reforms than does change in personal living situation. In other words, understandings of the past are just as important as personal economic evaluations in determining satisfaction with the economic reforms.
The results also support our argument that blame of the communist system mitigates blame attributed to the first-wave reformers. The direct effect of communist system blame on first-wave reformer blame is greater than the influence of change in one's living situation, even when indirect effects are taken into account. As expected, positive evaluations of change in living situation have a positive effect on satisfaction with economic reforms. Finally, personal economic assessments also are significant in shaping understandings of the transition history: Those who perceive an improvement attach more blame to the communist system and less to the first-wave reformers, while those who believe they are worse off focus more blame on the reformers while mitigating blame ascribed to the communist system.
Looking at the full model in Table 3, former membership in either Solidarity or the Communist Party is a strong determinant of where blame is fixed, but contrary to our expectations, former party members are not significantly more likely than Solidarity members to blame the first-wave reformers. Controlling for all else, income has a positive effect both on assessments of change in living situation and on satisfaction with reforms, as hypothesized, but the weak influence of occupation is somewhat surprising. As a group, the occupational variables were not significant. Age has a significant, curvilinear effect in all the equations, despite the fact that the t values for the separate coefficients do not always appear significant.(12) The interpretation of the age coefficients is that both older and younger people tend to evaluate their change in living situation more positively and are more satisfied with economic reforms than is true of those in the middle of the age range. People at both ends of the age spectrum also blame the communist system less and are more lenient in their assessments of the first-wave reformers.(13)
Education is not a significant predictor of communist system blame, but it has a significant effect in all the other equations. The results clearly show that those who have finished high school perceive that their living situation has improved since 1991, and they attribute significantly less blame to the first-wave reformers. Finally, the more religious people are, the more likely they are to blame the communist system and the less likely they are to blame the first-wave reformers, though the effect is more pronounced with regard to the communist system.
[TABULAR DATA FOR TABLE 3 OMITTED]
ECONOMICS, BLAME, AND VOTING BEHAVIOR
Although the 1993 Polish parliamentary election was commonly perceived as a referendum on the course of economic reforms, it also centered around many other issues, such as the legacy of communism and Churchstate relations (Sabbat-Swidlicka 1993).(14) Table 4 shows the major contenders for the Polish parliament in 1993, along with descriptions of their campaign appeal. We also include parties' left-right placement according to those surveyed in our sample, but this information should be approached with caution. The party space cannot be described in unidimensional categories of left and right because no single shared [TABULAR DATA FOR TABLE 4 OMITTED] conception of the terms exists in Polish politics (Jasiewicz 1995, Wojtaszczyk 1995).(15)
The post-1989 governments were dominated by members of the Solidarity "camp," but coalitions and party affiliations of the representatives shifted frequently. Support for each initiative was cobbled together from a different array of parties, depending on the measure being considered. The Democratic Union (UD) and six other parties from the Solidarity camp, including the Catholic Election Committee (KKW) and the Liberal Democratic Congress (KLD), comprised the minority coalition government. The Solidarity Trade Union (SOLID) was not officially in the coalition but provided soft support, although it instigated the vote of no confidence that led to the government's collapse and precipitated the 1993 elections.(16)
The Alliance of the Democratic Left (SLD) and the Polish Peasant Party (PSL), both reconstituted communist-era parties, were the largest opposition groups prior to the election, but their policy interests were not always congruent. Several parties with a Solidarity pedigree (KdR, PC, and UP) also opposed the government despite their common heritage with the coalition parties. The Confederation of Independent Poland (KPN), with a history of opposition predating the Solidarity movement, was a consistently shrill voice against the government.
The architects of the economic reforms (UD, KLD, and, to a lesser extent, KKW), along with the Union of Real Politics (UPR) and the political newcomer, Nonparty Bloc in Support of Reforms (BBWR), favored maintaining the status quo or accelerating the pace of reform. All other parties advocated some change; two in the Solidarity camp (UP and SOLID) made the strongest appeals for the protection of workers and society's weaker elements through subsidies or the slowing of privatization. The postcommunist SLD made general appeals for greater social protection, but its position was more ambiguous, since it had supported the government on issues of privatization.
To analyze voting behavior in the 1993 Polish parliamentary elections, we use a multinomial logit (MNL) model of unordered choice among several alternatives (see Appendix C for a discussion of this method). Table 5 shows the MNL coefficients, which represent the relative effect of each independent variable on the log-odds ratio of a given party, j, to a chosen reference party, J: ln[Prob(y = j)/Prob(y - J)]. We selected SLD as the reference category because it best illustrates the decision to vote for the postcommunists in comparison with all other parties.(17)
The precise interpretation of MNL coefficients is not intuitively obvious since the dependent variable (logodds ratio) itself is difficult to imagine. Nonetheless, looking at the parameter estimates, we can glean a basic understanding of which variables play an important role in determining the vote for the main postcommunist party (SLD) in comparison with the other parties in our analysis. Table 5 shows the overwhelming salience of communist system blame attribution and religiosity in the decision to vote for any one of the other parties relative to the SLD. People who blamed the communist system were more likely to vote for any party other than SLD, including the reformed communist Polish Peasant Party (PSL). Frequency of church attendance had the strongest effect on vote choice over SLD for parties that employ religious appeals (KKW, SOLID, and PSL). As we would expect, the more voters blamed the first-wave reformers, the more likely they were to choose SLD over one of the parties associated with implementing or advocating further economic reforms (UD, KKW, and BBWR).
Satisfaction with the economic reforms had a significant effect on the decision to vote for not only the main architects of the reforms (UD) but also for the Solidarity Trade Union (SOLID) as compared to SLD. This is an interesting finding, since Solidarity's campaign appeals were arguably more interventionist than those of the postcommunist SLD, but it supports our contention that interpretations of transition history are salient determinants of political behavior. Voters choosing SOLID over SLD were equally angry at first-wave reformers, but they were more satisfied with the reforms because of their intense antipathy toward the communist system and their continued identification with the opposition movement.
Controlling for all other variables, neither change in living situation nor income had an effect on vote choice for any party in comparison to SLD. Moreover, the influence of occupation was minimal. Admittedly, some economic effects may be masked by our decision to include both endogenous and exogenous variables in a single equation, particularly since we have already shown that the blame variables and satisfaction with reforms have a built-in economic component. In gaining the relative advantages of a multinomial logit analysis, we lose the ability to show the causal relationships among the variables that precede attitudes toward economic reforms. Therefore, we have removed the endogenous variables to show the full effects of income and occupation on vote choice. By adding back into the model one endogenous variable at a time, we can roughly replicate the causal order presented earlier in the paper. For clarity, in Table 6 we present the results only for the variables of interest and for four parties that adequately illustrate the differences across the party space: PSL, UD, KKW and SOLID in comparison to SLD.(18)
With all endogenous variables removed (Model 1), income is not a significant predictor of vote for any of the parties over SLD, and occupational effects are of quite limited significance. Given the speculation that imposition of a market economy will produce economic interests which will be expressed in the political realm (cf. Evans and Whitefield 1993, Kitschelt 1992), these results seem somewhat surprising. Our findings suggest that, with a few exceptions, parties are not mobilizing constituencies effectively along class or occupational lines. How people evaluate their change in living situation (Model 2) is an important determinant of vote for SLD over each of the reform-implementing parties (UD and KKW), but its addition to the model does not substantially improve the fit. For all parties, blame of the Communist system is highly significant, and its addition dramatically reduces the effect of change in living situation (Model 3). The inclusion of blame of first-wave reformers in the model does not substantially reduce the value or significance of blame of the communist system for any of the parties. When both blames are controlled, change in living situation remains significant for only one party, UD.
Since the MNL coefficients presented thus far still do not tell us the sensitivity of vote choice to changes in the independent variables, we calculated changes in the overall probability of voting for each of the parties at all levels of blame of the communist system and satisfaction with economic reforms. This is the most common way to interpret the results of MNL analyses.(19) Figure 2 illustrates the effects of communist system blame and satisfaction with economic reforms on the probability of voting for each of the parties when all other independent variables are held at their means.
The effects of communist system blame are particularly striking with regard to the reconstituted communist party (SLD). With all else equal, there is more than a 50% chance that someone who does not blame the communist system at all (value = 1) for Poland's [TABULAR DATA FOR TABLE 5 OMITTED] [TABULAR DATA FOR TABLE 6 OMITTED] problems will vote for SLD. Someone who attributes a high degree of blame to the communist system (value = 5) has only a 10% likelihood of voting for SLD. This effect is not as strong for the other communist-era party, PSL, which shows a curvilinear relationship; the probability of a vote for PSL drops off only at higher levels of communist system blame.
The likelihood of voting for the reform architects (UD) rises as communist system blame increases, but the changes in probability are much smaller than for SLD. A move from the lowest to the highest level of communist system blame is accompanied by a 44% decrease in vote probability for SLD and only a 12% increase in vote probability for UD. By means of comparison, there is only a 13.5% difference in vote probability for UD between the highest and lowest values of first-wave reformer blame. Since the post-Solidarity camp is more fragmented than the postcommunist camp, no single party benefits from high communist system blame in the same way that the postcommunist parties (SLD and PSL) benefit from higher levels of first-wave reformer blame.
Overall, the changes in vote probability for each of the parties over different levels of satisfaction with economic reforms are not as great as those for either blame attribution. As one moves from total dissatisfaction with economic reforms (value = 1) to total satisfaction (value = 5), the probability of voting for UD only rises about 10%, and the probability of voting for SLD drops by only a slightly larger 13.5%. All else being equal, at the highest level of satisfaction with economic reforms, voters are about equally as likely to vote for the party that spearheaded reforms (UD) as for its main postcommunist challenger (SLD).
DISCUSSION: LINKS BETWEEN ATTITUDES AND VOTING BEHAVIOR
Although evaluations of individual economic circumstances have a strong effect on satisfaction with reforms, one's view of the transition history is equally important, and understandings of the past, manifested as blame attributions, are essential determinants of voting behavior. Satisfaction with economic reforms has some influence on voting, but its effect is not as large as one might expect, given the prominence of market reforms and privatization. The greatly attenuated economic effects and the fact that, for many Poles, noneconomic issues are of paramount importance, suggest that the ascendancy of postcommunist parties is not so easily explained by economic dissatisfaction and punishment of incumbents.(20)
If economic circumstances and satisfaction with economic reforms did not play a large role in voter choice, then what accounts for the fact that the vote for postcommunist parties has increased over time? One explanation may lie in the transition history and, odd as it may seem, particularly the original goals of the Solidarity movement. After 1989, the post-Solidarity camp became increasingly fragmented both across and within parties by church-state issues, lustration (see footnote 15), and attempts to claim the "true" legacy of the movement. Politicians associated with this camp came to be regarded as exceedingly quarrelsome and absorbed with their own interests at the expense of society and the movement's goals.
In contrast, the major inheritor of the communist legacy, SLD, was not divided on issues of religion and lustration. Setting aside considerations of platform position for a moment, the SLD, at the very least, had to convince voters inclined to blame the communist system for Poland's problems that it had turned over a new leaf and was substantively different from its communist-era predecessors. Those who attribute little or no blame to the communist system and focus their blame on first-wave reformers could quite easily vote for the postcommunists. At the other extreme, people with the highest degree of blame for the communist system and lowest degree for the first-wave reformers would be very hard to persuade. They either would not be convinced of the postcommunists' competence, given their past errors, or they would see gains of power and privilege by postcommunists as unjust reward in light of their past. Voters with moderate or mixed blame attributions are harder to predict. People who blame both would have to decide not only that the new socialists were sufficiently distinct from the old in their competency, but also that a postcommunist attainment of power would not unfairly reward people from the old system.
As daunting as was the postcommunists' task, the first-wave reformers and, more generally, the post-Solidarity parties had an even larger image problem in some respects. They had to convince potential voters that they had not abandoned society in their struggle for power, and many of them also had to demonstrate they could manage to function as a party, would be able to work with other likely coalition partners, and could produce predictable policy outcomes. Given this scenario, it is easy to imagine how even those with a substantial degree of antipathy toward the communist system could decide that the reformers had outlived their usefulness for the moment and could be persuaded by the pragmatic appeals of the "new" socialists.(21)
In societies emerging from state socialism, where politics itself is discredited and people are often loath to identify with political parties, how one interprets the past fundamentally structures political attitudes and voting behavior. Our analysis of attitudes toward economic reforms in Poland demonstrates that although evaluations of personal economic circumstances strongly affect satisfaction with reforms, how one sees the transition is equally important. Prior understandings of the transition history, manifested as blame attributions, serve as a heuristic frame for sifting through the complicated political and economic situations that often characterize societies in transition. The importance of blame attributions is even more evident in our analysis of voting behavior in the 1993 Polish elections, which reveals that the effect of individual economic circumstances on vote choice was very weak.
What is the implication for future studies of political behavior in Eastern Europe or, more generally, in societies emerging from some form of authoritarian rule? Our findings suggest that we need to view democratic transitions (broadly construed) as critical founding events that should be understood as ordinary people apprehend the changes. Transitions have the potential to create a distinct political cleavage of lasting influence on the public consciousness, particularly when they involve an opposition movement that transcends or cross-cuts any existing social cleavages, as is the case in Poland. Many factors conceivably contribute to the likelihood and depth of a transition-based cleavage. Obviously, "bigger" events will have a larger effect, but certain general dimensions of transitions will determine whether these events play (and will continue to play) a salient role in shaping political attitudes and behavior.
First, protracted struggles are more likely to leave a lasting impression in people's minds than are rapid transitions. Second, transitions in which there is some element of mass participation, even of limited scope, are more likely to remain salient than are transitions involving only elites. People are more likely to feel psychologically invested if they participate (even if only as a spectator), and mass participation is likely to generate expectations that are part of an overt, public discourse.
Third, transitions in which clearly defined, distinct protagonists emerge are likely to crystallize and focus public debate in a way that creates a cleavage and perpetuates it in the future. Subsequent reconciliations and accommodations, such as those in Chile or Poland (the "thick line" demarcating past and present), do not necessarily eliminate distinctions between the opposition and the old regime; they may only recast the form of public discourse. If the truce seems hollow or inauthentic on some level, then it will not resonate with the citizenry and may intensify the cleavage.
Fourth, all else being equal, a violent transition is likely to demarcate stark lines of conflict and have a residual effect on public consciousness. If it is so violent that large segments of the elite, opposition, or society itself are decimated, however, then the "cleavage" may become a moot point.
Finally, the extent to which the old elite is discredited and whether it is removed from political life after the transition affects whether the transition history continues to be salient. For example, salience of a transition-based cleavage may be relatively low in the Czech Republic, where the small elite was thoroughly discredited and quickly removed from political and economic power. The existence of moderate forces or splinter groups in the old elite may blur the distinctions of past and present and make conflicts less pronounced. This seems to be the case in Hungary, where the political elite was less discredited and opposition to the regime not as continuous or well-defined as in Poland. Hungary's political arena remains bifurcated into postcommunist and postopposition camps, but the reformed Socialist Party and one of the main opposition parties have been able to form a coalition government, which has not occurred on the national level in Poland.
Although we have focused primarily on why understandings of the past may remain important for individuals, we also recognize that institutions must reinforce and perpetuate the salience of this division once the original transition stimulus has passed. In Poland, the continued presence of postcommunist parties as the most consolidated forces on the political scene will contribute to the maintenance of a discourse of blame in the immediate future. Although this may dissipate over time, it is also likely that political parties and key personages will continue to inherit and appropriate the legacies of the past. Furthermore, due to the dual nature of the political and economic transitions in Eastern Europe, the continued presence of former elites in economic life will tend to perpetuate a "scorekeeping" discourse centered around who is and rightfully should be profiting from the changes.
Poland's transition history is unique in some ways; perhaps this accounts for one of the most complicated patterns of political attitudes in Eastern Europe (see McDonough 1995 for this observation). Yet, its uniqueness should not be exaggerated. Our analysis of the Polish case suggests a more general framework that can be used to explain both how understandings of the past structure political attitudes and behaviors as well as why the transition history itself may be highly salient in some political cultures and relatively unimportant in others.
APPENDIX A: QUESTIONNAIRE ITEMS
Attribution of Blame
For some time - and especially in the election campaign - a debate has been running about who bears responsibility for the worsening state of our economy and the lowering of the population's standard of living. Do you agree with the opinion that:
[TABULAR DATA FOR TABLE A-1 OMITTED]
Blame Communist Regimes(*)(n = 1,847)
45 years of postwar PUWP and communist governments led the country to ruin and the people to the edge of poverty? (Decidedly so = 22.1%, Somewhat so = 25.2%, Hard to say = 10.9%, Rather not = 24.1%, Decidedly not = 17.7%)
Blame First-Wave Reformers(*) (n = 1,848)
The last 4 years of successive governments pushed our country into a state of chaos and poverty? (Decidedly so = 28.3%, Somewhat so = 35.8%, Hard to say = 10.5%, Rather not = 18.2%, Decidedly not = 7.1%).
Satisfaction with Economic Reforms(*)
In your opinion, are the changes that have occurred in the last four years in the area of economic life: (Very positive? = 4.1%, Somewhat positive? = 22.4%, Hard to say? = 8.8%, Somewhat negative? = 36.0%, Very negative? = 28.8%; n = 1,851).
Evaluation of Change in Living Situation
Since the last elections (1991) has your living situation generally: (Improved? = 6.9%, Not changed? = 27.8%, Worsened? = 65.3%; n = 1,810).
Reported Communist Era Memberships
Solidarity (n = 1,851)
Before the imposition of martial law on December 31, 1981 - in 1980-81 - did you belong to: (The Independent Self-Governing Trade Unions-Solidarity = 18.8%, branch unions = 12.4%, autonomous unions = .6%, Polish Teachers' Union = 1.5%, The Independent Self-Governing Trade Unions of Individual Farmers-Solidarity = .5%, other unions = 1.2%, do not remember = 2.1%, refused answer = .6%, did not belong = 62.3%).
Variable recoded as 1 = reported Solidarity membership, 0 = no reported Solidarity membership
Communist Party (n = 1,848)
Did you belong at any time to:
(The Polish United Workers' Party (PZPR) = 7.9%, United Peasants' Party (ZSL) = 2.9%, Democratic Party (SD) = .4%, Union of Socialist Youth or Union of Polish Socialist Youth = 8.0%, never belonged to any of the aforementioned parties = 80.5%)
ZSL and SD were satellite parties of the main communist party, PZPR. Variable was recoded as 1 = membership in PZPR, ZSL, or SD, 0 = no membership in any communist party.
National Particularism Scale
Do you agree with the following statements? [Scale: decidedly agree = 5, rather agree = 4, hard to say = 3, rather disagree = 2, decidedly disagree = 1]
Poland should not imitate Western patterns, but rely above all on its own traditions and experiences. (5 = 32.6%, 4 = 38.9%, 3 = 8.5%, 2 = 16.4%, 1 = 3.7%; n = 1,845)
Our schools should teach Polish history above all, and to a lesser extent world history. (5 = 22.4%, 4 = 39.0%, 3 = 8.3%, 2 = 24.4%, 1 = 5.9%; n = 1,845)
Foreign capitalists should not be permitted to buy up our Polish wealth - land, factories, houses. (5 = 39.1%, 4 = 30.5%, 3 = 10.4%, 2 = 14.6%, 1 = 5.4%; n = 1,845)
Left-Right Placement of Parties
Some parties taking part in the last elections to the Sejm are defined as "right," and others as "left." How would you situate the parties given below on a scale where 1 indicates a decidedly left orientation, and 7 decidedly right? (See Table A-1 above.)
Vote in 1993 Parliamentary Elections
For which list of candidates to the Sejm did you cast your vote? Please give the number of the list (20 response options = 92.8%, "do not remember" and "refused answer" = 7.2%; n = 1,142).
APPENDIX B: ALTERNATIVE MODEL SPECIFICATION
To test a model with reciprocal paths between blame of first-wave reformers and satisfaction with economic reforms, several control variables that were not significant in the original model were dropped to identify the model adequately. The revised equations are:
CHANGLIV (Y1)=[[Beta].sub.0] + [[Beta].sub.1] INCOME + [[Beta].sub.2] FARMER
+ [[Beta].sub.3] OWNER + [[Beta].sub.4] BLUESTATE + [[Beta].sub.5] WHITESTATE
+ [[Beta].sub.6] AGE + [[Beta].sub.7] SQAGE + [[Beta].sub.8] SEX + [[Beta].sub.9] SIZEPLACE
+[[Beta].sub.10] EDUC + [[Beta].sub.11] FINHS + [[Beta].sub.12] ED*HS + [E.sub.Y1];
BLAMECOMM (Y2) = [[Beta].sub.0] + [[Beta].sub.1] CHANGLIV
+ [[Beta].sub.2] INCOME + [[Beta].sub.3] FARMER + [[Beta].sub.4] OWNER
+ [[Beta].sub.5] BLUESTATE + [[Beta].sub.6] WHITESTATE
+ [[Beta].sub.7] FORMSOL + [[Beta].sub.8] FORMCOMM + [[Beta].sub.9] AGE
+ [[Beta].sub.10] SQAGE + [[Beta].sub.11] SEX + [[Beta].sub.12] SIZEPLACE
+ [[Beta].sub.13] EDUC + [[Beta].sub.14] FINHS + [[Beta].sub.15] ED*HS
+ [[Beta].sub.16] RELIG + [E.sub.Y2];
BLAMEFIRST (Y3) = [[Beta].sub.0] + [[Beta].sub.1] ECONSAT
+ [[Beta].sub.2] CHANGLIV + [[Beta].sub.3] BLAMECOMM + [[Beta].sub.4] INCOME
+ [[Beta].sub.5] FORMCOMM + [[Beta].sub.6] AGE + [[Beta].sub.7] SQAGE
+ [[Beta].sub.7] EDUC + [[Beta].sub.8] FINHS + [[Beta].sub.9] ED*HS + [[Beta].sub.10] RELIG
+ [[Beta].sub.11] NATION + [E.sub.Y3];
ECONSAT (Y4)= [[Beta].sub.0] + [[Beta].sub.1] CHANGLIV
+ [[Beta].sub.2] BLAMECOMM + [[Beta].sub.3] BLAMEFIRST
+ [[Beta].sub.4] INCOME + [[Beta].sub.4] FARMER + [[Beta].sub.6] OWNER
+ [[Beta].sub.7] BLUESTATE + [[Beta].sub.8] WHITESTATE
+ [[Beta].sub.9] AGE + [[Beta].sub.10] SQAGE + [[Beta].sub.11] SEX
+ [[Beta].sub.12] SIZEPLACE + [[Beta].sub.13] EDUC + [[Beta].sub.14] NATION + [[Beta].sub.Y4].
We analyzed the nonrecursive model using two-stage least squares, an instrumental variables regression method that corrects for the bias of OLS estimates in interdependent systems of linear regression equations. The results are consistent with our expectations that blame of first-wave reformers affects one's evaluations of the economic reforms, but not the other way around [ILLUSTRATION FOR FIGURE B-1 OMITTED]. Blame of first-wave reformers had a strong and significant effect at -.69, while the path in the opposite direction was statistically insignificant. The values of the remaining path coefficients did not change dramatically, but the path from blame of communist system to satisfaction with economic reforms became insignificant. We acknowledge, however, that blame of first-wave reformers and satisfaction with economic reforms are closely related temporally and appear to be very similar conceptually, given the high value of the path from the former to the latter. We also offer the caveat that the model is not strongly identified, since it is difficult to find instruments that are strong, unique predictors of both these variables. As a result, a thorough disentanglement of the separate effects of blame of first-wave reformers and satisfaction with economic reforms on each other may not be possible with these data.
APPENDIX C: EXPLANATION OF MULTINOMIAL LOGIT (MNL)
MNL is an appropriate method for analyzing vote choice in multiparty systems because it most closely replicates the choice faced by voters, and it yields results that are best for determining the relative bases of support among all the parties (Whitten and Palmer 1996). MNL models permit the simultaneous comparison of parties relative to one another in comparison to a single reference group rather than through multiple independent paired comparisons. MNL is appropriate for situations in which characteristics of the individuals are the primary determinant of the choice (Schmidt and Strauss 1975).
Despite the advantages of MNL models, the coefficients are somewhat difficult to interpret since they represent the relative effect of each independent variable on the log-odds ratio of a given party in comparison to a reference category. For this reason, we calculated the marginal effects for all independent variables and parties in the analysis. This was done by taking the partial derivative of Prob(y = j) with respect to each independent variable. The results confirm our presentations in tables 5 and 6 along with Figure 2, but they are too cumbersome to report here. Instead, they are available upon request from the authors.
One limitation of MNL is that the disturbance terms are assumed to be uncorrelated, commonly known as the assumption of Independence of Irrelevant Alternatives (IIA) (Greene 1993). When this assumption is violated, nested multinomial and multinomial probit (MNP) models are appropriate, because they relax the IIA assumption. In voting models, violation of the IIA assumption may arise as an issue in one of two ways. First, IIA implies that the removal of one party from the choice set (J) does not affect the relative probabilities of voting for the remaining parties. If one is trying to model the effect of the withdrawal or addition of a party to a choice set, then MNL is not an optimal method (Alvarez and Nagler 1995). Since we are interested in explaining vote choice at one point in time, this problem is not particularly vexing for us (Whitten and Palmer 1996).
Second, if subsets of parties (e.g., "government" or "opposition" parties) in the overall choice set are perceived similarly, this can cause the disturbances to be correlated. This is potentially more of a problem for our analysis, but the use of MNP is to date considered impractical due to the computational power required, and its inability to produce stable estimates for choice sets larger than three or four. Whitten and Palmer (1996), in comparing the results of their own multiparty vote analyses using both MNL and nested multinomial logit, found that the impracticalities associated with the latter outweighed its potential benefits. They conclude that MNL models produce sufficiently accurate results when the independent variables are individual-specific. For the equations of the general model and the partial effects as well as a more detailed treatment of these methodological issues, readers are referred to Greene 1993, Liao 1994, and Whitten and Palmer 1996.
Earlier versions of this paper were presented at the annual meeting of the American Political Science Association, New York, 1994, and at the Mellon Seminar on Collective Actors in Transitional Societies, Duke University, 1995. The survey was commissioned by the Working Group on Electoral Studies at the Institute of Political Studies, Polish Academy of Sciences, and supported by a grant from the Polish Committee of Scientific Research (KBN). The authors thank Stanislaw Gebethner, Krzysztof Jasiewicz, Anna Banaszkiewicz, and Radoslaw Markowski for allowing access to these data, which will be deposited at the Mannhelm Center for European Social Research at some time in the future. The authors also thank the participants in the Mellon seminar, Walter Davis, Barbara Hicks, Nicola Jones, Herbert Kitschelt, David Lowery, Mike Munger, George Rabinowitz, and particularly Stuart Elaine MacDonald for helpful suggestions and criticisms on previous versions. Denise Powers gratefully acknowledges the financial support of the National Science Foundation and the Joint Committee on Eastern Europe of the American Council of Learned Societies and the Social Science Research Council.
1 In contrast, Duch (1993) showed that negative economic evaluations have a positive effect on support for economic reforms in the USSR. His results, however, are consistent with and more appropriately comparable to research conducted in Poland prior to the implementation of market reforms in the late 1980s (see Mason 1992 and Przeworski 1993 for summaries of pretransition public opinion studies).
2 Sampling was done through a stratified cluster method and is representative of all eligible voters in Poland. Face-to-face interviews were conducted with 1,851 respondents.
3 Many studies of public opinion in Eastern Europe and the former Soviet Union have investigated the extent to which people embrace values and beliefs found in Western democracies and considered central to certain types of democracy and market economies (Evans and Whitefield 1995; Finifter 1996; Finifter and Mickiewicz 1992; Miller, Hesli, and Reisinger 1994; Miller, Reisinger, and Hesli 1996; Reisinger et al. 1994). Studies of postauthoritarian Spain have examined the relationship between economic evaluations and government support (McDonough, Barnes, and Lopez Pina 1986).
4 Although the transitions to a capitalist economy and democracy are arguably incomplete, we contend that the transition from state socialism in Poland has concluded. It is disputable when the "point of no return" was reached, but most see the decisive moment as the defeat of the Communist Party in the 1989 elections and the subsequent defection of some party members to help form a Solidarity-led government.
5 By 1993, Poland's unemployment rate was 16.4% (exceeding 30% in some regions), inflation was 36.9% per year, and per-capita income was below pre-1990 levels, despite the fastest rate of GDP growth in Europe - 3.8% annually (Economist Intelligence Unit 1994, Rocznik Statystyczny 1994, Rocznik Statystyczny Wojewodztw 1994).
6 A related line of research posits that people are more likely to support reforms if they fear a return of the communist system (Mishler and Rose 1994, 1996). Although the results of those studies show that perceptions of the communist past have an effect on support for reforms, we doubt that many people in East Central Europe actively feared a return of state socialism by 1991, when the data for the studies were collected. Political debate has centered on how reforms should proceed, not on how to stave off a reconsolidation of state socialism.
7 Although stabilization measures undertaken by the post-Solidarity government in 1990 were generally successful and reasonably popular, such Solidarity movement goals as greater participation in workplace decision making, an increased role for trade unions, and assurance of certain social supports were lost as the focus shifted to privatization and market reform. See Legters (1992) for reprints of the strikers' 21 demands and the resultant Gdansk Agreement signed by the government and the opposition in August 1980.
8 Rifts began to appear in the Solidarity movement in early 1981, and opposition activity from 1981-89 was generally limited to core activists and intellectuals in urban areas. When strikes again swept the country in 1988, the "Solidarity" that participated in the Round Table Accords in 1989 was comprised mainly of activists from the movement's center. The former rank-and-file and many leaders on the periphery, both of whom gave the early movement its mass character, were not an integral part of the opposition that negotiated the end of state socialism.
9 This supposition is loosely based upon each group's future expectations (cf. Kitschelt 1992) as well as shared experiences and social identities under state socialism. In Poland, the peasantry was acknowledged as a distinct social class, and owners began to emerge as a socioeconomic group in the later phases of state socialism. Whitecollar workers in state education, health care, and social services correspond roughly to a noncreative intelligentsia. Blue-collar workers in state-owned enterprises include both skilled and unskilled industrial workers. They are the most likely to have a worker identity, and they are still awaiting the effects of privatization on their place of employment.
10 For both groups, reported past membership in the sample is clearly lower than actual membership. In 1981, roughly half the adult working population in Poland belonged to Solidarity, but only 18.8% of respondents still claim they were members. The number of former Communist Party members is more difficult to gauge, since the party lost about one-third of its membership after 1981. In our sample, 11.3% of respondents said they had belonged at some time to the PZPR or one of its shadow parties. This corresponds roughly to membership estimates from the mid-1980s, but it underestimates the total number of people who were ever in the party.
11 In response to a reviewer's concern, we also tested a model in which reciprocal paths were run between blame of first-wave reformers and satisfaction with economic reforms. The results, which support our hypothesized model, are shown in Appendix B.
12 Age and [age.sup.2] naturally display a high degree of colinearity. Although multicolinearity does not bias the estimates of the coefficients, it makes it difficult to assess their significance. Multicolinearity tends to inflate the standard errors, which reduces the t values, and thus produces coefficients that appear insignificant even when this may not be the case. The presence of significant coefficients indicates a particularly robust effect, and F-tests of the null hypothesis that the aggregate effect of the age variables is zero was rejected for all the equations (i.e., age is significant in all equations).
13 The curves, however, do not all look the same. By taking the derivative of the binomial age expression, setting it equal to zero, and solving for x, we can determine the minimum (or maximum) of each curve. These calculations show that evaluations of change in living situation tend to worsen until about age 43, when they turn upward. Similarly, blame of the communist system is at its lowest among those in their mid-forties. In contrast, blame attributed to first-wave reformers peaks at age 35 and then tails off slowly with increasing age. Satisfaction with economic reforms bottoms out at around age 52, then rises among the oldest people in the sample.
14 This was the second parliamentary election since implementation of the structural adjustment program known as the Balcerowicz Plan in January 1990. By 1993, the effects of market reforms and privatization had become more widespread. The 1993 electoral system for the Sejm was based on proportional representation of 52 electoral districts, with 3 to 17 seats each. Parties had to surpass a 5% threshold nationwide in order to obtain seats; coalitions had to have 8% (Gebethner 1995).
15 "Left" is strongly identified with postcommunist parties, but it also connotes favoring increased state intervention in the economy, which is not entirely true for the main communist successor party. "Right" is often connected in general to post-Solidarity parties and therefore does not have a consistent association with free markets. Most often, "right" denotes religious and nationalist parties, those calling for the lustration (removal with the aim of purification) of former communist functionaries and having a certain shrill style of political conduct characterized by proselytization and combativeness (Pietrzyk-Zieniewicz and Zieniewicz 1995).
16 The Solidarity Trade Union Party initiated an unexpectedly successful vote of no confidence in May 1993 as a move to force government concession to wage increases for striking public sector employees. When no alternative government emerged, President Lech Walesa dissolved parliament and called for elections (see Vinton 1992 and 1993b for good English-language summaries of Poland's party system).
17 The choice of reference category is unimportant for estimation purposes and is based purely on substantive grounds (Liao 1994). Estimates produced with alternative reference categories are simply linear transformations of the same information.
18 Full results for all parties in the original analysis with all estimated coefficients are available from the authors. We have chosen to structure our analysis this way rather than group the parties into blocs based on economic platform for several reasons, including the lack of a clear economic program in some parties and the intricate manner in which economic positions are tied into other issues, thus masking whatever "pure" economic interests may exist.
19 See, for example, Gerber (1996), Whitten and Palmer (1996), and Cameron, Epstein, and O'Halloran (1996).
20 On an aggregate level, one can see that the main party associated with the reforms, UD, was not punished as severely as the two postcommunist parties were rewarded. That is, SLD and PSL increased their share of the vote by 8.4% and 6.2%, respectively, from 1991 to 1993, whereas UD's support slipped by only 1.7% (Gebethner 1995).
21 Two features of the electoral system also amplified the gains of the postcommunist parties: the newly implemented 5% threshold worked against the fragmented Solidarity camp, and a change in the method of seat allocation disproportionately benefited the highest vote-getters (Gebethner 1995).
* A middle category was created from unprompted "hard to say" responses.
Alvarez, R. Michael, and Jonathan Nagler. 1995. "Correlated Disturbances in Discrete Choice Models: A Comparison of Multinomial Probit Models and Logit Models." Political Methodology Working Papers (http:/wizard.ucr.edu/polmeth/working_papers95/alvar95e.html [as of 6/13/97]).
Cameron, Charles, David Epstein, and Sharyn O'Halloran. 1996. "Do Majority-Minority Districts Maximize Substantive Black Representation in Congress?" American Political Science Review 90(4): 794 - 823.
Cavarozzi, Marcelo. 1992. "Beyond Transitions to Democracy in Latin America." Journal of Latin American Studies 24(3): 667-84.
Coser, Lewis A, ed. 1992. On Collective Memory: Maurice Halbwachs. Chicago: University of Chicago Press.
Diamond, Larry. 1992. "Economic Development and Democracy Reconsidered." American Behavioral Scientist 35(4): 450-99.
Duch, Raymond. 1993. "Tolerating Economic Reform: Popular Support for Transition to a Free Market in the Former Soviet Union." American Political Science Review 87(3): 590-608.
Eckstein, Harry. 1988. "A Culturalist Theory of Political Change." American Political Science Review 82(3): 789-804.
Economist Intelligence Unit. 1994. Country Profile. Poland. London: EIU.
Eliade, Mircea. 1963. Myth and Reality. New York: Harper & Row.
Evans, Geoffrey, and Stephen Whitefield. 1993. "Identifying the Bases of Party Competition in Eastern Europe." British Journal of Political Science 23(4): 521-48.
Evans, Geoffrey, and Stephen Whitefield. 1995. "The Politics and Economics of Democratic Commitment: Support for Democracy in Transition Societies." British Journal of Political Science 25(4): 485-514.
Feldman, Stanley. 1988. "Structure and Consistency in Public Opinion: The Role of Core Beliefs and Values." American Journal of Political Science 32(2): 416-40.
Finifter, Ada. 1996. "Attitudes toward Individual Responsibility and Political Reform in the Former Soviet Union." American Political Science Review 90(1): 138-53.
Finifter, Ada, and Ellen Mickiewicz. 1992. "Redefining the Political System of the USSR: Mass Support for Political Change." American Political Science Review 86(4): 857-74.
Gebethner, Stanislaw. 1995. "System Wyborczy: Deformacja czy Reprezentacja?" [The electoral system: Deformation or representation?]. In Wybory Parlamentarne, ed. Stanislaw Gebethner. Warsaw: Wydawnictwo Sejmowe.
Gerber, Elisabeth R. 1996. "Legislative Response to the Threat of Popular Initiatives." American Journal of Political Science 40(1): 99-128.
Gibson, James L., Raymond M. Duch, and Kent L. Tedin. 1992. "Democratic Values and the Transformation of the Soviet Union." Journal of Politics 54(2): 329-71.
Greene, William H. 1993. Econometric Analysis. Canada: Maxwell MacMillan.
Haggard, Stephan, and Robert R. Kaufman. 1995. The Political Economy of Democratic Transitions. Princeton, NJ: Princeton University Press.
Hagopian, Frances. 1993. "After Regime Change: Authoritarian Legacies, Political Representation, and the Democratic Future of South America." World Politics 45(3): 464-500.
Halbwachs, Maurice. 1951. The Collective Memory. New York: Harper-Colophon Books.
Hartlyn, Jonathan. N.d. "Political Continuities, Missed Opportunities and Institutional Rigidities: Another Look at Democratic Transitions in Latin America." In Politics, Society and Democracy: Latin America, ed. Scott P. Mainwaring and Arturo Valenzuela. Boulder: Westview Press. Forthcoming.
Hartz, Louis. 1955. The Liberal Tradition in America. New York: Harcourt Brace Jovanovich.
Jasiewicz, Krzysztof. 1995. "Anarchia czy pluralizm? Podzialy polityczne i zachowania wyborcze w roku 1991 i 1993" [Anarchy or pluralism? Political divisions and electoral behavior in 1991 and 1993]. In Wybory Parlamentarne, ed. Stanislaw Gebethner. Warsaw: Wydawnictwo Sejmowe.
Karl, Terry Lynn. 1990. "Dilemmas of Democratization in Latin America." Comparative Politics 23(1): 1-21.
Karl, Terry Lynn, and Philippe C. Schmitter. 1991. "Modes of Transition in Latin America, Southern and Eastern Europe." International Social Science Journal 128(May): 285-300.
Kinder, Donald R., and David O. Sears. 1985. "Public Opinion and Political Action." In Handbook of Social Psychology, ed. Garndner Lindzey and Elliot Aronson. New York: Random House.
Kitschelt, Herbert. 1992. "The Formation of Party Systems in East Central Europe." Politics and Society 20(1): 7-50.
Legters, Lyman H., ed. 1992. Eastern Europe: Transformation and Revolution, 1945-1991. Lexington, MA: D.C. Heath.
Liao, Tim Futing. 1994. Interpreting Probability Models. Thousand Oaks, CA: Sage Publications.
Mason, David. S. 1992. "Public Opinion in Poland's Transition to Market Democracy." In Escape from Socialism: The Polish Route, ed. Walter Connor and Piotr Ploszajski. Warsaw: IFiS Publishers.
McClosky, Herbert, and John Zaller. 1984. The American Ethos. Cambridge, MA: Harvard University Press.
McDonough, Peter. 1995. "Identities, Ideologies, and Interests: Democratization and the Culture of Mass Politics in Spain and Eastern Europe." Journal of Politics 57(3): 649-76.
McDonough, Peter, Samuel H. Barnes, and Antonio Lopez Pina. 1986. "Economic Policy and Public Opinion in Spain." American Journal of Political Science 30(2): 446-79.
McIntosh, Mary E., and Martha Abele MacIver. 1992. "Coping with Freedom and Uncertainty: Public Opinion in Hungary, Poland, and Czechoslovakia 1989-1992." International Journal of Public Opinion Research 4(4): 375-91.
McIntosh, Mary E., Martha Abele MacIver, Daniel G. Abele, and Dina Smeltz. 1994. "Publics Meet Market Democracy in Central and East Europe, 1991-1993." Slavic Review 53(2): 483-512.
Middleton, David, and Derek Edwards, eds. 1990. Collective Remembering. London: Sage Publications.
Miller, Arthur H., Vicki L. Hesli, and William M. Reisinger. 1994. "Reassessing Mass Support for Political and Economic Change in the Former USSR." American Political Science Review 88(2): 399-411.
Miller, Arthur H., William M. Reisinger, and Vicki L. Hesli. 1996. "Understanding Political Change in Post-Soviet Societies: A Further Commentary on Finifter and Mickiewicz." American Political Science Review 90(1): 153-66.
Mishler, William, and Richard Rose. 1994. "Support for Parliaments and Regimes in the Transition Toward Democracy in Eastern Europe." Legislative Studies Quarterly 19(1): 5-32.
Mishler, William, and Richard Rose. 1996. "Trajectories of Fear and Hope: Support for Democracy in Post-Communist Europe." Comparative Political Studies 28(4): 553-81.
O'Donnell, Guillermo, and Philippe C. Schmitter. 1986. Transitions from Authoritarian Rule: Tentative Conclusions about Uncertain Democracies. Baltimore, MD: Johns Hopkins University Press.
Pacek, Alexander C. 1994. "Macroeconomic Conditions and Electoral Politics in Eastern Europe." American Journal of Political Science 38(3): 723-44.
Pietrzyk-Zieniewicz, Ewa, and Andrzej Zieniewicz. 1995. "Wizerunki autoprezentacyjne wazniejszych ugrupowan politycznych w telewizyjnej kampanii wyborczej w roku 1993" [Self-presented images of the more important political groupings in the television electoral campaign of 1993]. In Wybory Parlamentarne 1991 i 1993, ed. Stanislaw Gebethner. Warsaw: Wydawnictwo Sejmowe.
Przeworski, Adam. 1991. Democracy and the Market. Cambridge: Cambridge University Press.
Przeworski, Adam. 1993. "Economic Reforms, Public Opinion, and Political Institutions: Poland in the Eastern European Perspective." In Economic Reforms in New Democracies: A Social-Democratic Approach, ed. Maravall Pereira and Adam Przeworski. Cambridge: Cambridge University Press.
Reisinger, William M., Arthur H. Miller, Vicki L. Hesli, and Kristen Hill Maher. 1994. "Political Values in Russia, Ukraine and Lithuania: Sources and Implications for Democracy." British Journal of Political Science 24(2): 183-223.
Remmer, Karen L. 1991. "The Political Impact of Economic Crisis in Latin America in the 1980s." American Political Science Review 85(3): 777-800.
Rocznik Statystyczny. 1994. [Statistical yearbook]. Warsaw: Glowny Urzad Statystyczny.
Rocznik Statystyczny Wojewodztw. 1994. [Statistical yearbook of voivodships]. Warsaw: Glowny Urzad Statystyczny.
Sabbat-Swidlicka. 1993. "The Polish Elections: The Church, the Right, and the Left." RFE/RL Research Reports 2(40): 24-30.
Schmidt, Peter, and Robert Strauss. 1975. "The Predictions of Occupation Using Multinomial Logit Models." International Economic Review 16(2): 471-86.
Schuman, Howard, and Jacqueline Scott. 1989. "Generations and Collective Memories." American Sociological Review 54(June): 359-81.
Schwartz, Barry. 1982. "The Social Context of Commemoration: A Study in Collective Memory." Social Forces 61(2): 374-97.
Shils, Edward. 1975. Center and Periphery: Essays in Macrosociology. Chicago: University of Chicago Press.
Stepan, Alfred. 1986. "Paths toward Redemocratization: Theoretical and Comparative Perspectives." In Transitions from Authoritarian Rule: Comparative Perspectives, ed. Guillermo O'Donnell, Philippe C. Schmitter, and Laurence Whitehead. Baltimore, MD: Johns Hopkins University Press.
Sulek, Antoni. 1990. "Life Aspirations of Polish Youth: Intergenerational Stability and Generational Changes." Sisyphus 6(1): 189205.
Vinton, Louisa. 1992. "Poland's Government Crisis: An End in Sight?" RFE/RL Research Reports 1(30): 15-25.
Vinton, Louisa. 1993a. "Poland's Political Spectrum on the Eve of the Elections." RFE/RL Research Reports 2(36): 1-16.
Vinton, Louisa. 1993b. "Walesa Applies Political Shock Therapy." RFE/RL Research Reports 2(24): 1-11.
Weber, Max. 1968. Economy and Society. New York: Bedminster Press.
Whitten, Guy D., and Harvey D. Palmer. 1996. "Heightening Comparativists' Concern for Model Choice: Voting Behavior in Great Britain and the Netherlands." American Journal of Political Science 40(1): 231-60.
Wojtaszczyk, Konstanty A. 1995. "Prawica i lewica na polskiej scenie politycznej" [Left and right on the Polish political scene]. In Wybory Parlamentarne 1991 i 1993, ed. Stanislaw Gebethner. Warsaw: Wydawnictwo Sejmowe.
Zaller, John. 1992. The Nature and Origins of Mass Opinion. Cambridge: Cambridge University Press.
Denise V. Powers is a Ph.D. candidate in Political Science, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, and James H. Cox is Visiting Assistant Professor, Department of Political Science, SUNY, Binghamton, Binghamton, NY 13901.
|Printer friendly Cite/link Email Feedback|
|Author:||Powers, Denise V.; Cox, James H.|
|Publication:||American Political Science Review|
|Date:||Sep 1, 1997|
|Previous Article:||Citizenship and ethnicity: an examination of two transition moments in Kenyan politics.|
|Next Article:||Political constructivism in Rawls's 'Political Liberalism.'|