In August of 1989 the impossible happened in Poland: a non-communist
government under the leadership of Solidarity took power. Government
bodies, the parliament, the mass media, and other important centers of
power were populated with former political prisoners. Poland was a free
country. Optimism and support for the new government were massive.
In the ensuing months, however, it became clear that the economic
situation was not the sole problem the country was facing. The critical
political issue was whether the Solidarity movement could retain its characteras
broad popular coalition (Heyns & Bialecki 1991). During the spring
of 1990 the political fragmentation within the formerly united Solidarity
front became apparent. The political struggle within the movement resulted
in the announcement by two of its prominent leaders, the chairman of
Solidarity, Lech Walesa, and the Prime Minister. Tadeusz Mazowiecki. of
their candidacies for the office of president.
The purpose of this article is to identify different factors that influenced
voters in the presidential election that followed. The principal focus is
upon the impact of social characteristics and attitudes of voters on their
preferences for different presidential candidates. The analysis uses standard
socio-economic variables such as class. education. and economic standing.

Resumé

lri!'> article (¦(^¦ \;i 1111' K/- lu i'{ >r ti 11_1110!110! i' Hk! j:_1!'!1!'! -ii (o have innuenced pohtica! prc icrc ncc> (H introduction to the political and economic situation of Poland aftcr the rise to power of Solidarity. A short presentation of the main candidates. Walesa. Mazowiecki and Tyminski, and their political profiles is also provided here. Then results from a survey conducted one week before the first round of the presidential elections are presented and discussed. h is concluded that there were some syMemanc Jillercnces in socio-economic status and values oi the supporters of the main candidates. Specific features of Mazowiecki's electorate were: high social position, high level of education and high income. Typical values for this group were political and economic liberalism. Characteristic of Walesa's electorate were a low level of education. higher age, and an occupation as a small businessman. Conservative Catholic morality. anti-Semitism and pro-market attitudes were overrepresented among Walesa's supporters. Tyminski's supporters, tinally, were disproportionately working class and young peisons. His dcctoraic \va> aini-iihcral mi l'uuh iii;.' noliiica! .nul il'u economic mcaninii ot the word. His supporters were often opponents of traditional Catholic morality.

The attitudes studied include the appraisal of a market economy, of Catholic morality, and of liberal values in the political sense of the word. As an introduction, an overview is given of the economic and political development Poland since the takeover of power by Solidarity. The article continues vvith a brief description of the political profiles of the major candidates. The empirical part of the article then contains two sections. In the first section a description of social characteristics and attitudes of the constituencies of the various candidates is given. In the second section. the impact of social and attitudinal characteristics is systematically analyzed by means of multiple regression. The article ends with a discussion of the most important findings.

The Setting

At the point of the transfer of power, Solidarity lacked a specific overall economic program. The main approach advocated by the movement's leading economic advisers was a gradual transformation to a market economy a strong component of industrial democracy. The government's economic program as developed by the liberal economist Leszek Balcerowicz, a different direction, however. The explicit goal of the government program was the rapid introduction of a "Western"-style market economy in Poland. The program included two phases: a stabilization phase and a restructuring phase. The stabilization phase involved the establishment of "real" price mechanisms in the economy. Prices were to be deregulated and an austere financial policy was to help eliminate the budget deficit and dampen aggregate demand. It was hoped that these measures would halt inflation and transform the zloty into a stable and convertible currency. In the second phase, public companies were to be privatized and private ownership was to become the dominant feature of the economy. The goal was to break up the state's monopoly of the production of goods and services and to introduce competitive mechanisms - and, in this way, to stimulate production (Lipton & Sachs 1990; Ash 1991).

Effects of the new economic policy were not long in coming. Market adjustment of prices resulted in an inflational shock in the early months. Eventually, the rate of inflation became more moderate. Financial restraint and curtailment of demand led to a drastic drop in production. A decline in real wages occurred simultaneously with a dramatic increase in unemployment. developments included a stable exchange rate for the country's currency, a surplus in foreign trade, and, not least of all, an end to queuing.

It was obvious from the start that the economic upheaval would entail

severe hardships for the Polish people. It was not equally clear as to how the ensuing discontent would be politically articulated. This uncertainty was dispelled by Lech Walesa. Walesa appointed himself spokesman of the malcontents. This served as the kick-off to the presidential campaign.

Six presidential candidates ran in the first election round. Among these. Lech Walesa and Tadeusz Mazowiecki were generally considered the main contenders. No hope was given the remaining candidates. Least attention was paid to the completely unknown businessman, Stanislaw Tyminski. who, despite a motley past in Canada and Peru, was to leave his mark on the election campaign.¹

Walesa: "I Am For and Even Against"

The chairman of Solidarity, Lech Walesa, had ten years' experience leadins the opposition to the dictatorship. His skills in articulating and channeling the discontent of the people served him well during the election campaign. Walesa attacked Mazowiecki and the government with varying intensity and arguments. In response to the rising level of discontent, the election campaign slogan "acceleration" was adopted. The reforms, it was suggested, were certainly aimed in the right direction, it was just that the process was proceeding far too slowly. If elected president. Walesa promised stepped-up tempo for the reforms and a rapid end to the period of self-sacrifice. This was what could be called the positive phase of the campaign. Walesa did not deny that his and Mazowiecki's policies had much in common.

for the "eggheads from Warsaw" would be håndled successfully by the
electrician from Gdansk.

Mazowiecki: No Short Cuts

Tadeusz Mazowiecki was the opposite of the dynamic Walesa. It took a long time for Mazowiecki to decide to run for President. His campaign staff, in part consisting of well-known intellectuals, had difficulty agreeing on concrete strategies for the election campaign. The main focus of Mazowiecki's campaign ultimately became a defense of the incumbent government and a critique of Walesa and his political supporters. The more encouraging aspects of the economic situation were stressed. It was to the government's credit, it was noted, that the country was on the right track and that it had become a model country for economic reforms throughout Eastern Europe. Mazowiecki had never claimed that the reforms would produce immediate results. The country was still only at the "beginning of the road" and more sacrifices would be necessary. Mazowiecki as President, however, would be a guarantor for political stability and for the continuation of the reform work that should not be jeopardized by rashness. The strongest criticism of Walesa appeared in the election campaign's perhaps most renowned article by Adam Michnik (1990). Michnik wrote: "Are we on the way back to Europe, or on the way back to an old world of populist dictators, of tribal feuds, of unending instability. (. . .) Lech Walesa as President could be a catastrophe for the country. We risk hosting the first Peronist type of regime in central Europe" (author's translation). The wording of the article hints at the atmosphere which prevailed during the election campaign and at the apprehensions that arose in the minds of Walesa's political opponents.

Tyminski: Canada in Every Wallet

Stanislaw Tyminski, by comparison, presented himself as the independent candidate. He had been neither a member of the party nor of Solidarity. Instead, he had spent his time as a businessman in Canada and Peru. Tyminski's election strategy consisted mainly of attacking both Mazowiecki and Walesa. Solidarity was to blame for the declining standard of living in the country. The movement had nothing to show for its ten years of existence. (The fact that Solidarity had been illegal for much of those ten years was ignored by him.) Tyminski's quite apparent goal was to reap any discontent that was aimed at the whole of Solidarity.

economic situation within six months. The initial positive effects would become tangible within one month. How he planned to accomplish this, however, remained unclear. Critics contended that Tyminski's election campaign was a caricature of Walesa's. Where Walesa systematically undermined in the government. Tyminski accused Mazowiecki of treason. Walesa promised vague improvements as a result of "accelerated" reforms; Tyminski declared that this could be realized almost immediately. Populist promises that were aimed at channeling the voters' discontent eventually dominated the election campaign. Specific political issues and political programs, on the other hånd. received only limited space. According to the official results of the first round of the elections. 40 percent voted for Walesa, 23 percent for Tyminski. 18 percent for Mazowiecki, and 19 percent for other candidates. It should be pointed out that 60 percent of those eligible to vote did so in the first round of the elections. In the second round, Walesa won a landslide over Tyminski. Voter participation was about 53 percent.

Data

The following empirical analysis is based on data from a national representative conducted by the Center of Opinion Studies (CBOS) one week prior to the first round of the presidential election in Poland. In the analysis, persons who claimed to know which candidate they were going to vote for have been combined with those who indicated a probable candidate. The account is limited to findings for the three candidates who received the most votes in the first round. Persons who chose other candidates were collapsed into the category '"Others". The CBOS survey underestimated the constituency of Walesa by about 6 percent and overestimated constituencies of Mazowiecki and Tyminski by about 5 and 3 percent respectively.

Socio-Demographic Characteristics

The relationship between the political preferences and social position of voters is well documented in research on voter behavior in democratic countries. People who share similar social circumstances and life situations tend to seek similar political solutions to their problems. Roughly, it could be said that there is a tendency among lower social classes to vote for leftist parties and among higher classes to vote more conservatively. In social science debates on political ideologies and their social origins, such a relationship is taken more or less for granted. The issues debated of late are not whether such a relationship exists, but rather whether the correlation

between class and political preference has been weakened in recent decades, as well as whether other central aspects of the social structure are important explanatory factors of voter alignments (Korpi 1983; Heath et al. 1985; Holmberg et al. 1987; Marshall et al. 1988). In the former Communist countries, the situation is different. The political are often vague and cannot easily be located along a "Western" ideological left-right scale. The political parties are weak and only to a very limited degree promote specific group interests. Whether we can find a social base for the political preferences despite the lack of a more or less clear political map is the issue to be investigated below.

Social Class

In Table 1, the choice of presidential candidate is shown according to respondents' class position. A simple means of summarizing the results in the table is to divide respondents into three main categories: white-collar workers, manual workers, and self-employed persons. As can be seen, a relatively clear-cut class pattern is found in the material. White-collar workers primarily voted for Mazowiecki and secondly for Walesa. Walesa had strong support among manual workers and both groups of selfemployed Moreover, the voters in these latter two groups preferred over Mazowiecki. A more detailed analysis of the material yields further interesting findings. received his strongest support from the higher whitecollar Almost half (45 percent) of the people in senior business and administrative positions, intellectuals, and professional persons claimed that they intended to vote for Mazowiecki (for similar results see Bialecki 1991). The proportion of Tyminski sympathizers among higher white-collar workers, by comparison, was low. His support was stronger among other groups of white-collar workers. That the leader of Solidarity, Walesa, was widely supported by manual workers seems quite natural. More startling, though, was Tyminski's relative among manual workers. In all three of the manual worker categories shown, voters preferred Tyminski over Mazowiecki, despite Mazowiecki's roots in Solidarity and despite the support he received from many of the movement's leading figures. In this respect, Tyminski seems to have been successful in his attempts to exploit worker discontent with Solidarity. His success was primarily at the expense of Mazowiecki. Support for Walesa among farmers and small businessmen was also strong. Nearly half of all small businessmen preferred Walesa.² Tyminski's support among self-employed groups was similarly stronger than Mazowiecki's. it seems that the self-employed middle class to a great extent placed their confidence in candidates who promised rapid improve-

ments and who maintained that the current problems were due to faulty policies rather than to objective conditions beyond political control. Furthermore, for Tyminski was high among employees within commerce. This is a rather heterogeneous category which makes it diffkult to draw any definitive conclusions on the basis of our tindings.

Education und Economic Resources

In Table 2, we see preferences for presidential candidate broken down by the educational level of the voters. A clear correlation is found between education and political preference for two of the candidates. Support tor Mazowiecki was lowest among respondents with only a compulsory school

education, and rose with higher levels of education. Mazowiecki was clearly the most popular presidential candidate among those with higher education. This pattern was reversed for Walesa. The lower the level of education, the greater the support for Walesa.

The positive relationship between level of education and support for Mazowiecki seems to be stronger though than the inverse relationship found for Walesa supporters. The pattern was less clear-cut for Tyminski and the results difficult to interpret. Moreover, preference for "other" candidates was negatively correlated with level of education. The higher the level of education, the smaller the proportion who voted for other presidential candidates.

A relatively clear pattern also emerges when the choice of presidential candidate is broken down by income categories.³ Mazowiecki received his strongest support from high-income earners. People with relatively low incomes, on the other hand, tended to support Walesa. Support for Mazowiecki to a great degree with income, whereas in the case of Walesa, the variation appears to be weaker. For Tyminski, income seems to have played a subordinate role.

Attitudes

How did different attitudes influence voter preferences in the presidential election? To answer this question the effect on voting behavior of three types of attitudes is examined, namely those towards the market economy, towards some central aspects of Catholic morality, and towards liberal values respectively.

In a period when social gaps in the society are widening, it is undeniably a
drawback not to have included such measures.4

Market Economy

As several sociologists have pointed out, the Polish people have opposed the existing political system for decades, while still accepting the basic principles of the economic system. As late as in the middle of the 1980s, a majority of Poles expressed their support for State ownership of the means of production and for a planned economy (Nowak 1988; Ziolkowski 1990). But support for socialistic solutions in the economic sphere has decreased in recent years. This does not mean, however, that market-liberal solutions are embraced without reservation (Adamski 1991).

Voter attitudes towards some central aspects of the market economy are presented in Table 3. Not entirely unexpected is the fact that Mazowiecki's supporters displayed the strongest and most consistent support. Large proportions of his voters supported market-determined prices and a privatization State-owned companies. The acceptance of the negative consequences of the reforms was also most widespread among Mazowiecki's voters. Unemployment was relatively often seen as a necessary cost of the shift to a market economy. Among Walesa's voters, however, support for the liberal economic reforms was weaker. The difference between the two candidates' supporters was most pronounced with regard to the costs of the reforms (the inevitability of unemployment).

The lowest proportions of people who view a market economy favorably

are found among Tyminski's supporters.^s We can also see that support for market principles was, in general, moderate. Only among Mazowiecki's voters did more than 50 percent show favorable market attitudes on at least two of the analyzed questions.

Catholic Values

The Catholic Church views its role in society as the guardian of traditional morals. Periods of political and social unrest are often perceived by the Church as a threat to existing societal norms. In order to protect the faithful from the evils of modernity, control needs to be maintained over the moral standards of the society (Lipset & Rokkan 1967). After the fall of Communism in Poland, the powerful Catholic Church, which most Poles identify with, tried to enhance its traditional role. The Church's interest focuses on moral questions such as child-raising, family relations, and sexuality. In order to have an impact on these areas, the church is seeking to increase its influence within political life, the educational the mass media, and culture; almost all areas of the public arena have been targeted. Church participation in the public debate has been rather brutish at times. The country's bishops, for example, compared politicians who were calling for a referendum on abortion rights to the founders of concentration camps. Many observers consider today's church as an enemy of "the open society".⁶ As evidence from the survey underscores, attitudes towards the church, or more correctly towards Catholic conservatism, are an important political line of demarcation in today's Poland, and played a significant role for Poles during the presidential campaign. In terms of regular attendance at mass, support for mandatory Christian education in the schools, and support an absolute prohibition against abortions, the differences between Mazowiecki's and Tyminski's supporters, on the one hand, and Walesa's, on the other, appear to be extensive and systematic (see Table 4). Walesa's supporters displayed a higher degree of participation in religious activities as well as a greater acceptance of the edicts and prohibitions decreed by the Catholic Church. Attitudes towards Catholic conservatism, in short, appear to have influenced pattern of political preferences in Poland. Conservative Catholics preferred Walesa, whereas those who did not wish to see Catholic ethics imposed in the form of secular laws supported Mazowiecki or Tyminski.

Anti-Liberal Values

Among many observers the fall of Communism has evoked an image of
Eastern Europeans as freedom-loving people who overthrew a "sclerotic"

Stalinist dictatorship. But it soon became clear that Communism was not the only threat to freedom and democracy in these countries. Liberal values (in the political sense of the word) did not prevail uncontested. Racial prejudices, chauvinism, and authoritarianism that had coexisted symbiotically the Communist regimes rose explosively to the surface in some piaces. In Poland, anti-Semitism in particular took on startling proportions.⁷ The political significance of anti-Semitism became abundantly clear during presidential campaign (Gebert 1991: Krzeminski 19^C)1). It was claimed that Mazowiecki and other prominent politicians who supported his candidacy were Jews. A recurring demand of audiences at Walesa's election meetings was a purging of Jews from the government. The slogan "Jews to the gas'" was voiced on several occasions without any manifest counterreaction. While Walesa explicitly denied being an anti-Semite. he also stressed his "one-hundred percent Polishness" with "documents to prove it". In the face of open anti-Jewish sentiment, Walesa and his staff chose to hold a low profile. It is possible that they viewed anti-Semitism as a useful tool in their political struggle.

The national survey analyzed lacked direct measures of racist attitudes. To assess this dimension it is therefore necessary to employ a surrogate measure - i.e. attitudes towards selling stocks in prominent Polish companies foreign investors of various nationalities as proxies. There are, of course, several possible reasons for opposing the sale of stocks to foreigners. Opponents to large-scaie privatization would presumably also be opposed to excessive influence being afforded to foreign capital. Moreover, of privatization might perceive foreign capital as a threat to budding Polish capitalism. In addition, the sale of Polish companies to foreigners during an acute economic crisis might be viewed as both unprofitable and irresponsible.

In Table 5 respondents who oppose the sale of Polish companies to

foreign capital are distinguished from those who embrace xenophobic attitudes. More precisely, persons who oppose the sale of Polish companies to Americans are compared with those who oppose purchases of stocks by Jews. A negative attitude towards American capital in this context could be conceived as being grounded in the negative evaluation of the influence of foreign capital ingeneral. The differences in attitudes towards Americans and Jews that are found, in contrast, may be interpreted as a sign of anti- Semitic attitudes.

As we see from the table, anti-Semitic attitudes appear to be relatively common among the respondents. The proportion of respondents accepting American but rejecting Jewish capital is especially large among Walesa's and Tyminski's sympathizers.⁸

Another indication of anti-liberal values found in the data was people's attitudes towards capital punishment. The debate about the death penalty and its possible abolition played no role during the election campaign. Despite this, people's attitudes on this issue differed among supporters of the various candidates: 28 percent of Mazowiecki's, 33 percent of Walesa's, and 38 percent of Tyminski's supporters definitely supported the retention of the death penalty. Those absolutely opposed to the death penalty constituted 17 percent of Mazowiecki's, 12 percent of Walesa's, and 8 percent of Tyminski's supporters. The differences here were moderate, but would still, in combination with the analysis above, indicate that the most consistent support for liberal values is found among Mazowiecki's voters.⁹

The Determinative Factors of Voting Preferences

Up to this point, the analysis has shown the effect of various background characteristics and attitudes on choice of presidential candidate. It does seem reasonable to assume, though, that the values being examined are related to each other, and, further, that they are influenced by background

characteristics. It is difficult, therefore, to determine the relative importance the correlations presented so far. To gain some insight into this question the relations between background characteristics. values. and choice of presidential candidate have been investigated by means of multiple analysis. The assumption here is that the background characteristics may influence, both directly and indirectly, via attitudes, the dependent variable, that is the preference for different candidates. In the latter case, attitudes constitute an intermediate link between sociodemographic and the dependent variable. The purpose of the analysis is to find a pattern of life situations and fundamental values that could explain voting preferences.

In Tables 6 to 8, the findings of a series of regression analyses are presented.¹⁰ In the first model in Table 6 (Model A), the effect of the background factors on the propensity to vote for Mazowiecki is analyzed. Education and income are the variables most strongly related to support for Mazowiecki. Age and residence in the metropolitan areas also have some (positive) effect. Class differences seem to be less important for the outcome here as compared to what is indicated in Table 1. In Model A, there is a positive effect of belonging to the higher white-collar workers category. However, this effect is relatively weak which is due to the fact that educational level is also used as a predictor in the model." Moreover, a part of the class effect is mediated via income. The high educationaland can thus explain, at least to some degree, white-collar workers' strong preferences for Mazowiecki. Mazowiecki thus appears primarily to be the candidate of the well-educated people and of high-

income earners.

The correlations seen in Model B in the table largely follow the same pattern as those seen in Tables 3-5 above. Mazowiecki's supporters were liberal in both the political and economic sense of the word. They repudiate the death penalty and anti-Semitism more often than other voters do, and they are favorably disposed to a market economy. In addition, their attitudes towards Catholic conservatism are often negative. A positive attitude towards a market economy seems to be the most important explanatory here (see /3-values). A comparison of R2R² values in Models A and B reveals that the attitudinal factors have less explanatory power than the background characteristics.

Finally, Model C includes both background variables and attitudes. In general, it can be said that the effects of the background variables are somewhat weakened. (This can be seen when one compares /3-values in Models A and C.) This means that they influence voting preferences both directly and indirectly, via various attitudes. The significance of income, and especially of education, however, remains central. It should also be pointed out that (the negative) effect of belonging to the group of small

entrepreneurs is significant in this model. The parameter estimate is relatively though, which becomes apparent when the material is analyzed by means of logical regression. In any case, it must be concluded that small businessmen do not seem to be favorably disposed to Mazowiecki, his market-oriented policies.

When background characteristics are held constant, the effects of promarket are weakened, and the importance of attitudes towards transitional Catholic morals becomes negligible. This is largely due to the fact that educational level is strongly negatively related to Catholicism. In contrast, the significance of liberal values remains almost unchanged.

A partially inverse pattern emerges when Walesa's supporters are analyzed. analysis of the significance of background variables (Table 7; Model A) indicates that the level of education has a clearly negative effect on the propensity to sympathize with Walesa. The level of education is also the variable which explains the (relatively speaking) weak support from white-collar workers (cf. Table 1). The only significant class difference in the model indicates that Walesa enjoyed strong support among small businessmen. In total, class explains somewhat less of the support for Walesa than education does.¹² It is also shown that big-city residents, and particularly persons in the older age bracket, tended to express a preference for the leader of Solidarity.

In contrast to Mazowiecki's supporters, Walesa's exhibit anti-liberal values. Their attitudes towards Jews are negative (see Model B). Their propensity to vote for the Solidarity leader is furthermore strongly (positively) to a favorable attitude towards Catholic conservatism. In common with Mazowiecki supporters, Walesa supporters exhibit a positive attitude towards a market economy. This result diverges from what was shown in Table 4. When one considers the negative correlation between Catholic conservatism and pro-market attitudes, supporters of the Solidarity appear to be positive to the economic reforms. Catholicism is the variable which is far and away the most strongly related to support for Walesa (the /3-value). A comparison of R~ values in Models A and B indicates that the background properties and attitudes, each in their own right, are nearly equally strongly related to the propensity to vote for Walesa.

When both sets of variables are treated simultaneously (as in Model C) the direct effects of the background properties on the variable being studied do not change tangibly. Even the effects of the attitudinal variables remain largely the same. The only noteworthy changes here include the fact that the importance of a positive attitude towards a market economy is enhanced when one takes into account (the low) level of education, and that the importance of attitudes towards Catholicism is weakened when age and education are held constant.

Results of the regression analysis for Tyminski's supporters differ substantially both of the pictures presented above. Significant class effects (cf. Table 8, Model A) indicate a relatively strong propensity to vote for Tyminski among skilled manual workers, unskilled manual workers, and employees within the commercial sphere." Of the remaining effects studied, only age has a significant impact. Young voters display a tendency to favor Tyminski. The relationship among background properties otherwise significance here.

In contrast to Mazowiecki's and Walesa's adherents. Tyminski's supporters unfavorably inclined towards a market economy (Model B). Their attitudes towards liberal values are also negative. The factor that has the strongest correlation with a preference for Tyminski is a negative perception of Catholic conservatism. However, the variance explained (R:) in this model is less than in corresponding models in Tables 6 and 7.

From Model C, we can see that the direct effect of the background variables is somewhat weakened when they are analyzed in combination with values. Nonetheless, the main tendencies from Model A remain. The same applies to the influence of the attitudes on the variable studied. Of all the predictors studied, the effect of age emerges as the most powerful.

Overall, the studied correlations were moderate. If one compares R2R² values in the analyzed models with an R2R² value in a regression model depicting the choice of political bloc in Sweden, it is clear that standard predictors of political sociology explain voting behavior in Sweden better than in Poland. The ability of these predictors to explain preference for political bloc in Sweden, though, has decreased in recent decades.'⁴ Traditional political patterns are perhaps growing weaker in the old European At the same time stable patterns have not yet evolved in the new democracies.

Discussion

Despite the fact that the correlations shown here are not particularly strong. in light of the tendencies which emerge in the analyses reported above it is possible to draw some conclusions about the influence of the living conditions and values on people's choice of President in Poland in 1990. Mazowiecki's voters were characterized by high education and high social position. Mazowiecki represented the so-called intelligentsia. People with higher education viewed democracy and a market economy as pre-conditions individual development and success. Mazowiecki's policies were perceived as a serious and consistent attempt to break with the Communist legacy.

Furthermore, the economic crisis did not harm the privileged groups in

society to the same extent as it harmed other groups. High-income earners. for instance, did not experience the same insecurity as did low-income earners. The support for Mazowiecki was thus a natural consequence of one's economic well-being. Yet Mazowiecki seems to have failed in his attempts to spån conflicts between different interests in society. The social groups who were not immediately disadvantaged by the economic transformation his primary supporters. One can add that the impact of socio-economic factors on the propensity to vote for Mazowiecki was higher than for other candidates, which means that the position of his supporters within the socio-economic structure was relatively well dermed (see also Bialecki 1991).

Walesa's voters, on the other hand, were more likely to experience insecurity in the face of a systemic transformation. This insecurity bound less-educated people and older people from different social strata together in a search for a strong and effective leader. Older people in particular may have felt threatened by social upheaval, which has a destabilizing effect on their status in the labor market. In addition, the incomes of retired persons become less certain in an inflational economy. Walesa's promise of maximum change at minimal cost, despite its apparently contradictory character, obviously fell on fertile ground here.

Walesa's popularity among small businessmen can be explained as much by his generous promises as by the failure of the incumbent government's policies. Decreasing real wages reduced the demand for the goods and services produced by small companies. The high rate of inflation forced up interest rates, resulting in expensive bank credit. This reduced the opportunities for investment and development for small businesses. It can also be noted that social class explained a surprisingly small proportion of sympathy for Walesa. When the impact of other factors is considered, the standing of the Solidarity leader was only insignificantly stronger among manual workers than among farmers and white-collar workers.

Solidarity's greatest dilemma after taking power was how to represent the interests of the employees of large. State-owned, unprofitable companies at the same time support market reforms (Ziolkowski 1990). This dilemma has remained unresolved. No one has as yet succeeded in combining rapid economic reforms with a guaranteed social security. It is here that we probably find the explanation for Tyminski's relative success among manual workers. Where central problems remained unresolved, important needs were left unfulfilled, and disappointment with Solidarity was widespread, support for Tyminski was strong. Other research has also revealed that support for Tyminski was particularly strong in regions suffering from high unemployment, considerable structural problems, and high criminality. Tyminski seems to have attracted votes from the discontented from those who found the current politics and its symbols alien (Zukowski 1991).

According to a Polish sociologist, the reform program introduced by the first post-Communist government could be characterized as "'exhortations to build the social order from the grass roots ('everything is in your hands') without providing sufficient opportunities for the implementation of that requirement" (Kolarska-Bobinska 1990, 281). This gap between the ends and means is crucial for understanding how divergent social circumstances influenced voting behavior during the presidential election. A majority of Poles agreed more or less unambiguously on the goals. Those who also seemed to have the means (social position, education, and money) supported Those who lacked the means and/or suspected that they would not live long enough to see those goals attained (the old) voted for Walesa. Those who questioned the goals, for instance manual workers who disliked the idea of the market mechanisms determining whether their companies were profitable or not, and those who were less precise about the goals but who wanted to have fun along the way (the young) chose Tyminski.

The differences in attitudinal patterns that emerged in this study can be given both a historical and a current explanation. The difference between Mazowiecki's and Walesa's voters had its roots in Solidarity's history. There had always been people in Solidarity who primarily viewed the movement as a movement for democracy and civil rights. They advocated universalism and humanitarian values. Other members of Solidarity primarily saw it as the Polish national resistance movement against the Soviet, atheistic communism. They advocated a national particularism and a popular Catholicism. former image of Solidarity was better known outwards, due to its advocates' more pronounced literary skills. This tradition is primarily represented among Mazowiecki's voters. Their attitudes towards liberal values shown above could serve as evidence of this. Supporters of the second tradition are mainly found among Walesa's supporters: anti-Semitism and conservative Catholicism were overrepresented among them.

Finally, Tyminski's voters displayed a mistrust towards liberal economic reforms and against the growing influence of the Church. None of the more established politicians seem to have been able or willing to deal with the apprehensions that gave rise to this mistrust. That Tyminski's voters showed signs of anti-liberal values, moreover, is understandable in light of the atmosphere prevailing at his election rallies and the types of argument he used. The more committed supporters of democracy could not easily have voted for him.

Divergent social interests have in Poland not yet been "translated" into a stable party system. Even so, the fundamental hypothesis of political sociology - i.e. that social circumstances influence voting behavior - has a certain validity in post-communist Poland. In addition, the basic attitudes examined here - attitudes towards the liberal market ideology, conservative

Catholic morality, and anti-liberal political values - have significant effects
on people's choice of presidential candidate.

---

NOTES

1. The following account is based on Ajnenkiel et al. 1991: Bazylko & Wysocki 1991; Krzeminski 1991 and Winiarska-Maziuk 1991.

2. That relatively high proportions of agricultural workers and farmers supported other candidates is due to the faet that the agrarian candidate Bartoszcze is included in the category "Others".

3. The interviewees were asked to group themselves (or the family's head of household) in one of nine income categories. The relevant tahle is not given.

4. Considerinti that previous research has shown that cuahtanan .niiUide^ Lind support for the welfare system are relatively well-anchored among Polish people (Nowak 1988: Ziolkowski 1990: Kolarska-Bobinska 1991). this omission is quite startling. is about the same. Among Tyminski's supporters. though. the proportion who were clearly negative was significantly greater.

6. The above is based on interviews with Adam Michnik and articles by Barbara Stanosz and Dawid Warszawski in Politvka and Cazcta Wyborcza.

7. Anti-Semitism has been used for decades as an important instrument by the Communists their struggle against democracy. The student revolt of 1968 was succeeded by a centrally organized. conspicuous anti-Semiticeampaign. In the 198()s. anti-Semitic arguments were used against Solidarity.

8. Other research tindings support the results here: 25 percent of Mazowiecki's compared with 50 percent of Walesa's supporters thought that "Jews exert too much influence in Poland" (Gebert 1991). It may seem strange that anti-Semitism continues to play a political role in a country where there are only a few Jews left. The mechanism that makes this possible has been described by a Polish historian as a process of generalization anti-Jewish attitudes. Previously. Jews were perceived (in some circles) as enemies; today, enemies are claimed to be Jews (Krystyna Kersten in the weekly newspaper Polityka: 1991).

9. Using attitudes towards the death penalty as an indicator of liberal values is relatively common practice (see, for example. Heath et al. 1985). These authors show (p. 117) that attitudes to the death penalty constitute a line ol demarcation within British politics. 40 percent of Labour's voters and 27 percent of the conservatives clearly disapprove of the proposal to reintroduce the death penalty.

10. In each of the analyses, those who support one of the three candidates are given the value 1. whereas ali others who expressed some preference are given the value 0. The procedure means that Mazowiecki's supporters are given the value 1 in the tirst analysis. Walesa's in the second. Tyminski's in the third. In addition to social class. education. and income. we include age and town of residence as predictors of voting preferences. Class position is coded as a set of dummy variables in which self-employed farmers constitute the reference category. The town of residence is also measured as a dummy variable. Big cities with more than 500.000 inhabitants are given the value 1. Other towns are given the value 0. In the preliminary analyses, sex and marital status were also used as predictors. When they were found to be unrelated to voting preferences, they were excluded from all analyses here. The variables for market and Catholic attitudes are additive indices basedon the questions that are used in Tables 3 and 4. The variables that are included in these indices fell into two separate dimensions in a factor analysis. The indices have been standardized in such a way that lhev citn éissumc valucs i rom oto 1. The vaiue 1 on thc iru lt' \ ! n iti 1i ic iisuro^ (. cithohcism means that the respondent regularly attends mass. is in favor of mandatory Christian education in schools. and advocates an absolute ban on abortions. The value 1 on the variable market implies attitudes strongly in favor of market principles.

---

11. This cannot be seen from the table. To be able to make this statement, regression analyses were conducted where the significance of the class variable and education were studied separately. The results indicate that education explains more of the support for Mazowiecki than does class (R² value). For reasons of space, the results are not shown here. Moreover, the logistic regression shows that the parameter estimate for higher white-collar workers is relatively unstable. In the logistic regression, this effect was insignificant. It is important to mention that all results presented in Tables 6-8 have been tested by means of logistic regression. In practically all cases. the results of both approaches coincide. In those cases where some deviations were noted, this is discussed in the text.

12. It was also necessary here to enter the variables into the regression equation step-bystep. note 11.

13. It should be observed that the fact that sympathies for Tyminski are relatively high among manual workers does not imply that Tyminski was the primary candidate of manual workers (compare Table 1). What it instead implies is that manual workers distinguish themselves from the reference group (farmers) with regard to their sympathies Tyminski.

14. The comparison with Sweden is interesting here owing to the fact that class voting in analysis of the 1985 election, class, employment sector, area of residence, and religion were used as predictors. The R2R² value was 0.18 (Holmberg et al. 1987, 22\-221).

ACKNOWLEDGEMENTS

I thank the Center for Opinion Studies (CBOS) for generously granting me access to the
election study which serves as the basis for this article. Many thanks to Karen Leander who
tried to correct my English at the last minute.

REFERENCES

Adamski, W. 1991. "W obliczu prywatyzacji: poparcie i sprzeciw" (Facing privatization: For
and Against), in Adamski, W.,'ed., Polacy' 90 (Poles 90). Warszawa: Instytut Filozofii i
Socjologii PAN/Instytut Studiow Polityczynch PAN.

Ajnenkiel, A. et al. 1991. Prezydenci Polski (Polish Presidents). Warszawa: Wydawnictwo
Sejmowe.

Ash, T. G. 1991. We the People. The Revolution of '89 Witnessed in Warsaw, Budapest, Berlin
& Prague. Cambridge: Granta Books.

Bazylko, P. & Wysocki, P. 1991. "Anatomia kleski" (The Anatomy of Defeat), in Grabowska, M. & Krzeminski. 1., eds., Bitwa o Belweder (The Battle for Belweder). Warszawa: Mysl. Bialecki, I. 1991. "Pracownicze interesy i orientacje polityczne" (Employee Interests and Political Orientations), in Adamski. W. et al., eds., Polacy' 90 (Poles 90). Warszawa: Instytut Filozofii i Socjologii PAN/Instytut Studiow Polityczynch PAN.

Gebert. K. 1991. "Rola antysemityzmu" (The Role of Anti-Semitism), in Grabowska. M. &
Krzeminski. i., eds., Bitwa o Belweder (The Battle for Belweder). Warszawa: Mysl.
Heath, A. et al. 1985. How Britain Votes. Oxford: Pergamon Press.

Heyns, B. & Bialecki, I. 1991. "Solidarnosc: Reluctant Vanguard or Makeshift Coalition?",
American Political Science Review 85. 351-370.

Holmberg, S. et al. 1987. Vdljare och val i Sverige (Voters and Elections in Sweden).
Stockholm: Bonniers.

Inglehart, R. 1990. Culture Shift in Advanced Industrial Society. New Jersey: Princeton
University Press.

Kolarska-Bobinska, L. 1990. "Civil Society and Social Anomy in Poland". Acta Sociologica
33, 277-288.

Kolarska-Bobinska, L. 1991. "Poczucie krzywdy spolecznej a oczekiwanie opiekunczej roli
panstwa" (Sense of Social Injustice and Expectations of State Protection), in Adamski, W.

et al., eds.. Polacy' 90 (Poles 90). Warszawa: Instytut Filozofii i Socjologii PAN Insiytut
Studiow Politycznych PAN.

Korpi, W. 1983. The Democratic Class Struggle. London: Routledge and Keagan Paul. Krzeminski, I. 1991. "Wybrac lepszy swiat" (Choosing a Better World) in Grabowska. M. & Krzeminski, 1., eds., Bitwa o Belweder (The Battle for Belweder). Warszawa: Mysl. Lipset, S. M. & Rokkan, S. 1967, Party Systems and Voter Alignments: Cross-National Perspectives. New York: The Free Press.

Lipton, D. & Sachs, J. (1990) "Creating a Market Economy in Eastern Europe: The Case of
Poland", Brooking Papers 1:1990, 75-147.

Marshall. G. et al. 1988. Social Class in Modern Britain. London: Hutchinson.
Michnik, A. 1990. "Dlaczego nie oddam glosu na Lecha Walese" (Why I shall not vote for
Lech Walesa), Cazeia Wyborcza 251, 6-7.

Nowak, S. 1988. "Spoleczenstwo polskie drugiej polowy lat osiemdziesiatych" (Poles in the
Latter Half of the 1980s), Studio Sociologiczne 108, 23-55.

Winiarska-Maziuk, A. 1991. "Kampania wyborcza w telewizji" (The Election Campaign on
TV), in Grabowska, M. & Krzeminski, 1., eds., Bitwa o Belweder (The Battle for Belweder).
Warszawa: Mysl.

Ziolkowski, M. 1990. "Social Structure, Interests and Consciousness. The Crisis and Transformation
the System of 'Real Socialism' in Poland". Acta Sociologica 33. 289-304.
Zukowski, T. 1991. "Mapa wyborcza Polski" (Poland's Election Map), in Grabowska, M. &

Krzeminski, 1., eds.. Bitwa o Belweder (The Battle for Belweder). Warszawa: Mysl.