SEARCH

SEARCH BY CITATION

Keywords:

  • Political economy of transition;
  • support for reforms

Abstract

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Trends and evaluations of transitional reforms
  5. 3 The data, measurement issues and stylized facts
  6. 4 The empirical model
  7. 5 Who is against reforms?
  8. 6 Has support for transition increased, and if so, why?
  9. 7 Conclusions
  10. References
  11. Appendix

We study the dynamics of individual support for changes in the economic and political system, using a unique dataset for 12 transition economies over the period 1991–2004. We document that support for transition was initially lower in the CIS countries and that there has been a converging trend in the support for reforms between the CIS and the Baltic and Central and Eastern European countries. We suggest several explanations for the initial divergence and the post-98 convergence in support for transition between these three groups of countries, and show that economic growth, declining income inequality and improving quality of governance have contributed to increase the support for transition. In addition, we find that increased support for the market economy and democracy in the CIS is accompanied by a larger increase in trust towards the political institutions. Our results also confirm the implications of Aghion et al. (2010)'s model of a negative correlation between trust and the demand for government regulation.

1 Introduction

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Trends and evaluations of transitional reforms
  5. 3 The data, measurement issues and stylized facts
  6. 4 The empirical model
  7. 5 Who is against reforms?
  8. 6 Has support for transition increased, and if so, why?
  9. 7 Conclusions
  10. References
  11. Appendix

In the last two decades former socialist countries underwent the unprecedented experience of a parallel transition to a market economy and democracy. Although the paths of reform implementation and the sequence of the reforms differed across countries, transitional reforms soon produced both economic ‘winners’ and ‘losers’ (Brainerd, 1998; Terrell, 1999), and for those who were less ready or less able to face these changes, the costs of transition may well have outweighed, at least for some time, its benefits. At the same time, and in stark contrast with the strong economic performance until the recent economic crisis, dissatisfaction with the outcomes of transition remained widespread. In 2006, about one half of the Russian population was disappointed with transition and a large majority was in favour of high state intervention (Denisova et al., 2010). Russians were also found to have the most negative attitudes towards free market and democracy among 28 transition countries (Denisova et al., 2010). In 2007, 49 percent of respondents in 28 post-communist countries disagreed (and only 35 percent agreed) with the statement that the economic situation in their country was better than around 1989, with similar numbers corresponding to the political situation (EBRD, 2007a; Guriev and Zhuravskaya, 2009). Also, privatization, one of the most important transition reforms, received low support, with over 80 percent of respondents willing to revise it (Denisova et al., 2009; EBRD, 2007a).

To shed light upon the extent and evolution of public support for economic and political reforms, in this paper we employ a unique dataset, so far largely unexploited by economists, covering 12 transition economies for the entire period from the beginning of transition up to the first Eastern EU enlargement. We construct new measures of popular support for these systemic changes and analyze their dynamics and determinants, documenting how support for the economic and the political system has been evolving over this period. We analyze how individual determinants have influenced support and how their impact has changed throughout the period by differentiating between the earlier period of recession (1991–1998) and the later period of economic growth (2000–2004). In contrast to the majority of previous studies, we focus on the dynamics of public support for transition and explore the macro and institutional factors which may have influenced it. Our study is, to our knowledge, the first one to analyze the evolution over such a long time span and in a comparative framework of public opinion about economic and political reforms, and to examine the factors that may explain these changes.

As our aim is to deepen understanding of public support and satisfaction (or lack of it) with the economic and political process of transition and its evolution over time, as a first step we relate individual opinions about the transition to those characteristics that are expected to differentiate between potential winners and losers. This allows us to confirm results found in earlier studies and to confirm the validity of the dataset at our disposal. Second, we focus on the different trends in support observed in three different country groups: Central and Eastern Europe (CEE), the Baltic countries and the Commonwealth of Independent States (CIS), and on how they evolve over time. This allows us to explore at greater depth some specific hypotheses on the reasons for the lower support for transition in the CIS countries and for its increase in the more recent period. In this respect, Denisova et al. (2010) have well documented the low support for transition in the Russian population, which they relate to the poverty of social capital and the quality of the governance institutions. Their result can be considerably reinforced in the context of a multi-country study, which allows a proper comparison of countries with different levels of social capital and different development and quality of the governance institutions.

Our main findings are as follows: (i) The growth performance of the economy contributes to explaining the improved evaluation of the economic and political systems in the CIS and Baltic countries after 1998; (ii) Improving trends in income inequality contribute to explaining the increased support for the market economy, particularly in the CIS; (iii) Government effectiveness, rule of law and control of corruption contribute to explaining the increasing support for economic and political reforms in the Baltics, while regulatory quality matters most for changes in the CIS.

A theoretical framework that may be used to interpret our findings is that proposed by Aghion et al. (2010). They also provide a theoretical motivation for the work of Denisova et al. (2010). This model provides one way to rationalize the (apparently paradoxical) correlation between the lack of social capital and of trust towards the government with popular demand for more government intervention in the economy, even if this is expected to generate more corruption. To show this correlation, the model generates a good equilibrium with high trust and low regulation and a bad equilibrium with low trust and high regulation. In the latter equilibrium, people demand more government regulation, even if they know that the government is corrupt, since they would be even worse off in the absence of regulation, given the prevalence of corrupt behaviour adopted by ‘uncivic’ entrepreneurs.

Consistent with this model, in our data there is a strong positive correlation between different measures of trust towards the political institutions and the variables measuring the support for economic and political reforms (in other words, a negative correlation with the demand for regulation, as measured by distaste for the market economy and the demand for more state involvement). In addition, we show that an increase in support for market economy and democracy in the CIS countries is accompanied by a relatively larger increase of trust in the political institutions of these countries.

Our study is also related to several strands of the literature on public attitudes towards transition, in particular those focusing on the determinants of public support for reforms and for the market economy (Fidrmuc, 2000; Hayo, 2004; Kim and Pirttilä, 2006; Landier et al., 2008), on the ‘unhappiness in transition’ (Easterlin, 2009; Guriev and Zhuravskaya, 2009), on the determinants of popular dissatisfaction with privatization policies (Denisova et al., 2009), and also on the effects of communism on individual preferences towards state redistribution and state intervention (Alesina and Fuchs-Schuendeln, 2007).1 Our paper also relates to the literature on support for democracy and the market economy and on the sequencing of political and economic reforms. A large political economy literature has shown that voters’ opinions are crucial for the successful implementation of reforms, and that interest group coalitions may influence or even reverse the reform process (see Roland, 2002 for a comprehensive discussion). In particular, a recent study by Grosjean and Senik (2011) has shown, by exploiting within-country regional variations, the existence of a causal relationship, running from actual democratization to the popular support for a market economy. Our results are also in line with those findings in the literature that document a changing correlation between inequality and satisfaction with transitional reforms: in particular with Grosfeld and Senik (2010) who document that, in the case of Poland, in a first period increasing inequality was positively associated with satisfaction with the country's economic situation, while the correlation became negative after 1996 and was accompanied by increasing public sentiment that the process of income distribution was flawed and corrupt.

The remainder of the paper is structured as follows. Section 2 provides a brief overview of the transition-specific background. Section 3 presents the data, discusses measurement issues and outlines the basic empirical facts. The econometric model is presented in Section 4. Socio-economic determinants of individual attitudes towards economic and political reforms and their change over time are examined in Section 5. Section 6 examines the dynamics of the support for transition for the CIS, the Baltics and the CEE countries and suggests some explanations for the larger increase of support observed in the CIS and Baltic countries. Section 7 presents our conclusions.

2 Trends and evaluations of transitional reforms

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Trends and evaluations of transitional reforms
  5. 3 The data, measurement issues and stylized facts
  6. 4 The empirical model
  7. 5 Who is against reforms?
  8. 6 Has support for transition increased, and if so, why?
  9. 7 Conclusions
  10. References
  11. Appendix

The implementation of political and economic reforms began in the early 1990s in most countries in the CEE, CIS and the Baltics. But the process was not uniform. Differences across countries in the paths and sequences of the reforms were sometimes interpreted as examples of a distinction between a so-called ‘big-bang’ approach and ‘gradualism’.2 The transition process has been characterized almost everywhere by an initial deep recession, which in many countries also involved widespread unemployment. However, the pattern, depth and duration of this transitional recession and the speed of the subsequent recovery differed considerably across countries, with CEE countries, on average, recovering faster. A common feature of all the transition economies was the need to refocus the orientation of international trade, to restructure internal production, and to reallocate labour across regions, sectors and firms (Campos and Coricelli, 2002). Privatization, trade liberalization, macroeconomic stabilization and economic restructuring took place in a situation of institutional change, where many institutions that had hitherto provided social protection collapsed and others, such as taxation or banking, had to be introduced practically ex novo. The initial stages of transition brought about remarkable increases in income inequality in all countries, including those that had managed to avoid large increases in unemployment rates (Milanovic and Ersado, 2008).

One of the most important criteria for assessing the success of transition is a country's achievement in reallocating labour (Boeri and Terrell, 2002). As transition generated unprecedented economic insecurity, job insecurity became a crucial issue for many (Linz and Semykina, 2008). Low-educated, young, single individuals and women, especially married women, were more likely to become unemployed (Boeri and Terrell, 2002). Thus, transitional reforms soon produced both economic ‘winners’ and ‘losers’ (Brainerd, 1998; Terrell, 1999).

The adjustment patterns of the output and labour markets differed substantially between the CEE and CIS countries. With a few exceptions, all Central and Eastern European countries experienced a U-shaped pattern of GDP with a large fall in employment early in the 1990s and some decline in labour productivity leading to rapid structural change but also to high unemployment (with the exception of the Czech Republic), much of which was long term. In contrast, the CIS countries typically faced an L-shaped pattern of GDP during the 1990s and a relatively modest decline in employment with limited sectoral reallocations of labour. Here, however, there was a more pronounced deterioration in labour productivity and of real wages, as well as a significantly larger increase in inequality than in the CEE countries (Boeri and Terrell, 2002; Svejnar, 2002). Overall, while the labour market adjustment process took the form of larger declines in employment in the CEE countries, it typically occurred through real wage declines in the CIS. And only as transition progressed, unemployment began to increase gradually also in the CIS countries (Svejnar, 2002).

A large literature on the optimal speed of transition has studied the speed at which an economy restructures and destroys the old jobs in the state sector (see, for example, Aghion and Blanchard, 1994; and Boeri, 2000 for a review). However, by focusing on speed and thus distinguishing essentially between a ‘big bang’ vs. a more ‘gradualist’ approach, this literature fails to explain some key differences in the adjustment processes in the CEE and CIS countries (Boeri and Terrell, 2002). Alternative explanations relate the differences in performance to differences in institutions. In particular, social safety nets and non-employment benefits may have prevented the decline of wages in Central and Eastern Europe by setting floors to them (Boeri and Terrell, 2002). In addition, weaker legal systems and poor enforcement of laws and regulations in the CIS have probably encouraged both a profound lack of transparency in government and corporate recklessness, which in turn facilitated the spreading of corruption and rent-seeking behaviour (Roland, 2002; Svejnar, 2002). In general, the literature stresses the advantages of adopting a political economy perspective and of taking into account also the role of non-economic institutions, in order to explain the adoption and consequences of different policy models in each country (Roland, 2002). In this paper we follow this approach in order to study the determinants and evolution of public support for transitional reforms.

3 The data, measurement issues and stylized facts

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Trends and evaluations of transitional reforms
  5. 3 The data, measurement issues and stylized facts
  6. 4 The empirical model
  7. 5 Who is against reforms?
  8. 6 Has support for transition increased, and if so, why?
  9. 7 Conclusions
  10. References
  11. Appendix

The data used in this paper come from the New Barometer Surveys (New Democracy Barometers). These are representative surveys of the populations in transition countries consistently collected over time by the Centre for the Study of Public Policy (CSPP) at the University of Aberdeen and the Paul Lazarsfeld Society, Vienna.3

As each survey round contains a large number of common questions, which are maintained across time and countries, the set of available surveys constitutes a unique dataset that allows meaningful cross-country comparisons across several years. This allows us to identify trends in political and economic transformations and also, given the composition of the questionnaires, to analyze the determinants of individual attitudes in the face of such changes. Surveys are undertaken independently from governments and face-to-face interviews are performed by trained interviewers working for established national research institutes in the national language (with the exception of the Baltic countries, Belarus and Ukraine, in which cases Russian was also used). The survey includes nationwide multistage random samples of about 1,000 respondents (in Russia about 2,000) over 18 years old per country.

We have merged several waves of the New Europe Barometer, the New Russia Barometer and the New Baltic Barometer data. The result is a pooled cross-section dataset for 14 transition economies, with surveys taking place in several waves between 1991 and 2004. Ten countries in our sample became members of the EU with the 2004 or 2007 enlargements (Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia and Slovenia). Croatia is currently a candidate for EU membership, while three countries are members of the CIS (Belarus, Russia and Ukraine).

The set of explanatory variables employed in the regressions below includes standard socio-economic indicators, such as gender, age, education, marital status, urban residence, employment status and household income. We also use macrodata on economic variables, political institutions and governance quality, and individual data on the assessment of trust towards the political institutions. In the final sample we include all individuals with non-missing information on the key explanatory variables. We have also excluded Croatia from the final dataset due to the few observations available, concentrated in the years until 1998, and Belarus due the peculiar history of its non-reforms. Table A1 in the Appendix presents sample size by country. Definitions of the variables are given in Table A2.

The New Barometer Surveys include several questions on the degree of individual support (or opposition) towards the process of transition. For the purposes of this paper, we focus on the following sets of questions, which were included in all surveys:

Economic evaluation:

Q.1 ‘Here is a scale for ranking how the economy works (from +100 at top to100).

(a) Where on this scale would you put the socialist economy before the revolution of 1989/perestroyka?

(b) Where on this scale would you put our current economic system?’

Political evaluation:

Q.2 ‘Here is a scale for ranking how our system of government works (from +100 at top to100).

(a) Where on this scale would you put the former communist regime/political system before perestroyka?

(b) Where on this scale would you put our current system (with free elections and many parties)?’4

As a first step, we examine the patterns of responses to these questions across time and countries. Figure 1 shows the proportion of positive, negative and zero evaluations of past and present economic (left panel) and political (right panel) systems for 1993 and 2004. Regarding the current economic system, a majority of respondents valued negatively the current system in 1993, although they gave positive evaluations in 2004. And regarding the past economic system, a majority of respondents gave positive scores both in 1993 and 2004. The picture is somewhat different for the political system, as a majority of individuals evaluate positively both the past and the present system in both years, and the proportion of positive answers increases in 2004. Note also that neutral (zero) evaluations constitute only a small proportion in the overall poll.

image

Figure 1. Evaluations of the economic and political systems in 1993 and 2004

Source: Authors’ tabulations from the New Barometers data. Sample includes all individuals.

Download figure to PowerPoint

In principle, there are several alternative ways in which the evaluations presented above can be used to formulate an appropriate dependent variable for our analysis. For instance, should we focus only on individual judgements about the present system? Or instead on a comparison between evaluations for the present and the past? As we are interested in modelling the support for transition and reforms, a relative measure seems intuitively more appropriate, as it would reflect a comparison between the current and the past systems. Moreover, the answers to these questions may be related, inter alia, to whether the revision of opinions about the previous regime reflects a (selective) forgetfulness of the past or a delusion about the present or, indeed, a mixture of both. Our a priori position is that judgements about the past are meaningful, and that evaluating the past more favourably is part of the same process that results from a delusion about today's experience. Accordingly, a judgement about the past is not only a historical assessment, but it also conveys information about the evaluation of the present system. In other words, statements about the past and the present are not independent of each other, but rather reinforce each other. To take this into account we compute our dependent variable by taking the difference (or ‘distance’) between the responses to Question (b) (present) and to Question (a) (past) for the economic and political systems, respectively. Thus, a larger positive (negative) difference implies a larger positive (negative) assessment of the present regime relative to the former one (in the economic and governmental dimensions, respectively). The larger this distance is, the more an individual is positive about the current state of the economy or polity, relative to the past, and thus, we assume, the more supportive he or she is of the reforms that have been adopted.

In this context, it is important to note that differences in responses across countries may also arise due to different interpretations of the reference scale (−100; +100) in different countries and by different individuals, as they may be related to country-specific factors, such as culture. To this aim, we also standardize our dependent variable dividing it by its country (and year) specific standard deviation and control for country-specific effects in the regressions below. In this way we weight individual responses by a country and year specific variance, thus giving more weight to countries with relatively homogenous responses. In what follows, we shall refer to this variable as the standardized (economic or political) distance. A related problem that arises when using subjective data is that individual responses may be affected by several factors, such as the ordering of the questions in a survey, the exact wording of the questions or individual differences in the perceptions of the scale, which may introduce a measurement error (Bertrand and Mullainathan, 2001). Note that the questions on the economic and political systems in our survey are usually asked at the beginning of the corresponding sections on the economy and public affairs, before the questions on the personal (or family) economic situation or on political preferences. Note also that taking differences across individual answers for the same person may difference away individual-specific and evaluation-invariant factors such as pessimism or different individual perceptions of the scale, thus potentially reducing the associated biases.5

As we mentioned above, a potential criticism against using our measure of distance is that it does not take into account the ‘absolute’ evaluation of the current or of the past system given by the respondents. For instance, the same distance of 70 could characterize someone who likes both the past and the present (past = 30; present = 100), someone who dislikes them both (past = −100; present = −30) and someone who dislikes the past but is reasonably satisfied with the present (past = −40; present = 30). These absolute evaluations might contain additional information. To take it into account we also perform an additional exercise. We divide our sample into eight different sub-groups (that is, four each for the economic and the political evaluations) following a classification introduced by Rose and Mishler (1994) and used in their studies. Individuals who give positive evaluations to both the present and past economic (or political) systems are labelled as ‘positive’ (‘compliant’). Those who are neutral or negative about both the present and past economic (political) systems are ‘negative’ (‘skeptic’). Those who evaluate positively the present economic (political) system and negatively or neutrally the past are ‘pro-market’ (‘democrat’). And those who are negative or neutral about the present and positive about the past economic (political) system are called ‘nostalgic’ (‘reactionary’). Based on this classification, we estimate multinomial logit regressions for the probability of being in one of these groups.

Before proceeding to a formal analysis, it is useful to examine the data in more detail. In Figure 2, we show the evolution of the standardized distance for the economic (top panel) and political (bottom panel) systems for three different groups of countries, which are all present in the three years 1993, 1995 and 2004.6 In each case, the three groups are: the CIS (Russia, Ukraine), Baltic states (Estonia, Latvia, Lithuania), and CEE (Bulgaria, Czech Republic, Hungary, Poland, Romania, Slovakia and Slovenia). Both panels show the same, broad characteristics, which we summarize as follows:

  • A wide gap in 1993 in the extent of support for both economic and political reforms between the CEE and CIS countries, with the Baltics roughly in the middle;
  • A marked deterioration in both measures for the CIS in 1995, thus leading to an enlarged gap with the CEE;
  • An overall stationary or slightly declining support over time in the CEE countries, together with a strongly increasing support after 1995 in both the Baltics and the CIS. Interestingly, the decline in the CEE is more profound for the support for political changes …
  • … thus leading to a substantial convergence among the three groups towards the end of the period of observation.
image

Figure 2. Dynamics of support, 1993–2004. (a) standardized distance (economic support), (b) standardized distance (political support).

Download figure to PowerPoint

image

Figure Figure A1. Dynamics of support for individual countries

Download figure to PowerPoint

It is also interesting to compare the measures on the vertical axes of the two panels, which show that support for change in the political system is much stronger than for change in the economic system. This is perhaps not surprising, and is consistent with a political economy approach which suggests that more popular reforms should be implemented first, and with the observation that democratic reforms preceded economic reforms in CEE, since support for democracy was greater than for economic reforms (Roland, 2002). It is also consistent with the findings of Grosjean and Senik (2011) that democratization precedes, and hence causes, popular support for the market economy.

Summing up, on average citizens of many transition countries did not seem to give a favourable evaluation of the economic system they lived in, and they seemed to have regrets for the past. On the other hand, on average, they appeared reasonably satisfied with their current political system, but in some instances they still did not see it as an improvement over the past.7 This is true, in particular, for the CIS; however several other countries, such as Lithuania, Latvia, Hungary and Slovakia, also express negative evaluations, at least for certain years. On balance, however, we observe an increasing trend in the relative evaluations of the economic and political systems in many countries, which is particularly noticeable for the CIS. Nevertheless, the fact that support for transition is so low may appear puzzling, at least prima facie, if we compare these responses with the evolution of most standard macroeconomic and institutional indicators, especially in the new EU Member States. These aggregate differences, however, may be confounded by differences in individual characteristics and transition experiences. Moreover, country-specific macroeconomic policies and institutions may also have affected individual support attitudes and their dynamics. In the sections below we examine the role of these factors.

4 The empirical model

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Trends and evaluations of transitional reforms
  5. 3 The data, measurement issues and stylized facts
  6. 4 The empirical model
  7. 5 Who is against reforms?
  8. 6 Has support for transition increased, and if so, why?
  9. 7 Conclusions
  10. References
  11. Appendix

We model individual support for the economic and political transition assuming that it may be influenced by standard individual explanatory variables. These socio-economic characteristics may help to sort out actual or potential ‘winners’, who are likely to support the transition process, from the ‘losers’ who are unlikely to support it. These characteristics include gender, age, education, marital status, urban residence, labour market status and household income. We also include a ‘young cohort’ dummy, for those who were 18 years old or younger in 1990, in order to control for potentially different attitudes and experiences of those who were brought up during socialism and thus presumably had not experienced the labour market under socialism versus those who have acquired their skills and experience during transition and in a market economy. As the analysis of how socio-economic characteristics affect the individual assessments of the transition is well established in the literature, we do not expect this first exercise to provide original results; it is, however, useful in order to provide a confirmation of those earlier results in a broader sample of countries and periods, and also to validate the reliability of our dataset.

We begin with the following simple specification of a baseline model:

  • display math(1)

where Yijt is our measure of support for transition (standardized distance) for individual i in country j in year t, Xijt is a vector of standard individual socio-economic and demographic characteristics, μj are time-invariant country fixed effects and ϕt are year fixed effects and ɛijt is a random error term, which ideally should not be correlated with the rest of the variables.8

A complementary method of analyzing the characteristics of ‘winners’ versus ‘losers’, which has not been used previously in this literature, involves the estimation of a multinomial logit (MNL) model for the probability to be in one of the sub-groups defined in the previous section: positive, negative, pro-market or nostalgic in relation to the economic system, and compliant, skeptic, democrat or reactionary for the political system. We assume that, within each round of interviews, each individual identifies him/herself with one of those four economic and also political positions, and this identification may be modelled in relation to his/her characteristics:

  • display math(2)

where Uij defines the propensity of person i to declare him/herself as belonging to sub-group inline image and vijt is a random error term. The probability that individual i belongs to sub-group j is described by:

  • display math(3)

We estimate marginal effects for this model. In addition, to grasp the evolving dynamics of support for change in the economic and political systems, we split our analysis into two periods: ‘recession’ (1991–1998) and ‘growth’ (2000–2004).

In the last part of the empirical analysis we resort to our original dependent variables, the standardized distance between present and past support for the economic and political system, and study the support for transition before and after 1998 (the year of the Russian financial crisis) after grouping all the countries into three groups: CEE, Baltics and CIS. We thus introduce a time dummy (Post98) and two sub-group dummies (Balt and CIS) and also their interaction terms. These interactions capture the change in support attitudes in the CIS and the Baltics relative to the rest of the CEE. We also include year dummies to control for macroeconomic trends in these countries. The resulting specification is a sort of ‘difference-in-difference’ model, which has been used in many previous studies to estimate the effect of random treatments:

  • display math(4)

where ωijt is a random error term. In this exercise we shall be mainly interested in the coefficients of the interaction terms, and we shall test whether the introduction of several additional variables measuring the economic and/or institutional performance of the country groups ‘explain’ (by rendering the corresponding terms no longer significant) the evolving differentials in the support for the transition among country groups and through time.

5 Who is against reforms?

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Trends and evaluations of transitional reforms
  5. 3 The data, measurement issues and stylized facts
  6. 4 The empirical model
  7. 5 Who is against reforms?
  8. 6 Has support for transition increased, and if so, why?
  9. 7 Conclusions
  10. References
  11. Appendix

As a first step, we examine the importance of individual characteristics in affecting support for the transition, and how it might have changed between the initial and the later stages of transition. Transitional reforms undoubtedly generated economic ‘winners’ and ‘losers’ (Brainerd, 1998; Terrell, 1999). People who would not benefit from or could not adapt to the changing environment would probably not be in favour of the transition. For instance, as it has often been remarked, older workers, women and those already unemployed or with obsolete labour market skills can be expected to oppose the transition reforms, as they might fear decreased social security and increased unemployment risks. Also, individuals who had experienced the labour market under socialism probably have different attitudes than the younger cohorts. On the other hand, young, educated and wealthier individuals are more likely to support the process of transition, as they may benefit from the new opportunities that arise with it.

In this section, we seek a confirmation for these hypotheses. More specifically, we examine how the socio-economic characteristics of each respondent affect his or her evaluation of the system change and the probability of being in favour or against the transition. We do this using two complementary dependent variables: the first, reported in Table 1, is the standardized distance which measures (for either the economic or the political system) the ranking of the present system relative to the past one. The second group of dependent variables follows from a classification introduced by Rose and Mishler (1994) and used in their studies. As described in the previous section, we estimate a MNL model for the probability of belonging to one of the following groups in respect of the attitudes towards the economic (or political) system, respectively: positive (compliant), negative (skeptic), pro-market (democrat), and nostalgic (reactionary). Marginal effects from these regressions are presented in Tables 2aa and 2b.9

Table 1. Change in individual determinants of reforms evaluations before and after 1998, OLS
 (1)(2)(3)(4)
Economic reforms, 1991–1998Economic reforms, 2001–2004Political reforms, 1991–1998Political reforms, 2001–2004
Notes
  1. Dependent variable is the standardized distance between the rankings of present and past economic or political systems (see text). Estimation method: OLS. Standard errors clustered by country are reported in parentheses.

  2. a

    significant at 10 percent

  3. b

    significant at 5 percent

  4. c

    significant at 1 percent.

  5. Reference categories: male, age 20–29 years, cohort between 18 and 55 years old in 1990, less than secondary school education, married, living in rural or small town, employed, household income in the 1st quartile, Slovenia. In columns (1) and (3) reference year is 1998, in columns (2) and (4) reference year is 2004.

Female−0.151b−0.112b−0.098b−0.088c
(0.065)(0.038)(0.041)(0.028)
Young_cohort0.060a0.099c0.0280.098b
(0.031)(0.028)(0.024)(0.032)
Age 30–39−0.059c−0.045−0.063c0.011
(0.015)(0.032)(0.015)(0.037)
Age 40–49−0.173c−0.126b−0.153c−0.092b
(0.024)(0.048)(0.016)(0.038)
Age 50–59−0.179c−0.171c−0.159c−0.131c
(0.028)(0.034)(0.032)(0.038)
Age > 60−0.120b−0.163b−0.124a−0.113
(0.046)(0.067)(0.057)(0.064)
Secondary/vocational0.062a0.098b0.134c0.111b
(0.030)(0.042)(0.022)(0.044)
University0.253c0.329c0.338c0.321c
(0.033)(0.037)(0.021)(0.039)
Single0.060a0.134c0.0390.119c
(0.028)(0.023)(0.026)(0.018)
Divorced/Widowed0.043b0.0280.0090.052
(0.019)(0.019)(0.020)(0.033)
City0.155c0.0270.118b0.035
(0.041)(0.058)(0.048)(0.058)
Big town0.021−0.0220.009−0.037
(0.030)(0.027)(0.039)(0.041)
Unemployed−0.144c−0.145c−0.113c−0.091c
(0.037)(0.030)(0.026)(0.029)
Pensioner−0.0620.018−0.003−0.008
(0.036)(0.032)(0.025)(0.044)
Student/Housewife0.111c0.047a0.093c0.064b
(0.019)(0.022)(0.024)(0.021)
2nd household income quartile0.035a0.125c0.049b0.133c
(0.016)(0.020)(0.019)(0.037)
3rd household income quartile0.146c0.214c0.166c0.218c
(0.014)(0.027)(0.021)(0.039)
4th household income quartile0.303c0.471c0.265c0.438c
(0.029)(0.041)(0.025)(0.049)
Czech Rep.0.497c0.531c0.550c0.555c
(0.018)(0.010)(0.029)(0.014)
Slovakia−0.282c−0.149c−0.082c−0.084c
(0.013)(0.008)(0.018)(0.009)
Hungary−0.505c−0.038−0.421c−0.017
(0.013)(0.062)(0.019)(0.056)
Poland0.167c0.0240.196c0.056
(0.014)(0.060)(0.022)(0.054)
Estonia−0.251c0.376c−0.165c0.120a
(0.055)(0.079)(0.039)(0.064)
Lithuania−0.961c−0.218b−0.484c−0.026
(0.056)(0.076)(0.039)(0.055)
Latvia−0.722c−0.159a−0.434c−0.125b
(0.054)(0.075)(0.039)(0.056)
Bulgaria−0.421c−0.153b−0.009−0.086a
(0.009)(0.054)(0.010)(0.044)
Romania−0.043−0.130c0.441c0.112c
(0.043)(0.013)(0.031)(0.011)
Russia−0.756c−0.442c−0.790c−0.415c
(0.033)(0.076)(0.027)(0.053)
Ukraine−1.375c−0.302c−0.898c0.005
(0.023)(0.016)(0.026)(0.016)
19910.054 0.277b 
(0.071) (0.105) 
1992−0.107 0.076 
(0.064) (0.117) 
19930.078 0.240b 
(0.106) (0.089) 
19950.009 0.208a 
(0.108) (0.108) 
19960.150 0.245b 
(0.104) (0.082) 
2000 −0.356b −0.142b
 (0.130) (0.055)
2001 0.061 0.016
 (0.103) (0.088)
Constant−0.281c−0.490c−0.083−0.276c
(0.067)(0.060)(0.065)(0.068)
Observations44,68421,92343,70221,472
R-squared0.220.170.210.12
Table 2a. Change in determinants of support for the economic system, before and after 1998, MNL
 (1)(2)(3)(4)(5)(6)(7)(8)
1991–19982000–2004
PositivePro-marketNostalgicNegativePositivePro-marketNostalgicNegative
Notes
  1. Dependent variable is adhesion to one of the four alternative sub-groups (see text).

  2. Reported results are marginal effects from multinomial logit. Standard errors clustered by country are in parentheses

  3. a

    significant at 10 percent

  4. b

    significant at 5 percent

  5. c

    significant at 1 percent.

Female−0.022a−0.022b0.061b−0.017b−0.006−0.026c0.040b−0.008
(0.013)(0.011)(0.026)(0.007)(0.015)(0.006)(0.017)(0.005)
Young_cohort0.027b0.002−0.036c0.007−0.0080.028c−0.0320.013
(0.012)(0.008)(0.011)(0.012)(0.020)(0.011)(0.024)(0.015)
Age 30–39−0.001−0.016c0.0150.002−0.031a−0.0070.033b0.004
(0.004)(0.005)(0.010)(0.008)(0.016)(0.008)(0.015)(0.014)
Age 40–49−0.017c−0.030c0.070c−0.022c−0.028−0.027c0.061c−0.006
(0.007)(0.005)(0.009)(0.007)(0.027)(0.007)(0.023)(0.019)
Age 50–59−0.004−0.032c0.061c−0.025b−0.037a−0.039c0.077c−0.001
(0.006)(0.006)(0.014)(0.012)(0.020)(0.010)(0.023)(0.018)
Age > 60−0.011−0.021b0.042a−0.010−0.043−0.023a0.071a−0.005
(0.011)(0.010)(0.022)(0.013)(0.025)(0.013)(0.037)(0.019)
Secondary/Vocational−0.0010.015a−0.038c0.024c0.0180.023c−0.038a−0.003
(0.003)(0.008)(0.009)(0.009)(0.013)(0.009)(0.021)(0.008)
University−0.0090.053c−0.118c0.074c0.0200.091c−0.122c0.011
(0.011)(0.010)(0.008)(0.009)(0.016)(0.013)(0.021)(0.010)
Single−0.002−0.001−0.025c0.027c−0.0090.038c−0.047c0.018a
(0.008)(0.006)(0.008)(0.006)(0.012)(0.009)(0.008)(0.010)
Divorced/Widowed0.006−0.001−0.0120.007−0.021b0.013b0.0060.001
(0.005)(0.002)(0.009)(0.009)(0.008)(0.006)(0.011)(0.005)
City−0.017b0.032c−0.065c0.050c0.0060.006−0.0280.015
(0.008)(0.009)(0.014)(0.011)(0.024)(0.017)(0.019)(0.016)
Big town0.0010.007−0.0170.0090.008−0.002−0.004−0.001
(0.004)(0.006)(0.010)(0.008)(0.015)(0.012)(0.012)(0.006)
Unemployed−0.012−0.027c0.051c−0.011−0.014a−0.028c0.057c−0.015b
(0.008)(0.008)(0.011)(0.008)(0.007)(0.008)(0.009)(0.007)
Pensioner0.017−0.018b0.021a−0.020b0.030b−0.012a−0.004−0.013a
(0.014)(0.008)(0.013)(0.010)(0.014)(0.007)(0.017)(0.007)
Student/Housewife0.0040.019c−0.048c0.026c0.0030.007−0.0130.004
(0.008)(0.005)(0.010)(0.007)(0.012)(0.010)(0.012)(0.009)
2nd household income quartile0.0030.003−0.017b0.0110.032c0.025c−0.069c0.011c
(0.006)(0.005)(0.007)(0.007)(0.011)(0.007)(0.009)(0.004)
3rd household income quartile0.017b0.022c−0.056c0.017b0.042c0.049c0.097c-0.005
(0.007)(0.003)(0.007)(0.007)(0.008)(0.007)(0.009)(0.008)
4th household income quartile0.034c0.057c−0.129c0.038c0.041c0.120c−0.182c0.021a
(0.009)(0.010)(0.011)(0.007)(0.012)(0.013)(0.013)(0.012)
Czech Rep.0.059c0.157c−0.113c−0.102c−0.170c0.178c−0.133c0.124c
(0.016)(0.008)(0.008)(0.015)(0.002)(0.005)(0.006)(0.004)
Slovakia−0.012a−0.048c0.170c−0.111c−0.173c−0.026c0.157c0.042c
(0.006)(0.001)(0.004)(0.005)(0.003)(0.002)(0.005)(0.002)
Hungary−0.028c−0.069c0.235c−0.138c0.038−0.036c0.012−0.012
(0.010)(0.002)(0.006)(0.009)(0.032)(0.010)(0.033)(0.011)
Poland−0.0000.052c0.033c−0.085c−0.149c0.0130.098c0.038c
(0.013)(0.005)(0.008)(0.015)(0.022)(0.015)(0.030)(0.014)
Estonia0.021−0.025c0.166c−0.162c0.127c0.066c−0.167c−0.026b
(0.017)(0.009)(0.021)(0.006)(0.031)(0.017)(0.036)(0.011)
Lithuania−0.090c−0.094c0.362c−0.178c−0.057a−0.054c0.114c−0.003
(0.011)(0.003)(0.013)(0.005)(0.033)(0.012)(0.040)(0.011)
Latvia−0.107c−0.086c0.327c−0.133c−0.076b−0.035c0.082b0.028b
(0.009)(0.003)(0.012)(0.007)(0.031)(0.012)(0.039)(0.014)
Bulgaria−0.122c−0.057c0.282c−0.102c−0.147c−0.027b0.139c0.035c
(0.002)(0.001)(0.002)(0.002)(0.022)(0.011)(0.027)(0.013)
Romania−0.044c−0.020b0.091c−0.027c−0.165c−0.037c0.081c0.121c
(0.009)(0.009)(0.018)(0.010)(0.003)(0.003)(0.007)(0.007)
Russia−0.094c−0.107c0.330c−0.129c−0.076b−0.099c0.171c0.004
(0.008)(0.003)(0.009)(0.012)(0.033)(0.010)(0.039)(0.011)
Ukraine−0.172c−0.110c0.466c−0.185c−0.034c−0.066c0.100c0.002
(0.002)(0.001)(0.003)(0.003)(0.007)(0.002)(0.008)(0.004)
1991−0.050−0.021−0.0420.113    
(0.049)(0.018)(0.034)(0.078)    
1992−0.056−0.036a0.0130.078    
(0.057)(0.018)(0.039)(0.089)    
1993−0.001−0.005−0.0210.027    
(0.039)(0.024)(0.044)(0.052)    
19950.0190.009−0.010−0.019    
(0.052)(0.028)(0.050)(0.048)    
19960.0180.029−0.047−0.001    
(0.050)(0.032)(0.044)(0.065)    
2000    −0.184c−0.044a0.214c0.014
    (0.040)(0.024)(0.062)(0.013)
2001    −0.0680.0230.0120.032
    (0.049)(0.021)(0.054)(0.020)
Observations8,4755,79420,29610,1197,7433,4058,5922,183
Observations44,68421,923
Pseudo R-squared0.090.07
Table 2b. Change in determinants of support for the political system, before and after 1998, MNL
 (1)(2)(3)(4)(5)(6)(7)(8)
1991–19982000–2004
CompliantDemocratReactionarySkepticCompliantDemocratReactionarySkeptic
Notes
  1. See Table 2aa.

Female−0.003−0.033***0.033**0.0030.010−0.040***0.031***−0.000
(0.006)(0.012)(0.016)(0.009)(0.007)(0.008)(0.012)(0.006)
Young_cohort0.0140.011−0.019−0.0060.0250.019−0.055**0.012
(0.016)(0.012)(0.013)(0.014)(0.017)(0.012)(0.022)(0.016)
Age 30–390.006−0.025***0.020***−0.0020.010−0.0000.002−0.012
(0.006)(0.007)(0.008)(0.006)(0.021)(0.015)(0.018)(0.016)
Age 40–490.006−0.045***0.068***−0.028***0.028−0.031*0.029−0.026*
(0.009)(0.008)(0.010)(0.009)(0.022)(0.018)(0.022)(0.014)
Age 50–590.008−0.046***0.071***−0.034***0.023−0.046***0.055**−0.032**
(0.011)(0.010)(0.016)(0.009)(0.020)(0.015)(0.024)(0.014)
Age > 60−0.006−0.033*0.061**−0.022**0.007−0.0370.051*−0.021
(0.012)(0.017)(0.025)(0.011)(0.017)(0.024)(0.028)(0.018)
Secondary/Vocational−0.0090.049***−0.046***0.0060.0120.028**−0.046**0.006
(0.008)(0.012)(0.005)(0.009)(0.010)(0.014)(0.019)(0.008)
University−0.036***0.120***−0.108***0.024**−0.0020.106***−0.109***0.004
(0.007)(0.014)(0.006)(0.011)(0.013)(0.014)(0.015)(0.011)
Single−0.0010.003−0.020*0.019**−0.034***0.052***−0.024***0.007
(0.009)(0.013)(0.012)(0.008)(0.011)(0.011)(0.008)(0.010)
Divorced/Widowed−0.007−0.0020.0020.007−0.0160.012−0.0120.016***
(0.007)(0.006)(0.009)(0.008)(0.015)(0.013)(0.012)(0.004)
City−0.030***0.029−0.034**0.035***−0.0050.011−0.0240.018
(0.011)(0.019)(0.017)(0.013)(0.024)(0.025)(0.015)(0.022)
Big town−0.0130.002−0.0020.0120.002−0.0070.0040.002
(0.008)(0.015)(0.015)(0.010)(0.015)(0.015)(0.015)(0.013)
Unemployed−0.001−0.039***0.051***−0.0110.003−0.037***0.038***−0.004
(0.011)(0.009)(0.013)(0.010)(0.007)(0.007)(0.013)(0.008)
Pensioner0.025*−0.0070.000−0.018**0.030**−0.015−0.002−0.013
(0.015)(0.011)(0.013)(0.009)(0.015)(0.015)(0.019)(0.012)
Student/Housewife−0.033***0.042***−0.024*0.0140.019*0.023**−0.023−0.019
(0.009)(0.009)(0.013)(0.009)(0.010)(0.010)(0.014)(0.014)
2nd household income quartile−0.0030.014−0.013*0.0020.0090.042**−0.051***0.000
(0.006)(0.009)(0.007)(0.008)(0.010)(0.018)(0.011)(0.008)
3rd household income quartile−0.0040.053***−0.051***−0.002−0.0060.068***−0.070***0.008
(0.008)(0.009)(0.008)(0.008)(0.011)(0.015)(0.013)(0.008)
4th household income quartile−0.0100.091***−0.087***0.006−0.029**0.151***−0.139***0.017*
(0.010)(0.010)(0.009)(0.009)(0.014)(0.021)(0.013)(0.010)
Czech Rep.−0.034**0.235***−0.102***−0.099***−0.236***0.243***−0.131***0.123***
(0.014)(0.010)(0.008)(0.015)(0.002)(0.006)(0.005)(0.005)
Slovakia0.028***−0.035***0.063***−0.056***−0.216***0.010***0.102***0.104***
(0.005)(0.005)(0.007)(0.007)(0.001)(0.003)(0.005)(0.003)
Hungary0.036***−0.113***0.190***−0.113***0.005−0.0030.010−0.012
(0.013)(0.004)(0.008)(0.011)(0.026)(0.017)(0.027)(0.010)
Poland−0.0040.074***−0.004−0.066***−0.139***0.0300.0330.076***
(0.015)(0.008)(0.008)(0.016)(0.019)(0.019)(0.027)(0.011)
Estonia0.022*−0.039**0.093***−0.076***−0.0320.053**−0.0340.013
(0.012)(0.014)(0.014)(0.008)(0.031)(0.022)(0.030)(0.009)
Lithuania−0.011−0.160***0.205***−0.034***−0.147***0.0110.0450.091***
(0.010)(0.008)(0.014)(0.010)(0.025)(0.019)(0.028)(0.011)
Latvia−0.046***−0.149***0.193***0.002−0.143***−0.036**0.0380.141***
(0.009)(0.009)(0.014)(0.010)(0.023)(0.016)(0.028)(0.011)
Bulgaria0.023***−0.017***0.074***−0.115***−0.137***0.0080.089***0.040***
(0.003)(0.004)(0.003)(0.002)(0.018)(0.017)(0.024)(0.009)
Romania0.027**0.118***−0.104***−0.041***−0.204***0.025***−0.015***0.195***
(0.014)(0.017)(0.007)(0.011)(0.001)(0.004)(0.005)(0.004)
Russia−0.048***−0.245***0.314***−0.020−0.097***−0.122***0.150***0.069***
(0.008)(0.006)(0.012)(0.012)(0.028)(0.014)(0.031)(0.010)
Ukraine−0.099***−0.212***0.414***−0.103***−0.038***0.022***0.026***−0.010***
(0.005)(0.004)(0.010)(0.008)(0.004)(0.007)(0.007)(0.004)
1991−0.059*0.059−0.068***0.068    
(0.032)(0.045)(0.026)(0.051)    
1992−0.0490.003−0.0270.074    
(0.060)(0.041)(0.034)(0.079)    
19930.0030.054−0.070**0.013    
(0.033)(0.038)(0.029)(0.041)    
19950.0110.060−0.063**−0.008    
(0.038)(0.042)(0.032)(0.041)    
1996−0.0360.075−0.057**0.018    
(0.031)(0.041)(0.024)(0.042)    
2000    −0.072*−0.049***0.096***0.024**
    (0.041)(0.016)(0.037)(0.010)
2001    −0.0200.0050.0150.000
    (0.041)(0.027)(0.043)(0.019)
Observations9,06812,31111,93910,3847,0755,1746,2322,991
Observations43,70221,472
Pseudo R-squared0.080.05

For both sets of dependent variables, we split the regressions in two sub-periods, a period of ‘recession’ (1991–1998) and a period of ‘growth’ (2000–2004). Analyzing these two sub-samples allows us to assess how changes in the economic conditions and in the process of reforms may have affected the impact of individual characteristics on the evaluations of transition.

The degree of concordance between the findings from all regressions is quite high, despite the differences in the dependent variables and estimation methods. This similarity allows us to summarize our findings in a compact way. First, and focusing on the period of recession (1991–1998), the impact of individual characteristics on attitudes towards reforms confirms the earlier findings in the literature and is consistent with the ‘losers versus winners’ approach. The evaluations given by females, those older than 40 years and those who are unemployed are generally more negative (Table 1, column 1 for the economic evaluations, and column 3 for the political ones), and these groups are significantly more likely to express nostalgia for the previous economic system (Table 2aa) and to be politically reactionary (Table 2b). Instead, university graduates (and to a lesser extent also those with only secondary or vocational education), students, and those belonging to richer households are much more positive about both economic and political reforms (Table 1, columns 1 and 3) and more likely to be Pro-market (Table 2aa) and Democrats (Table 2b). The effect of urban residence is positive and significant in most regressions. There is also a positive and significant effect for the ‘young cohort’ dummy, thus suggesting a separate significant cohort effect for those who have experienced the labour market under different regimes.

Second, as transition reforms ‘settled in’, only a few individual characteristics appear to be linked to a change in the attitudes towards reforms, and these few changes are easy to interpret:

  • The ‘young-cohort’ becomes significantly more supportive of economic and political reforms (Table 1, columns 2 and 4), and also significantly more Pro-market (Table 2aa, column 6) and less Reactionary (Table 2b, column 7).
  • There is stronger Pro-market and Democrat support from those with higher education or higher incomes. Also those in the 2nd income quartile are more likely to support transition in the second period.
  • Those in the age group 30–39 become generally less negative (Table 1, columns 2 and 4) and less Reactionary (Table 2b, column 7).

Third, there are almost no relevant differences related to individual socio-economic characteristics between the support for economic and political reforms. This is broadly true both for the results reported in Table 1, and those in Tables 2aa and 2b. This fact is interesting, if we consider that, at the aggregate level, the support for the economic and the political transition appears quite different both between country groups and through time, but apparently these differences and changes are more related to macro-institutional variables than to individual socio-economic characteristics.10

This observation opens the way to the fourth and final point: the importance of the country-specific effects. Systemic country effects display a great variability, across countries and through time. Taking Slovenia as the reference country, we observe that support for both economic and political reforms is consistently much stronger in the Czech Republic, and to a much more limited extent also in Poland, in this case only for the first period (see Table 1, columns 1 and 2). All the other countries start off, in 1991–1998, with a negative support level (relative to Slovenia). The 1991–1998 country dummies are especially negative for the two CIS countries (Russia and Ukraine) and for two of the Baltics (Latvia and Lithuania). However, in the more recent period, 2001–2004, all these negative country effects appear less pronounced (as in the case of the four above-mentioned countries), or become less significant (as for Hungary and Bulgaria) or even positive for Estonia) (see Table 1, columns 3 and 4). These results are almost identically replicated in the MNL model in Tables 2aa (for the economic system) and 2b (for the political system), with the main difference being Poland, which in this model appears to be moving towards becoming less pro-market and more nostalgic in the economic dimension, and less democrat and more skeptic in the political one.

Overall, the examination of both OLS and MNL models points to the same fact: the observed convergence in the degree of support for both the economic and political transition between the CEE, Baltic and CIS countries cannot be explained in relation to changes in individual characteristics. Thus, this convergence should presumably be related to systemic, country-specific variables. In the next section, we turn to a more specific analysis of the role of these variables.

6 Has support for transition increased, and if so, why?

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Trends and evaluations of transitional reforms
  5. 3 The data, measurement issues and stylized facts
  6. 4 The empirical model
  7. 5 Who is against reforms?
  8. 6 Has support for transition increased, and if so, why?
  9. 7 Conclusions
  10. References
  11. Appendix

As we have shown in previous sections, the average level of support for economic and political reforms has been quite different across the transition countries, and reached its lowest level in the CIS countries. Given the diversity of the initial conditions, of the objectives and sequence of reforms, of the paths of political development and of economic performance, it would be surprising if citizens from different countries expressed the same evaluations of their countries’ experiences. But why were negative evaluations more concentrated in the CIS during the first decade of the transition? And why did this negative gap in part recede in the second decade? The responses to these questions should take into account several related aspects.

First, there were some important differences in the way socialism was implemented. While the CIS, Baltic and CEE countries all shared the experience of a socialist economy with relatively secure jobs, officially low inequality and equal pay, but also low motivation and low individual responsibility, their pre-socialist histories had been different, as were their democratic achievements before socialism (Svejnar, 2002). Most CEE countries had stronger historical and geographic ties and trade relations with Western Europe. These ties – which already provided a closer cultural proximity between CEE countries and those in Western Europe – were suddenly ‘rejuvenated’ when the perspective of adhesion to the EU became concrete. Also, the CIS countries have gone through a longer and more intense communist experience relative to most CEE countries: this experience lasted seven decades in the CIS, five in the Baltic and four in the CEE countries.

Second, after the beginning of transition, economic reforms were implemented using different strategies and policies, and as a result, the performance of the CIS and CEE countries also differed. At the same time, the path of political liberalization has varied so much so that in 2004 the Freedom House Ranking of political rights and civil liberties still ranged between ‘not free’ for Belarus and ‘partly free’ for Russia and Ukraine to ‘free’ for all the CEE countries. Several explanations for these differences in the process of democratization and in economic performance have been put forward in the literature. For economic performance, these include the role of larger safety nets and non-employment benefits in the CEE countries, better legal systems and enforcement of laws and regulations, and a lower degree of corruption and rent seeking than in the CIS. The reason for these differences may in part be traced back to the previous, pre-communist history of those countries, or also in present times to their closer links with the EU, which became possible with the signing of the first EU Association Agreements.

In this section, we examine specifically the changes that have occurred in the support for transition and hope to identify those factors that have influenced it. For this purpose and given the results in previous sections, we replace country fixed effects with group dummies, based on each country's average evaluations of the support for transition: the CIS with the lowest support, CEE with the highest support and the Baltic countries in between. We employ model (4) from Section 4 to compare the support for transition before and after the 1998 crisis in the CIS and the Baltics based on the interaction terms between the post-1998 dummy and the CIS and Baltics country group dummies. These interactions capture the change in support attitudes in the CIS and the Baltics relative to the rest of CEE.

We employ the same dependent variables and individual characteristics as in Table 1 to describe economic and political support for transition; we then introduce sequentially additional macroeconomic and institutional variables that we expect to explain the convergence in support across the country groups. Baseline results are presented in column (1) of Tables 3a and 3b for support for the economic and political reforms, respectively. The five dummies and interaction effects introduced in column (1) are all significant. They show lower support for reforms in the Baltic and CIS countries in the first part of the period,11 and how this has evolved in the second period. The coefficients in the fourth and fifth rows (Baltpost98 and Cispost98, which correspond to β4 and β5 in equation (4)) provide the most interesting observations: they are positive and statistically significant, implying that support for the economic (political) transition has increased in the Baltics by 0.57 (0.41) points of the standardized distance relative to CEE, and in the CIS by 0.47 (0.58) points, respectively. These coefficients thus document the extent of convergence in the assessments of transitional reforms in the Baltics and the CIS towards those in the CEE countries.

Table 3a. Impact of economic and political performance on individual evaluations of economic reforms
 (1)(2)(3)(4)(5)(6)(7)(8)
Notes
  1. Dependent variable is the standardized distance between the rankings of the present and past economic systems (see text). Estimation method: OLS. Standard errors clustered by country are reported in parentheses.

  2. a

    significant at 10 percent

  3. b

    significant at 5 percent

  4. c

    significant at 1 percent.

  5. The baseline model is presented in column (1) and also, for a reduced sample, in column (7).

  6. Additional controls include gender, age, young cohort, education, marital status, urban residence, labour market status, household income and year fixed effects (not shown, available on request).

Balt −0.574b−0.584b−0.435−0.560b−0.377−0.592b−0.430−0.340
(0.257)(0.230)(0.244)(0.252)(0.262)(0.229)(0.289)(0.306)
Cis −0.862c−0.962c−0.721c−0.781c−0.592b−1.084c−0.853c−0.664
(0.223)(0.237)(0.188)(0.213)(0.247)(0.260)(0.232)(0.379)
Post_98−0.262a−0.346b−0.393b−0.218−0.163−0.185−0.168−0.104
(0.144)(0.129)(0.162)(0.141)(0.173)(0.155)(0.109)(0.099)
Balt_Post980.567c0.571c0.472b0.551c0.3000.515b0.424a0.385a
(0.160)(0.159)(0.184)(0.157)(0.196)(0.173)(0.208)(0.202)
Cis_Post980.472a0.459a0.515a0.392a0.1120.497b0.469a0.362
(0.220)(0.252)(0.244)(0.215)(0.266)(0.185)(0.244)(0.297)
Unempl  −0.027      
 (0.016)      
GDP-pc  0.046a     
  (0.022)     
Inflation    −0.000b    
   (0.000)    
GDP-growth    0.026b   
    (0.011)   
Dem      −0.073  
     (0.055)  
Gini        −1.236
       (1.947)
Constant −0.1530.225−0.582b−0.196−0.402a0.443−0.299b0.008
(0.204)(0.293)(0.255)(0.205)(0.215)(0.518)(0.133)(0.537)
Observations66,60766,60766,60766,60766,60766,60746,40746,407
R-squared0.150.150.160.150.150.150.140.14
Table 3b. Impact of economic and political performance on individual evaluations of political reforms
 (1)(2)(3)(4)(5)(6)(7)(8)
Notes
  1. Dependent variable is the standardized distance between the rankings of the present and past political systems.

  2. All the other explanations are as for Table 3a.

Balt −0.436**−0.443***−0.402**−0.440**−0.367*−0.453***−0.284−0.249
(0.159)(0.139)(0.143)(0.156)(0.189)(0.143)(0.185)(0.178)
Cis −0.931***−1.015***−0.897***−0.956***−0.836***−1.152***−0.946***−0.874***
(0.127)(0.137)(0.112)(0.125)(0.189)(0.115)(0.136)(0.223)
Post98−0.431***−0.355***−0.471***−0.428***−0.586***−0.334**−0.139−0.127
(0.126)(0.112)(0.148)(0.125)(0.170)(0.125)(0.092)(0.097)
Balt_Post980.405***0.409***0.382***0.409***0.3110.354**0.2370.223
(0.116)(0.109)(0.115)(0.113)(0.181)(0.133)(0.150)(0.143)
Cis_Post980.576***0.564***0.586***0.600***0.449*0.601***0.687***0.647**
(0.143)(0.162)(0.150)(0.170)(0.235)(0.105)(0.191)(0.225)
Unempl  −0.022*      
 (0.011)      
GDP-pc  0.011     
  (0.024)     
Inflation    0.000    
   (0.000)    
GDP-growth    0.009   
    (0.011)   
Dem      −0.073*  
     (0.038)  
Gini        −0.470
       (1.496)
Constant0.315**0.485**0.2180.311**0.416**0.885**0.0040.133
(0.135)(0.171)(0.245)(0.134)(0.142)(0.290)(0.098)(0.427)
Observations65,17465,17465,17465,17465,17465,17445,66945,669
R-squared0.150.160.150.150.150.160.160.16

In the next three subsections, we investigate the factors that might have been affecting this remarkable evolution of attitudes towards transitional reforms. While the initial gap between country groups and its narrowing after 1998 are undoubtedly related to differences in the economic performance and the political process across countries, it remains to be seen which specific characteristics of those processes it is most closely related to. In this respect, we hypothesize that the convergence of support may be due to the following:

  • different quality of the reforms adopted in each country, or, more generally, different economic outcomes of the transition, either with respect to macroeconomic performance or to distributive justice and equality;
  • different degrees of government effectiveness, measured by the quality of governance;
  • different degrees of social capital and trust in political institutions in different countries, as also suggested by the model of Aghion et al. (2010).

6.1 The role of macro and institutional variables

In this subsection, we test whether the variables measuring the macroeconomic and political performance of countries contribute to explaining the positive trend in support for transition in the CIS and Baltic countries. The role of macro performance, institutions and policies in affecting individual attitudes in post-communist countries has been documented in the literature (see, among others, Denisova et al., 2009; Guriev and Zhuravskaya, 2009). For example output growth, lower income inequality, less corruption and better governance could open up opportunities for improvements for many individuals and thus could be associated with a greater support for reforms. On the other hand, as argued by Rodrik (1995) and Fidrmuc (1999), especially at the beginning of transition, high unemployment may actually signal the need for more radical reforms and thus, paradoxically, reinforce the support for reforms.

We introduce sequentially in the baseline model of columns (1) of Tables 3a and 3b the following systemic variables: unemployment rate, GDP per capita, inflation, GDP growth, democracy index from the Polity IV and the Gini index (see the Data Appendix for detailed descriptions). If these variables are indeed behind the positive trend in support in the CIS or the Baltics after 1998, the coefficients on the interaction terms β4 and β5 in equation (4) should become insignificant or lower in magnitude. Two interesting results are worth noting. First, in column (5) of Table 3a, following the introduction of GDP growth (which is significant and positive), four of the five dummies lose their significance: in particular, the Baltics are no longer significantly different from CEE either before or after 1998, and the Cis_Post98 dummy, which models the convergence of the CIS towards CEE after 1998, is no longer significant, which implies that macroeconomic growth performance explains a large part of the positive change in the extent of satisfaction towards the transition after 1998. Thus, the strong growth performance of the CIS and Baltic countries after 1998 and until the crisis in 2008 appears to have been important also to restore confidence in the transition. The same macro variable is not significant in Table 3b; however, after its introduction the coefficients on the interaction terms between the Post98 and the CIS or Balt dummies become smaller in magnitude and lose their significance (becoming marginally significant, however, for the CIS_Post98). Overall, we may conclude that the macroeconomic growth performance explains a large part of the improved evaluation of the economic and political system in the CIS and Baltic countries.

Second, introducing the Gini coefficient for net income inequality (see Table 3a, column (8)) makes the CIS and Cis_Post98 dummies no longer significant and reduces the magnitude of the Balt_Post98 dummy (these results have to be compared with those in column (7), which is estimated on the sample of countries with available Gini index). We thus tentatively conclude that inequality in income distribution may have also contributed to explaining the change in support for the market economy, particularly in the CIS.12

Before proceeding to the next section, it is worth commenting briefly on the impact of macro variables per se on the support for transition and, in particular, why many indicators are not or only marginally significant. There may be several reasons. One is that these variables are included in addition to country group and year-specific effects, which might better capture the changes in the macro environment. A second reason is that, in reference to unemployment, people might care more about their own performance than about aggregate unemployment. A third reason may be that left- and right-wing individuals might place different weights on unemployment relative to inflation. As argued by DiTella and MacCulloch (2005), left-wing individuals may care more about unemployment, while right-wingers care more about inflation, and these differences may cancel out when averaging across left and right-wing individuals, as is done here. A fourth reason is that high unemployment might actually signal the need for more reforms (as in Rodrik, 1995), rather than the failure of the past ones. On a different note, from Table 3b we observe that a higher level of democracy is negatively, although marginally, related to support for the political transition (the effect is not statistically significant in the equation for economic reforms). As this indicator has only a limited variation for a majority of countries, it acts almost as a dummy, essentially separating EU from non-EU members, and is thus collinear with the CIS country dummies.13 Alternative explanations could follow those proposed by Guriev and Zhuravskaya (2009), who find a negative relationship between democracy and the happiness index, and by Denisova et al. (2009), who show that, in more democratic countries, individuals who experience economic hardship during transition are more likely to favour re-nationalization.

6.2 The quality of governance

In this subsection, following the same empirical strategy as in the previous one, we introduce several World Bank Governance Indicators to the baseline model. The baseline equation, estimated over the reduced sample for which these indicators are available, is shown in column (1) of Tables 4a and 4b. Notice that, although the sample size drops substantially, the effects of the interaction terms remain positive and significant. In order to disentangle the differential impact of each Governance Indicator on support for the transition, we introduce sequentially the following six variables into the baseline model: voice and accountability, political stability, government effectiveness, regulatory quality, rule of law and control of corruption (see the Data Appendix).

Table 4a. Impact of governance indicators on individual evaluations of economic reforms
 (1)(2)(3)(4)(5)(6)(7)
Notes
  1. Dependent variable is the standardized distance between the rankings of the present and past economic systems.

  2. The model in column (1) is the same baseline model as in column (1) of Table 3a, estimated over a reduced sample.

  3. All the other explanations are as for Table 3a.

Balt −0.480−0.533*−0.471*−0.355*−0.828***−0.392−0.275
(0.284)(0.273)(0.251)(0.183)(0.230)(0.253)(0.263)
Cis −0.972***−0.557*−0.632−0.547**−0.142−0.386−0.519**
(0.242)(0.304)(0.405)(0.192)(0.239)(0.342)(0.234)
Post98−0.017−0.086−0.014−0.113−0.1400.055−0.026
(0.138)(0.149)(0.123)(0.132)(0.120)(0.128)(0.120)
Balt_Post980.477**0.525**0.363*0.279*0.579***0.3150.219
(0.191)(0.199)(0.187)(0.143)(0.149)(0.201)(0.203)
Cis_Post980.584**0.696**0.555*0.558*0.491*0.578**0.616**
(0.251)(0.285)(0.274)(0.269)(0.239)(0.246)(0.252)
Voice  0.013*     
 (0.007)     
Polstab   0.008    
  (0.006)    
Goveff    0.013***   
   (0.003)   
Regq     0.021***  
    (0.006)  
Ruleol      0.015* 
     (0.007) 
Cntrcorr       0.011**
      (0.004)
Constant−0.389*−1.268***−0.915*−1.194***−1.816***−1.395***−1.118***
(0.204)(0.406)(0.441)(0.127)(0.376)(0.439)(0.219)
Observations37,86037,86037,86037,86037,86037,86037,860
R-squared0.170.170.170.190.190.180.18
Table 4b. Impact of governance indicators on individual evaluations of political reforms
 (1)(2)(3)(4)(5)(6)(7)
Notes
  1. Dependent variable is the standardized distance between the rankings of the present and past political systems.

  2. All the other explanations are as for Table 4a.

Balt −0.302−0.313−0.299−0.265*−0.428*−0.266−0.252
(0.183)(0.189)(0.173)(0.146)(0.195)(0.167)(0.173)
Cis −0.993***−0.905***−0.865***−0.866***−0.694***−0.746***−0.881***
(0.133)(0.195)(0.212)(0.131)(0.190)(0.226)(0.148)
Post98−0.283**−0.298**−0.282**−0.312**−0.328**−0.254**−0.286**
(0.107)(0.126)(0.104)(0.122)(0.113)(0.098)(0.109)
Balt_Post980.278*0.288*0.2360.2190.315**0.2100.214
(0.148)(0.155)(0.148)(0.125)(0.132)(0.149)(0.155)
Cis_Post980.640***0.663***0.629***0.632***0.607***0.638***0.648***
(0.193)(0.203)(0.201)(0.200)(0.190)(0.190)(0.195)
Voice  0.003     
 (0.005)     
Polstab   0.003    
  (0.004)    
Goveff    0.004   
   (0.003)   
Regq     0.007  
    (0.005)  
Ruleol      0.006 
     (0.006) 
Cntrcorr       0.003
      (0.003)
Constant0.174−0.011−0.023−0.067−0.339−0.249−0.007
(0.144)(0.336)(0.280)(0.146)(0.303)(0.367)(0.173)
Observations37,17037,17037,17037,17037,17037,17037,170
R-squared0.150.150.150.150.150.150.15

We first note that introducing any Governance Indicators either reduces or renders non significant the CIS dummy in the equation for economic reforms.14 This suggests that a large part of the initial distaste for economic reforms in the CIS was due to the perception of the bad quality of governance in these countries, especially as measured by regulatory quality or the rule of law indicators (columns 5–6). Regarding the increase after 1998, the dummy Cis_Post98 remains significant even after the introduction of governance indicators in the equation for the political system (Table 4b) and, in the equation for economic reforms, regulatory quality reduces its coefficient, which becomes only marginally significant (Table 4a).

For the Baltic countries, regulatory quality increases considerably the absolute value of the negative coefficient of the Balt dummy. This implies that, were it not for the high quality of regulations, the evaluation of economic reforms in these countries would be much more negative. In addition, the positive changes in the economic and political evaluations in the Baltics seem to have been influenced mostly by the improvements in political stability, government effectiveness and, especially, rule of law and control of corruption. This is probably an outcome of the political and administrative requirements related to the EU accession in 2004.

6.3 Trust in political institutions

Our dataset contains responses to questions concerning the degree of individual trust in political parties, in the Parliament and (for a smaller subsample) in the country's President. Trust in most countries and years is measured on a scale from 1 (least trust) to 7 (most trust); however, in the Baltics in 1993 and 1996 and in Russia in 1993 the ranking is from 1 (trust a lot) to 4 (don't trust).15 We construct a homogenous binary indicator ‘trust’ that is equal to 1 if respondents rank their trust higher than 3 in the former group and for respondents who report ‘some’ or ‘a lot of trust’ in the latter group.

We introduce trust variables separately and also in interaction with the Balt, Cis, Post98, Balt_Post98 and Cis_Post98 dummies, in order to capture the potentially different effect of trust (and of its changes over time) on these country groups. Results are reported in Tables 5a and 5b. In columns (1), (3) and (5) of both Tables the baseline regressions are re-estimated for the appropriate subsamples. In the even-numbered columns, we add to the baseline model a trust variable and the related dummies: trust towards parties in column (2), towards the Parliament in column (4) and towards the President in column (6).

Table 5a. The role of trust in the individual evaluations of economic reforms
 (1)(2)(3)(4)(5)(6)
Trust partiesTrust ParliamentTrust President
Notes
  1. Dependent variable is the standardized distance between the rankings of the present and past economic systems (see text). Estimation method: OLS. Standard errors clustered by country are reported in parentheses:

  2. a

    significant at 10 percent

  3. b

    significant at 5 percent

  4. c

    significant at 1 percent.

  5. Additional controls include gender, age, young cohort, education, marital status, urban residence, labour market status, household income and year dummies.

  6. Regressions in columns (1), (3) and (5) are identical, except for the sample size, which is respectively chosen to match the availability of individual responses to questions relating to trust. These responses have been included in the regressions in columns (2), (4) and (6) respectively.

  7. Trust refers to the responses to questions about the degree of individual trust in political parties (Column (2)), in Parliament (column (4)), in the President (column (6)).

  8. The final five regressors in each column are the interaction terms between the relevant trust variable and the country and/or time effects.

Balt −0.615b−0.518a−0.612b−0.630b−0.612b−0.635b
(0.249)(0.243)(0.252)(0.226)(0.253)(0.217)
Cis −0.849c−0.712b−0.853c−0.708b−0.854c−0.798c
(0.268)(0.257)(0.265)(0.265)(0.265)(0.221)
Post98−0.271−0.234−0.283−0.276−0.284−0.263
(0.174)(0.171)(0.172)(0.168)(0.173)(0.174)
Balt_Post980.608c0.510b0.605c0.577c0.605c0.300a
(0.173)(0.175)(0.174)(0.159)(0.177)(0.159)
Cis_Post980.4500.3610.4570.3550.4580.371
(0.280)(0.272)(0.278)(0.269)(0.279)(0.244)
Trust  0.337c 0.322c 0.269b
 (0.095) (0.043) (0.089)
Trust_Balt −0.199a −0.009 −0.015
 (0.103) (0.072) (0.161)
Trust_Cis −0.401c −0.352c 0.084
 (0.095) (0.053) (0.172)
Trust_Post98 −0.026 0.008 −0.086
 (0.093) (0.053) (0.110)
Trust_Post98_Balt 0.198 0.119 0.413b
 (0.148) (0.079) (0.16)
Trust_Post98_Cis 0.249b 0.245c −0.050
 (0.094) (0.062) (0.194)
Constant−0.210−0.319−0.198−0.313−0.179−0.320
(0.233)(0.222)(0.234)(0.228)(0.240)(0.232)
Observations49,29749,29749,87749,87746,73546,735
R-squared0.140.160.140.160.150.16
Table 5b. The role of trust in the individual evaluations of political reforms
 (1)(2)(3)(4)(5)(6)
Trust partiesTrust ParliamentTrust President
Notes
  1. Dependent variable is the standardized distance between the rankings of the present and past political systems.

  2. All the other explanations are as for Table 5a.

Balt −0.500***−0.415**−0.484***−0.528***−0.482***−0.586***
(0.146)(0.134)(0.146)(0.127)(0.146)(0.157)
Cis −0.911***−0.762***−0.912***−0.770***−0.913***−0.874***
(0.142)(0.138)(0.141)(0.146)(0.140)(0.094)
Post98−0.398***−0.354***−0.398***−0.396***−0.401***−0.384***
(0.094)(0.095)(0.097)(0.106)(0.097)(0.088)
Balt_Post980.447***0.368***0.431***0.443***0.431***0.212
(0.103)(0.109)(0.104)(0.097)(0.108)(0.128)
Cis_Post980.563**0.440**0.568**0.435**0.574**0.405**
(0.201)(0.189)(0.202)(0.164)(0.207)(0.136)
Trust  0.376*** 0.327*** 0.310***
 (0.083) (0.054) (0.085)
Trust_Balt −0.108 0.042 0.085
 (0.096) (0.095) (0.215)
Trust_Cis −0.422*** −0.338*** 0.174
 (0.103) (0.086) (0.155)
Trust_Post98 −0.038 0.011 −0.099
 (0.078) (0.064) (0.105)
Trust_Post98_Balt 0.087 0.027 0.303
 (0.155) (0.071) (0.209)
Trust_Post98_Cis 0.370*** 0.347*** −0.004
 (0.076) (0.081) (0.229)
Constant0.2110.0880.2160.0990.2420.087
(0.154)(0.145)(0.157)(0.156)(0.161)(0.147)
Observations48,50048,50049,03149,03145,91945,919
R-squared0.140.160.140.160.150.17

We find a strong positive correlation between different measures of trust towards political institutions and support for the market economy and democracy. We interpret this correlation as being analogue to the negative correlation between trust and demand for government regulation or state involvement, which is the main theoretical implication of the model in Aghion et al. (2010).16 Our evidence is also consistent with their empirical findings and with those reported by Denisova et al. (2010) for Russia.17

Second, looking at the interaction terms provides additional, more interesting evidence. While the Trust_Balt interaction is never significant (thus indicating that the Baltics are homogenous to the Cee countries in this respect), the Trust_Cis interaction is significant and of the opposite sign as that of Trust in columns (2) and (4) of both Tables. This is consistent with the observation that, in the CIS, the degree of trust towards the political class is low and does not contribute in a positive way to the evaluation of the transition. Interestingly, this does not apply to the evaluation of the President, where also the Trust_Cis dummy is not significant. This indicates that trust in the President has the same effect on the evaluation of the transition in the CIS as in the other countries.

Third, we observe that introducing the trust variables and their interactions has important effects on other variables. For the Baltics, introducing Trust for the President has the effect of substantially lowering the value and significance of the Balt_Post98 dummy (in both Tables 5a and 5b): this implies that the increased support for the economic and the political transition in the Baltics after 1998 may be largely explained by the increased trust in the President. For the CIS, instead, it is interesting to observe that the interaction term Trust_Post98_CIS is strongly positive in columns (2) and (4), which indicates that the increased trustworthiness of the parties and the Parliament after 1998 has had a positive effect on the evaluation of the transition experience. Thus, while the lower trust in the political class in the CIS is overall a significant determinant of a more negative evaluation of the process of economic and political reforms, this negative effect is almost entirely compensated for in the second period. Hence, our results suggest that an increase in support for the market economy and democracy after 1998 in the CIS and Baltic countries is accompanied by a relatively larger increase in trust towards the political institutions, in particular, in the CIS.

7 Conclusions

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Trends and evaluations of transitional reforms
  5. 3 The data, measurement issues and stylized facts
  6. 4 The empirical model
  7. 5 Who is against reforms?
  8. 6 Has support for transition increased, and if so, why?
  9. 7 Conclusions
  10. References
  11. Appendix

The 2007 EBRD Life in Transition survey remarked that ‘17 years of transition have taken a toll’ (EBRD, 2007b). Even before the global recession of 2008–2012, ‘transition fatigue’ and discontent with transitional reforms had become apparent in many countries. In this paper, we have proposed a comparative systematic analysis of that discontent, in particular of its evolution over time for the 12 countries included in our study.

We confirm the findings from previous studies that the experience of transition obtained only a limited support from the older, less educated, unemployed and poor individuals and also from females, that is from those groups who were more likely to be among the ‘losers’ from the transition. These effects were generally stronger during the recession period in the 1990s. In addition, the initial support for transition was lower in the Baltic countries than in the rest of Central and Eastern European countries, and was the lowest in the CIS.

Our main results, however, are the findings of a significant increase in support for economic and political reforms in the CIS and the Baltic countries after the crisis of 1998, and an explanation of these changes. We document a converging trend in support for reforms between the CIS, the Baltics and the rest of CEE, which is mainly due to an increased support in the former two groups (mostly in the CIS) and also to a reduced support in CEE for changes in the political system. We also find that positive evaluations of the transition (that is, those expressed by ‘pro-market’ or ‘democratic’ individuals) have become more pronounced after 1998, especially among the younger individuals and those with higher education or income.

Focusing on the systemic factors, we find that GDP growth helps to explain the increased support for the transition in the CIS and Baltic countries after 1998. Also reduced income inequality becomes relevant for the increased support for the market economy in the CIS after 1998. The quality of governance, as measured by the World Bank Governance Indicators, is an additional important factor: political stability, government effectiveness, rule of law and control of corruption contribute to explaining the increasing trends in support for economic and political reforms in the Baltics, whereas bad regulatory quality and rule of law are related to the overall lower support for the market economy in the CIS.

To conclude, we note that our findings are also consistent with the model proposed by Aghion et al. (2010), in which individuals in societies with poor social capital demand more government regulation even when they know that the government is corrupt. We find a positive correlation between trust in political institutions and support for a market economy and democracy, which suggests a negative relationship between trust on the one hand and demand for regulation (distaste for market economy or democracy) or preference for more state involvement and responsibility on the other hand. We also show that an increase in support for the market economy and democracy in the CIS countries is accompanied by a relatively larger increase in trust towards the political institutions.

  1. 1

    See Rovelli and Zaiceva (2009) for a comprehensive review of related literature. Popular support for democracy and the market economy has been studied extensively by political scientists, using, in some cases data from the New Democracy Barometers (see, for example, Rose, 2007; Lazar et al., 2007; Mishler and Rose, 2000a, 2000b, 2002, 2008). Their findings indicate that, in general, evaluations of the economic system and the political regime are influenced by factors related to economic performance and political institutions, as well as by the communist legacy, and that these evaluations also influence each other.

  2. 2

    Although a simplification and generalization, these definitions are useful for a general description of the transition process. See, for example, Roland (2002) for a comprehensive discussion of the political economy of transition and a survey of studies on economic policy reform. Note that countries differed also in the initial conditions, a fact that must be taken into account when modeling the outcomes of transition.

  3. 3

    Earlier studies by political scientists using these datasets to study the evolution of popular support for post-communist regimes include Rose (2007), Lazar et al. (2007), Mishler and Rose, 2000a, 2000b, 2002, 2008). We refer readers also to these studies for the presentation of the sampling framework, methodology and representativeness of this dataset.

  4. 4

    Note that the questions have been framed in accordance with country-specific situations. For example, ‘free elections and many parties’ are not mentioned in the Russian questionnaire, and the questions are only about ‘the current system’ and the economic or political systems ‘before perestroyka’.

  5. 5

    In general, we have extensively tested the sensitivity of our main results to alternative definitions of the dependent variable. In particular, we have used as dependent variables evaluations only of the present or only of the past economic or political systems, as well as binary variables indicating higher ranking of the present relative to the past, or a variable indicating whether a respondent would like to return to communism (available upon request). Overall, our main results were robust to these changes.

  6. 6

    Figure 1A in the Appendix shows the dynamics of support for each individual country, for all the available years.

  7. 7

    These findings should not be interpreted as reflecting a desire to return to communism, as among the respondents who give positive evaluations for the past economic or political system, only about 30 percent would actually agree to ‘return to communist rule’.

  8. 8

    To analyze cross-country differences we estimate the model with country-specific effects and time dummies entered separately. In the earlier version of this study we have also estimated the model with country-year interactions. This did not affect the results for individual covariates (available upon request).

  9. 9

    Note that, by virtue of the estimation method, for each regressor the total effect across the four groups sums to zero.

  10. 10

    We have also estimated the impact of individual experiences, values and preferences related to transition on the support for economic and political system change. Although necessarily endogenous, these variables provide additional valuable information. As was expected and in line with existing studies (see, among others Guriev and Zhuravskaya, 2009; Roland, 2002; Alesina and Fuchs-Schuendeln, 2007) individual hardship experienced during the transition, the perceived ‘wrong’ speed of reforms (‘too high’ or ‘too slow’), preference towards dictatorship, the perceived extent of corruption among public officials, preferences for more redistribution and previous Communist party membership lower the support for transitional reforms (results not reported, but available upon request).

  11. 11

    The result for the CIS is consistent with that of Denisova et al. (2010), who find that Russians express the most negative evaluations towards the free market and democracy among 28 transition countries.

  12. 12

    This interpretation is consistent with the observation that income inequality has been declining in Russia recently (see Denisova et al., 2010), and with the finding by Guriev and Zhuravskaya (2009) that the Gini index has a negative relationship with the support for transition.

  13. 13

    This is consistent with the fact that including the Democracy Index increases the absolute value of the CIS country dummies.

  14. 14

    Note that the estimated effects of the six governance indicators on the support for transition are remarkably similar, which is to be expected given their very high correlation (all values higher than 0.8).

  15. 15

    Unfortunately, trust variables are available in our dataset only for a smaller number of countries and years than those reported in Tables 3a and 3b: they are missing for most countries in 1992 (trust in the President is not available in this year), for Russia and the Baltics in 1995, for the Baltics in 1998 and for Ukraine in 2001.

  16. 16

    Differently from Aghion et al. (2010), who use both a general indicator of distrust towards other people as well as specific indicators of distrust towards companies and civil servants, we use three indicators of trust towards political institutions. We believe that our indicators, while being consistent with the Aghion et al. analysis, are particularly appropriate in the context of our study, which examines the determinants of public support for the economic and political reforms.

    On the other hand, a measure of generalized trust towards other people (akin to the generalized trust measure adopted in the social capital literature) is available for most countries in our dataset only from 1998, and thus cannot be used in our regressions. A detailed study of social capital is beyond the scope of our paper (see in particular Fidrmuc and Gërxhani, 2008 and references therein for an analysis of the gap in social capital between eastern and western European countries). Moreover, while the correlation between generalized trust and trust in political institutions in our dataset is only around 0.2, the ceteris paribus correlation with the support for economic (political) reforms is consistent (that is, positive and significant) with the results reported in the text for the other trust variables.

  17. 17

    We have also estimated the regressions with preference for a greater state responsibility in the regulation of economic activity as a dependent variable (equals one if a respondent prefers a greater state responsibility versus individual responsibility). In line with the findings in Aghion et al. (2010) and Denisova et al. (2010) we find a negative relationship between different measures of trust and state involvement variable (available upon request).

References

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Trends and evaluations of transitional reforms
  5. 3 The data, measurement issues and stylized facts
  6. 4 The empirical model
  7. 5 Who is against reforms?
  8. 6 Has support for transition increased, and if so, why?
  9. 7 Conclusions
  10. References
  11. Appendix
  • Aghion, P. and Blanchard, O. J. (1994). ‘On the Speed of Transition in Central Europe’, in Fischer, S. and Rotemberg, J.J. (eds.), NBER Macroecomics Annual 1994, Vol. 9, Cambridge, MA: National Bureau of Economic Research, pp. 283319. Available online at: http://www.nber.org/chapters/c11012.
  • Aghion, P., Algan, Y., Cahuc, P. and Shleifer, A. (2010). ‘Regulation and distrust’, The Quarterly Journal of Economics, 125 (3), pp. 10151049.
  • Alesina, A. and Fuchs-Schuendeln, N. (2007). ‘Good bye Lenin (or not?) The effect of communism on people's preferences’, American Economic Review, 97, pp. 15071528.
  • Bertrand, M. and Mullainathan, S. (2001). ‘Do people mean what they say? Implications for subjective survey data’, American Economic Review, 91(2), pp. 6772.
  • Boeri, T. (2000). Structural Change, Welfare Systems and Labour Reallocation, Oxford: Oxford University Press.
  • Boeri, T. and Terrell, K. (2002). ‘Institutional determinants of labor reallocation in transition’, Journal of Economic Perspectives, 16(1), pp. 5176.
  • Brainerd, E. (1998). ‘Winners and losers in Russia's economic transition’, American Economic Review, 88 (5), pp. 10941116.
  • Campos, N. and Coricelli, F. (2002). ‘Growth in transition: What we know, what we don't, and what we should’, Journal of Economic Literature, XL, pp. 793836.
  • Denisova, I., Eller, M., Frye, T. and Zhuravskaya, E. (2009). ‘Who wants to revise privatization and why? The complementarity of market skills and institutions’, American Political Science Review, 103, pp. 284304.
  • Denisova, I., Eller, M. and Zhuravskaya, E. (2010). ‘What do Russians think about transition?Economics of Transition, 18 (2), pp. 249280.
  • Di Tella, R. and MacCulloch, R. (2005). ‘Partisan social happiness’, Review of Economic Studies, 72 (2), pp. 36793.
    Direct Link:
  • Easterlin, R.A. (2009). ‘Lost in transition: Life satisfaction on the road to capitalism’, Journal of Economic Behavior and Organization, 71 (2), pp. 130145.
  • EBRD (2007a). Transition Report: People in Transition, London, UK: European Bank for Reconstruction and Development.
  • EBRD (2007b). Life in Transition, London, UK: European Bank for Reconstruction and Development. Available online at: http://www.ebrd.com/pubs/econo/lit.htm
  • Fidrmuc, J. (1999). ‘Unemployment and the dynamics of political support for economic reforms’, Journal of Policy Reform, 3, pp. 139159
  • Fidrmuc, J. (2000). ‘Political support for reforms: Economics of voting in transition countries’, European Economic Review, 44, pp. 14911513.
  • Fidrmuc, J. and Gërxhani, K. (2008). ‘Mind the gap! Social capital, East and West’, Journal of Comparative Economics, 36, pp. 264286.
  • Grosfeld, I. and Senik, C. (2010). ‘The emerging aversion to inequality. Evidence from Poland 1992–2005’, Economics of Transition, 18(1), pp. 126.
  • Grosjean, P. and Senik, C. (2011). ‘Democracy, market liberalization, and political preferences’, Review of Economics and Statistics, 93(1), pp. 365381.
  • Guriev, S. and Zhuravskaya, E. (2009). ‘(Un)Happiness in transition’, Journal of Economic Perspectives, 23(2), pp. 143168.
  • Hayo, B. (2004). ‘Public support for creating a market economy in eastern Europe’, Journal of Comparative Economics, 32, pp. 720744.
  • Kim, B.-Y. and Pirttilä, J. (2006). ‘Political constraints and economic reform: Empirical evidence from the post-communist transition in the 1990s’, Journal of Comparative Economics, 34, pp. 446466.
  • Landier, A., Thesmar, D. and Thoenig, M. (2008). ‘Investigating capitalism aversion’, Economic Policy, Vol. 23(55), pp. 465497.
  • Lazar, O., Mishler, W. and Rose, R. (2007). ‘What is the effect of globalization on support for market economies in post-Communist Europe?’ Centre for the Study of Public Policy, Studies in Public Policy No. 421, University of Aberdeen, Aberdeen. Available online at: http://www.cspp.strath.ac.uk/.
  • Linz, S. J. and Semykina, A. (2008). ‘How do workers fare during transition? Perceptions of job insecurity among Russian workers, 1995–2004’, Labour Economics, 15, pp. 442458.
  • Milanovic, B. and Ersado, L. (2008). ‘Reform and inequality during the transition: An analysis using Panel Household Survey Data, 1990, Policy Research Working Papers No. 4780, The World Bank, Washington, D.C. Available online at: http://econ.worldbank.org.
  • Mishler, W. and Rose, R. (2000a). ‘Political support for incomplete democracies: Realist vs. idealist theories and measures’, Centre for the Study of Public Policy, Studies in Public Policy No. 333, University of Aberdeen, Aberdeen. Available online at: http://www.cspp.strath.ac.uk/.
  • Mishler, W. and Rose, R. (2000b). ‘Regime support in non-democratic and democratic contexts’, Centre for the Study of Public Policy, Studies in Public Policy No. 336, University of Aberdeen, Aberdeen. Available online at: http://www.cspp.strath.ac.uk/.
  • Mishler, W. and Rose, R. (2002). ‘Learning and re-learning regime support: The dynamics of post-communist regimes’, European Journal of Political Research, 41, pp. 536.
  • Mishler, W. and Rose, R. (2008). ‘A two-way street in popular support for new regimes: The political economy of post-Communist transitions’, Centre for the Study of Public Policy, Studies in Public Policy No. 441, University of Aberdeen, Aberdeen. Available online at: http://www.cspp.strath.ac.uk/.
  • Rodrik, D. (1995). ‘The dynamics of political support for reforms in economies in transition’, Journal of Japanese and International Economics, 9, pp. 403425.
  • Roland, G. (2002) ‘The political economy of transition’, Journal of Economic Perspectives, 16(1), pp. 2950.
  • Rose, R. (2007) ‘Learning to support new regimes in Europe’, Centre for the Study of Public Policy, Studies in Public Policy No. 424, University of Aberdeen, Aberdeen. Available online at: http://www.cspp.strath.ac.uk/.
  • Rose, R. and Mishler, W. (1994). ‘Mass reaction to regime change in eastern Europe: Polarization or leaders and laggards?British Journal of Political Science, 24, pp. 159182.
  • Rovelli, R. and Zaiceva, A. (2009). ‘Transition fatigue? Cross-country evidence from micro data’, Discussion Paper No. 4224, IZA, Bonn. Available online at: http://www.iza.org.
  • Svejnar, J. (2002). ‘Transition economies: Performance and challenges’, Journal of Economic Perspectives, 16 (1), pp. 328.
  • Terrell, K. (1999). ‘Worker mobility and transition to a market economy: Winners and losers’, in Birdsall, N., Graham, C. (eds.), New Markets, New Opportunities? Economic and Social Mobility in a Changing World, Washington, DC: Brookings Institution Press, Ch. 7, pp. 168191.

Appendix

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Trends and evaluations of transitional reforms
  5. 3 The data, measurement issues and stylized facts
  6. 4 The empirical model
  7. 5 Who is against reforms?
  8. 6 Has support for transition increased, and if so, why?
  9. 7 Conclusions
  10. References
  11. Appendix

Data acknowledgements and copyright

This research was based on the data from the New Europe Barometer (waves I–VII), New Russia Barometer (waves I–XIII) and New Baltic Barometer (waves I–VI). These data have been produced by the Centre for the Study of Public Policy, University of Aberdeen/University of Strathclyde and by Richard Rose (University of Aberdeen) and William Mishler (University of Arizona). The data have been supplied by the UK Data Archive, under Crown copyright. The original data creators, depositors or copyright holders and the UK Data Archive bear no responsibility for our analysis or interpretation of these data.

The following data were obtained directly from the UK Data Archive:

Rose, R., New Europe Barometer I–V, 1991–1998 [computer file]. Colchester, Essex: UK Data Archive [distributor], October 2005. SN: 5,241.

Rose, R., New Europe Barometer VI, 2001 [computer file]. Colchester, Essex: UK Data Archive [distributor], October 2005. SN: 5,242.

Rose, R. and Mishler, W., New Europe Barometer VII, 2004–2005 [computer file]. Colchester, Essex: UK Data Archive [distributor], July 2007. SN: 5,243.

Rose, R., New Russia Barometer, 2000–2001 [computer file]. Colchester, Essex: UK Data Archive [distributor], November 2003. SN: 4,550.

Rose, R., New Russia Barometer XIII, 2004 [computer file]. Colchester, Essex: UK Data Archive [distributor], August 2007. SN: 5,700.

Table A1. Sample size by country and year
 199119921993199519961998200020012004Total
Notes
  1. Final sample includes respondents with non-missing information on the key explanatory variables (as well as support for economic reforms).

    Source: Authors’ tabulations from the New Barometer Surveys.

Bulgaria89201,0351,043076601,0861,1305,952
Czech Republic6111,18799882200007684,386
Slovakia2645224589320777008483,801
Hungary7565948188760006504744,168
Poland9419627638190006295904,704
Romania9499560001,043007923,740
Slovenia83508106310610007533,639
Ukraine06248168340801001,5784,653
Russia019741,7411,7652,3101,5441,6861821182014,661
Estonia001,4741,05383905938267215,506
Latvia001,34685572906487327695,079
Lithuania001,68876685509601,0599906,318
Total5,2486,81911,94710,3964,7335,5413,8876,80311,23366,607
Table A2. Variables definitions and sources
Variable nameDescription and sources
Dependent variables:
Standardized distance (economic or political evaluations)Distance between present and past evaluations of the economic or political system. Constructed as the standardized difference between individual rankings of the functioning of the present economic/political system, and the past socialist economy/political system. Before standardization ranges from −200 and +200. It is treated as continuous. (See text in Section 3 for a more complete definition and discussion of properties).
Positive, Negative, Nostalgic, Pro-Market; Compliant, Skeptic, Reactionary, Democrat.

Two groups, each comprising four binary mutually exclusive variables defining whether an individual belongs to a specific group, based on his/her evaluations of the past and present economic or, respectively, political system (see text in Section 4.1 for a more complete definition)

Source: New Europe Barometers, New Russian Barometers, New Baltic Barometers.

Individual characteristics:
FemaleDummy equals to 1 if individual is female
Young_cohortDummy equals 1 if individual was 18 years old or younger in 1990
Age <30Dummy equals 1 if individual's age is less than 30 years old (Reference category)
Age 30–39Dummy equals 1 if individual's age is greater than 30 and less than 39 years old
Age 40–49Dummy equals 1 if individual's age is greater than 40 and less than 49 years old
Age 50–59Dummy equals 1 if individual's age is greater than 50 and less than 59 years old
Age >60Dummy equals 1 if age is greater than 60 years old
ElementaryDummy equals 1 if individual has elementary education (Reference category)
Secondary/VocationalDummy equals 1 if individual has secondary or vocational education
UniversityDummy equals 1 if individual has university degree
MarriedDummy equals 1 if individual is married or cohabiting (Reference category)
SingleDummy equals 1 if individual is single
Divorced/WidowedDummy equals 1 if individual is divorced separated or widowed
Small town/RuralDummy equals 1 if individual resides in a small town or rural area (with population less or equal to 5,000; in Russia <20,000) (Reference category)
Big townDummy equals 1 if individual resides in a big town (with population greater than 5,000 and less than 100,000; in Russia – between 20,000 and 1,000,000)
CityDummy equals 1 if individual resides in a city including capital (with population >100,000; in Russia >1,000,000)
EmployedDummy equals 1 if individual is employed (full-time part-time family helper apprentice or self-employed including and working pensioners in some countries) (Reference category)
UnemployedDummy equals 1 if individual is unemployed (including both with and without benefits in Russia)
PensionerDummy equals 1 if individual is a pensioner
Student/HousewifeDummy equals 1 if individual is a house-keeper or a student (in several countries it was not possible to disentangle these two categories)
1st household income quartileDummy equals 1 if household is in the first quartile of the country-specific income distribution (Reference category)
2nd household income quartileDummy equals 1 if household is in the second quartile of the country-specific income distribution
3rd household income quartileDummy equals 1 if household is in the third quartile of the country-specific income distribution
4th household income quartileDummy equals 1 if household is in the fourth quartile of the country-specific income distribution
Trust (Parties; Parliament; President)Trust variables are constructed as dummies equal to 1 if respondents report ‘some’ or ‘a lot of trust’ in Baltics in 1993 and in 1996 and in Russia in 1993 and rank their trust higher than 3 (on a scale from 1 to 7) in the rest of the countries (1992–2004). Source: New Europe Barometers, New Russian Barometers, New Baltic Barometers.
Macroeconomic variables and political institutions:
Unempl Unemployment rate (Source: EBRD. For Belarus the data are from IMF International Financial Statistics CD Rom for Ukraine – from World Development Indicators CD Rom for Estonia in 1990 and 1991 – from the World Development Indicators online database)
GDP-pcGDP per capita PPP (constant 2000 international $) (Source: World Development Indicators online database)
GDP-growthGrowth in real GDP (percent). (Source: EBRD Transition Report 2007)
Inflation GDP deflator (annual change, percent) (Source: World Development Indicators online database)
Gini Gini coefficient: distribution of per capita household net income (Source: Transmonee dataset, http://www.transmonee.org/). For Bulgaria Czech Republic Estonia Latvia Lithuania and Ukraine 1990 or 1989 is used instead of 1991. In Latvia 1997 is used instead of 1995 and 2000 instead of 2001. For Lithuania 1996 is used instead of 1995. For Russia instead of 1993 use 1994. For Slovakia instead of 1995 use 1996. For Slovenia and Ukraine instead of 2004 use 2002.
Dem Democracy Indicator, based on an additive eleven-point scale (0–10) (Source: Polity IV)
Voice, Polstab, Goveff, Regq, Ruleol, Cntrcorr World Bank Governance Indicators: Voice and Accountability, Political Stability, Government Effectiveness, Regulatory Quality, Rule of Law, Control of Corruption. (Source: http://info.worldbank.org/governance/wgi/sc_chart.asp). Note: data from 1996 are used instead of 1995, and from 2002 instead of 2001.
Country or Time Dummy variables:
Balt Fixed effect dummy, equals 1 for Estonia, Latvia, and Lithuania
Cis Fixed effect dummy, equals 1 for Russia and Ukraine
Post98Time dummy, equals 1 for years 2000, 2001 and 2004
Balt_Post98Interaction term between Balt and Post98 dummies
Cis_Post98Interaction term between Cis and Post98 dummies