Resumé

Regime change is closely connected to elite change, although classical theories of elite rule would predict considerable elite continuity. The basic question is how the downfall of communist regimes and introduction of democracy have affected elite patterns in the three Baltic states. The data indicate a combination of continuity and change: circulation of indigenous elites has taken place, while Russians are excluded from power in important state institutions. Elite continuity is explained by lack of "alternative elites" and a need for competent leaders in the state building process. Elite change is explained by ethnic cleavages which raise the issue of indigenous control over power positions.

Introduction

The regime transition process has been especially dramatic in the Baltic states. After 50 years of forceful integration in the Soviet Union, Estonia, Latvia (August 1991) and Lithuania (March 1990) declared their return to independence. During the Soviet period, a considerable number of Russophones immigrated to Estonia and Latvia, and to a lesser extent to Lithuania. One result was considerable recruitment of Russians into the local elites. Another consequence was local indigenous leaders taking part in the nomenclature career system, where membership in the Communist Party (hereafter CP and only referring to the pre-independence Communist Party) was mandatory. However, the middle and upper middle level positions of the state were often filled by "technical experts" who were not party members. This article focuses on whether and how regime change results in elite changes. Do we find elite continuity and recirculation, or are former elite members expelled from the new power positions? More specifically, we investigate how the new power elite is connected to the past by former CP

---

* Anton Steen, Department of Political Science, University of Oslo, P.O. Box 1097, Oslo, Norway.

membership, former positions in the status hierarchy, ethnic background, and generational changes. The main argument is that elite change is determined first of all by the new states' need for skilled leaders and indigenous control of political institutions. A comparison of the three Baltic states shows how differences in ethnic and socio-economic cleavages raised different political issues which may have influenced the elite patterns after the regime change.

From a situation with an ideologically unified and hierarchically organized elite based on Leninist principles during the communist period, with local loyal elites controlled from Moscow, the new states faced the task of establishing national elites to meet the new challenges of pluralist democracy and market economy. The new leaders had to cope with the basic problems of securing independence from Russian influence, establishing new political institutions, deregulating the centrally planned economy, and meeting the material demands of the population. A major challenge was how to establish a competent elite system that symbolized a break with the past while also representing major ethnic groups.

There is no doubt that elites are affected by political regime changes. Finding out how elites change, however, is a complex, although fascinating task for empirical investigations. Linz (1990), for example, focuses on how purging old regime elites as a part of democratization contributes to consolidation or creates other problems for the new regime. This article mainly focuses on the actual extent of purging of old elites and the preconditions that may have produced a pattern of combining elite recirculation with elite turnover. The question of how elite change has influenced the political stability is of minor importance here.

The main question is to which degree regime change from totalitarianism to democracy results in substitution of the former elites in political and social institutions, and how such an elite change may be explained. One proposition is that because basic state and social tasks have to be performed by elites also after a regime change, elite continuity will be prevalent when there is no alternative elite to take over. This general proposition of "state needs" must be supplemented by a specific contextual variable, where national "elite consolidation" became crucial to independence from foreign rule and Russian influence in the Baltics. The question of national independence became interwoven with the composition of the new elites, especially the control of state core institutions like parliament, bureaucracy and judiciary. Internal demands from large minority groups and external pressure from a neighboring state, in this case Russophones and Russia, are supposed to influence the configuration of the national elites. One dilemma for the new states was that the democratic ideal of proportional representation of minority groups was perceived as a direct menace to national and cultural independence.

The Concept of Elite

A broad concept of "elite" that covers several institutions is useful, especially in transforming societies where leaders often "float" between rather weak institutions. In Western pluralist countries, there is no single and comprehensive elite. Elite persons are specialized by institutions and interests and linked to each other in a complex way (Keller 1972). This makes it possible to make intra-national comparisons of how specific elite groups are changing and interacting. Some groups are more important than others in the decision making and distributive processes. Their decisions are more vital to society in terms of scope and impact. These may be called the ruling, or "strategic elites," to be distinguished from segmental, more specialized elites. Strategic elites will be the top political elite and the top administrative elite - in other words, the influential decision makers of the core state institutions. The boundaries between strategic and segmental elites are not clear in established democracies and are supposed to be even more fuzzy in transforming societies with new political and social institutions. Identifying elites by formal position is only one approach (Putnam 1976) and should ideally be supplemented by "reputation analysis" and "decision making" approaches (Dahl 1961). For example, the business elites and academic elites are formally more removed from political decision making but are expected to play an important role in defining the political agenda and influencing the political process, especially when lack of institutional traditions opens for informal inter-elite cooperation.

Studying elites in established democracies creates a different context. Here the institutional approach may be supplemented by an issue-specific identification of elites. The specific political issue at stake will determine which elites are activated and which are not (Schattschneider 1960). In a recent study in the USA, Heinz et. al. (1990) found that configurations of elites vary between policy areas. Safeguarding external national security and internal cultural consolidation are the main policy issues in the Baltic states. Under such circumstances, one would expect elite patterns characterized by continuity and integrated networks between indigenous elites.

National Characteristics

Comparing the different outcomes of elite transition in three rather
homogenous countries makes it possible to trace how different paths of
elite change may be related to specific national traits.

Estonia, Latvia and Lithuania are small states (1.6 mill., 2.7 mill., and 3.7 mill, inhabitants in 1989), with similar historic experiences during the Soviet period. One main difference is the large proportion of non-indigenous people in Estonia and Latvia. In Estonia, 30 percent were Russians, 9 percent

Belorussians, Ukrainians and others, and 61 percent were Estonians. In Latvia, 34 percent were Russians, 14 percent Belorussians, Ukrainians and others, while Latvians constituted only 52 percent of the population. In Lithuania, 80 percent of the population were indigenous, 9 percent Russians, and 8 percent Poles. Five years later, these numbers had only changed slightly.

While most of the previous elite were CP members, the reasons for membership varied. Some were idealists with strong beliefs in centralized leadership and collectivism, others were opportunists using the party as an entrance to a career, and a third group joined the party because of the opportunity to work for more republican self-determination and cultural autonomy within the Soviet system. Only a small proportion of the general population was CP members. In 1981, CP membership as a percentage of the total population was in Estonia 6.6 percent, Latvia 6.4 percent, and Lithuania 5.0 percent, while the average for the entire USSR was 6.5 percent (Harasymiw 1984). The Communist Party was mainly an arena for the elites.

Elite configurations after independence may have been influenced by the situation during the period of "national awakening." In Lithuania, the top elites understood the early popular protests better than in Latvia and Estonia, where the elite system was more Russian dominated (Taagepera 1993). This was reflected in the leadership style of the communist establishment in Lithuania and in Communist Party leader Brazaukas' balancing policy between national independence and Moscow ties. The Communist Party had become unpopular, but many people continued to prefer communist leaders over opposition leaders. This was demonstrated in the 1992 parliament election in Lithuania, where the former communists, now organized in the renamed Democratic Labor Party, won a landslide victory.

One interesting question is how structural differences between the countries constitute different conditions for elite change. Have the larger proportion of non-indigenous people, the higher CP membership rates, and greater Russian elite influence in Estonia and Latvia influenced elite patterns after the regime change?

The Data

The data is from structured face-to-face interviews with a total of 943 respondents in the Baltic states. The judicial system is not included in the totals due to a low number of respondents in Latvia. The number of respondents, excluding the judicial system, was in Estonia 282, in Latvia 298, and in Lithuania 303. "Don't know" and "no answer" are treated as missing values, which makes comparisons more reliable. In general, however, the proportion of respondents answering the questions is very

satisfactory. In some elite groups there are few respondents, and since these
percentages are less reliable, they are put into brackets.

The interviews in Estonia took place in winter 1994, in Latvia in spring 1993, and in Lithuania in fall 1993. This means that the deputies in Estonia and Lithuania had been elected according to post-independence election laws. In Latvia, the deputies were elected to the interim parliament, the "Supreme Council," before the declaration of independence in August 1991, while the interviews were done later. In all countries, the parliament samples were selected to be representative of the major political groupings. The sample sizes in the parliaments were as follows: Estonia 69 percent, Latvia 69 percent, and Lithuania 87 percent.

In addition to representatives in parliament, leaders from the following elite groups are included: state bureaucracy (political and administrative leaders), state enterprises and banks (leaders of state-owned activities), private business (leaders of larger firms, companies and banks), leaders of political parties/movements, intellectuals (leaders in higher education, mass media and culture), the judicial system (judges and prominent lawyers), and municipalities (local political and administrative leaders). It is more difficult to determine the universe of the other elite groups. The leaders were chosen from top institutional positions and sampled from comparable institutions in the three countries.

Theories of Elite Change

"Negotiated" and "pacted" transitions are argued to be a basic precondition of transition to democracy (O'Donnel & Schmitter 1986). Democratic institutions emerging from pacted transitions tend to have better chances of producing stable democratic developments. The "pacted solution" stresses negotiations between established and incoming new elites within a corporatist controlled and representative system. This context was most relevant in explaining transitions to democracy in Southern Europe and South America. While Higley and Burton (1989), O'Donnell & Schmitter (1986) and others have studied transition processes where "elite pacts" played a major role, the breakdown of totalitarian communist regimes creates another context, except, perhaps, in Poland (Zhang 1994). During the totalitarian period, the Baltic elites were extremely stable, unified and ideologically coherent, and oppositional elites were almost completely lacking. Therefore, pacted transition became irrelevant in the Baltic cases. The question is, what happens to the ideologically unified elite of a totalitarian regime, when that regime is transformed into a pluralized and democratic regime?

Ideologically unified elites and competitive elites constitute two fundamentally
different kinds of regimes with different recruitment systems. The

Baltic elites were earlier unified by ideological lines and tightly integrated into the Soviet elite system. Recruitment became salient to an elite career, which was a two-step process: first joining the Communist Party, then advancing in the nomenclature system. The communist establishment of the Soviet republics was carefully selected according to party loyalty. The local elites of the Baltic republics took an active part in making careers in the Soviet nomenclature system.¹

Social elites and particularly political elites that stay on after a regime change away from totalitarian societies have been regarded as barriers to transition to democratic regimes (Linz 1975). The vested interests of conservative elites in preserving a large, state-controlled economic sector may lead to support for an authoritarian regime (Nørgaard 1992). Elitedominated institutional processes make possible the reproduction of basic structural features, self-recruitment, and safeguarding of the interests of the former elites.

The question of why elites change has been examined from several angles. One angle is Pareto's (1901) theory of "elite circulation" caused by social decay among leaders. Here elite change is the result of competence, shrewdness, or violence of competing and ambitious alternative leaders who remove a decadent elite establishment by means of conspiracy or democratic elections. Another angle is Marx's socio-economic theory that the composition of the power elite is a consequence of economic changes and class structures in society. Mosca (1896) argued as Pareto that "elite continuity" was more common than sharp breaks due to the elites' capacity for political control by oligarchic organization. Common to these theories is that changes are gradual, and that the power elite, in a broad sense, constitutes a "ruling class" under a stable political regime. If elite continuity characterizes stable regimes, will regime change produce a comprehensive elite change?

Sudden and violent revolutions like in Russia in 1917 and China in 1949 produced rapid and comprehensive elite changes. Even if the transitions in Russia and Eastern Europe after 1989 were sudden, they were mostly nonviolent. Transitions to democracy will necessarily place more emphasis on consensus solutions and continuity than transitions to authoritarian rule.

The functional needs of the state explain how old elites may remain in power positions, even after a regime change. Certain tasks have to be performed to open possibilities for specialized former elites to fill power positions. A theory of "state needs" may be useful in explaining how old elites can stay in power. The argument is that certain vital functions must be continued after a regime change, because the permanence of bureaucracy is intrinsic to any modern society. Such functions are most easily carried out by persons with experience and competence from the former regime.

Another perspective on elite continuity is that the former power holders
have resources and networks, and they will support each other. The old elite

that once enjoyed powerful positions uses its privileges and networks to
continue in the same position or take on new functions in another elite group,
for example former nomenclature persons going into private business.

One may assume that in ethnically split societies like Estonia and Latvia, where state and nation building processes are closely related to achieving indigenous control over political institutions, the question of elite change will be different from more ethnically homogenous societies, like Lithuania. From the assumptions that state needs and ethnic situations are important to elite recruitment, we will propose the following hypotheses: (1) Since the new states need competent leaders to fill the new power positions, and a large proportion of them were related to the former regime, their former regime affiliations will not matter very much. (2) Since indigenous political control is paramount in newly independent states, considerable ethnic elite change will take place in ethnically divided new states. (3) Since the ethnic issue is largely absent in Lithuania, one may expect a more critical focus on former regime connections resulting in recruitment of more genuine new leaders. (4) Regime change will open career possibilities and access for young and ambitious persons aspiring to positions in the new power arena. States with a large ethnic elite change create lots of new opportunities for young indigenous careerists.

Elite transformation may be specified by changes in social background, recruitment pattern, personal contacts, interpersonal trust, personal characteristics and views on substantive policies and procedures. Here elite change will be related to four aspects of recruitment: are the new Baltic elites former CP members; have they changed position in their professional career; what is their ethnicity; and are they recruited from younger age groups?

Previous Membership in the Communist Party

To which degree have ideological supporters of the earlier regime achieved positions in the new elite system? We asked the elites in Estonia, Latvia and Lithuania about former CP membership. This will give an indication of how many persons committed to the former regime's ideology have continued an elite career under the new regime.²

In all three countries, a large proportion of the elites has been CP members: 53 percent in Estonia, 63 percent in Latvia, and 48 percent in Lithuania. When membership in the Young Communist League (Komsomol) is included, the figures are 73 percent for Estonia, 75 percent for Latvia, and 58 percent for Lithuania. The data clearly indicates that a substantial number of former party members have continued in elite positions. The Lithuanian elites have fewer former CP members and Young Communist League members than the two other countries, which indicates support for our

contention that absence of the indigenous issue may have made former party
affiliation a more pertinent question.

As mentioned before, it is difficult to say how many of the members were dedicated supporters of the former system, compared to the amount of opportunist careerists and nationally minded reformers. And even if continuity of persons with a "believing" communist background is common, this does not necessarily mean that these persons preserve their former leftoriented opinions. In Estonia and Latvia, former CP membership among the elites does not influence attitudes on the left-right scale. In Lithuania, such a tendency is more pronounced (Steen 1996). The explanation is probably that the national communists in Lithuania at an early stage expressed opinions that were not in line with Moscow and also claimed more autonomy, thus legitimizing left-wing orientations after the regime change. The fact that the former elites in Estonia and Latvia were more ideologically committed seems to be related to more elite continuity after the regime change than in Lithuania.

It may be argued that the introduction of democracy and free elections will result in more elite changes among politicians than among the more stable bureaucratic elites. Most deputies in the Latvian and Lithuanian parliaments have a communist background,³ unlike the Estonian parliament, which was elected in 1992. It must be added that the situation in the Latvian parliament has probably changed after the 1993 election. While there were relatively few former communists in the Estonian parliament at the time of the survey, their number increased considerably in the next parliamentary election in

1995. The "time factor" seems to have made former communists more acceptable as people's representatives, also among the critical Estonian electorate. The proportion of former communists in political parties and movements is high compared to the proportion among deputies in the parliament. Since leaders of parties and movements are not that exposed to direct popular evaluation by elections, their past seems to have been less of a hindrance to their political activity.

The opposite is the case in Lithuania, where the government is based on a parliamentary majority of the reformed Communist Party, renamed the "Democratic Labor Party," after the landslide election in 1992. Here, less than half of the representatives have a communist background, and only 20 percent in the political parties/movements, which is considerably lower than in the other two countries. The difference may be explained by a more distinct left-right polarization in this country. The left-oriented government in Lithuania attracts former members of the Communist Party and at the same time stimulates right-wing opposition party leaders without former communist connections.

The proportion of bureaucratic elites with a communist background is lowest in Estonia, except for elites in state enterprises and banks. In other public institutions like the judiciary and municipalities, the former Estonian communists are well represented. On the local level, there has been more elite changes in Lithuania than in the two other countries.

One rather interesting observation is that 50 percent of the Latvian and Lithuanian private business elites are former CP members. In Estonia, the number is 56 percent. If we include membership of the Young Communist League, the numbers are 80 percent for Latvia, 63 percent for Lithuania, and 70 percent for Estonia. Referring to Lieven (1993), some radical nationalists have alleged that the move into private business by former Communists and Komsomol officials is part of a strategy to take over capitalism and control the new states. A better explanation is probably clever use of opportunity in a situation without legal regulations. Many earlier managers of state enterprises have taken personal advantage of privatization of state property, often in a corrupt form (Lieven 1993).

Among intellectuals in Lithuania, the data indicates a split down the middle between members and those with no affiliation with the communists in Lithuania. Their Estonian and Latvian colleagues have been more active in the Komsomol. Including membership in the Young Communist League, 82 percent of the Estonian and 72 percent of the Latvian intellectuals have a communist background, compared to 50 percent in Lithuania. The early national communists in Lithuania were not the only opponents to Moscow rule. Intellectuals without communist sympathies, many of them active in the protest movement "Sajudis," seem to have been especially active in Lithuania.

Paradoxically, the issue of former regime affiliation became especially tense in Lithuania, the country that was least integrated in the Soviet system. During the communist period, the broader national conservative opposition, which was often rooted in the dissident Catholic church, put the "issue of the past" high on the political agenda. Furthermore, the ethnic situation has created different contexts for raising questions about the past. Due to the delicate ethnic situation, the first priority of the Estonian and Latvian indigenous elites was to oust Russians from power positions. Former CP membership therefore became subordinate to the ethnic background of the new power elite. The favorable ethnic structure in Lithuania made it less urgent to rally forces around ethnic issues. Because the nation-state was more consolidated, more "normal" left-right issues came to the fore. One conflictual issue became the "past ideology" of the Lithuanian elites.

One conclusion is that ideological supporters of the former regime have obtained advantageous positions also after the regime change. The tendency is that in all countries, former party members are well represented, not only in traditional state institutions but also in institutions erected after the regime change, such as parliaments and new political parties. On the other hand, there are considerable differences between the countries. Estonia has experienced the biggest changes in parliament and state administration, while in Lithuania, most changes have taken place in the political parties and local government.

Regime Change and Elite Career

Another indicator of elite change is how many of the elite persons with former leader positions continue in the new elite system. Let us therefore check if the current top elite is recruited from earlier top positions, or if the regime change became an opportunity for replacement and advancement for ambitious careerists in lower positions.

Table 2 shows which occupational group the respondent belonged to before assuming his or her present occupation. The high number of "no answer" to the question about earlier position in Latvia is interesting in itself and makes the conclusions tentative. Nonetheless, the main trend is clear in all countries: The new elites come from earlier high positions. In Latvia, however, about one fourth of the present elite had a low position before. These are mainly elites in political parties/movements and to some extent parliament and state enterprises.⁴ The tendency of elite continuity in the occupational hierarchy between the old and the new regime is strongest in Estonia and Lithuania. In Latvia there has been more upward mobility.

During the former regime, holding a high position in the professional
hierarchy often depended on membership in the Communist Party. Table 3

combines earlier occupation with CP membership. The question is how large a proportion of the present elite formerly had a privileged position in the political-administrative establishment in terms of party membership and occupation.

A combination of earlier membership and earlier occupation gives four possibilities: non-member/low position, non-member/high position, member/ low position, and member/high position. As mentioned, the term "alternative elite" is rather ambiguous when used in situations where no alternative elite was allowed. However, we will argue that even if no alternative elite existed before the elite change, it can materialize as a part of the transition process. Non-members in low positions before the regime change are newcomers and an "alternative" to the elites that were making careers during the former system. Non-members in former high positions probably have the greatest advantages in terms of personal resources without bearing the burden of the past. Members in low positions are also a kind of "alternative" to the last group of persons that represent the "purest" form of continuity, i.e. former members in high positions during the former regime.

Table 3 shows the combinations of elite background. A high degree of
"pure" elite continuity will concentrate in the bottom left cell.

About half of the current elite in the Baltic countries are recruited among people who held high positions in the previous regime and who at the same time were CP members. Estonia and Latvia have the clearest continuity of that kind of elite with as many as 54 percent and 55 percent. This is a rather strong indication of continuation of elites from one regime to another.

Non-party members who earlier held low positions have entered the present elite only to a smaller extent (Estonia 2 percent, Latvia 8 percent, and Lithuania 4 percent). The regime change did not result in upward mobility for persons without those two qualifications. But did party membership help

this group after the regime change? The combination of earlier low position and CP membership is somewhat higher in Latvia (15 percent) than in Lithuania (3 percent) and Estonia (1 percent). In general, the data indicates that earlier low status occupation has been prohibitive to entering the elites. CP membership has been of some use to this former low status group only in Latvia.

An interesting difference is found in the last cell combination, high former occupation/non-membership. In Lithuania, 49 percent of non-members who earlier held high positions have continued into the new elite. The percentage is somewhat lower in Estonia at 44 percent and especially in Latvia at only 22 percent. It seems that persons in Lithuania with earlier high positions but not tainted by a communist background have had especially good opportunities to enter elite positions after the regime change. Probably, this group was rather marginal during the totalitarian period but was given wide access to high positions after the regime change. In Estonia, and in particular in Latvia, the "ideological innocence" seems to have been less relevant, since the main concern was the ethnicity of the new ruling elite.

The reason why the "pure" alternative elite is larger in Lithuania may be related to the more nationally oriented communist party which allowed more liberal recruitment of elites during the pre-democratic period and to the more polarized transition to independence (the "Landbergis approach') that purged earlier communists from elite positions during the first stage of independence. In connection with purging, we should expect a larger proportion of non-CP members in the state bureaucracy and enterprises in Lithuania. Table 1 shows that in Lithuania, the highest percentages of earlier

non-CP members is found among participants in the political discourse - politicians (parliament, parties/movements) and intellectuals. This means that the elite change in Lithuania is mainly a result of elections and formations of parties/interest groups creating opportunities for non-CP members. The high ratio of non-CP members among intellectuals supports the proposition that a more national, independent communism in the preindependent period in Lithuania opened for a more generous recruitment of non-CP members to elite positions in liberal and academic professions.

Because a lot of new positions had to be filled, the group of dissidents was too small to fill all elite positions. If some kind of negotiated or "pacted" transition (O'Donnell and Schmitter 1986) has been going on in the Baltic states, it may be due to an understanding between the non-tainted alternative elite and the earlier ideologically committed elite. This is not a "pact" where former elite persons receive compensation to withdraw from positions, although those from the former elite who switched to business activities did receive such "compensation." This is how mainly Russians found a new arena accepted by the indigenous elites: they contributed to building the economy and at the same time had no direct political influence.

One tentative conclusion is that a pragmatic co-existence has been established between incoming and former elites in order to secure certain basic state functions in the Baltic countries. A theory of "state needs" must be supplemented with an explanation that emphasizes the interests and abilities of the former elites to remain within the elite system. Because only few of the indigenous elites were able to transfer to business activities, most were dependent on a job in the state in order to survive economically. The "state need" explanation is therefore most relevant in the bureaucracy, enterprises, judiciary and local government and to some extent among the intellectual elite. Additional mechanisms are probably at work among political elites. The burden of the past seems to be heaviest in political institutions.

Elite Change as National Consolidation: De-Russification

The old elites in Estonia and Latvia were more dominated by Russians than Lithuania's elite was. This situation reflects the proportion of Russians among the general population. Furthermore, the Baits were deeply integrated in the Soviet ruling system. From 1940, Latvia experienced a considerable flow of Latvians born or educated in Russia into local leader positions. These were ardent communists who were loyal to the Soviet federal state. They occupied key positions and without them it would have been difficult to execute Soviet policies (Silde 1987).

The purge of nationally oriented Latvian communists in 1959 and the following installment of Russians and Moscow-loyal Latvians accentuated the Russian domination. By 1986, the principal posts in Latvia were still held by Russians. As Levits (1987) remarks, the distribution of leading positions functioned as a mechanism for control. It was important to avoid "national bases" within the state. One technique was to integrate indigenous leaders into the Soviet system. During the Soviet period, attaining a Moscow education and marrying a Russian was the ideal means of gaining support from the high rank nomenclature and making a fast career after returning home.

An elite change had to consider four different groups who had been active in the Baltic elite system: First of all, ethnic Russians with middle and high level status; second, indigenous high rank leaders who had benefited from and been integrated in the Soviet nomenclature system; third, indigenous "technocrats" who were party members and strived for a career, but who were not in top positions; and fourth, national communists with middle or high level positions who at an early stage expressed opposition to Moscow. Tables 4 and 5 indicate to which degree Russians have been removed from elite positions.

Comparing all elites, only 4 percent in Lithuania are non-indigenous. Even if the Russians are the largest minority group at 8.7 percent of the population, they are represented by only 1 percent among the elites. The Poles have a higher rate with 3 percent among the elites, but 7.1 percent of the population.

In Estonia, 90 percent of the elites are indigenous, and 8 percent of the elites are Russians out of 30.3 percent of the population. The Latvian elite is the most heterogeneous with 82 percent indigenous and 10 percent Russians. 33.4 percent of the population are Russians.

One conclusion is that the Russian-dominated leadership of the Soviet
period is broken. We have no figures that illustrate the situation before
independence, so we must rely on more indicative material. Still, there is no

doubt that a substantial ethnic change has taken place among the elites. Latvia, the Baltic country that ethnically was most integrated in the Soviet Union and most dominated by Russian elites from an early stage, still has the largest proportion of Russians in the new elite. How then is the ethnic composition among the elite groups?

In the Estonian parliament, no deputies in the sample were Russians, but in the parliament as a whole, two non-Estonians out of 100 deputies were elected in 1992. After the parliamentary election in March 1995, the number of non-indigenous deputies increased to six, mainly Russians. In parties and movements, as many as 14 percent are Russians. In Latvia, the situation is

almost the opposite with a substantial group of Russians in the first parliament. The official record of the first parliament elected in March 1990 shows that of 201 representatives, 69 percent were Latvians, 22 percent Russians, 6 percent Ukrainians and Belorussians, and 3 percent others. However, few Russians are leaders of political parties and movements. The next parliament elected in June 1993 had 12 percent non-indigenous deputies out of 100 representatives, and after the 1995 parliamentary election, the proportion of other ethnic groups had decreased to 9 percent. In other words, the proportion of indigenous Latvian representatives increased from 69 percent in 1991, to 88 percent in 1993, and to 91 percent in 1995. Political consolidation seems to imply increasing indigenous control over political institutions.

Lithuania has a relatively high proportion of Polish representatives in its parliament and parties/movements, and very few Russians. Out of 141 deputies in the Lithuanian parliament after the 1992 election, there was one Jew, three Russians, and seven Poles, amounting to 8 percent of the representatives. For all countries, there is a very good match between the actual distribution and data from this investigation of ethnic proportions in the parliament.

From a democratic representative point of view, the situation is special. The large Russian minority in the population has few ethnic spokesmen among the national elites, although as we have just seen, in representative institutions such as parliament and political parties/movements, the situation is somewhat different.

The national ethnic control of the state bureaucracy is overwhelming. Practically no Russians have top positions in the state administration. The idea of a bureaucracy representing several population groups (Lægreid & Olsen 1978) has no validity. The figures indicate a policy of protecting the core activities of the state from influence by other groups. As a strategy for nation building and promoting national interests, this is rational. The longterm effects on the decision making process and conflict solution of closing out decision makers from other groups still remain to be seen. The same applies to the judicial system. In Estonia and Lithuania, there are very few non-indigenous people represented in the judiciary. The core institutions of the state, the executive, the legislative and the judiciary powers, are well protected from minority influence.

In Estonia and Latvia, large state enterprises and banks have traditionally been dominated by Russian leaders. The Russian working population is mainly employed in huge factory plants, and the process of privatization has been hampered due to old nomenclature interests in preserving these industries (Andersen et. al. 1993). Still, by 1993/94 there was a considerable number of Russian industrial leaders in these two countries, while this study found none in Lithuania.

Several observers have noted the influx of Russians, mostly former
nomenclature persons, into private business (Lieven 1993). This is also
reflected in our data, particularly in Latvia.

Among intellectuals, the data indicates a larger proportion of Russians in Estonia and Latvia. The higher percentage in Estonia is somewhat surprising, and may be explained by national circumstances. Also among municipal elites in Estonia, a high proportion is Russian.

In Estonia, the Russian elite was removed from central state institutions like parliament, bureaucracy and judiciary, but to some extent they are still active in other institutions. In Lithuania, very few Russians remain in top positions, while the Poles, relative to their population share, are well represented. Latvia is somewhere in between. One feature of all the Baltic countries is that there are almost no Russians in the top state bureaucracy and in the judiciary, i.e. the core institutions of the state. Consolidating core institutions by national expertise, even if most of them have a communist background, seems to have been an important strategy.

The new democracies reveal a special pattern of post-communist pacted transition. Former Russian elites are needed, but they are politically pacified because they are only allowed to keep their positions in institutions under national control, such as state enterprises, or to move into private business activities. Here they can contribute to market development and maintain or improve their income and social status, thereby compensating for the loss of political influence. When Russian leaders become more involved in business than politics, the scope for political mobilization among ordinary Russians is reduced, which in the short run may stabilize the political process. What will happen in the long run, when large proportions of the population are not represented by their ethnic leaders in the state institutions is still to be seen.

The Young Career Elite

Even if a major part of the present Baltic elites prior to the regime change were members of the Communist Party or the Komsomol, and many also had leader positions, they can be characterized as a new elite. An important reason why they have become a new elite is that they belonged to a younger generation than the pre-independence power elite. The young generation was an important force behind the changes in communist countries. They were well educated, well informed urbanites with fewer psychological and social inhibitions than their parents' generation (Higley & Pakulski 1992). There is no available data about age structures in the Baltic elites before independence. In the former Soviet Union, the average age among the elites was high (Harasymiw 1984). In 1917, the average age in the Politburo of the Soviet Communist Party was 39, and it increased gradually to 58 in

1947 (Schueller 1965). In the Central Committee of the Soviet Communist Party, the average age was 67 in 1981, gradually decreasing to 61 in 1986, and to 56 in 1991 just before the breakdown of the Soviet Union (Lane and Ross 1995). The dissolution of the Soviet regime was accompanied by a remarkable generational shift starting in the early 1980s.

The age structure of the new Baltic elites indicates a generation change. The average age of the bureaucratic elite in Estonia is 43 years, and 46 in Latvia and Lithuania. The percentage of "younger" top bureaucrats, i.e. between 22-44 years of age, is 52 in Estonia, 50 in Lithuania, and somewhat lower at 37 in Latvia. Putnam (1975) reports that 10 percent of the senior civil servants in Britain were between 26-47 years, in West Germany 33 percent, in Sweden 32 percent, in the US 37 percent, and in Italy no senior civil servant was younger than 46 years.

The West German elite study from 1981 (Hoffman-Lange 1987) found an average age of 53 years for nine elite groups (political, administrative, business, interest groups, etc.). The average age for all elite groups is 45 years in Estonia, 46 years in Latvia, and 48 years in Lithuania. However, recruitment of relatively young persons, many of them starting their career under the former system, has been an important element in the regime change process. One consequence is that the Baltic elites are mostly younger than their colleagues in Western countries.

Conclusion: Elite Recirculation and Change

According to the classical theories of Mosca (1896) and Pareto (1901), elite changes are not abrupt. Elites have a common interest in gradual change by "circulation" in order to safeguard the power of "the ruling class." O'Donnell and Schmitter (1986) and Huntington (1991) have argued that a change from authoritarian to democratic rule implies a large degree of elite continuity. Our study supports these theories, but only partly and on a rather abstract level. Since change can take place along several dimensions, in this study specified to membership in the Communist Party, occupational position, ethnic background and age, the conclusion is that a combination of continuity and change gives the best description of the Baltic elites. Going back to our hypotheses, they seem to get considerable support and can be summed up in two main conclusions.

(1) While the nomenclature was largely removed from power, the younger, well-educated, mid-level leaders from the former regime are continuing and are now occupying most of the top positions. A considerable part of the elite who came into position after the regime change is recruited among former CP members, but mostly they are not from the tainted top nomenclature from the Soviet period. They are the ambitious young

generation who took their first steps in the Soviet hierarchy, and who after
independence saw the opportunity to improve their position by stressing the
value of an independent nation state.

There was next to no alternative national elite outside the party apparatus. Because most members of the young and ambitious generation had joined the party in order to make a career, the "alternative elite" could be found among the younger strata of earlier party members and high level "technocrats." The main criteria for belonging to the "alternative elite" seem to be an indigenous ethnic background and professional competence. Common to this group of persons were new career opportunities, nationalist orientations, and the desire for indigenous control of important state institutions. The exclusion of Russians from elite positions was therefore not only important as a strategy for nation building, but also facilitated recirculation and career opportunities of the ambitious and nationally oriented indigenous elites. Since there was no alternative elite outside the earlier power system, the new ruling elite had to be recruited from within the existing structure, and here the young generation saw its chance. In this sense, elite recirculation occurred mainly as upward mobility, and only to a smaller extent as recirculation of the old nomenclature.

(2) The ethnic composition among the elites has changed dramatically. The ethnic structure of the society seems to influence the composition of the elites directly but contrary to what should be inferred from normative theories of democratic representativeness. Indigenous elites almost completely dominate the power institutions in countries with large minority groups. The regime change meant not only the breakdown of totalitarianism, but also the break from foreign rule and Russian cultural domination. Therefore, the Estonian and Latvian elites were recruited along ethnic lines among younger persons with minor responsibilities during the former regime.

Why continuity? Here we have proposed a "state need" explanation where the expertise of the former elite is indispensable for the continuous running of the state and not least as leaders of new institutions in the state building process. We have also proposed a supplementary "individual resource" explanation where the young generation's capability to utilize networks and contacts is vital for success in the new elite system.

In comparison, the three countries exhibit many similarities in elite patterns. In Lithuania, however, the recruitment pattern deviates to some extent, with fewer earlier communists recruited into the new elite system. The communist elites had a more clearly defined national platform, and the process of independence began earlier with organized opposition against Moscow. Therefore, we might expect that former communists were largely accepted in the new elite system. One explanation of the relatively modest elite continuity is that the more pronounced left-right political cleavage

stimulated a more critical focus on the past than in the other two countries. The reformed national communists returned to power, and this event consolidated the position of some nomenclature persons, but it also stimulated the right wing opposition that claimed a break with the past and thereby indirectly influenced the recruitment process.

The ethnic cleavage in Estonia and Latvia with the internal pressure from a large Russian population was the key issue that led to exclusion of Russians from elite positions. Indigenous and nationally oriented young people, many of whom had been recruited recently into the former system through the Communist Party and Komsomol, were ready to fill the positions. Why Estonia has gone further than Latvia in excluding Russophones from core state institutions will not be discussed in detail here. As shown in Steen (1996), it is consistent with more skepticism among Estonian leaders towards Russians, less integration with Russians during the Soviet period, specific historic experiences related to Russian domination, and cultural peculiarities.

The general conclusion is that the issue at stake produces specific elite patterns. The more politicized socio-economic cleavage in Lithuania led to more confrontations between left-right opponents, thus creating rather fragmented elite configurations. In Estonia and Latvia, ethnicity as the main political cleavage resulted in a more integrated elite pattern, not only in terms of ethnic background but also in terms of more interconnected behavior and attitudes (Steen 1996). The integrated indigenous elite pattern in these two countries can be explained as a response to the insecurity that arises when large minority groups are perceived as a threat to national survival.

Several theorists have emphasized the relationship between elite structure and regime stability. In pluralized states, elites take on an accommodating function for social forces like economic development, class formations, ethnicity and religion (Higley and Burton 1989). What will be the "filtering function" of elites that are mostly recirculated from the previous system and who do not represent large minority groups? Returning to Linz's (1990) introductory question about consequences of elite change for political stability, it is clear that the Baltic regimes have gone through a process of national consolidation by recirculating indigenous and competent leaders. The long-term consequences of an elite structure without ethnic pluralism are, however, still to be seen. One possibility is stable political development if minority needs are taken into account by the ruling elites. Conventional wisdom would predict instability and confrontation between ethnic groups when social groups are not represented in the decision making structure. Hence the most stable development would come from restructuring the elites by recruiting representatives for minority groups into influential elite positions. But this has to be a gradual process. Baltic nationalists will probably

react with intensity against Russians in power positions. The "specialpacted" indigenous solution in the Baltic states has the potential for confrontation both with nationalists and Baltic Russians. The basic dilemma, especially in Estonia and Latvia, is that including Russians into the elite structure is seen as a real threat to national culture and independence. The collective trauma of fifty years of occupation and Russification is probably the best explanation of why it is difficult for the elite structure to mirror ethnic realities. However, whether elite exclusiveness is a rational response to traumatic experiences of the past remains to be seen.

---

NOTES

1. Fanner (1992) found that in particular Latvian and to some extent Estonian nationalities were overrepresented in the Central Committee of the Communist Party of the Soviet Union.

2. We have obtained data about former positions in the Communist Party in Lithuania. 14 percent (22 respondents) of those who say they have been members, say they have held a higher position in the Party.

3. The data for Latvia is for the first elected parliament, and for Estonia and Lithuania the second elected parliaments.

4. The more detailed table for elite groups is not shown here.

ACKNOWLEDGEMENTS

I would like to acknowledge useful comments to an earlier version from Knut Heidar and Tore
Nyhamar. The earlier version of this article was presented at the workshop "National Political
Elites and European Integration," ECPR, Madrid, April 1994.

REFERENCES

Andersen, G. 8., Thuesen, L. & Willumsen, H. 1993. "En komparativ studie af de økonomiske
reformer i de baltiske lande," Politica 4, 399^16.

Dahl, R. A. 1961. Who Governs? Democracy and Power in an American City. New Haven:
Yale University Press.

Farmer, K. C. 1992. The Soviet Administrative Elite. New York: Praeger.

Harasymiw, B. 1984. Political Elite Recruitment in the Soviet Union. London: MacMillan
Press.

Heinz, J. P., et.al. 1990. "Inner Circles or Hollow Cores? Elite Networks in National Policy
Systems," Journal of Politics 52, 356-390.

Highley, J., et. al. 1991. "Elite integration in stable democracies: a reconsideration," European
Sociological Review 7: 1, 35-53.

Higley, J. & Burton, M. G. 1989. "The Elite Variable in Democratic Transitions and
Breakdowns," American Sociological Review 54, 17-32.

Higley, J. & Pakulski, J. 1992. "Elite Transformations in Eastern Europe and Russia," Paper,
XVI World Congress of IPSA, Berlin.

Hoffmann-Lange. 1987. "Surveying national elites in the Federal Republic of Germany," in
Moyser, G. & Wagstaffe, M., eds, Research Methods for Elite Studies. London: Allen &
Unwin.

Huntington, S. 1991. The Third Wave: Democratization in the Late Twentieth Century.
University of Oklahoma Press.

Keller, S. 1972. "Elites," in International Encyclopedia of the Social Sciences, vol. 5,
Macmillan Company London.

Lane, D. & Ross, C. 1995. "The CPSU Ruling Elite 1981-1991: Commonalities and
Divisions," Communist and Post-Communist Studies 28: 3, 339-360.

Levits, E. 1987. "National Elites and their Political Function within the Soviet System: The
Latvian Elite," The Journal of Baltic Studies XVIII: 2, 176-190.

Lieven, A. 1993. The Baltic Revolution. Estonia, Latvia, Lithuania and the Path to
Independence. New Haven: Yale University Press.

Linz, J. J. 1975. "Totalitarian and Authoritarian Regimes," in Greenstein, F. I. & Nelson, W.
P., eds, Handbook of Political Science. Vol 3. Reading: Addison-Wosley.

Linz, J. J. 1990. "Transitions to Democracy," The Washington Quarterly, Summer, 143-164.

Mosca, G. 1896. The Ruling Class. Translated and revised by Kahn, H. D. & Livingston, A.
1939. New York: McGraw Hill.

Nørgaard, O. 1992. "The Political Economy of Transition in Post-Socialist Systems: the Case
of the Baltic States," Scandinavian Political Studies 15: 1, 41-61.

O'Donnell, G. & Schmitter, P. C. 1986. Transitions from Authoritarian Rule. Tentative
Conclusions about Uncertain Democracies. Baltimore: Johns Hopkins.

Pareto, V. 1901. The Rise and Fall of the Elites. An Application of Theoretical Sociology.
English version. 1991. With an introduction of Hans L. Zetterberg. New Brunswick:
Transaction Publishers.

Putnam, R. 1976. The Comparative Study of Political Elites. New Jersey: Prentice Hall.

Putnam, R. D. 1975. "The Political Attitudes of Senior Civil Servants in Britain, Germany and
Italy," in Dogan, M., ed., The Mandarins of Western Europe. The Political Role of Top Civil
Servants. New York: John Wiley & Sons.

Silde, A. 1987. "The Role of Russian-Latvians in the Sovietization of Latvia," The Journal of
Baltic Studies XVIII: 2, 191-198.

Schattschneider, E. E. 1960. The Semisovereign People. New York: Rinehart & Winston.

Schueller, G. K. 1965. "The Politburo," in World Revolutionary Elites. Studies in Coercive
Ideological Movements. Lasswell, H. D. & Lerner, D., eds, Cambridge: MIT Press.

Steen, A. 1996. Elites, Democracy and Policy Development in Post-Communist States. A
Comparative Study of Estonia, Latvia and Lithuania. Research Report 02/96. University of
Oslo: Department of Political Science.

Taagepera, R. 1993. Estonia. Return to Independence. Boulder: Westview Press.

Zhang, B. 1994. "Corporatism, Totalitarianism and Transitions to Democracy," Comparative
Political Studies 27: 1, 108-136.