Marked differences in socio-economic background, ethnicity, and religion, largely, but not exclusively, between west and east

While the Soviet form of politics, economics, and social life was applied broadly speaking uniformly across the 11 countries examined here, major differences, often inherited from the past, did remain. Variations are large in the annual per capita gross domestic product of these countries: around 2007–8 they ranged from 14,400 in Russia and 10,740 in Belarus to under 2,000 in Kyrgyzstan (1,950) and Tajikistan (1,710) (these and the following are in purchasing power parity (PPP) in US dollar equivalent). The contrast between the more Western members of the group and the Central Asian group of countries is not total, however. Ukraine is only fourth in the range; with an annual per capita GDP of US$6,810, it is behind Kazakhstan, which has a GDP per capita of US$9,700. While Uzbekistan has a low GDP per capita of US$2,400, Turkmenistan's GDP per capita was US$4,350, not much below Georgia, which had a GDP per capita of US$4,770. The proportion of the active population in agriculture follows a similar pattern: it is highest in Kyrgyzstan (34%), Uzbekistan (24%), and Tajikistan (22%), and lowest in Russia (6%), Ukraine (7%), and Belarus (9%); but, by being as low as it is in Russia (6%), the percentage of the Kazakhstan population engaged in agriculture is in sharp contrast with that of the other four Central Asian countries (Table 1).

Table 1. General characteristics of countries of the CIS

This situation is partly the consequence of the population movements that have taken place since the fall of the Soviet Union: Russia, Belarus, and especially Ukraine have lost population. In Ukraine the population fell by 10% (from 51 million in 1989 to 46 million in 2008), in Belarus by 5% (from 10.1 to 9.7 million), and in Russia by 3% (from 147 to 142 million). On the other hand, except in Kazakhstan, where it declined from 16.5 to 15.8 million), the population grew in the Central Asian states, especially in Uzbekistan, where it almost doubled during the period (from 15.4 to 27.2 million).

The major socio-political problems that emerged after the break-up of the Soviet Union did not have a merely economic origin however, they were connected to the ethnic and/or religious background of the newly independent post-Soviet states. Admittedly, the ethnic composition of each state varies, each having a majority ethnic group, except Kazakhstan where the Kazakhs are scarcely half the population (53%) while the Russian minority constitutes 30% of the population. Elsewhere, the major ethnic group constitutes 60% to 80% of the population; indeed, in Armenia, Armenians are 93% and, in Azerbaijan, Azeris are 91%.

Especially between these two countries, but also at the borders between the Central Asian states and Iran or Afghanistan, the religious cleavage has added to tensions – and indeed war – by being combined with the ethnic cleavage. There is, as expected, orthodox supremacy in Russia, Belarus, and Ukraine, though less so in the last two countries, as Catholics are a sizeable proportion of the population in the western areas in both cases. Meanwhile, Islam, primarily of the Sunni variety, dominates in four of the five Central Asian states, the exception being Kazakhstan, where the large proportion of Russians accounts for a substantial Orthodox minority. Azerbaijan is also predominantly Muslim, while the other two Transcaucasian republics are primarily Christian but not Orthodox.

Major political conflicts and indeed outright wars

As a matter of fact, neither the ethnic cleavage nor the religious distinctions fully account for the amount of violence prevailing in the post-Soviet states. Large-scale demonstrations taking place over a substantial period, if not almost continuously, as well as assassinations, are endemic in some of the states. This is the case, for instance, in the three Transcaucasian countries, even in Armenia, despite the fact that, by and large, internal politics in that country has followed many of the rules of liberal democracies. Meanwhile, neither of the first two presidents of Georgia left office in a ‘regular manner’ in the first 20 years after independence, and the stability of Azeri politics since the mid-1990s has been largely based on the clientelistic and patrimonial rule of the father and son who succeeded each other as presidents.

Moreover, the main source of conflicts was not internal to each of the new states, except in Georgia (with respect to the Abkhazian and South Ossetian minorities) and in Ukraine (with respect to the division between the more ‘Russian’ east and the more ‘Ukrainian’ west of the country). By far the most dramatic conflicts were among some of these states or at the border between these states and non-Soviet countries. Armenia and Azerbaijan were the one instance in which war occurred between two ex-Soviet Union states; when outright war stopped, tension remained as no solution acceptable to both sides could be found for the status of the Armenian Nagorno-Karabakh enclave in Azerbaijan.Footnote ⁵ Meanwhile, ‘regular’ politics experienced great difficulty in much of Central Asia, and above all in Tajikistan, where civil war took place in the early 1990s and where the leaders of neighbouring Turkmenistan and Uzbekistan, jointly with Russia, took hard ‘precautionary’ measures which affected political life in the country.

Presidentialism comes to all 11 countries, but with variations in method and timing

Despite differences in the socio-economic fabric of the 11 countries, however, ‘presidentialism’ was adopted everywhere, albeit with variations in the extent to which authoritarianism prevailed. The timing and method of the introduction of presidential government also differed. The title of ‘executive president’ had already been adopted while the Soviet Union still existed for the new position held by Gorbachev in May 1989, as Gorbachev and his immediate entourage wanted to dissociate themselves from the Communist Party: the position of ‘executive president’ was not filled by popular election, however, but by the ‘Congress of the People's Deputies’. Election by universal suffrage came two years later, in 1991.

Three different moves towards a fully fledged ‘presidency’ took place in the 11 ex-Soviet republics. In four countries, Kazakhstan, Armenia, Azerbaijan, and Russia, the appointment of a president, first approved in 1990 by the Supreme Soviet, was ratified by the people in 1991, in Russia in June, Azerbaijan in August, Armenia in October, and Kazakhstan in December, indeed before that country had even proclaimed its independence.

In four other countries, three Central Asian states, Uzbekistan, Kyrgyzstan, and Turkmenistan, and Ukraine, an ‘executive president’ was first appointed, a move which occurred in 1990 in the first three of these cases and in July 1991 in Ukraine. This arrangement seemed designed to placate those more accustomed to the idea of a ‘chairman’ or ‘party secretary’. The adjective ‘executive’ was abandoned when ratification by the people took place, in October or December 1991 for three of the countries concerned and in June 1992 in the case of Turkmenistan.

The last three countries, Georgia, Belarus, and Tajikistan, had a more tortuous evolution towards presidential government. Georgia's Supreme Soviet appointed Gamsakhurdia as its chairman in November 1990; the position of ‘executive president’ was then set up and Gamsakhurdia was elected by the people with 87% of the votes in May 1991. There was then armed conflict against him as a result of which he fled the country and was deposed in January 1992; the position of president was then abolished. It was re-established three years later in 1995 after Shevardnadze had returned to run Georgia, where he had been secretary of the Communist Party under Soviet rule and had been popular; yet he was first ‘merely’ Chairman of the Supreme Council and became president only when a fully presidential constitution, with no position for a prime minister, was adopted.

In Belarus, the position of Chairman of the Supreme Soviet was maintained up to March 1994; but the case for ‘presidentialism’ had by then gained ground and the 1994 constitution embodied the principle. A popular election was organized in July of that year to fill the position: the candidate who won was an outsider, Lukashenka, who had acquired some popularity for having set up in 1991 an organization known as Communists for Democracy. The newly elected president was indeed to benefit markedly from the new arrangement!

In Tajikistan, ‘presidentialism’ went through a difficult early phase. The First Secretary of the Communist Party, Makhkamov, was elected ‘executive president’ by the Supreme Soviet of the state in August 1990: he was forced to resign a year later, as he had not been against the August 1991 coup in Moscow. His successor, Nabiyev, was elected by popular suffrage in November 1991, but he also had to resign a year later after violent demonstrations – indeed truly a civil war – had started. The office of president was then abolished in November 1992, and the new leader, Rahmon, who was to rule the country for the next two decades, indeed as president, was originally elected as Chairman of the Supreme Council. The first presidential election took place only two years later, in November 1994: Rahmon won, but by ‘only’ 58% of the popular vote: he was to do much better afterwards!

The presidents of the 11 post-Soviet republics

Twenty-eight men (no woman was appointed during the period) occupied the post of president between 1990 and 2009, admittedly sometimes after a period as ‘Chairman of the Supreme Soviet’, as we just saw.Footnote ⁶ In two countries, Kazakhstan and Uzbekistan, the same person ruled throughout the whole period, Nazarbayev in the first case, Karimov in the second. In two countries, Kyrgyzstan and Turkmenistan, two people became president successively, the first of them having been in office about 15 years, Akayev in Kyrgyzstan and Niyasov in Turkmenistan: Akayev was ousted by major demonstrations and fled into exile in Russia in 2005,Footnote ⁷ while Niyasov died in 2006. In six countries, Armenia, Belarus, Georgia, Russia, Tajikistan, and Ukraine, three people were successively president: only in Armenia, Russia, and Ukraine was the duration of at least two of these presidents relatively normal (about eight years) – on the assumption that Medvedev is considered the ‘real’ president of Russia. In the other three countries, Belarus and Tajikistan in particular, and in Georgia, but less so, one president was in office for a long period – ten years or more – and the other two were present for a relatively short period. Finally, in Azerbaijan, there were four presidents successively, the last two being Aliyev senior and junior; as he was about to die, the father was replaced by his son, who was elected in 2003.

The story of presidentialism in the 11 post-Soviet republics is thus varied and complex; however, it has been characterized across the area by one fundamental element, the election of the leader by universal suffrage. The fact that Gorbachev's legitimacy and power remained limited as he had not been ‘anointed’ in this way surely had a major part to play in this respect. Yeltsin realized that his legitimacy and his power depended on such an ‘anointment’, and it duly took place in June 1991, before the move had occurred in the other republics, with the exception of Georgia, where the first presidential election had taken place a month earlier. As Yeltsin had succeeded in his gamble to acquire legitimacy and power by popular election, first the majority of the other countries (seven), and eventually all of them, followed the model.

Yet the specific nature of presidentialism in each of the post-Soviet republics was closely adapted to the circumstances of the various countries which emerged out of the USSR. The case of Russia – with Yeltsin, Putin, and perhaps Medvedev – constitutes what might be regarded as a ‘middle-of-the-road’ arrangement, since, to adopt the expression of Lydia Shevtsova, the country has been ‘lost in transition’ since the end of communism.Footnote ⁸ Presidents have been more powerful in the Central Asian republics because the experience lived through by these countries in the two decades after 1990 was different from that of Russia, and, in the case of Tajikistan, traumatic. Meanwhile, at the other end of the spectrum, successive Ukrainian presidents appeared obliged to follow a tortuous path between ‘east’ and ‘west’ without being able to find a solution which would simultaneously satisfy both extremes of the divide. Similar difficulties occurred in Armenia and in Azerbaijan, even if the political processes in the two countries were rather different. Thus, presidentialism, ‘post-Soviet style’, has fitted all kinds of political situations, and the object of the last two sections of this paper is to explore these variations. Before doing so, however, one must first return to the origins of presidential government and consider both the American presidency and what followed, and thus place post-Soviet presidentialism within the worldwide context of the development of that form of government.

The presidency: an ‘invention’ of the American Constitution

There was never a ‘presidency’ in any state before the Founders of the United States created that position in the 1780s, having decided (after some discussion) that they did not want a monarchy. Admittedly, ‘regular’ monarchical systems had not been the only form of government in existence: there had been ‘usurpers’ who overthrew the monarch who was in place, although those who began as ‘usurpers’ often then adopted a monarchical title. There had also been republics, typically in very small states, unless one went as far back as the Roman Republic. Yet whether in ‘ancient Rome’ or elsewhere, republics did not have ‘a’ ‘president’, that is to say (a) a single leader, (b) appointed for a truly substantial period to be in charge of steering the ‘ship of state’. The Roman Republic had two consuls and a number of other (elected) officials; all of these were appointed for short periods (a year) and typically could not stand again, at any rate not immediately. The republics established later, especially in Italy, were based on the same principles of multiple appointments and quick rotation: Florence was a clear case in point. Venice had a different system, but the overall idea was similar, namely that a republic could not function properly if it was in the hands of one man for a substantial period.

The Founders of the American Constitution had to overcome this challenge. They overcame it by deciding that there would be a single president to be elected by the people (at least indirectly) for a substantial period (four years, but indefinite re-election was allowed).Footnote ⁹ Thus, the American Constitution went as far as possible to retain the benefits of the monarchical system (namely continuity) while rendering change possible, avoiding the risk run by the monarchical system that because it did not allow for change, a ‘usurper’ might emerge.

The path was narrow, however. The president was to be in office long enough to have an impact on the polity and give strength to the ‘new nation’ in the process:Footnote ¹⁰ the corollary was, however, that he might want to retain power by unconstitutional means. The gamble succeeded in the United States, but rarely elsewhere, whether in Spanish America from the 1820s or in France in 1848, as the success of the ‘invention’ required genuine agreement by the president to abide by the constitution and to rule on the basis of a wholly new arrangement, which might have been called ‘popular leadership’ but which has typically been referred to simply as leadership.Footnote ¹¹ What was needed were new attitudes and new modes of behaviour among both the leaders and the led.

The President is central to ‘presidentialism’, not the legislature or Congress

The setting up of the presidency in the American Constitution is not just a key ‘innovation’ but the key innovation of the presidential system. Shugart and Carey state in Presidents and Assemblies that the ‘spirit’ of presidentialism is to be found in the coexistence of the two ‘powers’, both autonomous and strong, the executive power and the legislative power. Yet in practice the ‘invention’ brought about by presidentialism has not been extended to the legislature or the separation of powers: parliamentarism was to bring about these developments more smoothly and more effectively. What has truly characterized presidentialism over the last two centuries has been the existence of a direct link between president and ‘people’.Footnote ¹²

‘Presidentialism’ spread primarily to ‘new’ countries

‘Presidentialism’ has spread primarily to new countries, and to begin with, in the early part of the nineteenth century, in the newly independent colonies of Spanish America, since, unlike Brazil, these states did not have a monarch at their disposalFootnote ¹³ (the ill-fated Mexican episode of the 1860s was indeed to show that monarchical rule was simply out of the question). But the American gamble did not succeed in the region, as ruthless presidents often came to power and takeovers by coups (especially by the military), as well as interventions by ‘caudillos’, occurred frequently. In Europe, the attempt to introduce ‘presidential’ government was also ill-fated: the Second French Republic was presidential; it was replaced after only four years by the Second Empire of Napoleon III in 1852. Meanwhile, the other countries of the area followed Britain in moving gradually from constitutional to ‘parliamentary’ monarchies. ‘Presidentialism’ was indeed viewed as less satisfactory than parliamentarism: the Third French Republic which emerged in the 1870s was also the first ‘parliamentary’ republic in the world.

Except for a few attempts, after World War I, to introduce forms of presidential government in Central and Eastern Europe, all of which collapsed with World War II, it was the 1960s before the end of colonization, in Africa in particular, led to a second and very large wave of presidential government, the parliamentary governments often introduced by departing colonial powers having rarely lasted, especially in Africa.Footnote ¹⁴ Parliamentary governments remained only in some Commonwealth countries, above all in India, but also in the Caribbean. Thus, the 11 post-Soviet presidencies which emerged in the 1990s constitute the third phase of the spread of a type of government which had been widely adopted by the many new states of the second half of the twentieth century. While parliamentary government appeared complex to organize, presidential government seemed to provide a clear-cut and straightforward solution to problems of legitimacy and of nation-building in new countries.

‘Presidential’ systems and ‘republics with presidents’ in the contemporary world

A survey of regimes in existence in 2009 shows the extent to which ‘presidentialism’ has come to prevail at the beginning of the twenty-first century; but it also shows the marked variety and therefore complexity of what is entailed by ‘presidentialism’. Out of 180 countries with a population of over 100,000, 37 were monarchical, 29 of them being ‘constitutional’; these were either European or mainly drawn from the Commonwealth in other parts of the world. Thus, 143 states were republics, but eight of these did not formally have a president: four were Communist states (China, Vietnam, Laos, Cuba), one had a leader without any formal appointment (Libya), one was still constitutionally incomplete (Nepal), and two were highly decentralized (Switzerland and Bosnia-Herzegovina). So 135 republics were headed by a president.

The variety of the arrangements shows itself first in the nature of the appointment mechanism. In a large majority (103) of these 135 countries, the president is elected by universal suffrage, but in the substantial minority of 32 countries the president is appointed otherwise, typically by the legislature. That division does not coincide closely with the distinction between ‘presidential’ and ‘parliamentary’ republics, however. On the one hand, 11 of the 103 presidents elected by universal suffrage are the nominal heads of ‘parliamentary’ republics, Ireland and Austria for instance, and on the other, 14 of the 32 presidents not elected by universal suffrage are strong presidents, as in Moldova, the Democratic Republic of Congo, Ethiopia, Laos, or Surinam.Footnote ¹⁵ Putting it differently, only in 18 of the 29 ‘parliamentary republics’ is the president not elected by universal suffrage: presidents elected by universal suffrage are thus a substantial minority in the group. Conversely, in a small group of 14 countries in which the president is not elected by universal suffrage, ‘presidential’ government obtains in view of the part played by the head of state in political decision-making. This state of affairs does not arise among the republics stemming from the former Soviet Union, but one cannot state categorically that, worldwide, ‘presidentialism’ exists only where the head of state is popularly elected. One must therefore adopt a somewhat more flexible definition of ‘presidentialism’ in relation to the mode of election of the head of state.

Moreover, if ‘presidential government’ prevails among contemporary republics, this is not true of ‘limited’ executives faced with strong and separate legislatures. There are at most 60 ‘presidential’ countries in which the head of state is elected by universal suffrage and in which the constitution does not stipulate whether the executive is in any way connected to the legislature. Among these 60, furthermore, only for 33 is the United States ‘the’ model, a model often followed only in theory, and in practice adopted mainly in the 17 Latin American countries which belong to that group. In the other 27 countries belonging to the 60, there is a prime minister, unlike in the United States, presumably rendering the government more compact than where there is no prime minister; 22 of these countries are found in Africa: the constitutional relationship between executive and legislature is left unclear.

Meanwhile, in the other 32 countries in which a strong president is elected by universal suffrage, a link – not a separation – between executive and legislature is specified by the constitution: either the government is subjected to a confidence or censure vote and/or, vice versa, the president may dissolve the legislature. Countries which are typically described as ‘semi-presidential’ fall into this category, but these make up less than half of the group.Footnote ¹⁶ Nine of the ex-Soviet republics and nine African countries are in that group. In these systems, not only is there no ‘separation’ between executive and legislature, but whether the executive or the legislature exercises the greater power is likely to vary appreciably. The American blueprint of ‘presidential government’ has thus become profoundly modified, not just away from liberal democracy, as in the first wave of new states from Latin America in the early nineteenth century, but from the very idea of a separation of powers, with the second wave of presidencies in Africa and elsewhere and with the third wave in the ex-Soviet republics (Table 2).

Table 2. Presidential systems across the world in 2009

The need for presidents to acquire legitimacy in the context of nation-building

‘Presidentialism’ thus provided, accidentally to some extent, the ‘modern’ instrument that made it possible to challenge the established colonial order. The American ‘invention’ was adopted in other countries since it appeared that the sheer presence of a president would help to maintain in existence a new nation, a new state, and perhaps also, but only occasionally, a new regime.

For such a development to have the effect of creating a stable new polity, something else was required, however, for the president to be able to maintain a close link with the people, and at the limit to embody, in the eyes of the people, the nation the presidency had helped to create or sustain. This meant that the president had to be the ‘midwife’ enabling a new legitimacy to be born, and the protector who would sustain the continuity of the political entity that had emerged: the unwritten understanding was that the needed legitimacy would exist if the bulk of the population was not merely passive, but actively recognized that the president was helping the nation to flourish. There had to be marked support for the president, as the president was effectively guaranteeing the future of the ‘new polity’.

It may be felt that the problems of nation-building and even of building legitimacy apply only marginally to the 11 post-Soviet republics that emerged in 1991. Yet there are marked similarities with the ‘decolonized’ ‘new’ nations of Latin America and of the ‘Third World’. Russia is an ‘old’ state, enjoying a high level of legitimacy among most of its population, admittedly, but there are serious problems in some of its peripheries. Armenia has a strong sense of nationhood, but events related to Nagorno-Karabakh suggest that this sense of nationhood needs to be buttressed. Georgians tend to have great pride in their nation, but this led to a somewhat unrealistic determination to control those parts of the country that had defected, Abkhazia and South Ossetia. In the other ‘new’ republics, the problems are often very large, even in Ukraine, as a result of the east–west divide, let alone in Azerbaijan, Central Asia, and indeed Belarus.

The part the presidents of post-Soviet republics have played towards nation-building and building legitimacy must therefore be examined. Some of these presidents may not have had any chance to grasp, let alone seize fully the opportunities the political system offered them, in part because they did not have the time to act, in part because circumstances may have drastically limited their ability to do so. As in all cases involving leadership, there has to be an adequate combination of ‘situation’ and ‘personality’. It is to the nature of such combinations that we need to turn in the next section, before considering, in the final section, the extent to which the actions of these presidents did fit, fitted in part, or did not fit at all the two requirements: of a liberal-democratic polity and of effectiveness, in effect of popular leadership.

The duration of presidents in office in post-Soviet republics

We saw in the first section that the 28 men who, at some point, led one of the 11 post-Soviet republics were confronted with markedly different situations. We also saw that a number of them remained in office for many years, in two cases even throughout the whole period between 1990 and 2010. This is despite the fact that all the constitutions adopted by the post-Soviet republics limited duration to two mandates (at any rate originally) of four or five years and exceptionally of seven years. The constitutional rule was overridden in all five Central Asian states, but also in Belarus and Azerbaijan, by special referendums, typically adopted some years in advance of the constitutional end of the tenure of the president concerned.Footnote ¹⁷

All seven short-term presidents were in office during the early period

A quarter of the 28 presidents (seven) were appointed in the early 1990s but remained in office for only one or two years. These early short-term presidents came from only four countries. Two were from Belarus: Shushkevitch and Hryb (who lasted only two months); a popular election took place in 1994 as a result of which Lukashenka was elected at the second ballot. There were also two short-term presidents from Tajikistan: Makhkamov and Nabiyev. One, Gamsakhurdia, came from Georgia and he fled the country in January 1992 and was quickly dismissed. The last two cases of short-term presidents are from Azerbaijan: Mutalibov between August 1991 and March 1992, when he resigned, and Elchibey between June 1992 and August 1993, when he left Baku. There were thus short-term presidents in the three countries seen in the first section above to have been less ‘enthusiastic’ about the formal move from a ‘chairmanship’ to a ‘presidency’. Azerbaijan is the fourth case: it took time for Heydar Aliyev to head the country again. He had previously been party secretary, but had left Baku to become first deputy chairman of the USSR Council of Ministers, a post from which he was dismissed for corruption in 1987. He returned to Azerbaijan and seemed to have been behind some of the opposition to the previous two short-term presidents of the country, Mutalibov and Elchibey; the latter, however, being unable to handle the war with Armenia, asked Aliyev to join the government in 1992 and left Baku, only to be formally dismissed from the presidency by 97% of the voters in a referendum in August 1993.

The long-term duration of eight of the 28 presidents

Some of the presidents appointed in the early period were thus short-term. At the other extreme, eight presidents stayed in office markedly longer than their ‘normal’ two-term mandate allowed. Four of these eight were still in office in 2010 after having been president 15 years or more: these are Nazarbayev of Kazakhstan, Karimov of Uzbekistan, Rahmon of Tajikistan, and Lukashenka of Belarus. A further two had been in office 15 years but had ceased to be president: one died in office in 2006 (Niyazov of Turkmenistan) and the other (Akayev of Kyrgyzstan) was confronted with such serious demonstrations that he resigned and fled to Moscow in 2005. Finally, two long-term presidents were in office ‘only’ ten or 11 years, Shevardnadze of Georgia, who was forced to resign under pressure in 2003, and Aliyev senior (Heidar), who resigned to leave the post open for his son when he was very ill and he did indeed die shortly after having resigned, also in 2003. Thus, the ‘instability’ of the early leaders of Georgia, Tajikistan, Belarus, and Azerbaijan was more than compensated for by the length of the tenure of their successors in all five Central Asian states, in Azerbaijan, Georgia, and Belarus.

The 13 presidents whose duration was average, though not always ‘normal’

Between the short-term and the long-term presidents, almost half the incumbents – 13 – had a more ‘normal’ career. Seven of these were still in office at the end of 2010: in four cases, they were in their first term (Medvedev of Russia, Yushchenko of Ukraine, Sarkisian of Armenia, and Berdimuhammedov of Turkmenistan); in three cases, they had completed their first term and had been re-elected (Aliyev junior of Azerbaijan, Saakashvili of Georgia, and Bakiyev of Kyrgyzstan)). However, one of the first group, Yushchenko, was so badly defeated at the first ballot in late 2009, at the end of his first term, that he could not even stand at the second ballot. Little can be said about the future of the other three (who include Medvedev of Russia) as they were elected late enough in the period analysed here (in 2005 or afterwards) to be entitled to stand for re-election for a second term. Of the second group of second-term presidents, one was Bakiyev of Kyrgyzstan, who was ousted by a revolt of the people in April 2010, as his predecessor, Akayev, had been five years earlier; on the other hand, Aliyev junior was confirmed in anticipation by referendum for a third mandate before he had finished the second, as his father and the other seven long-term presidents had been. There is therefore uncertainty only about one ‘second-term’ president, Saakashvili of Georgia, about whom the question arises as to whether he is to be the first Georgian president to leave the presidency at the end of his regular two-term tenure.

There remain six cases of presidents who fully completed their careers at the top and subsequently returned to ‘ordinary’ life. Kravchuk of Ukraine, left office because he was defeated at the polls in 1994; Yushchenko, also of Ukraine, was badly defeated at the first ballot in 2009 in his bid for a second term. Kravchuk served less than a full four-year term, but did not attempt to stand again. Two presidents left office in a ‘regular’ way after the end of their second term as they could not legally stand again: they were Kuchma of Ukraine in 2004 and Kocharian of Armenia in 2008. These are in reality the only cases of presidents who completed their presidential careers entirely regularly, as the cases of the last three of the group of six are somewhat dubious. Ter-Petrossian of Armenia resigned in February 1998 in the middle of his second term, almost certainly under pressure. Yeltsin resigned slightly before the end of his second term for private reasons which were not wholly elucidated. Putin did resign the presidency at the normal end of his second term but only to be appointed prime minister by his successor; there is therefore a suspicion that he remained the real leader and, indeed, that he might try to be re-elected president when constitutional provisions allow him to do so in 2012.

Overall, out of 28 men who were appointed to the presidency in the 11 post-Soviet republics, seven leaders of the early period chose to, or had to, resign. The fates of three of the first-term presidents and one second-term president who were in office in 2010 are still open: some of these may or may not be re-elected to a second term and Saakashvili in Georgia may or may not ask for a change in the constitutional provisions of his country after 2010, as one of his colleagues, Aliyev junior of Azerbaijan, successfully did. Only in at most four cases did incumbents strictly abide by the rules, two of them being defeated in their second-term bids, so that only the other two completed their two terms. In two cases, the second term was rather brutally interrupted (Ter-Petrossian of Armenia and Bakiyev of Kyrgyzstan), while there is some doubt about the way in which the tenure ended with respect to the last two (Yeltsin and Putin).

Meanwhile, eight of the presidents did remain in power beyond, and in most cases much beyond the ‘term limits’, and a ninth, Aliyev junior of Azerbaijan, may be in a position to join these eight. We will have to return to the question of the ‘liberal’ or ‘authoritarian’ character of these presidencies in the next section. Meanwhile, the very long tenure of these eight presidents as well as the tenure of some of those who remained in power rather less time, both suggest that they will have had an opportunity to make a profound impact on the countries they ruled, whether in view of their own personal ‘appeal’ or of the instruments which they – and possibly their opponents as well – used to achieve their policies (Table 3).

Table 3. The presidents of the ex-Soviet republics

The powers given by constitutions and their limited effect in restraining presidential control

In the large majority of the 11 countries, constitutions provided presidents with powers that American presidents do not possess, even assuming that the political forces in existence in the country, and in particular the political parties, were in a position – which was rarely the case – to resist successfully attempts made by the president to control fully the political system.

The constitutions of the ex-Soviet republics are much closer to the semi-presidential model, at any rate juridically, than to the American model. Specific arrangements vary, but in general the government is headed by a prime minister, except in Turkmenistan, where the president leads the government directly, and in Georgia between 1995 and 2005, under Shevardnadze. The prime minister is appointed by the president, but typically needs to obtain a confidence vote from parliament; only in Turkmenistan has parliament no power over the cabinet at all. A majority of the constitutions also give the president the power to dissolve the chamber, except that in four countries (in Russia and, following the Russian arrangement, in Azerbaijan, Georgia, and Kyrgyzstan) this is confined to cases in which parliament has successively rejected three prime ministers proposed by the president. In Belarus, Kazakhstan, and Turkmenistan, on the other hand, the president is not constrained in this manner, while in Armenia, Tajikistan, Ukraine, and Uzbekistan the constitution is silent on the subject. Yet, in practice, the constitutional provisions that formally give some powers to the legislature have not constituted serious limitations on the ability of the presidents of the ex-Soviet republics to achieve what they wished to achieve. There have been only two episodes of dramatic confrontations between president and parliament, and the president won in both cases, in Russia in 1993 and in Kazakhstan in 1995. There were also conflicts elsewhere, but of a less dramatic character, especially in Armenia and Ukraine, but they had a relatively limited impact on the ability of presidents to act in the way they wished. In one case, that of Ter-Petrossian in 1998, the president resigned before the end of his term, but this was caused more by a ‘conspiracy’ within the cabinet than by opposition in the legislature.

Whatever the constitutions may have decided in the ex-Soviet republics, in all except one, Russia, presidents were confronted with new countries and therefore with large nation-building problems. In many of the states, some action had to be taken with respect to language; foreign affairs have tended to be paramount, however, especially given the wars, internal or external, in which these countries were involved.

Nation-building and the role of language

There could have been serious language difficulties in the Central Asian republics, as it is clear the Russian language will remain for a long time the main communication medium among the political and social elites everywhere. The case of Kazakhstan was ostensibly particularly difficult, since there is a large Russian minority in that state and it is concentrated in one part of the country (the north). The matter was cleverly handled by Nazarbayev. The problem also arises in Ukraine, and possibly in Belarus, in particular because of the existence of large numbers of mixed-language families.

Nation-building and foreign policy

A sense of national identity had also to be built through foreign policy action. Major foreign affairs problems have played a part – and were to an extent played up – in three of the five Central Asian states, genuinely in Tajikistan, less so in Turkmenistan and Uzbekistan (while little occurred in this respect in either Kazakhstan or Kyrgyzstan). The success of a Russian-Uzbek military coalition in blocking attempts from Afghanistan to destabilize Tajikistan seemingly had the effect of raising the morale of the ‘loyal’ parts of the Tajik population with respect to the new republic, although problems remain as support for the Republic of Tajikistan is far from being equally shared across the whole country.Footnote ¹⁸

The issue of Nagorno-Karabakh did boost the morale of Armenians, given Armenia's military successes against an unprepared and, originally at least, relatively uninterested Azerbaijan. The issue led to the resignation of Azeri presidents in the early period, as well as probably to that of the Armenian president Ter-Petrossian in 1998, while the personal and almost tribal support for the two presidents Aliyev of Azerbaijan was strong enough to prevent difficulties emerging for these leaders. Foreign affairs have impinged in a different, but equally serious manner on Georgia, as a result of the proclaimed secession of Abkhazia in the north-western tip of the country and subsequently of South Ossetia in the central north, both matters being connected with Russia, which has been repeatedly accused of defending the de facto independence of Abkhazia and South Ossetia. Transcaucasia is thus likely to remain the most unstable of the three regions – the level of national fervour in at least two of the three states being ostensibly the main reason for permanent tension and periodic conflict.

In Ukraine and Belarus, relations with Russia have been more than occasionally the source of major conflicts, both on some general issues probably related, deeper down, to the fact that Russia remains uneasy about the independence of these two states, and more specifically over the distribution of oil and gas. Yet the three countries of what we can term ‘Old Russia’ remain generally less willing to use – and do not seem to feel public opinion requires them to use – the more violent measures that have been adopted in Transcaucasia and in some of the Central Asian states.

Little building of institutional support in most countries: emphasis on traditional control rather than on the development of genuine parties

While the building of nationhood was a necessity for the ex-Soviet republics, except in Russia itself, in most of these little was done to achieve the desired result by building political parties that would aim at linking citizens to the state. Admittedly, the leaders of the post-Soviet republics did inherit their republic's Communist Party; indeed, several presidents had been leaders of their Communist Party or of the Council of Ministers under Soviet rule.Footnote ¹⁹ In a number of Central Asian states, though not in all, that ‘inheritance’ had a positive effect: it made it possible to maintain, if not legally but in fact, a (pure or at least wholly dominant) single-party system, although the party's name was changed, as in Kazakhstan and Turkmenistan. There has also been a dominant government party (but not a single party) in Azerbaijan: it is closely linked to the Aliyev family and built on a strong network of personal relationships. In Georgia, under Shevardnadze, the dominant party was also created with no particular programme except to support the president, although eventually some of its prominent members became part of the operation which led to Shevardnadze's downfall.

As a matter of fact, few members of the political elite of the post-Soviet republics, including that of Russia itself, as is shown by the examples of Gorbachev and Yeltsin, felt that, for a relatively open modern society, ties between citizens and the state needed to be institutionalized by means of political parties. In Russia, Putin was to be the first to seize the opportunity of building a party, to his great advantage, though that party was too closely and directly dependent on him to acquire more than a semblance of autonomy. The fact that these views on parties prevailed while leaders had a ‘personalized’ style, in Russia and elsewhere in the region, shows that the role of such institutions as parties – as well as other groups – was generally regarded as secondary. Lukashenka of Belarus appears to have adopted a somewhat sophisticated version of this model: the parties outside parliament exist but are weak, while the legislature behaves as if a non-party system sustained the president.Footnote ²⁰ Thus, presidents have tended to avoid strengthening parties, with the result that these have emerged more as cliques or, at best, as organizations of ‘notables’, as appears to a large extent to be the case in Armenia.Footnote ²¹ Meanwhile, the complexity of the east–west division in Ukraine has led successive presidents to remain rather unclear (deliberately or by accident) about the overall line they plan to follow; they have not felt it valuable to help to set up parties that would adopt clear-cut policies in that country either.

Had such a view not prevailed, serious political difficulties that resulted in the use of ‘strong’, including unconstitutional, measures may well have been unnecessary. This was the case, for instance, when the president of Kazakhstan closed parliament for a year; even in the more pluralistic countries such as Ukraine and Yeltsin's Russia, unconstitutional measures might have been avoided had presidents embarked from the start on setting up and nursing parties designed to muster genuine support for their actions.

Admittedly, what appeared to be an inability to want to build parties with deep roots in the population may have been due, at any rate in Russia and possibly elsewhere except in parts of Central Asia, to the fact that there was no tradition under Soviet rule for parties to form a link between government and people, while there was perhaps a need to ‘purge’ the country of the widely felt perception that parties were instruments of authoritarianism. Thus, street ‘demonstrations’ seemed to be regarded as a more ‘democratic’ technique: this would account for the fact that such a technique was used so commonly and for such long periods in Belarus, Azerbaijan, Georgia, Armenia, Kyrgyzstan (both in 2005 and in 2010), and in the last resort in Tajikistan. Yet street demonstrations have the double negative effect of creating further instability and so providing a justification for presidents to use repressive instruments of government.

Political life was thus rendered harder in many of the post-Soviet republics, and even brutal in some, if not all, in part as a result of the absence or the weakness of parties, whether of those supporting the government or of those opposing it. Yet this situation does not altogether alter the fact that there have been differences across the area in the size of the opposition and the character of its behaviour, a point we address in the next section.

The Freedom House findings

For some years, Freedom House has provided rankings of the extent to which liberal democracy prevails in all the countries of the world. These rankings became more precise in 2009 when separate scores were provided for political rights and civil liberties.Footnote ²² Two conclusions relating to the ex-Soviet republics emerge from these findings. First, as in previous years, these countries are ranked much lower on average on the two scales of political rights and of civil liberties than the countries of the world are on average. While worldwide, 42 (22%) of the countries were classified as ‘not free’, seven (63%) of the 11 republics’ fall in that category, four of these being from Central Asia, the other three being Azerbaijan, Belarus, and, perhaps somewhat surprisingly, Russia. Kyrgyzstan, Georgia, and Armenia are ranked as ‘partly free’, and Ukraine as ‘free’, although the scores reached by that republic are only just above the middle of the range of ‘free’ countries. Meanwhile, second, a substantial difference emerges between the scores of the Central Asian republics and those of the other six countries of the group.

Overall, Armenia and Russia are ranked somewhat lower than expected, especially by comparison with Azerbaijan and perhaps Georgia. There is no reason to doubt the broad conclusions however, namely that there is rather more freedom outside the post-Soviet republics than inside, and in particular that there is markedly less freedom in Central Asia (Table 4).

Table 4. Distribution of countries according to freedom house (2009)

The Freedom House findings do not indicate, however, whether presidents are in any way responsible for the rankings of the post-Soviet republics in general, and of the Central Asian republics in particular. To assess that, we must therefore look elsewhere and, to begin with, examine a number of key characteristics displayed by these presidents. Four features are obviously relevant: the duration of their tenure; the posts occupied under Soviet rule, if any; the percentage of votes obtained at the first ballot of the presidential elections; and whether the tenure ended according to constitutional rules or not.

The duration of presidents in office was appreciably longer in Central Asian states than elsewhere in the group. In Central Asia, every leader ruled on average at least three-quarters of the first two decades after independence; in the other six countries, every leader ruled, also on average, only about half of these 20 years, there being little difference between Transcaucasia and ‘Old Russia’ in this respect. Moreover, while there were some long-time presidents in Belarus, Georgia, and Azerbaijan, this was not so in the remaining three countries, especially in Ukraine and Armenia, Russia being difficult to assess, given the ambiguous position of Putin. If long-standing leaders are more likely to dominate the polity, especially in rather authoritarian regimes, it would follow that presidents were stronger in the Central Asian republics than in the other post-Soviet countries.

Second, the presidents of Kazakhstan, Turkmenistan, and Uzbekistan moved directly from a top position in the Soviet hierarchy to the presidency of their ‘new’ country.Footnote ²³ This was not the case for the first president of Kyrgyzstan, while the president of Tajikistan had occupied only a low-level position in the Soviet hierarchy. On the other hand, after 1992, both Georgia and Azerbaijan came to be ruled by former top officials of their country under Soviet rule. By contrast, Lukashenka, who was to become the leader of Belarus from 1994 onwards, never belonged to the Soviet hierarchy, while Putin owed his position to Yeltsin. The first president of Ukraine, Kravchuk, had been part of the hierarchy under Soviet rule, but this was not the case for the other two presidents of the country, who claimed power because they had been at one point prime minister of the previous president, a practice which was also followed in Armenia after the resignation in 1998 of the first president, Ter-Petrossian, who had been an outsider.

Third, there are marked differences among the three groups of countries in the percentage of votes obtained by successful candidates at presidential elections. Those from ‘Old Russia’ (Russia, Belarus, and Ukraine) obtained 57% of the votes at the first ballot, those from the three Transcaucasian countries 71%, and those from the five Central Asian countries 86%.Footnote ²⁴ Even though lack of fairness and even fraud have been repeatedly pointed out about most of these elections, the gap between the three groups is sufficiently large to indicate that there have been greater opportunities to vote relatively freely in ‘Old Russia’, and even in Transcaucasia, than in Central Asia (Table 5).

Table 5. Presidential election results

Fourth, the question of the end of tenure does not arise for the four presidents who were still in office in Uzbekistan, Kazakhstan, Tajikistan, and Belarus in 2010; a further two must be added, one each from Turkmenistan and Azerbaijan, as they died in office. By contrast, the tenure of Akayev in Kyrgyzstan and of Shevardnadze in Georgia was ended by the forcible removal of these leaders (Akayev's successor having met the same fate). Meanwhile, the end of the tenure of presidents was constitutionally correct in all cases in Ukraine and in Armenia, except for the fact that the first president or Armenia, Petrossian, did resign, apparently under pressure, before the end of his second term. The case of Russia is more complex, given the somewhat peculiar arrangement between Putin and Medvedev.Footnote ²⁵

Five presidential pairs and one separate case in the post-Soviet republics

On the basis of the four characteristics that have just been described, except for one president who stands alone, all the others form pairs in terms of the extent of the power they exercised. (1) Power was particularly strong in a first pair of dominating and repressive presidents, which includes Turkmenistan and Uzbekistan, while Kazakhstan stands on its own, as its president was dominating but much less repressive. (2) The second pair is composed of manipulating presidents in Tajikistan and Azerbaijan, where some opposition was manifest but became gradually tamed and rather ineffective as a result of the actions of the presidents. (3) The power of presidents came into question in the third pair, that of declining presidents, composed of Kyrgyzstan and Georgia, where the opposition was able to oust one president in one case and two in the other. (4) By contrast, in the fourth pair of presidents of increased strength, composed of Belarus and Russia, presidential power was at first limited but subsequently increased and was associated with ostensibly substantial popular support. (5) The fifth pair, characterized by complex political combinations, includes Armenia and Ukraine, where constitutional practices were followed, often with difficulty. Obviously, within each pair, each president differed to some extent in his modes of behaviour from the other, but the amount of power exercised and the evolution of that power were similar.

Power and repression in Turkmenistan and Uzbekistan, but power with little repression in Kazakhstan

Observers rank the president of Turkmenistan, Niyazov, who died in 2006, and the president of Uzbekistan, Karimov, still in power in 2010, as particularly repressive, although in this case, a balance had to be struck among various clans. On the other hand, Nazarbayev was able to control Kazakhstan throughout the two decades after independence without engaging markedly in repression. While exercising violence against opponents, Niyazov practised a personality cult on a grand scale – for instance by having arches built in his honour and decreeing that his birthday was to be a national holiday.Footnote ²⁶ Meanwhile, the Uzbek president's regime was defined as ‘command-authoritarian’.Footnote ²⁷ The campaign fought against Islamists may have provided a further ground for authoritarianism, with tension high and bombs exploding periodically, while riots occurred in the Fang'onn Valley, especially in 2004. Thus, the rule of the presidents of these two countries was clearly dictatorial.

The ‘climate’ was different in Kazakhstan. Nazarbayev enjoyed the same amount of power, but remained more in tune with the population. J. Ishiyama (Reference Ishiyama and Cummings2002: 51–2) has suggested that his regime was ‘neopatrimonial’, like those of Niyazov and of Karimov, but was of ‘a competitive one party’ character, while the regimes of the other two presidents were based on a personal dictatorship or on an oligarchy.Footnote ²⁸ The Kazakhstan president both enjoyed prestige in the population and exercised very strong power. Early in his tenure, in 1995, rather as Yeltsin had done, he did not hesitate to dissolve parliament without having the constitutional power to do so and subsequently to govern by decree for a year; but in general he did attempt ‘to conduct negotiations with all the political groupings and preserve relative inter-ethnic and religious tolerance in the country’ (Petrov and Gafaly, Reference Petrov and Vassiliev2001: 40). Political arrests were few and typically lasted for a rather short period. Overall, there seem to have been ‘good feelings’ for the regime on the part of a population that is also economically better off than the citizenry of other Central Asian republics, as was noted above. Nazarbayev was thus able to exercise dominant power without having to adopt the repressive measures of his colleagues from Turkmenistan and Uzbekistan.Footnote ²⁹ He has thus been closer to displaying ‘popular leadership’.

Political manipulation in Azerbaijan and Tajikistan rather than straightforward political control

The problems faced by Rahmon, the president of Tajikistan, and, to a lesser extent, by the two presidents Aliyev of Azerbaijan have resulted in more complex and tortuous behaviour on the part of the holders of the post than that of their colleagues who controlled and dominated Turkmenistan and Uzbekistan. The president of Tajikistan came to be in charge of the country in 1992, as speaker of the parliament, after having been a sovkhoz (state farm) director and the governor of a province. He was not even a compromise candidate: he was appointed because the various protagonists could not find any way to agree on a satisfactory solution (Atkin, Reference Atkin and Cummings2002: 99–102). Thus, Rahmon had scarcely any truly personal support; he could govern only by using his skills. Rather unexpectedly, he remained in office uninterruptedly afterwards, but his policy had to play hot and cold in a country that was deeply divided geographically and where there was a strong opposition movement, Islamic Rebirth (IRP). Rahmon attempted to proscribe the IRP in 1993, a decision that led to a temporary occupation of the capital by insurgents. By 1994, at his first popular election, the president had already half succeeded, although he received ‘only’ 58% of the votes, a result held to have been fraudulent (he was to do much better afterwards); yet his main opponent, Abdullojonor, who had been prime minister, had received 35% of the votes. Rahmon then engaged in discussions with the opposition (he even went to Iran in 1995 to meet its leader), but fostered police action at other times. There were periods of fighting and of guerrilla warfare (Rahmon himself was wounded in 1997), interspersed with peace agreements, throughout most of the first decade of the twenty-first century. The president's party, the Democratic Party of Tajikistan, obtained 74% of the votes in the parliamentary election of 2005, but the IRP did obtain 8% of the votes and the Communist Party (which was also allowed) obtained 13%. Only by manipulation could Rahmon succeed in remaining continuously at the top: in the process this ensured that Tajikistan was led in a somewhat less authoritarian manner than Turkmenistan or Uzbekistan.

The situation was not markedly more favourable in 1992 to Heydar Aliyev, who had come back to Azerbaijan from Moscow after being dismissed from the functions he had held in the USSR government. Demonstrations and other forms of violence occurred, in a context in which the war with Armenia over Nagorno-Karabakh was turning very strongly against the Azeris. Elchibey's resignation in favour of Aliyev senior meant that the new leader was obliged, as Rahmon was at the same time, to turn hot and cold against his opponents, the main one, Husseinov, receiving a heavy prison sentence in 1996. Only in 2000 did the party which Aliyev had created, the New Independence Party (NAP), win handsomely in parliament, the election result being condemned as fraudulent by the Organization for Security and Co-operation in Europe. The legislature had by then become based on a quasi single–party system, a feature perpetuated when Aliyev junior, Ilham, was elected in October 2003 to replace his father; he was re-elected in October 2008 with the even larger majority of 89%, in combination with a referendum which allowed the president to bid for a third term at the end of his second mandate. The clientelistic character of Azeri politics seemed by then to have become entrenched, with little effective challenge, apart from demonstrations, to the rule of the president. What the Aliyevs had succeeded in elaborating was what might be called a ‘soft authoritarianism’ in which opposition was muffled, as might have been the case at the beginning of the nineteenth century in many parts of the slowly awakening Western world.

The declining presidencies of Kyrgyzstan and Georgia

We noted earlier that both Kyrgyzstan and Georgia were ranked as ‘partly free’ by Freedom House. The case of the presidency of Kyrgyzstan was indeed very different from the cases occurring in the other Central Asian republics; but the background of the president was also very different from that of Shevardnadze in Georgia. Shevardnadze's past resembles that of the other presidents in Central Asia in that, before moving to Moscow, he had been the head of the party in Georgia and had been appreciated as leader. Akayev of Kyrgyzstan was a complete outsider, having been elected by the Supreme Soviet in December 1990 as ‘executive president’ precisely because it was felt that he would open a new era: being a mathematician and the head of the local Academy of Sciences, he was assumed to be in no way tainted with the past. Although he was the only candidate at the presidential election of October 1991, he wanted to introduce democracy; he supported what was the rather ‘parliamentary’ constitution adopted in May 1993. He did not see the need to build a party devoted to him: his parliament was in fact essentially composed of independents.

However, because some scandals emerged, Akayev subsequently demanded greater powers and successfully asked the people to grant these by referendum in 1996. By then he seemed no longer to have the ‘Cincinnatus-like’ modesty he had previously displayed. After having said that he would not do so, he felt the need to circumvent the constitution about term limits and in 1998 obtained, again by referendum, the right to serve a third term. He was re-elected with 75% of the votes in 2000, but the spell seemed over: demonstrations began to occur in the capital, especially in favour of the ex-mayor of Bishkek, who had been arrested for embezzlement, acquitted, and re-arrested. Akayev's regime was not authoritarian, but he failed to give the people what they had (perhaps unrealistically) hoped for. He never built a party on which he could count; but nor did the opposition build any institution that would ensure that the departure of the president took place according to pre-established rules.Footnote ³⁰ Major street demonstrations occurred from 2002 and in the following three years. Akayev resigned in April 2005 and left for Moscow; Bakiyev was to be Akayev's successor: he was to suffer the same fate as Akayev five years later in 2010.

After the bizarre episode of Gamsakhurdia's presidential rule and departure as a result of the civil war, Shevardnadze returned to Georgia. He was the only candidate in October 1992 for the chairmanship of the Supreme Council, for which he obtained 95% of the votes, and at his first election as president in 1995 he was elected with 75% of the votes, a majority which even increased to 80% in 2000. Such results were buttressed by similar scores for the party he had created, the Citizens’ Union. Yet conspiracies were already beginning to form to replace the president, who had also been the target of two assassination attempts, in 1995 and 1998, with the Minister of Finance alleged to have been involved in the 1998 attempt. From 2001 onwards, the political climate began to be truly sombre: the Minister of Justice, Saakashvili, resigned and created a National Movement of his own, while, in the following year, the speaker of the parliament, Zhvania, decided to lead a faction of Shevardnadze's party against the president at a time when demonstrations demanding fair elections were taking place. Thus, the scenario was rather similar to the one that Akayev was subjected to: in 2003, Saakashvili demanded the resignation of the president, and this duly took place in November of that year as demonstrations became increasingly large.

Thus, in both cases, the president was at first given a great deal of power and was truly popular; after a number of years, in both cases, demonstrations took place, partly supported by politicians who had been close to the president, and the president left in disgrace. Yet it is not clear that Shevardnadze's successor, Saakashvili, who received 96% of the votes in the presidential election of 2004, contributed to an increase in the strength of the Georgian presidency. He obtained a large majority in parliament for the party he had set up, but strong claims were made that the results had been falsified. The new president sponsored a constitution which replaced the American-type presidential system (without a prime minister) brought about under Shevardnadze, and was semi-presidential: Zhvania became prime minister under the new system, but he was assassinated early in 2005. Suspicions of coups and allegations of electoral irregularities led to arrests in 2006, these being followed in 2007 by accusations of criminal behaviour levelled at Saakashvili. Yet the president was re-elected, but with only 54% of the votes, in January 2008 – the first instance when the result of Georgian presidential contests was not overwhelming.Footnote ³¹ This was not to be the end of the difficulties for Georgia and its president, however, since a war with Russia over South Ossetia took place in 2008 in circumstances in which Saakashvili's part was somewhat unclear. Demonstrations against the president then began to take place on a substantial scale.

Both Shevardnadze and Saakashvili organized, shortly before each election, a party devoted to them, but that party, rapidly put together and without real roots in the population, did not provide effective support for either president in the long run. The context was thus not vastly different from that prevailing in Kyrgyzstan in 2005 – indeed assassinations, attempted or effective, did not occur in that country. Nor did the Georgian opposition prove better able to channel discontent through effective parties. Thus, the two decades between 1990 and 2010 showed that, in both countries, the presidency declined in strength, but without leading to a political situation in which liberal democracy could be said to have been truly established in the process.

The increased strength of the Russian and Belarus presidencies

While the power of the presidency declined in Kyrgyzstan and in Georgia, it increased markedly and swiftly in Belarus from the moment Lukashenka was elected president in June 1994, and more subtly, by small stages, in Russia, when Yeltsin resigned after having appointed Putin as prime minister and as his likely successor. The case of Belarus could be regarded as surprising: there was a period of uncertainty up to 1994, during which a confrontation occurred between the Communists who still controlled parliament and the supporters of the Belarus National Front (BNF), who wanted the country to move towards liberal democracy. The constitution adopted in March 1994 provided for a president with relatively limited powers: it seemed that there would be no change with the election in the second ballot of Lukashenka, an outsider from the previous regime and a member of parliament. One of his planks was to improve the economic conditions of the country, which had markedly deteriorated during what had been a kind of interregnum.Footnote ³²

A few months after his election, meanwhile, Lukashenka quickly issued presidential decrees, which were turned down by the constitutional court. ‘It appeared therefore that even if the new President failed to live up to expectations, there were limits that could be placed on his authority’ (Marples, 2003: 28). By 1996, however, Lukashenka had obtained from the people by referendum the right to dissolve parliament and to revamp the constitutional court, a referendum which was marked by ‘gross improprieties in the voting procedure’ (2003: 28, 29): the markedly authoritarian character of the regime and the strength of the presidency had thus been put in place.

Steps towards a stronger presidency went at a much slower pace in Russia; they had also remained more limited by 2010. While Lukashenka fully controlled the media, arrested many opponents, and made it very difficult for demonstrations to take place, Putin never acted in such a broad sweep. Reforms were introduced in the relationship between central government and member states of the Federation, which resulted in the direct appointment of governors by the president; there were moves against top businessmen who were showing too much independence; there were murders, of journalists in particular, which remained mysteriously unsolved; there was some harassment of opposition parties and candidates: but there was never a blanket endeavour to abolish political parties. As a matter of fact, unlike Lukashenka, who ensured that elections were fought as much as possible on a non-party basis, Putin launched the Party of Russian Unity to win the parliamentary elections and to rely on a safe majority in the legislature. Yet, as in Belarus, the power of the president was markedly increased in Russia over the first decade of the twenty-first century (Fish, Reference Fish2005: 270–1).

The question therefore arises in both cases as to how the two presidents could achieve such an increase in their power without encountering major opposition, as experienced by the presidents of Kyrgyzstan and Georgia. There were some demonstrations, typically dispersed without too much difficulty, and generally rather small, more so in Russia than in Belarus. Yet the explanation of the ‘stability’ of the two countries arises from a situation that may be regarded as surprising, especially for Belarus, but was none the less real, namely that, in both cases, the president was popular, indeed markedly more popular than anyone else on the political horizon. Support for Lukashenka declined, admittedly, from 54% in 1999 to 42% in 2000, but no other politician had a rating even as high as 5% (Rotman and Danilov, Reference Rotman, Danilov, Korosteleva, Lawson and Marsh2003: 102). Interestingly, the New Democracies Barometer showed that support for Lukashenka was not the result of a positive attitude towards a military regime or to an authoritarian government (Haepfer, Reference Haepfer, Korosteleva, Lawson and Marsh2003: 91–3). As C. W. Haepfer states: ‘the Belarusian electorate does not fulfill the popular expectations nourished by international publications and media, which suggest that Belarus and, by implication, its population, are non-democratic and somewhat politically backward’ (Reference Haepfer, Korosteleva, Lawson and Marsh2003: 96). The same conclusion is valid for Russia: Putin would not have received the vast support at presidential elections nor would his party have done so well at parliamentary elections if the population had not been, by and large, favourable.

This substantial support within the electorate is accounted for by the stability these leaders were regarded as having brought to the political system and the impression that the socio-economic difficulties encountered since the end of the Soviet Union would be surmounted. Thus, the increased strength of the president results in both cases from the general impression that the country had found a ‘leader’ who would bring about at least a partial return to what was regarded as highly satisfactory in the Soviet system.

Where deals and compromises prevail: Armenia and Ukraine

With Armenia and Ukraine, we enter a rather different world, one in which there are problems of coalition building and of relations with parliament, or at least with the parliamentary representatives in the government. There are some differences between the two countries, admittedly, in that the political initiative was more in the hands of the president in Ukraine, including the selection of his possible successor, while, in Armenia, government members and in particular the prime minister have had a major say.

These characteristics suggest that the two countries are liberal democracies: yet serious worries have been raised about the fairness of elections in both of them. In Armenia, while the 1991 presidential election was felt to have been fair, ‘in every election thereafter . . . elections have become or have been perceived as having become steadily more corrupt and less free and fair’ (Ishkanian, Reference Ishkanian2008: 36). In Ukraine, the widely recognized unfairness of the 2005 presidential election led to major demonstrations which culminated in the ‘Orange revolution’; however, although the fairness of the subsequent 2009 presidential election was contested for a while by the major defeated candidate, the result was eventually accepted as having been, by and large, free and fair.

The view that Armenia and Ukraine are liberal democracies is further supported by the fact that there is press freedom,Footnote ³³ that there are parties in parliament, and that presidents have to negotiate with these to be able to build a majority government or obtain a majority for the government. Yet, in Armenia, the impression prevails that ‘most political parties are composed of small groups of individuals or of intellectuals who are opposed to the government or to the leaders of the parties or organizations to which they previously belonged’ (Dedeyan, 2008: 686; my translation). In Ukraine, while the parties are somewhat less based on small groups of individuals, their discipline in parliament leaves much to be desired. Finally, perhaps the more worrying aspect of Armenian politics is the fact that violence has been endemic among politicians: the worst episode occurred in 1999 when armed men entered parliament and assassinated the prime minister and other members of the political elite.

The distance is large between the presidentialism of Nazarbayev of Kazakhstan and the (semi-)presidentialism of Armenia and Ukraine. Many characteristics of the political life of Armenia suggest that the political system of that country is close to being parliamentary rather than presidential. Even though this is less true in Ukraine, the fact that the president of that country often has to negotiate with members of parliament about the fate of the government, and in particular the appointment of the prime minister, suggests that the president has little autonomy in relation to the legislature. While presidentialism on the United States model operates with a separation that gives autonomy to the composition and internal management of the executive, the semi-presidentialism of Ukraine and even more of Armenia, based as it is on parties lacking discipline and well-defined programmes, renders the government – and through the government, the president – closely dependent on the vagaries of the legislature. If presidents do not or cannot exercise what was referred to earlier as ‘popular leadership’, the specificity of semi-presidentialism seems to disappear and the slide towards parliamentary government seems inevitable.

The strength of post-Soviet presidents and the location of the regimes on the authoritarian–liberal democratic continuum

There is thus a sharp divide among ex-Soviet republics in terms of the strength of their presidents. There are states in which the president is strong, even very strong, or has gained strength; on the other hand, there are states in which the president is rather weak – even if not as weak as are, at least in theory, parliamentary presidents – or has become weaker over the two decades since independence. Moreover, while some of the longer-term presidents have been particularly strong, this is not a universal characteristic: the strength of two longer-term presidents declined so much, after many years during which they were popular and (too) much was expected of them, that they were effectively forced to resign, as a result of popular demonstrations, while ‘conspiracies’ among the elite had also taken place.

If we broadly accept the Freedom House rankings as accurately reflecting the reality of the spread of the 11 republics along the authoritarian–liberal democratic continuum, it appears that the strength of the presidency is negatively correlated with the extent to which the regime is liberal democratic. First, and with the partial exception of Kazakhstan, the post-Soviet republics whose presidents have been very strong throughout their whole tenure are the least liberal-democratic ones (Turkmenistan and Uzbekistan). Second, there has been a small opportunity for popular involvement where presidents have been highly ‘manipulative’ (Tajikistan and Azerbaijan). Third, in those republics in which presidents gained in strength (Belarus and Russia), opposition has been harassed, albeit not wholly silenced.

On the other hand, first, there is at least some pluralism where presidents have been or have felt obliged to abide by constitutional rules (Ukraine and Armenia), even if the liberal-democratic character of both polities – and perhaps somewhat more in Armenia than in Ukraine – leaves much to be desired. Second, there is an unclear future for those polities (Georgia and Kyrgyzstan) in which presidents gradually lost so much power that they were forced to resign: given that, as was mentioned earlier, expectations were probably too high among many in the population, the danger is either that instability will continue to prevail or that an elected president, as in Russia since 2000 or even as in Belarus since 1994, imposes a strong rule which can scarcely be described as liberal-democratic.

What have therefore been found at the level of ex-Soviet republics are signs of incompatibility between presidential strength and liberal democracy in that region: the kind of ‘popular leadership’ that presidential government requires if it is to be liberal democratic as well as effective has not (as yet) emerged. Yet what cannot be said at this point is whether that incompatibility is primarily the consequence of specific characteristics of the group of countries studied here, and in particular whether the very authoritarian past constitutes an element of the problem. The answer to such a question has to wait until the analysis of presidencies is undertaken generally and ex-Soviet republics can be examined in this broader context.

The presidential system has been essentially adopted in new countries, although it has also been introduced, by what is perhaps a strange analogy, in countries which have found it difficult, as France did, to maintain a stable parliamentary system. Over the years, of course, the American presidential system, because of its extraordinary stability over two centuries, has naturally come to be regarded as a, or even as ‘the’, model: but it is a model that few countries have come close to imitating: even Costa Rica has not had an ‘unblemished’ record, while Chile's party system has still not fully recovered from the trauma of the long Pinochet dictatorship, and Venezuela has suffered a ‘relapse’ after several decades in which ‘regular’ constitutional practices had been followed. Only time will tell whether the Latin American presidential systems are moving towards an era of liberal-democratic stability, but, even if Brazil and perhaps Mexico seem to be doing so, the twenty-first century examples of Venezuela, Bolivia, Ecuador, even of Argentina, suggest that prognostications about the future of liberal democracy in the region have to be made with great caution.

The study of presidentialism has suffered, perhaps surprisingly, but none the less profoundly, from the fact that the United States was the very first case of presidential government in world history, and that presidential government in the United States has been such a success – especially in terms of its stability. It therefore seemed reasonable to conclude that presidential government could be equally successful elsewhere, especially since it seemed to be the first form of government that appeared truly democratic: hence the surprise that the extension of presidentialism in Latin America has not been a success. The very peculiar conditions of the United States were simply not considered; nor was it noted, of course, that presidentialism was likely to emerge almost exclusively among new nations given that there were so few new nations between the 1820s and the 1960s!

It is therefore only if one takes into account the experience of new nations – the attempt being made here being only a very limited fragment of the whole universe of that experience – that it is permissible to hope to discover what might be described as a ‘theory’ of presidentialism. It is useless to try to develop such a theory on the basis of the almost unique case of a truly working liberal-democratic presidential government, that of the United States; what has to be done is to examine the range of situations shaping the general picture of presidentialism.

As was noted earlier, analysis of the forms taken by presidentialism in the post-Soviet republics showed that the strength of the presidency, in that group of countries, has been negatively correlated with liberal democracy, even though the condition of a popularly elected president was met in these countries. What must therefore be discovered is whether the felicitous case of the United States has to be accounted for by the highly peculiar circumstances of that country, or whether, on the contrary, the ex-Soviet republics are the ones that are peculiar. When a general study of the characteristics of presidencies has been undertaken, but only then, can the question that has been a key problem in comparative government analysis at last be answered, namely whether the presidential system can be, for all nations, and especially for new nations, the solution that it did appear to be when the American polity was ‘invented’ in the late eighteenth century.