social capital, organizations, corruption, variables, market economy, political participation, civic activities, Russia, World Bank, Aarhus School of Business, the transition, Department of Economics, Latin American countries, WN, totalitarianism, dictatorship, London Granato, London Olson, Eastern Europe, Journal of Economic Perspectives, Economic Transition, Eastern Economic Journal, American Journal of Political Science, European Political Science, London Nove, Soviet Economic System, goods and services, control system, control systems, American Journal of Sociology, World Development, barter networks, Glasgow Schneider, European Democracy, Conni Paldam, participation, Putnam, totalitarian control, Hannah Arendt, totalitarian regimes, Paldam & Svendsen, Stalin, Communist rule, dictators, totalitarian system, Latin-American countries, official organization, civic participation, total control, social welfare services, Column, Richard Rose, Latin countries, measurement errors
Missing social capital and the transition in Eastern Europe Martin Paldam Department of Economics University of Aarhus Gert Tinggaard Svendsen Department of Economics Aarhus School of Business *) Abstract The transition of the »Old Communist« countries of East and Central Europe has been disappointingly slow given the amount of physical and human capital
available at the start of the transition. We argue that this slowness is caused by the lack of social capital, which is an important factor of production. The Communist system replaced it with an official organization of society. Further, the communist system needed a set of grey/black networks of »fixers« to give it the necessary flexibility. These networks were tolerated, but controlled. When the Communist regime ceased the official organizations collapsed and so did most of the control systems. This allowed a flourishing of the grey/black networks, which can be harmful to the operations of a market economy. The available data are still scanty, but they confirm the argument. JEL Classification: A12, C71, D23, D70 Keywords: Social capital, network, communism, Eastern Europe, trust, corruption, political participation. ___________________________ *. The authors want to thank the discussants at the IMAD conference in PortoroУ, especially Ron Wintrobe. At the conference the first version of this paper was presented together with Paldam (2000b) surveying the social capital discussion. We also thank Conni Paldam and Ebbe Yndgaard. We use the abbreviation OC for »Old Communist« countries now in transition from socialism. The word »Communist« is used for countries ruled by pre-reform communist parties
. It should be noted that the article is not written with the mild transition of Slovenia in mind. Addresses of the authors are, MP: Department of Economics, Aarhus University, 8000 Aarhus C, Denmark. Phone: +45-8942-1607. E-mail: [email protected]
, home page: http://www.martin.paldam.dk. GTS: Department of Economics, Aarhus School of Business, 8210 Aarhus V, Denmark. Phone: +45-8948-6408. E-mail: [email protected]
Written for the Journal for Institutional Innovation, Development and Transition »During the past 70 years, a new man has been created who is obedient and easily frightened. What has been created over the decades cannot be undone in a day.« Bulat Okudzhava, 1989.
I Introduction: A missing link? The collapse of the old centrally-planned (Communist) system in East and Central Europe - from now the OC-countries - took place only one decade ago. Much has happened, and when one walks down a street in central Moscow on a sunny spring day, it looks as a typical European capital.1) Wealth is apparent, shops are booming, people are well dressed, and private cars are everywhere. However, the data tell another story. The economic development of the countries in transition of the OC-country has been disappointing in three ways: (d1) The output collapse just after the big change in 1989/90 was on average 40% of GDP. (d2) The upswing has been slow. In some cases it has hardly even started.2) (d3) The transition has caused an increase in corruption, rent-seeking and crime. The main reason that most economists expected a small adjustment crisis (not d1) and a quick upswing (not d2), was the predictions from the theory of economic growth
. It speaks of steady state growth paths determined by the stocks of human and physical capital. The OC-countries did have large stocks of both capitals before the transition. The existing stocks of the two capitals were not made for a market economy. So they should not be valued at reproduction costs. However, even after writing off say 25%, a lot would have been left. Further, the standard of living in the post transition OC-countries was much lower than it should have been given the existing stocks of the two capitals, so things looked promising. The reasons for both (d1) and (d2) remain hotly debated. The logical explanation is that something intangible is missing. Maybe the 70 to 45 years of socialism destroyed some X, which is hard to rebuild. Often it is argued that X is »initiative«, »business knowledge« or even the »spirit of capitalism«. However, such proposals seem without a theoretical basis. We argue that X is social capital, . Item (d3) is easier to understand, though it was still unpredicted. Below we treat it as a consequence of the rise in negative social capital ( ), and use the available data on corruption
One of the authors spent half a year visiting most of the OC-countries around 1970s and has recently
However, there seems to have been a large relative growth in the unrecorded »shadow« economy since
then, as discussed in Section IV. Some calculations suggest that the increase is no less than 25% of GDP
in the average OC-economy. If this is true, the development is much less disappointing than argued in
perceptions as an indicator of the size of the phenomenon. The interaction between »normal« (positive) social capital ( +) and »negative« social capital ( ) is not well known, but will be an important subject below. The main theory behind the study is the dictatorship theory of missing social capital. It argues that dictatorships - especially totalitarian ones - destroy social capital. Some dictatorships such as the Communist ones even create conditions that favor the building of negative social capital. When dictatorships are abolished, social capital may develop negatively and act as a brake on economic development. Standard textbooks on the transition - as Gros & Steinherr (1995) and Lavigne (1999) do not even mention social capital. We try to provide the reverse case and see how much social capital can potentially explain, and to assess if the available data support this explanation. We do not attempt to compare »our« explanation with alternative explanations. Section II reviews the main definitions of social capital and relates it to the economy. Section III proposes three hypotheses concerning the level of social capital in Eastern Europe arguing that the old communist systems destroyed normal social capital, while feeding negative social capital. Consequently, social capital may be a major part of the explanation for (d1), (d2) and (d3). Section IV presents some data from the World Value Survey and Transparency International in a preliminary attempt to test the argument. In addition, a few scattered observations will be considered. Section V is a brief conclusion. II Social capital theory: definitions and the relation to other matters This section will be dogmatically short, as we have developed our views - with the necessary qualifications - elsewhere3) - we concentrate on normal (positive) social capital: (+). II.1 Two definitions, a reformulation and a proxy Most definitions of social capital are summarized in Table 1. It gives two definitions,4) and a reformulation in the language of game theory. Also, an easy-to-use proxy named Putnam's Instrument is given.
See Paldam & Svendsen (2000), Svendsen & Svendsen (2000) and Paldam (2000b).
The origins of the network definition is hard to trace, but something like it is found in Coleman (1987)
and the trust definition may originate in Fukyama (1995). Both authors quote earlier thinkers. The main
problem is that social capital is a joint name given to synthesize old thoughts.
Table 1. Two definitions, a reformulation and a proxy
Network definition N is the density of peoples informal network
T is a measure of trust among people
Excess cooperation a) The excess propensity to play the cooperative solution in prisoners dilemma games Putnam's Instrument is the density of voluntary organizations b) Note. The definitions are discussed and some classics reprinted in Dasgupta & Serageldin (1999), see also Putnam (1993, 2000), Deth et al (1999). a. Taken as a reformulation of the trust definition in the »language« of game theory. b. The density is often weighted by the amount of contact the individual has with the organization.
We assume that the two definitions are closely related, so that given a proper calibration to the same common scale, we have: (1) = N + JN, where JN is small and random (2a) = T + JT, where JT is small and random (2b) T = + J, where J is small and random
One argument below is that the definitions are likely to deviate in the OC-countries, precisely because they create problematic networks, bordering on the corrupt. If the J's are small and random we have:
Social capital dream:
= N = T =
The dream is thus that all definitions have the same underlying »basis«. It will take some time before we know if the dream contains some truth. Social capital is in the early stage of development. No definition is generally accepted. It is clearly difficult to aggregate social capital, and many problems of measurement remain to be solved. Still, we may learn that something basic and robust lies beneath the idea of social capital, making it a useful variable.5) Many other concepts routinely used by economists are problematic. They are unsharply defined and easy to »dissolve« into parts, which can only be aggregated for certain purposes. However, as we know the limitations of the concepts, they are still useful. The last variable, , is Putnam's Instrument. The voluntary organizations listed by Putnam (1993) are comprehensive. That is, the list includes choral societies, bridge clubs and bird watching societies. Clearly, it is not important for the development of society and the economy that people get together to sing and look at birds. However, the fact that such societies exist and are widespread in some places, but not in others, tells us something about the society. We conclude that Putnam's Instrument can be a proxy only. Finally, we return to the point that cooperation and networks come in positive and negative forms. People may cooperate about socially desirable goals or about goals decreasing
It did take a long time for the profession to come to grip with physical and human capital, and we have
barely started to use social capital.
aggregate welfare. For example, they cooperate in mafias and similar gangs. Hells Angels have plenty of esprit de corps
, but they unite against the rest of the world and like to fight »wars«. Normally, such negative networks and organizations are secret, and hence not recorded by polls and other methods of measurement, but it does not, of course, mean that they do not exist. So what we measure is: (1) (+) II.2 Relating social capital to the economy 6) Behind the growth argument - presented in the introduction - is the idea of a macro production function: (4) yt = Yt/Nt = Y(t, Lt, Kt, Ht), where . = (Y/t)/Y> 0, Y/x > 0, 2Y/x2 < 0, for x = L, K, H Y is GDP, N is the population, t is time, L is employment, K is physical capital, H is human capital, and . is the rate of technical progress. A huge literature discusses (4). It is fraught with problems, but it is widely agreed that it makes enough sense to allow us some insight into the process of growth. The »optimistic« predictions about the transition of the OC-countries mentioned in the introduction, relied upon the idea that the levels of K and H in the OC-countries correspond to Y's, which were larger than the Y's actually observed. The missing social capital idea can be expressed, simply by including social capital t as a factor of production.7) This gives us equation (5). It is much less researched than (4), simply because t is much less known than the other variables, but let us for the moment assume that it makes sense. (5) yt = Y(t, Lt, Kt, Ht, t), where Y/ > 0, 2Y/
2 < 0, and is normal social capital. The negative social capital is negative precisely as Y/ () < 0. The missing social capital idea is simple: While the old communist regimes built K and H they destroyed , so that after the transition was much too small. If is a substantial factor of production, and if the elasticities of substitutions between and the other factors of production are limited, then the low produces a Y, which is (too) low. The optimistic prediction of economists (in the terminology of the introduction non-d1 and non-d2) was thus based on overlooking the low . Several of the micro-studies in World Bank
(1999) suggest that is just as important as H to the income of people, and hence to the generation of the product Y. The evidence is still
We concentrate on the relation between social capital and the economy. Much of the social capital
literature is written by politologists concentrating on the relation between social capital and democracy
and good governance, see Putnam (1993, 2000), Deth et al (1999) etc.
Paldam & Svendsen (2000a) discuss this simplification and present alternatives.
scanty and disputed,8) but the existing evidence seems to support the idea that is a substantial factor of production. This is also the claim of Putnam (1993), Olson (1996) and other writers in the »social capital school«.
III Dictatorship and social capital: A negative causal link Dictatorship comes in many forms. We will distinguish among the three types listed in Table 2. The most extreme form is totalitarianism. Wintrobe (1998; p 7-11) surveys a set of definitions of totalitarianism, and ends by endorsing the definition of Hannah Arendt (quoted in Table 2). Wintrobe stresses that the goal of total control - in part - is achieved through the »atomization« of human relationships
. That is, of course, the opposite of social capital.
Table 2. Three main types of dictatorship
Totalitarianism is a system that aspires to the total domination of each single individual in each and every sphere of life.
Ordinary dictatorship Rule by individuals (mainly military) who conquer power.
Rule by Monarchs, who inherit power according to historical principles.
Often very durable.
Note: Both monarchs and ordinary dictators normally »only« try to control political life, the
instruments of power and the media. Communist rule is in principle totalitarian, but its later
versions have been softer, more ordinary dictatorship.
III.1 The core observation: dictatorship destroys social capital All dictators know that when people cooperate outside the control of the regime they may also cooperate against the regime.9) Therefore, most dictators are against the uncontrolled cooperation of their subjects. Public choice theory teaches that organizations have a strong inborn tendency to take on a life of their own. The leaders - however appointed - often want to use the organization for their own purposes. That means to rally their members behind themselves, but then they have to provide something in return. Sometimes the central leadership turns so slack that the members manage to seize the organization or a part of it and turn it into their own. This is well known by all dictators, and they normally strive to build safeguards to secure that all - or the main - organizations of society stay under full control, and that their leaders have no possibilities to build their own power base. Leaders are often changed, or care is taken to
Several of the leading researchers in empirical growth theory - as Dale Jorgenson - have argued (at least
verbally) that little space is left in our knowledge of the growth process for social capital to explain.
This has often happened, eg, with choral societies in the Baltic Countries during the Soviet rule
, and in
Denmark during the German occupation.
make them incompatible with their members.10) Social control is crucial for all dictators who want to remain in power. Totalitarian regimes are those who are most extreme in these matters. Putnam's example (in his 1993 book) is the great difference of social capital in the North and South of Italy, where the division is along a line somewhere between Rome and Naples. In the South, the Kingdom of Sicily ruled for 700 years, till the unification in 1861. This Kingdom is described as an example of absolutism that deliberately prevented the building of social capital. It was not a particularly »bloody« regime, but it did try to control its subjects, and it had a secret police, etc. Putnam stresses that the long time the regime lasted was crucial for the destruction of trust and cooperation. Also, it led to the creation of the Mafia, which started as a secret mutual aid organization among the peasants.11) It is contrasted with the city-states of the North. They did have their tyrants and ruling families too, but they kept changing, and in between the citizens ruled. The Kingdom of Sicily appears much like Russia of the Czars. An example of ordinary dictatorship (of shorter duration) is Spain during the period, 193975.12) The study of Torcal & Montero (1999) observes that social capital (by available measures) was low compared with similar Latin countries. The mechanism is most easy to observe, when we consider the most extreme case: In a totalitarian system the dictator wants to mold society into »his system«, with himself in the apex and where everybody else is controlled. Totalitarian systems are one-party states, where the dictator derives his power from being the leader of the party, and where the party chooses the successor, when the dictator dies. Everybody has to be organized, under the control of the party, and with the dictator as the head of the party. All organizations are thus integrated as parts of the system. They are used to generate loyalty and control. Many studies have been made of such systems. Wintrobe tries to generalize a vast material dealing mostly with the regimes of Hitler, Mussolini and Stalin. Of these systems the Stalinist/Communist system proved the most durable. The most thoroughly researched study of this system at its heyday, at the grass-roots level, is no doubt Fainsod (1958).13) It documents the 10. The Empire created by Stalin did have one generation of leaders (normally the 2nd) appointed by Stalin in 1946-48, ruling in his spirit (such as Ana Pauker, Mбtyбs Rбkosi, Jakub Berman and Klement Gottwald).They were typically appointed instead of leaders with a local following, making them strongly dependent upon their foreign boss. Soon after Stalin's death most disappeared and a divergence of the vassal states into more normal countries started. However, even in 1989 all the countries were still run as Communist one-party states. 11. The biographical essay (novel) The Leopard (Lampedusa, 1958) brilliantly describes attitudes and values in the Kingdom of Sicily at the time of the unification (1860-63). Perhaps the most illuminating is Cpt 4 where the Sicilian prince explains the deep level of distrust and inaction in Sicily to a »new« Italian from the north. 12. The story of the regime of General Franco
is much like the one of his admirer General Pinochet, though all numbers (see below) are larger, corresponding to the relative sizes of Spain and Chile and the duration of the regime. Also, Spain went through a civil war (lasting 3 years), while Chile experienced a coup lasting 10 hours. Note that even when Spain did have democracy before Franco, it was shortlived. 13. The study builds on an investigation of a railroad car full of archive material for 1917-37 from Smolensk Gubernaya/Oblast. It was occupied by Germany during the War and they loaded the car with party archives, and sent it to Germany. After the War the archives became available to a group of US 6
extent of the control system. Even when it was both clumsy, slow and somewhat arbitrary, it did extend into all spheres of life and nobody could be sure to do anything uncontrolled. In the late DDR, the archives of STASI have been opened and subjected to much research already. Once more the extent of control was staggering, as files on more than 10% of the adult population existed. The control was kept till the end, even when the system gradually mellowed a little. The studies made all stress that the creation of fear and distrust is an important element in totalitarian control. All totalitarian regimes have created competing instruments of repression and fear, and some of these instruments have been made highly visible. The descriptions agree that an atmosphere of intimidation is easy to create and most totalitarian regimes have greatly succeeded in this endeavor. They do not need to kill and torture very many to be convincing.14) However, some totalitarian regimes have killed and jailed a sizable fraction of the population, often in the form of wavelike purges that took a dramatic momentum of their own. This has created a strong atmosphere of fear and distrust leading into paranoia. Those who live through such periods tend to be changed. They have thoroughly learned to trust nobody, to obey authority, and take no initiative of their own. The resulting atomization of social relationships
clearly involves destruction of social capital. The official main aim of the 70 years soviet system was to create a new socialist man and eliminate capitalist man. In order to do so, all voluntary organizations were brought under the leadership control of the communist party. By doing this, the party told people what to do, and de facto abolished all unofficial voluntary organizations. Great efforts were made to root out non-system initiatives and even the boy scouts were replaced by official party scouts (pioneers). Strong incentives were given to restrict all activities to the (relatively) safe one of obeying orders. The state made almost all decisions and left no room for entrepreneurship, experiments and voluntary organization into social groups (Paldam and Svendsen, 2000). This can be summarized as hypothesis one: dictatorship and especially totalitarianism destroys normal social capital. III.2 Network under communism15) The preceding section hints that strict totalitarianism is rare. Hitler, Stalin (and his vassals), Mao sovietologists, who made the study. 14. The Pinochet regime ruled Chile for 17 years. It tried to have absolute power, without developing a truly totalitarian system. It killed almost 3000 people. More than half were killed within one month after the coup. In the remaining period less than 100 were killed per years, while the population increased from 9 to 13 mio inh. It is about 10% of the number killed in the traffic. Even when the number of tortured was larger, it was still small by comparative standards. However, all evidence indicates that this was enough to thoroughly intimidate the population. A rather similar case is the DDR regime. It did intimidate without much killing especially during its last 30 years. 15. We treat the Stalinist model as the »core« model of Communist totalitarianism. Many other softer versions of the system have been tried. Some versions, as the Titoist model, and the Chinese model of today are rather different from the core model, even when they had/have the Communist Party as the one and only political party. 7
and other extreme cases can be mentioned, but they are rare. Stalin's successors gradually turned softer. The system mellowed, even when it stayed totalitarian in principle and in many details till the end. Two economic features of the communist economic system
gave a big premium to nonsystem cooperation, namely (1) supply constraints and (2) plan fulfilment pressures. The ensuing mechanisms - see Table 3 - are often described (see, eg Nove, 1961, 1977). They created informal networks bordering on the illegal. Especially the fixer-connections between firms were grey and often corrupt. Official channels existed to change the plan when something went wrong. However, they were very cumbersome and it was much better to »fix« something through grey channels and middlemen (tolkach). The government and the party knew that such networks provided some flexibility to the economy, so they were tolerated, if they did not go too far. In fact, directors who fulfilled plan targets were rewarded, and those who stuck to the rules and failed were punished. Many stories have been reported in the press where such networks got rather far (involving hundreds of people protected by local top-politicians making millions of roubles) before they were stopped.
Table 3. Two features of communist systems giving non-system cooperation
Feature Plan fulfilment pressures and
Supply constraints causing
slowness of planning apparatus queuing and quality problems
Fixer and barter networks. Often large scale deals.
Shopping and barter networks. Premiums on good connections.
Grey, often tolerated. Often close to corruption
Partly legal or moderately grey. Sometimes close to corruption
Households were supply-constrained (ie pressures of excess demand existed) and thus had forced savings available if provisions of goods and services could be obtained - one way or another. On the »white« side households were known to help each other with queuing, and bits and pieces of »open free« markets were always permitted. However, much was in the »grey« or even »black« sectors of the economy. In Poland from the late 1960s an open grey/black market for foreign exchange was tolerated. Traders were not allowed to set up exchange shops with permanent premises, but they walked the central streets and squares in the main towns in a highly visible way with bags of money. The Central Bank and the planning authorities kept careful track of the parallel rate of exchange and would openly admit that this was considered a safety-valve for the excess demand. Finally, it should be noted that the true costs of goods and services were hard to know in the communist economy and rather irrelevant to most decision makers. People who had the right to distribute goods and services could often do so with little regard to the costs. Hence, values for goods and services in different uses varied and differed from the true cost structure. In other 8
words the price system had many irrationalities, and rents appeared in mysterious ways to be picked by the audacious.16) Once again, connections and »grey« fixer networks came to mean a lot. The above picture of the communist economy thus describes a society, where the grey/black sector has a big potential and is allowed some leeway, but it is still kept under control by several control systems. It consisted of several police corps, the banking system had financial control as its main job, and in the end everybody was controlled by the party. Column 1 on Figure 1 shows the social capital consequences of model described. The grey/black networks were necessary for the system. In a sense they represented normal social capital, but they were precariously near to negative social capital. Clearly such networks are harmful in a market economy, where the trading rules are to be respected not bypassed. With the softening of communism the safeguards and controls of the system weakened, allowing the grey/black networks to bloom. Also, the official organizations gained some independence. We have seen and heard many stories from Eastern Europe describing such developments.
Figure 1. The model of social capital development in the OC-countries presented
Column 1 Communist Controls
Column 2 Transition economy
Column 3 NW European situation Controls
Grey/black nets* Positive social capital Official organization * Necessary, thus ac- cepted, but controlled
Grey/black nets* Positive social capital * Harmful, but hard to control
Grey/black nets Positive social capital
16. A well-known example was the wild prices for western jeans. Another example was Georgian traders flying to Moscow to sell as many tomatoes as they were allowed to carry onboard and still making a profit. 9
Figure 1 shows three columns. Column 1 for the communist society and Column 3 for the developed market economy are very different. The latter has a moderate control system, a small grey/black sector, and much social capital. The communist model has much more control, a larger grey/black sector, much official organization and less normal social capital. The left-hand side graph also illustrates the trade offs between the size of the control apparatus and the grey/black sector. This can be summarized as: Hypothesis two: The communist model generates grey/black networks, which are kept at bay by a large control apparatus. When a dictatorship ends, social capital is low. The official organization made by the old regime crumbles. People have learned distrust and to take no initiative of their own. At the same time many other institutions weaken. This is particularly important for a transition from communism: Hypothesis three: The transition switches the grey/black networks from necessary into harmful, and the weakening of institutions allowed this negative social capital to grow and become a barrier to a normal economic development. IV Post communism: The problematic inheritance The development from Column 1 to Column 2 on Figure 1 shows - in a highly stylized way - our model of the transition of social capital. It did not develop in the direction of Column 3 of the developed market economy, but in a much less favorable way. The control system weakened and the grey/black networks - already in place - grew and turned decisively negative. At the same time the official organizations collapsed, and only much smaller voluntary organizations remained. IV.1 The normalization horizon: Putnam's pessimism and other evidence Putnam's study of Italy deals with the development of institutions of governance in Italy,17) as it analyzes the efficiency of new regional administrations set up after a municipal reform in 1970. It shows large differences in the efficiency of the new administrations - especially between the North and the South. This is ascribed to the difference in social capital in the two parts of Italy due to the past. However, the two areas were united in 1861 - no less than 109 years before the reform. The institutions of the Kingdom of Sicily were not dissolved the day Italy was united, but allowed to fade away. However, in any case the time lag of at least half a century failed to close the social capital gap, and it is supposed to remain till this day. This is a deeply pessimistic perspective when we turn to the OC-countries. Fortunately, other evidence exists. The study of Spain already mentioned (Torcal & Montero, 1999) demonstrates that existing evidence shows that Spanish social capital ceased to 17. In a corresponding study Helliwell & Putnam (1995) demonstrated that the conclusions for institutionbuilding holds for economic development as well. The connections in the Italian data are further discussed in Granato, Inglehart & Leblang (1996) and Jackman & Miller (1996). 10
differ from other Latin European countries already 1-2 decades after democracy was reintroduced.18) However, as we will see below Latin DCs have low social capital compared with other DCs. This also applies to France with its old tradition for democracy. The study by Stolle & Rochon (1999) compares available measures of social capital in the USA, Sweden and Germany. It shows no clear difference, even when Germany had democracy a century later than the other two countries, and, in addition, had a short, but nasty, spell of totalitarianism 1933/45.19) We thus conclude that the adjustment period does not need to be more than a few decades. The reason why a large difference appears between Italy, Spain and Germany may be that social capital has a propensity to stay in certain equilibria. Below we argue that social capital instead of developing good dynamics may go bad and get stuck in a bad equilibrium.
IV.2 The replacement of public mass organizations with voluntary organizations The study of Siisiдinen (1999) compares social capital in three Nordic countries (Norway, Sweden and Finland) and the three Baltic countries using Putnam's Instrument. The main results are cited in Table 4. The comparison is made using a set of polls done in the late 1990s.
Table 4. The number of voluntary organizations the average citizen in 6 countries belongs to
Number Finland Norway Sweden Estonia
Source: Siisiдinen (1999). Polled data covering 1-2000 people in each country during the late 1990s.
Siisiдinen makes the point that the frequency of membership in organizations before 1989/90 was much the same in the two country groups. However, here the organizations in the Baltic's were relatively few, but large, state-controlled mass organizations, while there were many more, but often smaller, voluntary organizations in the Nordic countries. Since 1989/90 small changes only have occurred in the Nordic countries. The system controlling the mass organizations in the Baltic's collapsed in 1990, so did the most of the organizations. All three Baltic polls report a reduction of two-thirds of member-
18. The reader may remember that the Spanish Republic lasted one decade from 1923 till the civil war started, while General Franco ruled from 1936 till 72. The story of Spain the last half century before 1923 is rather complex, with shifting monarchies, dictators and some democracy. 19. German history before the Weimar Republic (1918-33) was that of a weak absolutism after the unification 1864-7, as the reader is likely to remember. 11
ship in the old mass organizations. In the process some organizations disappeared while others managed to change and stay alive, and some of them even started to grow. Also, new organizations were formed. However, there is still much less social capital than in the Nordic countries. Much the same findings appear in Rose's (1999: p.151) study of Russia. He finds that Russia today continues to suffer from a »missing middle« of organizations linking informal grass root networks and modern organizations. Rather, the gap is filled by antimodern enterprises run by former officials or by Mafia organizations. We return to Rose's theory of antimodern society below. IV.3 Measures of generalized trust and corruption The World Value Survey brings numbers for many polled items, see Inglehart et al (1998). One is trust, measured in a general way - that is, the respondents are asked if they »trust« other people.20) This is perhaps the best measure available of social capital. Figure 2 compares the answers with this question with the corruption perception index, , for the same countries.21) Note that the -index is really an honesty index, as low values show corruption and high values honesty.22) We have argued that this is the best measure available for negative social capital. The country observations on Figures 2 to 4 have signatures dividing them into the following groups: OC is the OC-countries and Or are the oriental countries. Two groups of (high-income) DC's are shown: WL is Western-Latin and WN is Western-North, as they are dominated by German-British cultures. Belgium has an intermediate marking to show that it is half way in both groups. Finally, Res is the group of the remaining countries. This group contains 4 Latin America
n countries. Argentina, Brazil, Chile and Mexico. 20. The exact question is: »Generally speaking, do you believe that most people can be trusted or can't you be too careful in dealing with people?«. It is discussed and many aspects are analyzed in Uslaner (2000). 21. The -index used is the one published by Transparency International, a German based NGO with the net address: . It is an aggregate of all available cross country measures of corruption. 22. Treisman (2000) and Paldam (2000a) are the two most recent studies explaining the cross-country pattern in the -index with economic and cultural variables. The main explanatory variable is the GDP- level, but also inflation is a strong explanatory factor. The regulatory system in the country also matters. A wave of high inflation quickly increases corruption. The OC-countries had probably lower corruption before the transition than now. However, it was still relatively high thanks to the high rent seeking potential in the old system. Most of the increases can be explained by falling real income, and the wave of high inflation in connection with the change of system. 12
Figure 2. Generalized trust and corruption in 40 countries as discussed in text Correlation r = 0.65 The first observation from Figure 2 is that the points follow a clear pattern. The two series are both hard to measure, and thus likely to have considerable measurement errors. The errors are unlikely to be correlated as the two series are reached by different methods. So the series are so closely correlated as one could wish. Secondly, we note that the OC-points are totally separate from the WN-points: as a crude rule of thumb trust is twice as high in the WN-countries and so is honesty. Interestingly, the WL-countries are in between. So are the Latin American countries. Two countries of interest stick out: It is China and Russia, both having relatively high trust. The result for China is probably a serious measurement error (see Uslaner, 2000). The result for Russia is less extreme, but it still causes Shleifer & Treisman (2000; p 99-105) to downplay the importance of social capital. However, other studies find a low level of trust in Russia. World Bank (1996; p 94) report that government credibility is low in Russia and Eastern Europe in general. Rose and Mishler (1998) found similar results. A battery of questions about trust in the institutions of Russian society showed that most Russians distrust every major institution, especially the representative institutions of governance. IV.4 The perceptions of political participation A second variable from the World Value Survey is closely related to social capital, as shown by Deth et al (1999). It is the density of civic participation.23) Again it is a difficult variable to 23. Respondents were asked whether they participate in different civic activities, i.e. voluntary, activities, including: (a) social welfare services for the elderly and deprived; (b) education, art and cultural 13
measure, and the data reached for two countries differ very much from the others. It is available for slightly fewer countries. Figures 3 and 4 compare the answers with the two variables of Figure 2. Both graphs support the view that we are dealing with aspects of something robust. The WN and the OC-points are always separate with the WN-countries in the top-right corner and the OC-countries in the bottom-left corner, indicating low values of the social capital related variables. The difference is large in relation to the scale used. Given that the OC-countries have too low GDP relative to their level of physical and human capital (their (K, H)-values) we conclude that we have found evidence suggesting that they are held back by their low social capital. The OC-countries have the same social capital as the Latin-American countries, which have the same GDP-levels as well. Figure 3. Generalized trust and perceived political participation in 33 countries Correlation r = 0.62 Things thus hang nicely together until one looks at the WL-countries. They are closer to the OCcountries than to the WN-countries in social capital according to most of the indicators tried. Especially the low level of trust (22) in France is puzzling. The WL-countries are as rich as the WN-countries, especially France and the northern parts of Italy and Spain. Note, however, that the lowest scores of the social capital related variables are the ones to perception of participation. In spite of recent reforms, the Latin countries are relatively centralized, with powerful central activities; (c) local community affairs; (d) activities related to conservation, environment and ecology; and (e) work with youth. The density of civic participation is measured as the percentage of these civic activities in which an average respondent in a country is involved. 14
administrations controlling weak local governments. Clearly, much about social capital is puzzling.24) Figure 4. Perceived political participation and corruption in 33 countries Correlation r = 0.71 A similar paradox is provided by Putnam (2000) demonstrating by a wealth of statistics that (nearly) all indicators of social capital have gone down in the last two decades in the USA. This large reduction has occurred during a period of particularly rapid economic growth in the country. These developments contradict the idea that social capital is an important factor of production. However, disregarding the contradictory evidence, we conclude that the OC-evidence speaks strongly of a low level of normal social capital and a high level of negative social capital. We would very much have liked to have had time series for the relevant variables, but only bits and pieces are available. IV.5 The growth of the »shadow« economy Several studies try to put some time series evidence together. Much evidence is found in a recent conference volume (Feige & Ott, 2000) and surveyed by Schneider & Enste (2000) and most recently by Schneider (2000). Given the inherent difficulties of measurement the evidence is 24. Several of the leading researchers (notably Anirudh Krishna) stress that social capital is a »latent« variable that has to be measured by measures which are relevant for the particular environment. While Putnam's instrument seems relevant in an »European« cultural environment, it does not work in India, where the level of voluntary organizations are low. In fact, the level of voluntary organizations in India is lower than in most African countries. However, in other dimensions India scores (much) higher. 15
surprisingly coherent. The studies quoted by Schneider suggest that the shadow economy has grown from about 16% of GDP in the average OC-country to about 40% from the late 1980s to the mid 1990s. The growth is reported to be above average in Russia, and below average in the smaller, westernmost countries as Estonia, the Czech Republic and Hungary. In Poland, the shadow economy appears to be shrinking, but it was relatively large before the transition started. The numbers now reached are much like the ones calculated - by similar methods - for the Latin American countries and much higher than the ones for the WN-countries. Once more the WL-countries are in between. The numbers for the shadow economy given have thus much the same pattern as the other social capital indicators, even when they are calculated by models using standard economic variables. IV.6 Two related ideas A whole set of studies by Richard Rose (see, eg Rose and Mishler (1998) and Rose (1999)) develop a theory of antimodern social capital. His main point is that social capital networks (used to produce goods and services) are distinctive in a society characterized by organizational failure and corruption of formal organizations. In response against the State, individuals can invoke informal networks: begging or cajoling public officials, using connections to »bend« rules or paying bribes that break rules; social networks compensate for organizational failure. In this way, Rose (1999: p.156) shows that citizens in Russia have formed many private networks for getting things done. Rose (1999: p.148-49) furthermore argues, that the persistence of such networks is a formidable barrier to Russia's transition from an antimodern to a modern society with a wellfunctioning market economy. These informal interest groups maintain status quo by exploiting their personal contacts, barter or bribes to get what they want within the existing system and they range widely from ordinary households to the mafia. In all cases, mutual cooperation is based on the morality of face-to-face groups that Max Weber characterizes as Binnenmoral. The complement is »outsider moral« (Aussenmoral) which cannot stand on its own and therefore justifies formal organizations (ibid: p.150). In fact, these informal groups are utilizing the same potential for »rent-seeking« as are formal interest groups such as state monopolies. They trust each other when undertaking these unofficial and hidden activities. The second theory is due to Chand & Moene (see 1999a, and especially1999b). They have developed a rather general corruption model with two equilibria. One has low social capital, while the second has high social capital, ie they correspond to the OC bottom-left corner and the WN top-right corner on the graphs. The second paper applies the model to Russia. The variables used in the theoretical argument are not all observable, but everything looks much as our verbal argument. It is reassuring that the logics pursued in this paper can be so neatly formalized. 16
IV.7 Back to Figure 1: some additional mechanisms 25) When we compare Columns 1 and 2 on Figure 1 they imply a demand theory of control. Nobody likes to be controlled, so when the system mellowed, the controls were reduced, and when the system broke down in 1989/90, the controls were quickly reduced. The police systems turned into normal »civic« police. The banking system turned into a Western banking system dealing with credit no longer acting as a controller for the state. Finally, the all-over controller, the Party, lost power. It should be added that at the same time the public salaries dropped making the new police and the personnel in other institutions more vulnerable to the temptations of corruption. Also, the first decade after the system change was one of large scale privatization. It is always difficult to privatize, and here it was done by new and inexperienced regimes during a period of chaos. So much arbitrary redistribution - sometimes known as rent grabbing - took place. Arbitrary redistribution was also a side effect of the waves of high inflation, which followed the big system change. Arbitrary redistribution is surely demoralizing for the population to experience. For all these reasons it is understandable that the low post communist social capital has been slow to grow into the normal size fitting with the other factors of production. It also helps to explain that instead it deteriorated into the negative. V Concluding remarks It is important to start the conclusions by stressing our limited knowledge. Systematic theorizing about social capital is recent. We understand little about the way social capital is built. Data about its size and its development over time are still scanty. The following conclusions are therefore tentative and made to encourage further research. We have presented an argument summarized in three hypotheses: Hypothesis 1 argued that the communist states - as all other dictatorships - destroyed social capital. The data presented showed that it is as low in the transition economies of East and Central Europe as it is in the Latin American countries. We have further argued that this may explain why the GDP-level is low compared with the levels of physical and human capital. Hypothesis 2 claimed that the communist system had the characteristics that it needed a certain amount of grey/black networks to function. Such networks were consequently tolerated, even when they were kept under control by massive control systems. Hypothesis 3 maintained that the big change in 1989/90 switched the old necessary grey/black networks into (harmful) negative social capital, and the reduction in the control systems (and the transition problems) allowed these networks to grow. The data presented were weak on the time series evidence, but certainly supported the idea that the level of negative social capital is high in the former communist countries. When we turn from the past to the future, the two main problems are: (a) How fast is the 25. The additional mechanisms mentioned in this section are treated in much detail in the transition textbooks as Gross & Steinherr (1995) and Lavigne (1999). 17
building process for (positive) social capital going to be? (b) Can something be done to speed up this process of social capital building? Evidence about point (a) is not plentiful, but once again scattered observations exist. We have argued that a couple of decades may be sufficient. On point (b), it is important to distinguish between active interference and passive support. The logic of social capital warns us against active interference. However, it seems likely that passive support may help. That means that governments can do much by creating the proper enabling environment for social capital generation and by fighting negative social capital. 18
References: Chand, S.K. & Moene, K.O., 1999a. Controlling fiscal corruption. World Development 27: 1129-40 Chand, S.K. & Moene, K.O., 1999b. Rent grabbing and Russia's economic collapse. Working Paper, Department of Economics, Oslo University Coleman, J.S., 1987. Social Capital in the Creation of Human Capital. American Journal of Sociology 94: 95120. Reprinted in Dasgupta & Serageldin (1999) Dasgupta, P. & Serageldin, I., eds., 1999. Social Capital: A Multifaceted Perspective, World Bank: Washington, DC Deth, J.W.v., Maraffi, M., Newton, K., Whiteley, P.F., 1999. Social Capital and European Democracy. (ECPR Studies in European political science) Routledge: London Fainsod, M., 1958. Smolensk Under Soviet Rule. (The Rand Corporation) Vintage Books: NY Feige, E.L., & Ott, K., 2000. Underground Economies in Transition. Unrecorded activity, tax evasion and organized crime. Ashgate: Aldeshot, UK. Fukuyama, F., 1995. Trust. The Social Virtues and the Creation of Prosperity. Hamish Hamilton: London Granato, J., Inglehart, R. & Leblang, D., 1996. The effect of cultural values on economic development: Theory, hypotheses, and some empirical tests. American Journal of Political Science 40 Special Issue: 607-716. Gros, D., & Steinherr, A., 1995. Winds of Change. Economic Transition in Central and Eastern Europe. Longman: London Helliwell, J.F. & Putnam, R.D., 1995. Economic growth and social capital in Italy. Eastern Economic Journal 21: 295-307 Inglehart, R., Basaсez, M. and Moreno, A., 1998. Human values and beliefs: a cross-cultural sourcebook. Political, religious, sexual, and economic norms in 43 Societies. Findings from the 1990-93 World Values Survey. University of Michigan Press: Ann Arbor, MI Jackman, R.W. & Miller, R.A., 1996. A renaissance of political culture. American Journal of Political Science 40 special issue: 607-716 Lampedusa, G.T.di, 1958. The Leopard. (Italian original, translated into many languages, many editions) Lavigne, M., 1995, 1999. The Economics of Transition. From Socialist Economy to Market Economy. Macmillan: London Nove, A., 1977 (1961). The Soviet Economic System (The Soviet Economy). George Allen and Unwin: London Olson, M., 1996. Big Bills Left on the Sidewalk: Why Some Nations are Rich, and Others Poor. Journal of Economic Perspectives 2: 3-24 Paldam, M., 2000a. The big pattern of corruption: Economics, culture and the seesaw dynamics. Pt working paper available from 19
Paldam, M., 2000b. Social Capital: One or Many? Definition and measurement. To appear in the Journal of Economic Surveys. Pt working paper available from Paldam, M., & Svendsen, G.T., 2000. An essay on social capital: Looking for the fire behind the smoke. European Journal of Political Economy 16: 339-366 Putnam, R.D., 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton UP: Princeton, NJ Putnam, R.D., 2000. Bowling Alone. The Collapse and Revival of American Community. Simon & Schuster: NY Rose, R., 1999. Getting Things Done in an Antimodern Society: Social Capital Networks in Russia. In Dasgupta & Serageldin (1999: 147-71) Rose, R., & Mishler, W., 1998. Untrustworthy Institutions and Popular Response. Studies in Public Policy No. 306, University of Strathclyde, Glasgow Schneider, F., & Enste, D.H., 2000. Shadow Economies: Size, Causes, and Consequences. Journal of Economic Literature 38: 77-114 Schneider, F., 2000. Illegal Activities, but still value added ones (?): Size, Causes, and Measurement of the Shadow Economies all over the world. Arbeitspapier 0010, Department of Economics, University of Linz, May Shleifer, A., & Treisman, D., 2000. Without a Map. Political Tactics and Economic Reform in Russia. The MIT Press: Cambridge, MA Siisiдinen, M., 1999. Voluntary associations and social capital in Finland. Pp 120-43 in Deth et al (1999) Stolle, D. & Rochon,T.R., 1999. The myth of American Exceptionalism: a three-nation comparison of associational membership and social capital. Pp 192-209 in Deth et al (1999) Svendsen, G.L. and Svendsen, G.T., 2000. Measuring Social Capital: The Danish Cooperative Dairy Movement. Sociologia Ruralis 40: 72-86 Torcal, M., & Montero, J.R., 1999. Facets of social capital in new democracies: the formation and consequences of social capital in Spain. Pp. 167-191 in Deth et al (1999) Treisman, D., 2000. The causes of corruption: a cross-national study. Journal of Public Economics 76: 399-457 Uslaner, E.M., 2000. The Moral Foundations of Trust. Book manuscript preliminary version available from Wintrobe, R., 1998. The Political Economy of Dictatorship. Cambridge UP: Cambridge, UK World bank, 1996. From Plan to Market. World Development Report, 1996. The World Bank: Washington DC World Bank, 1999. Local Level Institutions: Social capital for Development. Conference at the Chr Michelsen Institute, May 5-7, Bergen, Norway. Conference volume forthcoming. 20
M Paldam, GT Svendsen