1. Economic outcomes before and after the break-up of Yugoslavia
What are the costs of nationalistic policies? The expectations may be more optimistic than is warranted as the example of the breakup of Yugoslavia suggests. Assuming that nationalists expected that economic results will be better than in federal Yugoslavia, it makes sense to get some idea of what an alternative to a nationalistic strategy would have achieved in the last 20 years. The alternatives are nationalism vs. integration, not necessarily independence vs. federal Yugoslav state.
That allows Slovenian development to be the basis for counterfactual simulations of where would have other Yugoslav states been had they followed its strategy of transition. This because (i) Slovenia opted for gradual transformation of the inherited Yugoslav institutions to those characteristic of the European Union (this is sometimes seen as a type of economic nationalism though in view of the next characteristic it is more of a gradualist strategy of transition); (ii) chose to integrate with the EU and indeed the EMU, both as soon as possible, rather than pursue a protectionist, nationalistic strategy; and (iii) fared well, though not miraculously well. Comparing Slovenia with the other Yugoslav countries it is possible, because of (i), to indirectly evaluate the quality of the starting position, i.e. the Yugoslav institutions, to assess, because of (ii), the costs or benefits of the alternative nationalistic strategies, and thus to get an idea of the costs of nationalism, in view of (iii).
I will rely on four indicators: GDP growth and GDP per capita levels, employment and unemployment, openness in terms of exports, and industrial development. There are two points to stress before actual comparisons are made.
First, Slovenian economy grew faster than those of most other transition countries in the last 20 years, except for Poland and Slovakia when the last two crisis years are included. Still, that growth was not exceptionally high, in real terms about 2.3% per year. As a consequence, Slovenian GDP in 2010 stood at 157.2% of that in 1990 (at constant prices).
How important is this growth rate for the other Yugoslav countries? That is the other point. In the Yugoslav times, regional differences in terms of GDP per capita were practically constant (there were of course cyclical variations). In other words, there was no convergence in terms of GDP per capita. Indeed, there has not been much of a convergence with the EU in the case of Slovenia since independence and there was quite a divergence in the case of other countries in the last 20 years too.
So, it is not completely arbitrary to suggest that if other Yugoslav countries had followed Slovenian strategy of transition, they would have achieved Slovenian growth and kept their distance, in terms of GDP per capita, where it had been throughout the Yugoslav period. Assuming that, in terms of Slovenian GDP per capita, the first column gives GDP per capita in 2010 euro, the second the distance from that of Slovenian GDP per capita in 2010, the third the same just before the break up, the fourth the hypothetical GDP per capita in 2010 if regional differences would have stayed as they were in column three, while the fifth column gives the shortfall in 2010 euro. The sixth column reports average annual real growth rates from 1990 to 2010, while the last column gives the distance of countries’ GDP per capita from the Slovenian GDP per capita in euro purchasing power standard in 2010.
Bosnia and Herzegovina
Note: 2010A is actual GDP per capita in 2010, GDPpc and GDPpc/Y are comparisons with Slovenia now and in the Yugoslav period respectively, 2010H is hypothetical GDP per capita in 2010 (if regional differences had been as in GDPpc/Y rather than GDPpc) A-H is 2010A – 2010H, ∆GDP is average growth rate per year, and GDPpc/P is 2010 GDP per capita at purchasing power parity with Slovenia as 100.
Sources: Gligorov (2002a), wiiw, own calculations.
The differences are striking especially in the case of Serbia. This is in part due to the difference in the growth of nominal GDP in euro in comparison with the real GDP growth (column six). For instance, in the case of Croatia, real GDP has increased by only 11% since 1990, and in the case of Macedonia only 15%, which is reflected in their annual average real GDP growth rates of 0.5 and 0.7 respectively. Other countries have recorded negative average real GDP growth rates. However, price levels have diverged a lot, mostly due to different exchange rate policies, so that Croatia’s GDP per capita looks better than its real GDP growth would indicate while Serbia’s looks worse.
In terms of GDP per capita in purchasing power parity euro (PPP, last column in Table 1), the deviation from the exchange rate is much smaller in the Croatian case than in the case of all the other countries. Given the growth record, that suggests that the exchange rates are probably overvalued and relative growth of consumption to that of investments much faster, except perhaps in the case of Macedonia. There is no PPP data for the Yugoslav period, but the stability of the divergence in GDP per capita over longer period of time would suggest that probably the price level changes were not all that different across Yugoslav regions and have clearly increased since.
What has happened with the openness of these economies? In Table 2, in the first two columns, exports of goods to GDP in the Yugoslav period (i.e. just before the dissolution) can be compared with those now. These include intra-Yugoslav and now intra-regional trade and the change is striking. Slovenia has increased its exports of goods (as percent of GDP) and Bosnia and Herzegovina has almost kept its level of openness while all the other states now have more closed economies than in their Yugoslav period. The worst performers are Montenegro and Croatia with Serbia close by. Macedonia does somewhat better, but these are small economies and the level of openness was not all that high before the break up and it is quite low now.
Bosnia and Herzegovina
Note: X/GDP/Y is export as % of GDP in Yugoslavia, X/GDP/N is the same now, Un%/Y is unemployment rate in Yugoslavia and Un%/N is the same now.
Sources: Gligorov (2002a), wiiw, own calculations.
This is mostly the consequence of the decline of industrial production. Even if the effects of the crisis in the last two years are put aside, in 2008 industrial production of Slovenia was just short of 30% higher than in 1990 while that of Croatia was only slightly above 90% of what it was in 1990 and that of Macedonia a trifle over 60%. Industrial production in other countries in 2008 was around 50% or below that achieved in 1990. Of course, the crisis has led to further decline. However, even in the last ten years, when GDP growth was rather strong, industrial production grew rather slowly (2010/2000): Slovenia, 118.8; Bosnia and Herzegovina 187.3%; Croatia 122.6%; Macedonia 100.9%; Montenegro 90.2; Serbia 106.1. Speedy growth in Bosnia and Herzegovina is of course just the indication of the extremely low level from which its industrial production is recovering after the war and other destructions.
Even in the case of Slovenia, growth of industrial production is less than impressive and is quite a bit slower than that of its GDP growth, though the performance of the latter has not been stellar too. Overall, the rest of the post-Yugoslav region has gone through quite significant process of deindustrialization. That, of course, is also reflected in the already noticed low level of exports of goods.
In Table 2, the last two columns show rates of unemployment. Unlike in the other socialist countries, employment was to an important extent market determined in Yugoslavia; it was voluntary and not compulsory as in soviet type socialist countries, which also supported significant intra-Yugoslav mobility and also continuous outward migration. Thus, significant unemployment existed and regional differences were large. Comparing employment and unemployment, however, is a bit difficult because figures for the Yugoslav period are drawn from registration, while those for today are based on labor force surveys (which include informal employment). Even now, registered unemployment is higher or even quite higher than survey unemployment, so it can be assumed that actual unemployment rates were if anything lower, and employment rates higher, than those recorded. So, the comparison is most probably biased against the state of the labor markets in the Yugoslav period. Also, these figures are from the time just before the break up and are higher than at any time before, and somewhat similar is the case of 2010. However, the comparisons of other years or with some kind of an average would not lead to different qualitative conclusions.
As can be seen in Table 2, Slovenia has a significantly higher unemployment rate than in 1989 – 7.5% rather than 3.2% (and below 2% from 1952 onwards). Croatia’s unemployment rate was 8% just before the independence and was 12% in 2010 (around 5% from mid-1960s till the pre break up hike). Serbia’s (without Kosovo) unemployment rate was less than 15% and is about 20% now. Macedonia and Bosnia and Herzegovina had unemployment rates around 20% and about 10 percentage points higher now. Montenegro is an exception because it’s unemployment rate jumped strongly just before the breakup of Yugoslavia and is practically at the same level now, again jumping somewhat as a consequence of the current crisis. Demographic factors played significant role in these labor market results, but that aside, labor market outcomes have been inferior to those achieved in Yugoslav times. The latter was especially beneficial to Slovenia and Croatia as they tended to have rather low or quite low unemployment rates.
So, what can be concluded about nationalism vs. integration? Yugoslav institutions were not supportive of regional convergence and those less developed suffered from high unemployment, in part because of faster growth of population (e.g. in Kosovo). However, if Slovenia is at all representative, the transformation of Yugoslav institutions to those characteristic of the EU could have been achieved gradually and with much better results than the alternative nationalistic policies. That counterfactual is not conditional on the continuation of the existence of federal or any other Yugoslavia, as the EU integration process is also an alternative to nationalistic strategies. Again, if Slovenia can serve as an example, even without GDP per capita convergence with the EU and within the Yugoslav group of countries, as has been the historical record of integrative processes so far, the outcomes would have been vastly superior to those achieved with one or the other variety of nationalistic strategies, political and economic, pursued by the Yugoslav successor states.
2. Yugoslav lesson for EU
What are the institutional or rather constitutional similarities between Yugoslavia and the EU if any? The key similarity is the idea that economic and political union is sustainable without a fiscal union. Yugoslavia gradually fiscally devolved, especially from mid-1960s onwards, while EU was expected to gradually increase its fiscal integration in the course of gradual political integration. These comparative developments open up a fundamental debate about political stability in alternative fiscal arrangements or, to put it differently, about political fragility of alternative fiscal arrangements.
Few political economy models of state fragility have been inspired, in part to be sure, by the disintegration of Yugoslavia. Perhaps the most well know are those by Alesina and Spolaore (1997 and 2003) and Bolton and Roland (1997). The main point is that regional inequality and in particular growing inequality (e.g. in income per capita) supports secessionism for distributional reasons while too much intra-national redistribution supports integration. Put differently, fiscal integration may not be sustainable in countries with large regional differences while market integration may be desirable in countries that follow excessively nationalistic economic policies, i.e. hurt efficiency with too much redistribution. Clearly and ideally, federal or other types of complex states or political unions (e.g. EU) could combine fiscal decentralization with market integration and thus provide for the right level of redistribution and optimal efficiency.
In the Yugoslav case, after the constitutional reform of 1974, the federal budget, apart from transfers to the less developed regions, basically consisted of expenditures on the military and on the federal administration, while tariffs were for the most part the only direct source of revenues; the rest were contributions from the member states and provinces. The structure of sources of public revenues was quite similar to that of EU, though the key part of public expenditures was devoted to the common military, which is not the case with the EU. This lack of fiscal integration was substituted, or that was the intention, with tax harmonization and regulation in general, out of budget transfers to less developed regions, and the monetary union – which is also similar to the EU.
Transfers to the less developed regions (something like cohesion funds, which were not formally part of the federal budget) were unpopular and can be seen as an example of what the cited political economy literature has identified as the problem of different regional preferences for levels of taxation – the rate of contribution to the cohesion funds (which was the same for everybody) being too high for more developed regions and too low for less developed ones. In other words, less developed regions thought transfers were too low while the more developed regions thought they were way too high. These transfers did support inter-regional trade and trade deficits of less developed regions with the more developed ones, which was also the reason for never ending political debates. They were discontinued before the country actually disintegrated though.
The behavior of the central bank and of the financial system is also quite illuminating. Since mid-1960s, commercial banking developed while the public chose to save and price in German mark, D-mark, so that the central bank chose to manage the exchange rate with a preference for an adjustable peg with the D-mark. Central bank’s monetary policy was somewhat complex, but real interest rates tended to be negative for the country as a whole and at different levels across the regions (as regional inflations differed). In addition, banks were exposed to currency risks because their liabilities were to a very large extend in D-marks, while assets were in dinars. Putting a lot of things aside, the central bank accumulated a lot of contingent liabilities and increasingly functioned as a quasi-fiscal instrument. It bought mostly corporate debt at preferential discount rates and thus eventually needed to be recapitalized. That proved to be difficult given the non-existence of a fiscal union. Indirectly, the restructuring of foreign debt within the program with the IMF (in 1982 and again in 1988) introduced common and separate responsibility for the debts on the federal and regional levels, but that proved to be difficult to implement, and in any way, after a brief attempt to reform and start transforming the system in 1990, the country broke up.
Thus, the main Yugoslav lesson for the EU is that fiscal harmonization and monetary union are no substitute for a fiscal union. The other way to see this point is to consider the interplay of monetary and fiscal policies. In a closed economy, there is no difference or rather there is no inherent conflict between the two. In an open economy, however, monetary policy dominance is required by external sustainability while fiscal policy dominance is required by internal sustainability (Leeper 1991). Once disequilibria develop, fiscal policy will tend to dominate and if it is decentralized it will strain monetary policy to the point of monetary disintegration. That does not need to lead to nationalistic policies taking over, but that is certainly one of the possible outcomes of monetary disintegration.
3. Representation and taxation
In political economy models, democratization supports disintegration of non-homogeneous political systems because it allows each region or member state to choose its preferred level of redistribution (e.g. tax rate, but even more level and structure of public expenditures). This is based on the assumption of regions being in politically relevant sense more homogeneous separately than when integrated. However, political homogeneity is more an outcome of the political system than the other way around. This is because, as I argued twenty or more years ago, in general, one can always ask the question, “Why should I be a minority in your state, when you can be a minority in mine?” At a point when democratization raises that question, nationalistic alternative may lead to internal as well as international conflicts. That is indeed the lesson on which European Union was based and justified and which has been forgotten or is in the process of being forgotten with the rise of nationalism and the rejection of multiculturalism in contemporary Europe.
Drawing on the historically closer lesson of the breakup of Yugoslavia, there are two points that may be of interest. One is about the causes for the rise of nationalism. The other is about one necessary condition for sustainable democratization in complex states.
Disintegration of Yugoslavia came at the end of a decade long constitutional and economic crisis. As constitutional solution was not found, nationalism proved popular because it generally seemed to offer solutions to prolonged economic crisis or stagnation. Nationalism looks like a solution to the distribution of a shrinking pie or of the pie that is not increasing because it tends to decrease the number of those that can legitimately claim a piece of the pie or ration its distribution discriminatingly. It is straightforward from that to infer that the implied redistribution will also tend to shrink the pie additionally, which is what the Yugoslav experience, if no other, exemplifies. That even if no other conflicts are assumed.
The other point is that rules are no substitutes for representation. They are a form of taxation, or rather a substitute for it, and thus require democratic representation, which is the lesson being learned in the current crisis. EU rules, fiscal and most other, are only implemented and observed if they are, in the jargon, incentive compatible, otherwise they are overridden by domestic democratically adopted decisions. Yugoslavia produced constitutions and legal and other rules on a large scale, but those that these rules taxed were not properly represented. In addition, and looking at the inverse of “no taxation without representation”, representation without taxation is not binding and does not support the democratic game. The EU has national and even democratic representation, but taxes by rules and not fiscally. Thus, it risks national breach of rules and no democratic interest in EU representation.
In general, a political union needs both: representation and taxation.
4. A conclusion
Slovenian transition has been a repetition of the process of EU integration: market integration, followed by financial and monetary integration. Yugoslav example is useful at the point when the choice is made whether to go forward with the fiscal integration or to reverse the integration process and go back to nationalistic polices. The costs of reversal were high in the case of Yugoslavia even without counting bodies and valuing the sheer physical destruction.
Addendum: A very short note on Kosovo
Data for Kosovo are still rather scarce and because of that it is left out here. I have looked at its development over time (Gligorov 2000b, 2002b, 2007) and concluded that it does not represent a special case in the context of the strategic choice between nationalism and integration. One important difference is demographic, so in terms of GDP per capita, Kosovo fared much more poorly than the other Yugoslav political entities and regions. However, in terms of GDP growth, that is not true, as Kosovo’s economic growth was for the most part faster than that of the rest of Yugoslavia. In the last twenty years, the costs of Yugoslav disintegration have been exceptionally high in Kosovo, though nationalistic expectations there as in other nations were different. In any case, it is clear that integration is advantages to Kosovo’s development.
Alesina, A., E. Spolaore (1997), “On the Number and Size of Nations”, The Quarterly Journal of Economics 113: 1027-1056.
Alesina, A., E. Spolaore (2003), The Size of Nations. The MIT Press.
Bolton, P., G. Roland (1997), “The Breakup of Nations: A Political Economy Analysis”, The Quarterly Journal of Economics 113: 1057-1090.
Dornbusch, R. (1992), “Monetary Problems of Post–Communism:Lessons from the End of the Austro–HungarianEmpire”, Weltwirtschafliches Archiv 128: 391-424.
Eichengreen, B. (2007), “The Breakup of the Euro Area”, NBER Working Paper No. 13393.
Feldstein, M. (1997), “The Political Economy of the European Economic and Monetary Union: Political Sources of an Economic Liability”, Journal of Economic Perspectives 11: 23-42.
Gligorov, V. (1994), Why Do Countries Break Up? The case of Yugoslavia. Acta Universitatis Upsaliensis, 1994 (second English and first Serbian edition, Pescanik, 2010).
Gligorov, V. (2000a), “Delaying Integration: The Impact of EU Eastern Enlargement on Individual CEECs Not Acceding or Acceding only Later”, wiiw Research Report 267.
Gligorov, V. (2000b), ‘The Economics of Kosovo: Does Viability Matter?’, The International
Gligorov, V. (2002a), „Southeast Europe: History of Divergence”, wiiw Balkan Observatory Working Paper 24.
Gligorov, V. (2002b), ‘The Economic Future of Kosovo’, European Centre for Minority Issues Working Paper.
Gligorov, V. (2007), “Costs and Benefits of Kosovo’s Future Status”, wiiw Research Report 342.
Leeper, E. M. (1991), “Equilibria Under ‘Active’ and ‘Passive’ Monetary and Fiscal Policies,” Journal of Monetary Economics 27, February 1991: 129-147.
Vladimir Gligorov, WiiW (The Vienna Institute for International Economic Studies), 14.07.2011.
- On the causes and reasons for dissolution of Yugoslavia, almost in real time see Gligorov (1994 and 2010). On the possible effects of delaying integration see Gligorov (2000a). ↑
- wiiw (2011) p. vii. ↑
- For details see Gligorov (2002a). ↑
- wiiw (2011) p. vii. Figure for Serbia, Montenegro and Bosnia and Herzegovina are author’s assessments. ↑
- Comparison is between the peaks of crisis in terms of employment and unemployment then and now. If averages were to be used, the qualitative conclusion would be the same, though the levels would be lower in both periods. ↑
- I discuss that and related work in Gligorov (2000a). ↑
- For a discussion of another example see Dornbusch (1992). ↑
- For those to erupt, territorial disputes have to emerge, which was indeed the case in the breakup of Yugoslavia. Feldstein (1997) and Eichengreen (2007) have argued that breakup of EMU would be quite costly in economic, political, and indeed in terms of security. ↑