Carles Boix* The Institutional Accomodation of an Enlarged Europe I n May 2004, Poland, Hungary, the Czech Republic, Slovenia, Slovakia, Malta, Cyprus, Estonia, Latvia and Lithuania joined the European Union bringing the number of its members to 25. In a not too distant future, this number may well reach over 30, as several countries in the Balkans, such as Romania, Bulgaria and Croatia, as well as Turkey are accepted into the fold of the EU. Can the current European institutions and the Union itself survive the process of enlargement? Can the European Union maintain its democratic and social nature as it includes some of the poorest countries of the continent? Must the European Union reform its institutions in a more drastic way than what has been accomplished in the current constitutional draft? If so, how? Should the EU tighten its decision-making procedures or should it decentralize its policy-making mechanisms? Should a two-speed Europe be put in place? What are the institutional and social preconditions that must be met to sustain and deepen the process of European integration? To answer these questions this paper anchors its analysis in current empirical democratic theory and applies the latter’s theoretical propositions to assess the future development of the European Union. More specifically, this paper is built as follows. In its first section the paper surveys the set of social and economic conditions under which democracies(mainly understood as a decision-making procedure but also as an institutional structure associated with advanced welfare regimes) are politically viable. In the second section the paper applies these insights to describe the most appropriate constitutional arrangements to resolve political disagreements between territories that differ in the nature of their interests. To put it briefly, a central implication of this analysis consists in showing that the degree to which political actors, regions or, for that matter, countries can pool their resources and policymaking procedures together in a feasible manner is a function of the extent to which their interests(and, particularly, their standards of living) are relatively homogeneous. The third section then proceeds to use the preceding theoretical discussion to shed light on the case of the European Union. It accordingly outlines * The University of Chicago, Department of Political Science. several of the consequences of the current enlargement and makes a few recommendations about the type of constitutional configuration that may be most adequate to sustain a working European Union. Democratic Theory To determine which are the most appropriate procedures to govern the European Union, that is, to establish the extent to which its members may be willing to pool their resources and authority together, I first proceed to delineate the conditions under which a democracy is a viable system of government for a given territory and population. Let me start by noting that a democracy consists of a procedure through which its citizens decide by casting a vote or a sequence of votes how to govern themselves- that is, what rules should bind their collective life, which should be the optimal distribution of assets and so on. This mechanism of decision implies that, once a vote is conducted over a certain issue, the majority of the population determines the position(or welfare) of each member of the population and therefore of the minority that has not agreed with that majority. In other words, in a democracy the minority is subjected to the will of the majority. If that is the case, that is, if in a democracy the minority remains at the mercy of the majority, it is plain that a democratic regime will only become possible if the minority nonetheless accepts the electoral outcome in which it participates. Now, since the definition and composition of the minority may vary with each issue or decision put to a vote, we can restate the same idea in more general terms. We will say that a democracy will only be possible if their participating agents accept the possibility that the outcome generated by a popular vote may differ from their most preferred alternative. To shed more light on this question, consider it in a slightly more specific manner in the context of a representative democracy where two candidates compete for a given political office, i.e. the presidency of the state. After both candidates engage in an electoral campaign and voters cast their ballots, the candidate

Carles Boix The Institutional Accomodation of an Enlarged Europe Europäische Politik (06/2004) 2 with most votes is proclaimed winner and therefore liament, the use of plural voting, a second chamber in assumes the presidency. The loser must wait for new the hands of the minority, the distribution of powers elections to be held in the future(after a given period to different territorial units and so on. More radically, of time) to have a chance to be elected. In the mean- the losers of elections may attempt to establish an autime, he has to accept the decisions and the policy thoritarian regime. program of the elected politician. The extent to which all citizens will accept or reject Notice that the electoral process carries no guaran- the democratic order is a function of two things. In the tees, in itself, that any of the two politicians will re- first place, the smaller the differences in ideal policy spect the terms and continuity of the procedure. The positions between majority and minority, the higher loser may abide by the election, accept the defeat and the incentives everyone will have to comply with the wait till the new electoral contest takes place. But, if it democratic outcome since the losses experienced by is too unsatisfactory for him to behave well, that is, if the minority will tend to be negligible. The variance in the current benefits of the office he is forsaking are policy preferences may have different sources. too large, he is likely to denounce the results and even- It may reflect diverse interests over the distribution tually stage a coup to grab the presidency by non- of wealth in a given economy. The distribution of inelectoral means. In turn, the winner may have as well come may affect the chances of introducing and susan incentive to use her tenure of the presidency to taining a democratic regime as follows. A more unshift resources in her favor to boost her future electoral equal distribution of wealth increases the redistributive chances, to alter the rules of electoral engagement and demands of the population and the ultimate level of even to postpone or cancel the new election. taxes in a democratic system. As the potential level of A stable or successful democracy, that is, the unin- transfers becomes larger, the authoritarian inclinations terrupted use of free and fair voting mechanisms to of the wealthy increase and the probabilities of democarrive at political decisions and to select public officials, ratization and democratic survival decline steadily. will only take place if both the winner and the loser Conversely, as the less well-off grow richer and their (or, if you will, the majority and the minority) have an income becomes closer to that of high-income voters, interest in complying with the outcomes of the peri- economic tensions decline, and the rich are increasodic votes they employ to decide how to govern them- ingly inclined to accept a democratic regime- elections selves. 1 have a marginally minor impact on the wealthy’s conWhether the political candidates, or, more gener- sumption level. 2 ally, the citizenry, will comply or not with the electoral Preferences may vary also according to religious and procedure will depend on the alternative value re- ethnic preferences – as fragmentation along those ported by each one of the political strategies they may lines increases, heterogeneity should go up as well, follow. If the gain they expect to obtain from an elec- and the chances of democracy should decline. 3 toral victory is larger than the net benefit they would In the second place, the likelihood of a stable deaccrue from rejecting the democratic outcome, they mocratic outcome will increase with the costs of overwill accept the democratic game. Otherwise, they will turning democracy – in more general terms, the probcontest the democratic procedure and try to curtail the ability of a stable democracy will rise as the political space of policies in which the majority may govern. To and organizational resources of both the majority and put it differently, losers may abide by the election, ac- the minority become more balanced. cept the defeat and wait till the new electoral contest takes place. But, if it is too unsatisfactory for them to behave well, that is, if the current benefits of the office they are forsaking are too large, they will be likely to denounce the results and try to block the adoption of 2 This insight is proved more rigorously, both formally and empirically, in Boix 2003. This book also shows that, in addition to growing equality, democracy is fostered by a decline in the extent to which capital becomes either more mobile or harder democratic procedures. The protection of minority interests may be achieved through the introduction of vote aggregation proto tax(or both). As the mobility of capital increases, taxes rates necessarily decline since otherwise the capital holder would have an incentive to transfer his assets abroad. Similarly, whenever capital can be easily hidden from the state or cedures that curb down the powers of the majority. Those mechanisms may take many forms: redistricting, the overrepresentation of certain constituencies in parit becomes of a kind that can only be used by its owner, the temptation to confiscate it also declines. As the redistributive pressures from non-capital holders on capitalists diminish, political conflict levels off and the likelihood of democracy rises. 3 As shown in Boix 2003: 83-84, although religious and ethnic heterogeneity have a negative on democratization and de1 On democracy as a self-enforcing mechanism, see Boix 2003. mocratic consolidation, their impact is much lower than inPrzeworski 1991 and Weingast 1997. come inequality.

Internationale Politikanalyse International Policy Analysis Unit Federalism and Secession as trade integration at the world level goes up, the incentives to achieve trade integration through an instiHaving sketched the conditions under which a democtutional mechanism specific to the two regions of our ratic regime is feasible, I now turn to explore the impliexample declines since the latter do not need to estabcations for the joint governance of several territories lish any particular joint institutions to secure a level of that differ in preferences and income levels. This trade openness(or military cooperation) that is already should lay down the basis for a discussion of the prosin place at the continental or world level. pects of the European Union, its future enlargement On the other hand, integration comes at a cost, parand the choice of its governing institutions. ticularly for the would-be minority territory. It gives the poorer region the capacity to engineer significant transfers from the wealthier to the less developed terriTerritorial Divergence tories. The degree to which those transfers will be set and Economic and Political Unions up will vary with the income differences across regions. The more divergent the standards of living across Assume a continental area with two regions or councountries, the higher the incentives of low income vottries, A and B , which entertain the possibility of estabers to tax the richer areas(or to establish common lishing some kind of economic and political union. regulations to homogenize the distribution of wealth). Whereas A is a rich region, that is, most of its inhabiThe threat of more redistribution reduces the willingtants and, particularly, its median voter, are wealthy, ness of richer regions to join into an economic and poregion B is a low-income region and most of its inhabilitical union with less developed areas. Accordingly, as tants are relatively poor. If the two regions decide to the differences between territories(regions in a nation integrate their economies, they will pool both their asor countries in a continent) increase, integrating the sets and their economic authority. Asset-pooling imlatter under a common state, that is, inducing everyplies that the two states will agree to reduce or remove one to pool their assets and authority under a joint debarriers to exchange that exist between them. They cision-making organization, should be harder. will also establish an institutional procedure in which their populations vote together over policies that affect both regions. If we further assume that B has a larger Decentralization as a Partial Solution population than A , once the two regions decide to join in a single country, the electoral majority would be in If the costs of taxation exceed the benefits of trade the hands of low-income voters. Accordingly, the latter and peace to the point of deterring one or more terri(and specially B voters) could impose high taxes on the tories from joining the union, a partial solution to wealthy individuals and, particularly, on region A . achieve some form of integration would consist in limIntegration will happen whenever the net benefits iting the degree to which all the regions pool their asof the union are positive for the two regions involved. sets and authority together. This strategy would imply On the one hand, there are two types of benefits that maintaining most political decisions in the separate should push them to vote for integration: economic or hands of each region that joins the union and then trade gains, which result from their specialization in enumerating, in a relatively strict manner, the policy the production of particular bundles of goods; and domains(such as free movement of labor and capital) peace gains, which accrue to them as a result of a rein which common decisions are taken and the proceduction of the military threat that each region poses to dures according to which they should be taken. the other one or as a result of A and B pooling their To put it differently, as the level of heterogeneity army and military resources to balance against a third rises, and holding constant trade and peace gains, the territory, C . As is well know, these types of benefits level of political decentralization(and the strength of underlie most of the existing theories explaining the the guarantees to be given to every territory vis-a-vis current process of European integration. 4 As trade and the other members of the union) should rise to make peace gains increase, the incentives to integrate go up. the political and economic union feasible. Notice that those gains are relative to the level of trade Naturally, this is a partial solution since although the integration and peace at the world level. For example, low-income regions would rather have this type of weak union to no union at all, they would prefer complete political integration(with tax and transfer powers 4 On economic explanations of European integration, see Moin the hands of all the union) over any other alternative see Rosato 2004. ravscik 1998 and Mattli 1999. For a security-driven account, constitutional arrangement. 3

Carles Boix The Institutional Accomodation of an Enlarged Europe Europäische Politik (06/2004) 4 Table 1: Observed Probability of Democratic Breakdown by Federalism and Economic Conditions, 1850-1990 Non-Federal Systems (In parenthesis: Parliamentarian regimes) Federal Systes (In parenthesis: Parliamentarian regimes) Annual Observations Observed Probability of Failures Breakdown a Annual Observations Observed Probability of Failures Breakdown a Per Capita Income US$ of 1985(1850-1990) 0-1,999 397(203) 25(13) 6.30%(6.40%) 97(67) 3(0) 3.09%(0.00%) 2,000-3,999 690(518) 14(7) 2.03%(1.35%) 163(93) 3(1) 1.84%(1.08%) 4,000-5,999 283(246) 1(0) 0.35%(0.00%) 148(87) 2(0) 1.35%(0.00%) 6,000-7,999 191(173) 0(0) 0.00%(0.00%) 96(63) 1(0) 1.04%(0.00%) 8,000 and over 340(316) 0(0) 0.00%(0.00%) 185(133) 0(0) 0.00%(0.00%) Total 1,901(1,456) 40(20) 2.10%(1.37%) 689(443) 9(1) 1.30%(0.23%) Average Percentage of Non-Agricultural and of Urban Population(1850-1980) 0-24.9 125(78) 11(7) 8.80%(8.97%) 39(28) 1(0) 2.56%(0.00%) 25-49.9 599(426) 18(7) 3.01%(1.64%) 234(143) 1(0) 0.43%(0.00%) 50-74.9 447(377) 1(1) 0.22%(0.27%) 284(195) 5(1) 1.76%(0.51%) 75-100 63(63) 0(0) 0.00%(0.00%) 17(11) 0(0) 0.00%(0.00%) Total 1,234(944) 30(15) 2.43%(1.59%) 574(377) 7(1) 1.22%(0.27%) Gini Index (1950-90) Above 50% 47(24) 2(0) 4.26%(0.00%) 15(0) 1(0) 6.67%(0.00%) 35%- 50% 270(165) 6(2) 2.22%(1.21%) 89(63) 0(0) 0.00%(0.00%) Below 35% 162(151) 0(0) 0.00%(0.00%) 93(67) 0(0) 0.00%(0.00%) Total 479(340) 8(2) 1.67%(0.59%) 197(123) 1(0) 0.51%(0.00%) a Ratio of observed failures to annual observations. The introduction of politically decentralized institutions to manage inter-territorial differences is not exclusive of potential international unions. It is well extended among countries that exhibit some non-negligible degree of inter-territorial heterogeneity. Political and tax decentralization are rather straightforward strategies to stabilize democracy in territorially diverse countries(Beramendi 2003). Again, in the presence of considerable income differences across territories, a completely centralized policy authority fuels political tensions and makes those populations concentrated in the richer areas more likely to contest democracy. The political resistance of the latter should decline as political and fiscal sovereignty is devolved to smaller units – although political demands for more centralization may probably increase among the majority. Table 1 shows the probability that a democracy will break down as a function of its social and economic conditions(per capita income, level of urbanization and industrialization, and income inequality measured through the Gini index) and of the type of constitutional arrangement, either federal or unitary, in place. The data are collected for the period between 1850 and the late 20 th century and consist of country-year observations about the type of regime in place and the number of democratic breakdowns that occurred over that period. The probability of democratic breakdown is then calculated as the ratio of the total number of cases of democratic breakdown over the total number of annual observations of democracy of the universe under analysis. 6 The table also shows, in parenthesis, the data for parliamentary(as opposed to presidential) regimes. 6 The definition of democratic political regime is taken from Boix/ Rosato 2001). All sovereign countries from 1800 to 1999 are coded there as either democratic or authoritarian. Countries are coded as democracies if they meet three conditions: elections are free and competitive; the executive is accountable to citizens(either through elections in presidential systems or to the legislative power in parliamentary regimes); and at least 50 percent of the male electorate is enfranchised.

Internationale Politikanalyse International Policy Analysis Unit By level of per capita income, federalism works much better in countries whose per capita income is below$4,000. Between$4,000 and$8,000, federalism is slightly worse, basically because the data reflect the Argentinian crises of 1962, 1966 and 1976. Above $8,000 the type or territorial structure has no impact. Notice that among parliamentary systems federalism is much more stable than unitary systems – there is only one democratic breakdown. The same pattern of higher stability among federal systems emerges for different levels of urbanization and industrialization. At high levels of inequality(a Gini above 50 percent or few family farms), the probability of breakdown is slightly higher in federal systems. Yet for middle levels of inequality, democratic breakdowns fluctuate around 2 percent in non-federal countries against 0 percent in federal systems. Again, the most stable regime, with very few or no breakdowns, is one that combines federalism and parliamentarism. The European Union Let us now examine the extent to which the previous discussion on the feasibility of democratic regimes and on the use of federal structures may have any bearing on the governance of European institutions. wide disparities of income would engender harsh political tensions and lead to the introduction of an authoritarian system or to violent revolutionary explosions. According to the estimations I have made elsewhere, in a country with a Gini index of 66 percent, the annual probability of a transition to democracy is 0 and the annual probability of an authoritarian coup is 0.76(Boix 2003: 156). In the third place, democracy survives in part of the globe as a result of a system of territorially divided sovereignties. Today's system of separate nations makes sure that the‘South’ does not effectively impose redistributive mechanisms on the‘North’. The lack of truly unified supranational institutions in fact secures the maintenance of stable democracies in the developed world as well as in certain developing countries. Similarly, the survival of relatively democratic regimes in the North Eastern and Western areas of the United States in the 19th century was dependent on the maintenance of a de facto confederate system- where states enjoyed nearly complete sovereignty over taxes and the legality of slavery. With a very centralized state, those units would have been affected by the harsh inequalities of the South and a democratic system would have been harder to sustain. Indeed, it was the assertion of the federal government, under an administration opposed to slavery, that led to the American civil war. Economic and Political Integration at the World Level Regional Integration Differences in per capita income across the globe are Political integration, which seems impossible at the extraordinary. Figure 1 shows the distribution of world world level, may still be feasible among certain subsets population according to average per capita income of of countries, provided they at least meet two condithe country where they lived in 1993. If we apply the tions: first, that sufficient trade(and security) gains can insights of the previous section, three facts stand out. be derived from their union; and, second, that their In the first place, the incentives to establish a single living standards are relatively similar and so the level of political union at the world level(with a roughly de- internal conflict under the union may be managed mocratic procedure of a majoritarian bent) are simply with little difficulty. absent, at least among developed countries. It is worth stressing again the importance of the In the second place, if all countries were to unite second condition(i.e. relative income homogeneity) to under a single political authority, democracy, that ap- secure trade agreements and political and economic pears to be sustainable in separate countries with dif- integration for the following reason. Traditional trade ferent income levels(provided each one of them is intheory predicts free trade(and, accordingly, the creaternally homogeneous and politically independent), tion of some joint institutions to oversee and sustain it) would simply collapse. The world Gini index in 1993 particularly among countries that can realize substanwas 66 percent – equal to the maximum national Gini tial trade gains, that is, among countries that are very across the world since 1950. 7 In a unified country, such dissimilar because they are specialized in different types of industries. By this token, we should see numerous North-South or developed-developing integra7 Figure 1 and the world Gini index are taken from Milanovic tion schemes because the factor endowments of the Zimbabwe in 1990 according to Deininger/ Squire 1996. 1993. The maximum Gini index(of 66.43) corresponds to two areas are so dissimilar that all customers should 5

Carles Boix The Institutional Accomodation of an Enlarged Europe 6 Europäische Politik (06/2004) Per capita income as a ratio of EU-15 Figure 1: Distribution of Population according to Average Per Capita Income of Country of Residence 100 90 80 70 60 50 40 30 20 10 0 1970 EU-15 1975 Spain,Portugal,Greece 1980 1985 Year Czech Republic and Hungary 1990 1995 2000 Poland Baltic Republics Southern Balkans Figure 2: Per Capita Income in Selected European Countries, 1970-99

Internationale Politikanalyse International Policy Analysis Unit gain from the creation of a common market. Yet, with the exception of NAFTA, the prediction of standard trade theory is not born at all. The only possible reason is that the very dissimilitude in factors and industry types that exists between countries would lead to strong distributional imbalances and harsh inter-territorial pressures in the context of a joint political structure. This would in turn push public opinion, especially in rich countries, to block any free trade or integration arrangement or, at least, to make it as loose or decentralized as possible. Table 2: Regional Gini coefficients in 1993 (distribution of persons by$PPP income per capita) pean Union. Greece, Portugal and Spain, which joined in the mid eighties, have a per capita income of about 70 percent of the EU average. All the new countries that either will become members in 2004 or have applied or plan to apply for membership have much lower standards of living. Per capita income in the most advanced countries, such as Hungary and the Czech republic, stand at between 50 and 60 percent of the European average. In Poland and the Baltic republics, per capita income does not reach 40 percent of the European average. The difference is even more abysmal among western Balkan states: their per capita income is equal to about a fifth of the European Union average. Africa 47.2 Asia 61.8 Latin America and Carribbean 55.6 Eastern Europe and Former Soviet Union 46.4 Table 3: Income, Inequality and Poverty in the EU Population (a) Average income Median income Gini index %Low income (b) Western Europe(EU-15) 32.4 EU-6 222 9,326 7,892 31.0 12.5 Source: Calculated based on the data set“Household survey data used to derive the first‘true’ world income distribution- version1,” gathered by Branko Milanovic at the World Bank and available at http://www.worldbank.org/research/inequality/data.htm. The regional Gini index is calculated by treating each individual equally- simply as an inhabitant of that region EU-9 EU-12 EU-15 EU-25 289 9,343 7,892 32.1 14.2 348 8,633 7,166 34.2 19.9 370 8,622 7,274 34.2 19.9 444 7,685 6,231 38.0 29.6 EU-27 476 7,314 5,959 39.9 33.3 Holding peace and trade gains constant, income difEU-28 535 6,793 5,426 42.3 38.5 ferentials across countries should be a good predictor of the chances of building supranational democratic bodies in each area. Table 2 displays the Gini index for EU-28+ 550 EU-28++ 620 6,662 4,973 43.0 40.0 6,138 4,633 45.4 45.6 several regions in the world in 1993. The level of income inequality ranges from a Gini of 61.8 in Asia to a Gini index of 32.4 in Western Europe. In line with the theory just considered, Western Europe, that is, the USA 258 Australia 18 Canada 29 12,381 9,924 39.4 10.0 9,083 7,600 34.5 10.0 11,716 10,082 31.0 10.0 continental area that exhibits the lowest level of income differences across its countries, has been the only region where there has been a sustained process of political integration in the last four decades. India 901 521 443 32.8 100.0 (a) in millions;(b) income below 50% median of EU-6 EU-27: EU-25+ Bulgaria and Romania; EU-28: EU-25+ Bulgaria and Romania and Turkey; EU 28+: EU-28+ Western Balkans European Integration What are the implications of the theory for the European Union, its expansion and its internal constituEU 28++: EU-28+ W. Balkans+ Belarus+ Ukraine+ Moldova Source: Calculated based on the data set“Household survey data used to derive the first‘true’ world income distribution- version1,” gathered by Branko Milanovic at the World Bank and available at http://www.worldbank.org/research/inequality/data.htm. tional structure? Consider first the evolution of living standards among EU members. Figure 2 shows the evolution of per capita income across several European countries from 1970 to 1999 as a proportion of the average income of the current members of the EuroExamine next the actual distribution of income among European individuals. Using World Bank data, I calculate and report in Table 3, measures of per capita income and inequality for past(EU-6, EU-9, EU-12, 7

Carles Boix The Institutional Accomodation of an Enlarged Europe Europäische Politik (06/2004) Population (Thousands) 8 60,000 50,000 Figure 3. Distribution of Population by Per Capita Income (in each$500 segment) 50% of median income of EU-6 40,000 30,000 20,000 10,000 0 0 2,500 5,000 7,500 10,000 12,500 15,000 17,500 20,000 22,500 25,000 PPP$ of 1993 EU-15 EU-25 EU-25, Balkans, Turkey, Ukraine, Belarus, Moldova EU-15), present(EU-25) and potentially future European Union areas. Table 3 shows the average and median average income, the Gini coefficient and the percentage of low-income population. Low income is defined as any income below 50 percent of the median income of EU-6. For purposes of comparison, Table 3 also reports measures for four other big federations in the world: the United States, Australia, Canada, and India. Inequality in the European Union remained relatively low until the entry of Southern European countries in the mid 1980s. The Gini index grew slightly from 32 percent to 34 percent at that time and did not change after the entry of Sweden, Finland and Austria. With the last enlargement, the Gini index has risen to 38 percent and the percentage of low-income people has jumped to almost 30 percent of the population. Any new enlargement should lead to even higher levels of inequality. The entry of Bulgaria and Romania would push the Gini index upward to a higher level than the United States. It would also increase to about a third the proportion of the population with low incomes, or about three times the percentage in the United States. The entry of Turkey and western Balkan countries puts the Gini at 43 percent – higher than the Philippines or Venezuela and close to Nigeria. Inequality and poverty are even more extraordinary in an expansive concept of Europe that includes Ukraine, Belarus and Moldova: almost 50 percent of the population would live under poverty and the Gini index would reach 45.4 percent. Figure 3 reinforces this fact by showing the distribution of the population by income level in EU-15, EU-25 and the broadest definition of Europe(without Russia). In the light of the theory employed in the paper, the integration of less developed countries cannot but have substantial consequences for the European Union. Unless the trade and security gains of any new enlargement wave are considerable, the European Union will be forced to delay any plans for tighter institutional integration. New enlargements may even result in a movement toward relatively more decentralized structures of power. This latter solution may adopt different forms. It may simply consist of passing new clauses that give additional voting powers or jurisdictional guarantees to certain countries in the Union. More radically, it may finally press European politicians to embrace a two-speed institutional structure. Conclusions and Policy Recommendations The European Union is the result of the decision by several European countries to establish a common economic area overseen by joint political institutions. This decision is in turn the result of at least two conditions: first, the realization that European countries can accrue significant trade and security gains through the

Internationale Politikanalyse International Policy Analysis Unit construction of a union; and, second, the fact that difnative strategies. On the one hand, it could raise the ferences in standards of living are not excessive and volume of funds to be transferred to incoming memthat a political and economic union should not result in bers to accelerate their growth rates. This redistributive inordinate transfers from the wealthy to the poor strategy faces, however, two problems: first, it looks countries. politically unfeasible since any transfer scheme of sigThe level of political and economic integration of nificant size would be met with the resistance of the the European Union, that is, the degree of asset- and European core of wealthy nations; second, it seems authority-pooling, has been far from perfect or comeconomically ill-advised since the European structural plete. This may partly be due to the absence of a funds spent on the countries that joined the Union in commonly perceived idea of European citizenship or the 1980s do not appear to have sped up the former’s legitimacy among European citizens. But it is mostly economic catch-up with the European core in any conthe result of a diversity of interests that, as in any other siderable manner. On the other hand, the Union could federal or federative system, call for the maintenance press the incoming members to rapidly modernize their of barriers between its state members. political and institutional infrastructure. This should As a matter of fact, the success of the European Unthen make them attractive to European business, ion in part derives from the decision made by its memquicken the flow of capital from Europe’s core to its bers to establish a relatively‘soft’ set of overlapping periphery, and result in higher levels of economic coninstitutions(the European Commission, the Council of vergence in the medium run. Only then it would be Ministers, the European Parliament, the European possible to push forward any stronger federation Court of Justice) in line with the relative degree of inscheme. ternal heterogeneity of the Union. Those institutions are‘soft’ in the sense that the set of policies over which they can act is limited(although growing as a References result both of new treaties among the states that form the Union and of some policy and institutional spillovers). They are overlapping in that more than one of them are required to make fundamental policy decisions affecting all the members of the Union. Although these two characteristics make them prone to the criticism that the European Union suffers from some sort of‘democratic deficit’, they are the best guarantee that none of the states will impose its interests over the rest of the members of the European Union. This guarantee in turn ensures that all members have an Beramendi, Pablo(2003):“Political Institutions and Income Inequality: The Case of Decentralization”: WZB, unpublished paper. Berlin Boix, Carles(2003): Democracy and Redistribution . Cambridge University Press, New York. Boix, Carles and Sebastian Rosato(2001):“A Complete Data Set of Political Regimes, 1800-1999,” University of Chicago. Deininger, Klaus and Lyn Squire(1996):“A New Data Set Measuring Income Inequality,” The World Bank Economic Review 19. incentive to maintain their commitments to the Union. Mattli, Walter(1999 ): The Logic of Regional Integration: The enlargement of the European Union imposes Europe and Beyond. Cambridge University Press, New clear limitations on the extent to which Europe can York. push for deeper levels of integration or establish more centralized structures in the short and medium run. At a time when the heterogeneity within the European Union is growing, any attempt to make the pooling of assets and authority much more complete could jeopardize the whole project of European integration. Tied up by the growing diversity of interests at the continental level, the most likely horizon is one in which the Union will hardly move away from its current institutional status quo. A possible, although less probable, outcome would consist in the creation of a two-speed Milanovic, Branko(1993):“True world income distribution, 1988 and 1993: First calculation based on household surveys alone,” World Bank Policy Research Working Paper , no 2244(November 1993). Moravscik, Andrew(1998): The Choice for Europe . Cornell University Press, Ithaca. Przeworski, Adam(1991): Democracy and the Market . Cambridge University Press, New York. Rosato, Sebastian( 2004): The Logic of Integration . The University of Chicago, Ph.D. dissertation. Weingast, Barry R.(1997):“The Political Foundations of Democracy and the Rule of Law,” American Political Science Europe, which would again put in place significant barReview 91(June): 245-63. riers between the core and the periphery of Europe. To speed up the process of economic convergence and then achieve a deeper political integration of Europe, the European Union could consider two 9 alternative strategies. On the one hand, it could raise

Carles Boix The Institutional Accomodation of an Enlarged Europe T his paper examines the impact that the process of enlargement may have on European Union institutions in the light of both current democratic theory and the literature on the political economy of trade. The paper assumes that any set of countries will only establish a continent-wide economic and political union if they all benefit in net terms from jointly pooling together their economic assets and political authority. On the one hand, the gross benefits of integration increase with trade and peace gains. On the other hand, integration comes at the cost of losing political sovereign. With complete political integration, the most populous countries can establish regulatory mechanisms on the smaller nations. Thus, the more divergent the standards of living and the distribution of interests and preferences across countries, the more burdensome a political union will be for would-be minority territories and integration will be harder to achieve. If the costs of taxation and regulation exceed trade and peace gains to the point of deterring one or more countries from joining a union, a partial solution to achieve some form of integration would consist in limiting the extent to which all countries pool their assets and authority together. More generally, holding trade and peace gains constant, the depth of political integration would be correlated with the homogeneity in income and interests of the potential members of the union. These insights can be applied to the evolution of the EU. Employing recent data on income distribution, the paper shows, in the first place, that the relatively low level of inter-state inequality in Western Europe explains why this is the only region at the world level where there has been a sustained process of political integration in the last four decades. The paper investigates, in the second place, how this relatively homogeneous distribution of income may change as the EU expands. The paper estimates that the level of inequality has increased from low levels(a Gini index of 32 percent for EU-9 and 34 percent for EU-15, similar to Canada and Australia respectively) to much higher levels(a Gini of 38 percent for EU-25, close to the United States’ Gini index). If the EU proceeded to include the Western Balkans and Turkey, the Gini would shoot upwards to about 43 percent. Based on these results, the paper concludes that, unless the trade and security gains of any new enlargement are very substantial, the EU will not achieve any tighter institutional integration and, in fact, may have to accept more decentralization in the future. The paper ends by exploring possible strategies to reduce income heterogeneity and increase political integration in the medium run.