Are democracy, sovereignty, and globalization compatible? Is it possible to have any combination or even all three simultaneously? In many ways, this is the most pressing political economy question of our times.
An accelerated pace of economic change increases the strains on political governance. Even as the world continues to consist of independent sovereign states, it is undergoing a massive integration of markets. Revolutions in transportation and communication combine with political and institutional changes to create unprecedented global flows of capital, goods, and even people.Footnote 1 At the same time, we have witnessed an enormous growth in the number of states with democratic institutions. Yet, the sovereign state system remains the dominant form of political organization.
Each of these features – sovereignty, globalization, and democracy – remains an essential yet contested feature of world politics.Footnote 2 The world consists of sovereign states and numerous secessionist movements underline the continuing political desirability of sovereignty. Yet, even as we observe pressures for fragmenting existing states in order to accommodate the birth of new ones, there are dramatic efforts to achieve political integration, to transcend the state in some form of transnational organization. At the same time, movement toward greater economic integration has been met by sustained opposition. Finally, even as the number of democratic states grows, there is both a resurgence of autocracies and forces within democracies reducing the scope of accountability and citizen control.
In different quarters, the incompatibility, culmination, or demise of each of these features of governance is seen by some as a historical inevitability. As described more fully below, some see globalization as indicating the end of the sovereign state, whereas others see globalization as leading to the end of democracy, and finally, some argue that the combination of these three forces is problematic. The arguments are empirical as well as analytic, and they are muddied by competing and incommensurate characterizations of each of the three features of governance.
This paper explains the conflicting strands and frames the reasoning in precise materialist terms. It explicates the case for a trilemma – the argument that of globalization, democracy, and sovereignty, only two can coexist in their pure forms at the same time. It also delineates the critical assumptions underlying the trilemma.
The paper also develops the implications of the trilemma – pressures to constrain all three even as the sovereign state system witnesses the expansion of democracy and the growth of globalization. Many of the conflicting currents of world politics – concomitant pressures to strengthen and weaken the state, simultaneous efforts to press forward with economic integration and to restrain it, and the coexistence of new global institutions that are not democratic and the expansion of democratic rule – derive from the trilemma.
Movement along any two dimensions of global organization generates pressures to restrain the third. The combination of democratic rule and state sovereignty results in efforts to constrain globalization. Maintaining democratic rule and globalization leads to pressures to constrain sovereignty. Finally, furthering economic integration in a system of sovereign states is presumed to require limiting representative government. The argument developed here is that, given the incompatibility of all three, the coexistence of any two generates pressures to constrain the third. Contemporary politics regarding globalization is driven by the conflict over which of the three to restrain.
The paper notes that this argument is rooted in narrow materialist terms and with key assumptions underlying the trilemma, and thus points to what can make them compatible. The substantial changes that would be required are some of the key contested features of global governance.
The paper delineates the trilemma, the two critical economic and political assumptions underlying it, and its political implications. The next section begins with globalization and how it interacts with democratic rule and a sovereign state system. This is followed by the development of the trilemma, which is followed by a discussion of assumptions and implications.
Globalization and admixtures
Globalization
Globalization is one of the widely used terms of the day and it has many meanings across the disparate disciplines and subfields in which it is used.Footnote 3 Indeed, ‘globalization studies’ has emerged as a field (Turner Reference Turner2011), but one that defies definition and classification:
Notoriously slippery and expansive, potentially encompassing almost everything under the sun, interrelating apparently disparate elements into wholes so complex, multi-dimensional, and open-ended as to defy presumptive encapsulation in terms of particular theories or perspectives, the study of globalisation is less a sub-field than an overdetermined meta-field (Rupert Reference Rupert2005, 457).
All the features of globalization that scholars discuss derive from its enormous force in economic life. Thus, globalization is used here in its economic sense and refers to the flows of the factors and output of production and the construction of a global marketplace.Footnote 4
Used in this economic sense, globalization refers to the integration of national economies, a phenomenon which has come to include most of the advanced industrial countries of the world and, increasingly, growing numbers of developing nations. In recent years, the forces of economic integration have grown rapidly. Developments in communications and information storage have made possible larger scale enterprises. Capital now flows in vast amounts across borders. The volume of foreign exchange transactions grew from about $600 billion a day in April 1989 to $1.5 trillion a day in April 1998 and to $5.3 trillion a day in April 2013 (U.S. President 2000, 205; Bank of International Settlements 2013, 9). The proportion of many countries’ economies that move across borders has grown significantly. World trade in goods (exports and imports combined) totaled ~20% of world GDP in 1960, 32.5% in 1990, 41.5% in 2003, and 51.8% in 2013 (UNCTAD 2004, 50; World Bank 2005, 324; World Bank 2013, 100). The growth of economic interdependence, noted in the 1970s, remained a regional rather than global phenomenon (Rosecrance and Stein Reference Rosecrance and Stein1973; Stein Reference Stein1984). The opening of China and the end of the Cold War ended two of the great economic barriers and expanded the domain of integrative forces (Yardeni Reference Yardeni2000).Footnote 5
A cottage industry devotes itself to assessing when globalization began, how far integration has proceeded, and whether it has yet had the impact that people have long touted.Footnote 6 Some point out that the market integration evident today is not new – that trade, capital, and even labor flows were substantial in earlier eras – and that globalization can be said to have begun in earlier centuries.Footnote 7 The evidence is clear that the levels of trade seen right before World War I were not reached again until the early 1970s. But they have grown, and in some cases quite dramatically, since. Still other scholars argue that globalization is more slogan than reality – that the majority of production remains domestic, that allegedly international firms retain a domestic orientation, that there is more regionalization than globalization, and that globalization has not generated the convergence in various economic indicators that had been predicted.Footnote 8
Some of the roots of these disagreements lie in the definition and measurement of integration. Economic integration is a process, one that begins with policies that permit cross-border flows, entails actual flows of both the mobile factors of production and the goods produced, and that creates pressures for further policy harmonization and eventual convergence in factors and goods prices.
Those who measure economic integration by the volume of cross-border flows and the relative size of domestic and foreign economic transactions find that international exchange is rising in importance.Footnote 9 It is growing faster than domestic economies. International trade accounts for an increasing proportion of domestic product for most countries. Financial flows slosh across borders in increasing and dizzying amounts.
Others assess globalization by its putative economic effects, by whether the convergence that is expected is occurring. The evidence is mixed. In part, conflicting findings reflect the particular focus of individual studies. Some look narrowly at economic phenomena such as the expectation that prices will converge, whereas others focus on the convergence to be expected in regulatory regimes or in macroeconomic phenomena or welfare policies.Footnote 10
That integration has not yet had certain consequences is subject to multiple interpretations. Some presume such a tight link between globalization and its consequences that the failure to observe those effects is evidence that globalization is not occurring. This is problematic and depends on the purported inexorability of the effects of globalization and their presumptive timing. Economic integration is a process and the harmonization and convergence to be expected from growing volumes of exchange occurs after some lag. More important, states may both foster exchange and pursue policies to ameliorate and attenuate the consequences of integration (discussed further below). The ability of states to grapple with the consequences of globalization is at issue, as are the actual consequences themselves.Footnote 11
The focus here is narrowly material, defining globalization as economic integration consisting of growing volumes of exchange that generate pressures for policy harmonization. That such integrative forces are at work is clear. All agree that economic integration has been growing dramatically and has either reached unprecedented heights or is qualitatively different than the integration seen in the past (Bordo, Eichengreen and Kim Reference Bordo, Eichengreen and Irwin1998; Baldwin and Martin Reference Baldwin and Martin1999; Bordo, Eichengreen and Irwin Reference Bordo, Eichengreen and Kim1999; Garrett Reference Garrett2000; Frieden Reference Frieden2006). Also clearly evident are the pressures for policy harmonization, whether those involve the treatment of intellectual property, accounting standards, environment standards, and other practices. The question is how this growing integration and the concomitant pressures for convergence interact with other aspects of global governance.
Globalization and sovereignty
Globalization is proceeding apace in a world of sovereign states and the implications of one for the other have been debated. The debate is rooted in the reality that globalization entails cross-border flows and common policies whereas sovereignty implies independent autonomous states.Footnote 12
A continuing essential element of international politics is that it consists of separate states exercising control within some territory. They are sovereign in that they are autonomous and make political choices reflecting some mechanism for arriving at public policies. Sovereign states make their own decisions that bind the citizenry within the territorial domains they control.Footnote 13 Technological changes may increase or decrease their ability to control various forms of cross-border movements, yet states retain the ability to control the movement of goods, people, and even capital.Footnote 14
States may become interdependent, they may constrain one another, but they retain autonomous decision-making ability.Footnote 15 Given the focus of this paper, the key aspect of sovereignty is the ability autonomously to control or shape cross-border exchanges. Given an ability to make independent decisions, the form of governance matters (democracy, discussed below); and an ability to control cross-border flows implicates the sovereign state in the prospect for, and development of, globalization.Footnote 16
Despite the growth of transnational organizations and the impact of other forces, international relations remains the domain of interacting states. The presumption here is that territorial states remain the preeminent political actors setting public policy and controlling cross-border flows.Footnote 17
Although the focus has changed over time, there have been arguments that globalization and sovereignty are inherently in tension with one another. In the past, the argument was that sovereignty precluded globalization. In the current age, it is commonly argued that globalization constitutes an assault on sovereignty.
Classical mercantilists felt that sovereignty implied policies of closure that precluded globalization. A concern with the national interest in a presumed constant-sum world implied foreign economic policies that precluded economic integration.
More broadly, mercantilism, and even more so realism, equated sovereignty with a search for self-sufficiency and a rejection of a division of labor that implied dependence on others. Sovereignty, especially for great powers, meant a search for independence and not the interdependence (or the dependence) that globalization entails (Waltz Reference Waltz1970, Reference Waltz1979).Footnote 18
Yet, sovereignty did not preclude some degree of economic integration. During the 19th century, even great powers adopted policies that increased their openness, that led to some division of labor, and an interdependence that led to growing economic integration. The economic nationalism of the 19th century did not entail a reversion to mercantilism but maintained a recognition that the international economy was not a constant-sum game.Footnote 19
Technological advances of the last two centuries were important spurs to globalization. Exchange over substantial distances depends on both an ability to transport over the distance and to do so without excessively degrading the product (food spoilage, for example) and without the transportation cost destroying the market. Revolutions in transportation, especially in the last two centuries, made possible exchange on an unprecedented scale (Cooper Reference Cooper1995, 363–64). Exchange also depends on an ability to communicate. Revolutions in communication made possible the development of multi-locational enterprises that make use of economies of scale and scope.Footnote 20
Although technology makes possible certain developments, it does not assure them. Globalization also depends on states pursuing policies of economic openness. Globalization is both a product of state policy and in turn affects policy. Both equations are essential to an understanding of the dynamics of globalization. States choose to open their borders to trade (and on what conditions and to trade with whom over what categories of goods) and capital flows and the movement of people (which states continue to control and direct). Globalization reflects the policies of states. Globalization is not yet a global phenomenon, in part, because not all states have made those choices.
One important lesson of the realization that there was extensive globalization in the era preceding World War I is the recognition that the policies of governments dramatically reduced, and, in periods, effectively killed, economic integration (O’Rourke and Williamson Reference Obstfeld, Shambaugh and Taylor1999a; James Reference James2001; Horowitz Reference Horowitz2004). The post-World War II recovery and reconstruction took three decades to reachieve the levels of integration preceding World War I. There is then nothing that requires globalization to continue. There is no historical imperative of globalization.
Despite the reality that globalization is a product of state policies, most contemporary discussion focuses on the conflict between globalization and sovereignty (Berger Reference Berger2000). It is often put in terms of the tension between states and markets. But markets both require states to enforce property rights and assure competition, yet also depend on some freedom from state control and interference. States depend on markets to obtain their resources. Yet, states and markets compete for they sometimes provide the same goods, and the result of their conflict determines the relative share of public and private goods available (states providing both and markets only the latter).
Further, the growth of markets poses problems for governments. In most societies, private markets dwarf the size of government and governments face some difficulty in affecting market processes. Governments thus find themselves in the unenviable position of being looked to increasingly for jobs and growth and price stability while their ability to shape such outcomes declines. The impact of globalization simply makes shaping domestic outcomes harder.
It is the reduced scope for government action that is sometimes characterized as globalization’s assault on sovereignty.Footnote 21 Globalization affects the ability of sovereign states to control internal policy outcomes and limits the policy instruments available to governments. The issue is how significant these constraints are and how important states remain in such a setting.
The constraints on governmental policy generate no inherent incompatibility between a global marketplace and a system of sovereign states. The compatibility of globalization and sovereignty has to do with the choices of states and whether the set of states in the sovereign system simultaneously adopt policies of openness. The theoretical presumptions of modern economics is that openness is Pareto improving for absolute material welfare of all states and on that basis there is no incompatibility of sovereignty and globalization.Footnote 22 There is also no basis for presuming that openness is a policy characterized by congestion, that is, that its simultaneous pursuit by all is not possible.Footnote 23
The tension between globalization and sovereignty and its resolvability has been formalized in the area of monetary policy. Robert Mundell and J. Marcus Fleming showed that fixed exchange rates, open capital markets, and national monetary autonomy were not all possible.Footnote 24 That countries could only have two of the three. Open capital markets reflect the dimension of globalization and national monetary autonomy reflects sovereignty. These are inconsistent only if there is a commitment to fixed exchange rates. Both sovereignty (monetary autonomy) and globalization (open capital markets) are possible in a world of freely floating exchange rates.Footnote 25
Still a different formalization of the relationship between a globally integrated economy and a sovereign state system is provided by Casella and Feinstein (Reference Casella and Feinstein2002). In their model, jurisdictions are endogenous to the choices of traders who are also voters. They decide on a jurisdiction to belong to, vote on the level of public goods the jurisdiction will provide and the taxes they will pay. They want public goods which affect their ability to trade but they also want to minimize taxes and they can shift jurisdictions. They conclude, ‘A move to larger markets per se is neither necessary nor sufficient for unification of jurisdictions: there is no logical link between the two’ (Casella and Feinstein Reference Casella and Feinstein2002, 30). Further, they argue, ‘The integration of markets and the integration of institutions devoted to public good provision are interdependent but different processes; they need not proceed together’ (Casella and Feinstein Reference Casella and Feinstein2002, 33).Footnote 26
More broadly, the presumption that globalization maximizes the material welfare of states combined with the presumption that states maximize their power and wealth result in a view that sovereignty and globalization are compatible. Moreover, history tells us both that states, even great powers, adopted policies of openness and allowed economic integration to grow. Yet, there was no technological imperative and states undid in the early part of the 20th century the very integration they had allowed to develop at the end of the 19th and beginning of the 20th centuries. In the last half century, and increasingly in the last two decades, states have again chosen openness and economic integration (though with constraints, as discussed below). But there is nothing either necessary or irreversible about it. No analytic relationship exists between sovereignty and globalization that precludes both simultaneously, nor one that necessarily conjoins them. In short, both a historical and analytic assessment suggest that there is no necessary link or incompatibility between sovereignty and globalization.
Globalization and democracy
The continuing course of globalization must contend not only with a sovereign state system but also with a growth in the number of democracies. Democratic governance has triumphed in recent decades and has been adopted around the world.Footnote 27 The percentage of the world’s independent countries characterized as electoral democracies grew from 25% in 1974 to 41% in 1989, to 62% in 1996, and to 63% in 2015 (Karatnycky Reference Karatnycky2000; Puddington Reference Puddington2015; Freedom House 2016).
In addition, its practice has come closer to its spirit in nations long considered democracies. Japan and Mexico, for example, have had democratic governance for many years, but single party rule led to doubts about democracy’s actual impact. Both have now experienced the election of an opposition government and peaceful transitions of power.Footnote 28
Democracy comes in many forms and is defined in a variety of ways (Przeworski Reference Przeworski2010). Among the forms of participation and involvement that have been analyzed are elections, interest groups, associations such as parties, civic engagement, and community involvement. Labels applied include electoral democracy, participatory democracy, deliberative democracy, and liberal democracy. To capture the range of features that constitute democratic governance, Dahl (Reference Dahl1971) developed an alternative term, polyarchy, that has seven features: inclusive suffrage, free and fair balloting, elected officials, the right to run for office, freedom of expression, alternative information protected by law, and associational autonomy. Others add still other features, including respect for the rule of law and an independent judiciary. The most widely used metric of democraticness (Polity IV) consists of six component measures that are aggregated into a 21-point scale with a recommended truncation that generates five different levels of democracy.Footnote 29
Central to democratic rule is that it attaches equal weight to the preferences of all its citizens and that outcomes reflect the will of a majority. An electorate of all adults, rather than a select subset (i.e. a selectorate), either chooses policies directly in a referendum or selects representatives and rulers who exercise legislative and executive functions making and enforcing laws.Footnote 30 These choices are made in elections that permit electoral competition and allow access to information. Such direct or indirect control of public policy through periodic elections with majority rule should imply that the preferences of the median voter are pivotal.Footnote 31
The core political economy feature of democracy is the presumption that it reflects the preferences of the median voter, or what is sometimes referred to as the referendum model (Mayer Reference Mayer1984). As elaborated below, this assumption is at the heart of work both about the requisites of democracy and of its implications, and a great deal of intellectual effort has been expended assessing this core assumption.
My focus here is on the material implications of democratic governance. Globalization is an economic phenomenon with economic consequences, but it is also the product of state policy to allow cross-border exchanges, and it affects state policy in generating pressures for policy harmonization. In assessing the compatibility of democracy and globalization, it is critical, therefore, to focus on the material impact of globalization in a democracy and the policy impact of democratic governance on globalization.
Just as some see an incompatibility of globalization and sovereignty, there are those who argue that globalization is antithetical to democracy.Footnote 32 Much of this is a rather unfocused reaction to the presumed impersonal and large-scale forces at work in the process of globalization.Footnote 33
Globalization and democracy can readily coexist.Footnote 34 Historically, the world has witnessed economic integration conjoined with democratic political integration. The United States, for example, consisted of 13 autonomous colonies and independence generated a weak confederacy, in which ‘there was no general executive and no general judiciary, no standing army, no power to control commerce’ and in which the legislature had no power to tax. It was ‘far from being a government’, but was ‘a mere diplomatic assembly of states’ (Tomasi Reference Tomasi2003, 225). The Constitution created a federal structure that conjoined economic integration with political integration consisting of democratic governance.Footnote 35
Similarly, one can imagine an integrated world economy and a democratic world government. Planetary governance would reflect the interests of the median voter and there would be an integrated world economy.Footnote 36 Just as individual countries have integrated national economies and are governed democratically, so too, can the world have an integrated economy and be governed democratically.Footnote 37
The great trilemma
The problem that many feel with current globalization is that democratic governance is national in character and the forces of economic globalization are transnational in character. The combination is what is troubling – that democracy functions within units which are less relevant in world affairs and that the constraints on sovereignty have meant the growth of supranational institutions that are more removed from, and less accountable to, individual citizens and voters. The problem is that globalization is occurring within the context of a sovereign state system whose individual members are increasingly ruled democratically.Footnote 38
For democracy, sovereignty, and globalization to be compatible, the forces of economic integration that are the hallmark of globalization must be consistent with a world of independent states each governed democratically. As states still have the power to control cross-border flows (certainly of people and most classes of goods, though less so of capital), globalization depends on the policies adopted by sovereign states. And if each state is a democracy, then the policies that sustain globalization would have to reflect the preferences of a majority of the citizenry in each state.
The classical liberals presumed that the emergence of democracy within sovereign states would lead to policies of openness that would bring economic integration (Stein Reference Stein1993). They opposed mercantilism as a system which reflected the interests of absolutist states rather than individual subjects. Economic reform that would allow competition domestically would also generate foreign economic policies of openness. Individuals interested in maximizing their wealth would embrace competition and accept a division of labor that would allow all to increase their absolute wealth.
But the classical liberal vision came a cropper by the end of the 19th century. Restraining monarchs and empowering individuals did not bring an end to protectionism. Newly enfranchised workers looking to government to protect them from the vagaries of the marketplace eagerly sought protection. It turned out that protection was not just a policy interest of rapacious monarchs. Protectionism, once seen as solely in the interest of autocrats and monarchs, came to be recognized also as a product of the pressures and preferences of a mass electorate. Mercantilism’s top-down protectionism was replaced by bottom-up protectionism. Protectionism could reflect the preferences of rulers, capitalists, or workers. Political liberalization did not necessarily result in economic liberalization. Neither competition nor protection was inherently and unambiguously adopted by either capitalists or workers.Footnote 39 The endogenous theory of protection focuses precisely on domestic political pressures within representative political systems for protection (Trefler Reference Trefler1993). Representative political governments face pressures to deal with those suffering adverse consequences, especially in economic downturns.
The argument that there is a problem with globalization in a world of sovereign democratic states has recently been made by Rodrik (Reference Rodrik2011) but for different reasons than implied above and developed below.Footnote 40 His argument is that ‘hyperglobalization impinges on democratic choices’, and more precisely that ‘deep globalization’ implies a competition between governments that precludes their adoption of the regulations, standards, and other policies their citizens want (Rodrik Reference Rodrik2011, Ch. 9). But this need not create an incompatibility of sovereignty, democracy, and globalization if the citizens in every democracy share regulatory preferences or if democratically elected governments can agree on a set of common standards supported by a majority of their citizens. For him, the incompatibility derives from the inability of the majority of citizens to have their preferred regulatory policies adopted and thus generates a reaction to hyperglobalization. The ability to thwart the preferences of the median voter is what makes sustained liberalization possible in a world of democratic sovereign states. For Rodrik, it is the problem of convergent regulatory policies and the thwarting of national regulatory preferences that is the problem. In short, for Rodrik, globalization has already defeated democracy in a way.Footnote 41
The argument here is different and somewhat opposite, the incompatibility derives from the ability of citizens to get their way when the preferences of a majority of citizens are opposed to continued economic liberalization and integration. And the problem is that in some democratic states a majority of citizens will be directly opposed to liberalization, quite apart from pressures to converge regulatory regimes. This implies a trilemma that emerges from an analytic basis in which democratic preferences diverge in such a fashion as to preclude globalization in a state system consisting of all democratically elected governments, and this is developed below.
For globalization to encompass the entire world, were it composed only of democracies, it must be the case that the policies that sustain it are in the interests of the median voter within every sovereign state. Whereas globalization could succeed in a sovereign system by being Pareto improving for the states in the international system, globalization can only succeed in a sovereign system of democratic states if it is in the interest of the median voter in every individual democratic country.Footnote 42
This is problematic, if not an impossibility. Even if we assume the possibility that all countries increase their economic welfare through integration (that globalization is Pareto improving for countries), it cannot be the case that globalization by itself improves the welfare of the median voter in each country.
Globalization generates pressures for adjustment that have distributional consequences. It generates winners and losers within every country. Openness increases the returns to certain factors of production and reduces those of others.Footnote 43 These consequences are the basis for political alignments regarding both the desirability of greater openness and the maintenance of current exposure to the global marketplace.Footnote 44 Losers will oppose greater openness and support rolling it back. The winners will support greater openness and oppose any backtracking.
As globalization is a product of state policy and affects domestic economic conditions, its viability is dependent on supportive government policies. Even though globalization generates winners and losers, it is presumed to increase national welfare. An interest in maximizing national wealth and an ability to control domestic politics makes globalization and a world of states compatible. But democracies who are responsive to their constituencies face a greater challenge.
In a democracy, the issue will be decided by a majority and thus reflect the preferences of the median voter. The economic winners in many societies, though not in all, will be more numerous. As most voters are workers, ‘Democratization will lead to more liberal trade policies in countries where workers stand to gain from free trade and to more protectionist policies in countries where workers will benefit from the imposition of tariffs and quotas’ (O’Rourke and Taylor Reference Obstfeld and Taylor2007, 195). The standard view is that labor in labor-abundant countries will gain disproportionately (Krueger Reference Krueger1983; Bhagwati and Srinivasan Reference Bhagwati and Srinivasan2002), and such societies should be the ones in which democracy lends strong support for globalization (Milner and Kubota Reference Milner and Rosendorff2005; Milner and Mukherjee Reference Milner and Kubota2009). But at any point in time, there will always be some countries in which a majority of voters will favor constraining globalization. These are the countries in which the scarce factor which loses from openness is also numerically superior. Assuming that voters vote their economic interests, globalization will come under assault in these countries.Footnote 45 If globalization worsens the position of the median voter in a democratic polity, then policies that sustain globalization are in jeopardy if voters select among politicians on the basis of their material interests (assuming that politicians remain free to adopt policies that constrain globalization, on which see the discussion on constraining sovereignty below).
The analytic implications derived from economic theories of liberalization (the nature of the winners and losers) and from political theories of democracy (policy reflects the preferences of the median voter) have begun to be assessed empirically. One strand of research focuses on the distributional consequences of globalization within nations. The evidence for the 19th century fits the expectation that the impact of democracy on protection depends and demonstrates ‘the power of the Heckscher–Ohlin factor endowment theory in understanding late nineteenth-century trade and politics’ (O’Rourke and Taylor Reference Obstfeld and Taylor2007, 213; also see Baldwin Reference Baldwin2008). Similar results have been found for the late 20th century as well (Dutt and Mitra Reference Dutt and Mitra2002).
Another path of empirical verification consists of studies of changes in inequality and their relationship to globalization. The record seems to be reasonably clear that globalization before World War I reduced inequality in ‘resource-poor, labor abundant, agrarian economies around the European periphery’, but rose ‘in resource-rich, labor scarce New World countries’ (Williamson Reference Williamson1998a, 170; also see Williamson Reference Williamson1997; O’Rourke Reference Obstfeld2001). Williamson (Reference Williamson1998a, Reference Williamson1998b) is even prepared to argue that there was a policy backlash against globalization as a result.Footnote 46 The contemporary debate is about inequality and the relative importance of globalization in observed changes in wages. The empirical consensus is that trade is increasingly responsible for changes in US inequality.Footnote 47 The most recent scholarship develops a more differentiated view of the economy, focusing on firm and worker heterogeneity (Kanbur Reference Kanbur2015).Footnote 48 The result is a more refined sense of the characteristics of the winners and losers from trade. But that there are many losers who in some societies constitute a numerical majority is not really questioned.
Another focus for empirical work is the basis for people’s trade policy preferences and whether individuals’ policy preferences reflect their jobs and skills, and the implication of liberalization for their income and wealth. Studies have looked at peoples’ sense of economic insecurity (Scheve and Slaughter Reference Scheve and Slaughter2004), their support for particular trade agreements (Balistreri Reference Balistreri1997), on their support for particular parties in which the elections were effectively referenda on free trade (Beaulieu Reference Beaulieu2002), on general attitudes toward free trade (Scheve and Slaughter Reference Scheve and Slaughter2001; O’Rourke and Sinnott Reference Obstfeld2002; Kaltenthaler, Gelleny and Ceccoli Reference Kaltenthaler, Gelleny and Ceccoli2004; Baker Reference Baker2005; Mayda and Rodrik Reference Mayda, O’Rourke and Sinnott2005; Mayda, O’Rourke and Sinnott Reference Mayda and Rodrik2007; Hoffman Reference Hoffman2009; Owen and Quinn Reference Owen and Quinn2016). In general, they find broad support for the expected link between labor market attributes and trade policy.Footnote 49
That there are winners and losers from globalization means that there are those pressing for it and those opposed to it.Footnote 50 The relative size of these factions within any country and the specific circumstances (the nature of the technology, the liberalization being contemplated, the state of the economy, and the institutional mechanisms for amelioration) determine the political balance.Footnote 51
Historically, openness and liberalization have been neither self-reinforcing nor self-defeating. The world, and most of the nations in it, have typically been at neither extreme, of complete closure nor complete openness. Were self-sufficiency and closure optimal and viable, and were those hurt from openness always a clear majority, then, in a democracy, either openness would never be attempted or any attempts at openness would immediately be extinguished. Policies of openness would either not be attempted or would be self-defeating. Alternatively, at the other extreme, were openness always preferred by a majority because it benefitted a clear majority then openness would be self-reinforcing in a democracy. Openness would always be favored to closure and the results of liberalizing measures would be merely to enrich and further empower a majority who would press for more openness.Footnote 52 Liberalization would then be self-reinforcing and states would move toward complete liberalization. Yet neither outcome has been the historical norm.Footnote 53 Few states have pursued either complete autarky or complete openness.
Assuming that states retain autonomous decision making, and that as democracies their policies reflect the will of the median voter, then globalization is incompatible with democracy in a world of sovereign states for those countries in which the losers constitute a numeric majority.Footnote 54 Ironically, the pursuit of globalization presumes a preeminent concern with material well-being, but the political consequences of globalization’s material impact renders it self-defeating in some independent democratically governed states. A focus on the material bases and consequences of politics in a system of independent states leads to the conclusion that globalization, sovereignty, and democracy are not compatible. The implications of this are discussed next, and then followed by a discussion of assumptions and the requisites for alleviating the trilemma.
Contested domains
The full force of the trilemma becomes more and more apparent as the number of democracies grows and as globalization continues apace, all in a system of independent states. Then pressures arise to restrain each of the three. Sustaining two of the features comes with pressure to constrain the third. In some countries, those in which a majority of voters are losers, maintaining representative government and state autonomy generates public policy to constrain globalization. A preference for continued economic integration and state autonomy generates a demand for constraining democracy. Finally, a preference for democracy and globalization becomes the basis for limiting sovereignty and political integration. These are core implications of the trilemma and become an additional way to assess the argument. As globalization has proceeded apace and as the number democracies has grown, the world has witnessed pressures to constrain each of the three.
Constrained democracy
The requisites of economic efficiency and integration in a system of sovereign states have generated pressures to limit the domain of democratic governance in societies in which the losers numerically exceed the winners. Democracy itself is under assault and being constrained. There have been and continue to be pressures to dedemocratize domains of economic policy, to remove them from popular control or at least to limit such control.Footnote 55
Although undertaken for reasons other than globalization, democratic limitations have long existed in the area of monetary policy in the form of independent central banks. It is now commonplace to argue that independent central banks generate more stable economic outcomes.Footnote 56 The insulation of monetary policy from politics is now widely accepted and championed.Footnote 57
Globalization has led to some insulation of trade policy from politics. Although not subject to the same depoliticization as monetary policy, trade policy has witnessed its share of constraints on democratic governance. Tariffs are typically set by legislatures but it is the executive branch that negotiates trade agreements, which have historically been the vehicle for trade liberalization. Negotiating trade treaties subject to legislative approval has long been a problem for the executive branch.
Movement toward trade liberalization in the United States, for example, has led to successive limitations on legislative control.Footnote 58 In the 19th century, Congress often rejected trade treaties negotiated by the president or so amended them as to nullify the treaty and any prospect of negotiating one. Trade liberalization required amending the political process for negotiating and ratifying agreements. First, Congress provided advance authorization to the president to negotiate lower tariffs. When even that procedure generated excessive protection, presidents asked for ‘fast track authority’, which provided both prior authority to negotiate percentage cuts in tariffs and committed Congress to vote on any agreements up or down in toto without amendment. The institutional history of the evolution of US trade policy involves repeated reform efforts in search of the institutional arrangement that will deliver liberalization with the least onerous protectionism (Destler Reference Destler1986, Reference Deudney2005; Hody Reference Hody1996).
Whether such limitations on the role of legislatures makes trade politics less democratic can be debated (Hudec Reference Hudec1993). After all, the executives that negotiate these agreements are also elected and the problem of constructing public policy is that of different sized constituencies with different median voters.Footnote 59 Nevertheless, it is clear that the particular character of governance in the United States has had to be modified to deal with the specific problem of negotiating trade agreements, in a way in which it has not in other policy domains.Footnote 60
Constrained sovereignty
Where it has proved difficult to constrain democracy, the alternative has been to constrain sovereignty. Concerned about the implications for economic integration of disparate sovereign jurisdictions whose democratic majority may shift, political and economic elites have gone about constructing supranational arrangements and transferring decision-making authority from the national level. Regional integration schemes such as the EU and NAFTA, along with global arrangements such as the WTO, are full of such arrangements. The ‘collusive delegation argument’ made by EU scholars holds that European policymakers ‘chose to centralize trade policymaking in order to insulate the process from protectionist pressures and, as a result, promote trade liberalization’ (Meunier Reference Meunier2005, 8).
Concern has been voiced both about the loss of sovereignty entailed and that these institutions are not democratically accountable.Footnote 61 This has been dubbed the democratic deficit in the EU case. But these issues are separable. Constraining sovereignty through political integration, of any kind, need not, and typically does not, come with any reduction in democratic accountability. First, it is possible to create supranational representative institutions, such as the European Parliament (and as the US Congress). Second, states themselves retain a vote, and thus for states which are democratically governed, all that has occurred is the creation of an additional layer of representation.Footnote 62
Note that the focus here is not on the ways in which sovereignty is reduced by the forces of globalization and the reduced ability to achieve specific outcomes. Rather, it is on the constraints on the decisional autonomy of sovereignty that states willingly agree to in order to further the forces of economic integration.Footnote 63 And they do so because of the concern about the implications of retaining such authority at the domestic level, especially when they are subject to democratic pressures.Footnote 64
The implications of constrained sovereignty are best understood by looking at domestic society as an analogy for international developments.Footnote 65 The United States is an integrated market containing multiple jurisdictions within it. Market integration within the United States applies not only to capital and goods but even to people who are free to migrate across borders within the country. Yet, there remain local jurisdictions who provide club goods to those living within their communities. These jurisdictions have different levels of taxation and provide different levels of public services. Despite free mobility, a multiplicity of jurisdictions thrive providing different levels of club goods. Each of these localities is democratically governed. In this case, market integration and democracy in subnational entities are compatible because these entities are not sovereign. State and localities retain decisional autonomy in some matters, but they are fundamentally constrained from imposing barriers to flows (a core feature of the Constitution). This case, too, illustrates the compatibility of democracy and globalization even with multiple jurisdictions, as long as their decisional autonomy (sovereignty) is constrained.Footnote 66
In short, sovereignty is constrained to facilitate globalization when the supranational arrangements are both binding and not democratic. In such cases, states can retain their democratic character and not restrain globalization because particular instruments of state policy have been shifted to a supranational level.
Constrained globalization
Despite the trumpeting of the process of globalization, it remains a dynamic simultaneously being pressed forward and restrained. As noted above, states have adopted policies of partial openness but have not moved either individually or collectively toward complete openness (with the possible exception of projects for complete economic integration such as the EU). As classical barriers to trade have dropped, they have adopted a variety of measures to manage their adjustment. Over time, even the domestic measures they used to cushion the shock of international competition came under assault as non-tariff barriers.Footnote 67 The pressures to harmonize are evident, as are the array of governmental policies used to deal with the pressures of economic integration. Much of the observed slowness to witness convergence has precisely to do with public policies used to forestall the consequences of globalization.Footnote 68
Concerns with the pace and scope of integration have been apparent all along as most international monetary and trade agreements typically include prolonged phase-in periods as well as escape clauses for exigent circumstances.Footnote 69 States negotiating these agreements recognize they will need a period in which to adjust to increased integration and that there will be situations in which they will simply not be able to sustain their liberalizing commitments. Long phase-in periods are intended to deal with immediate opposition whereas escape clauses are intended to deal with economic downturns and future crises. Yet, these have been insufficient to preclude continuing efforts to manage exchange and sustained and successful opposition to globalization.
When extended phase-in periods have proved insufficient to secure political support for further liberalization, the price of liberalization has entailed protectionist buyouts and side payments for firms and industries and workers who would be hurt. In the 1960s, textiles received special treatment, by the 1980s it was steel and autos. Movement toward liberalization, even after institutional reform, has still entailed protectionist buyouts as a price for further liberalization (Stein Reference Stein1993).
The restraints on globalization are also evident in the area of international capital movements. On the one hand, there are steps states take autonomously, as evident in the continued use of capital controls and renewed discussion of their viability, and interest in capital controls reflects distributional considerations (Alesina and Tabellini Reference Alesina and Tabellini1989; Alfaro Reference Alfaro2004). On the other hand, there are discussions about measures states can take together to manage global capital movements. The International Monetary Fund (IMF) has shifted its position on the acceptability of capital controls under certain conditions.Footnote 70 Economists are now suggesting the desirability of limiting capital movements. The implications of heightened sensitivity to capital movements has resurrected an idea associated with James Tobin of taxing international capital transactions.Footnote 71
Despite heightened living standards and increased wealth generated by greater economic openness, attacks on globalization seem omnipresent. And they are driven by the reactions of economic losers. There is a great deal of evidence about the reactions of economic losers. As expected in an argument focused on the preferences of individual citizens that derive from their economic prospects, survey evidence suggests that views of protection flip between higher-income and lower-income countries: protectionism is preferred by low-skill individuals in the former and high-skill individuals in the latter (Dutt and Mitra Reference Dutt and Mitra2002; O’Rourke and Sinnott Reference Obstfeld2002; Baker Reference Baker2005; Mayda and Rodrik Reference Mayda, O’Rourke and Sinnott2005; Mayda, O’Rourke and Sinnott Reference Mayda and Rodrik2007; Dhingra Reference Dhingra2014).
The globalization that has developed reflects technological change but also state policy. States have tried, with different degrees of success to channel and manage globalization. Significant restraints on economic integration remain and individual states continue to constrain the forces of globalization.Footnote 72
A sovereign state system and democratic rule have generated pressures to restrain globalization. Successive trade rounds have taken longer to negotiate, have faced greater hurdles in industrial democracies in which losers outnumber winners, and in some cases have failed. In effect, the pace of globalization has slowed sufficiently to reduce the number of losers sufficiently to sustain extant levels of integration and sustain small further increments.
Assumptions
The foregoing analysis depends critically on a set of assumptions, about the nature of citizen preferences, about the consequences of globalization, and about the workings of democracies. First and foremost, the preceding analysis has been relentlessly materialist. In part this reflects the central consequence of globalization as consisting of economic integration. Globalization is first and foremost an economic phenomenon and its economic consequences matter for governments and citizens. This assumption reflects the centrality for governments of assuring economic performance. In the modern world, even autocratic rulers are concerned with the state of the economy and its implications for their continued rule. It also assumes that people are most concerned with the economic rather than social and cultural implications of globalization. And it presumes that electoral choices are driven by material considerations (more on this below) and reflect the median voter or a national referendum. The presumed trilemma rests critically on these key economic and political assumptions. Absent these assumptions, democracy, sovereignty, and globalization would be compatible.
The analysis rests on a set of assumptions rooted in economic theory for which there is some empirical validation. International trade and finance are assumed to improve the material well-being of all states (even as it may increase the degree of inequality between them). Without this assumption, there would be states opposed to globalization even without their being democratic. An additional critical assumption is that there are economic losers as well as winners from increased openness and that these are by and large known.Footnote 73 That in some cases the state as a whole can benefit while a majority of its citizens do not is what underlies the trilemma.
The argument further assumes the centrality of material factors in the political process, that people vote their economic interests, that if the median voter is a loser to globalization then she votes for policies to contain it. It assumes that losers from economic integration will use the political process to redress their deteriorating economic fortunes and that political contestation will reflect such concerns.Footnote 74 As noted, there is ample evidence that individuals vote their economic interests and that their preferences regarding trade and economic integration reflect those interests.
The trilemma is obviated if economic losers do not vote their economic self-interest. For example, if people vote sociotropically, on the basis of what is best for the country even if it hurts them, then there would not be a problem given that we assume that globalization does benefit society as a whole.Footnote 75
Even if voters focus on their narrow material self-interest, it may still be that other concerns dominate economic ones both as issues in specific campaigns and as the basis for individual political choices. Policies that further economic integration will nevertheless be sustained by democracies in which a majority are hurt economically by globalization if non-material concerns dominate the political calculations of the economic losers. The importance of such non-material considerations has been proffered as a more general explanation for the limited degree of redistribution in democracies and why a poor majority does not vote to redistribute wealth from a rich minority (Roemer Reference Roemer1998).Footnote 76 It should be noted that the role of non-economic factors is a dual-edged sword and may be the basis for opposition to globalization by others who would otherwise benefit by it materially. Cultural and ideological factors, such as patriotism and chauvinism, can affect individual trade preferences for protection (O’Rourke and Sinnott Reference Obstfeld2002). Opposition to globalization’s social and cultural consequences is another manifestation (Margalit Reference Margalit2012).
Both formal models and empirical assessments provide support for an argument that there is less redistribution than would be expected on purely self-interest grounds.Footnote 77 Lindert (2004) describes a ‘Robin Hood paradox’ in which democracies with highly unequal income distributions redistribute less than would be expected from a median voter model. The reason typically proffered, in both formal models and empirical studies, is that parties mobilize voters on grounds other than income and that individuals make political choices based on these other factors.Footnote 78 The common argument for why they do so is religion, or moral values (Roemer Reference Roemer1998; Frank Reference Frank2004; Scheve and Stasavage Reference Scheve and Stasavage2006; Huber and Stanig Reference Huber and Stanig2007; De La O and Rodden Reference De La O and Rodden2008; Lee and Roemer Reference Lee and Roemer2008; Finseraas Reference Finseraas2010).Footnote 79
The existence of a trilemma is also critically based on an assumption of the political process. The focus here has been on electoral democracies because, contrary to autocracies, we have an expectation of policy as subject to the preference of the majority. No such expectation exists in autocracies. The relationship between autocracy and globalization depends on the societal groups to which autocrats are beholden. For example, autocrats beholden to the minority interests of capital in labor-abundant countries would oppose globalization. Similarly, populist autocrats in labor scarce countries would also oppose globalization. In short, the policies of autocracies regarding globalization is not derivable from the materialist implications of economic integration, but also depend on the particular coalitions sustaining any particular autocracy. There is no systematic relationship between the winners and losers from globalization and the policies regarding globalization of autocrats.
The case for a trilemma for democracies not only presumes the basis of the political behavior of losers from globalization, it also presumes that policy in democracies reflects a referendum or the interest of the median voter. Alternatively, scholars of the political economy of trade policy have developed models focusing on lobbying and campaign contributions rather than the median voter (Rodrik Reference Rodrik1995). These models are typically used to explain how protection is for sale (Grossman and Helpman 1994), but they are inherently indeterminate in that there is no necessary analytic basis for the relative level of lobbying by protectionist and liberalizing interests nor for their relative political effectiveness.
Yet, even a characterization of democracies as reflecting the median voter should not result in a trilemma, absent certain assumptions about the resultant public policies. After all, in a material sense, democracy is a commitment device to accept the median voter’s preferences on redistribution, and if the gains of the winners exceed the losses of the losers then redistribution should result in a democratic commitment to liberalization even if the presumptive losers are more numerous.
For globalization to be a Pareto choice, in such a setting, simply requires the losers to be compensated by the winner so as to make them barely prefer openness or at least be indifferent (Kemp Reference Kemp1962; Samuelson Reference Samuelson1962). The winners’ gains are reduced but the losers are compensated.Footnote 80
The problem is that redistribution and the compensation of losers is problematic. First, the optimal form of redistribution is through lump-sum transfers but these have insurmountable informational requirements and are not incentive compatible (Dixit and Norman Reference Dixit and Norman1986; Ichino Reference Ichino2012). That is, this form of redistribution requires precise information about the characteristics of individuals who have strong incentives to misrepresent the impact of globalization. Economists thus shifted their attention to non-lump-sum transfers but these have their own informational requirements and their own inefficiencies. The problem is that a Pareto distribution requires targeted transfers with precise knowledge of the winners and losers and the requisite level of compensation.
The compensation schemes adopted by governments are either inadequate or problematic in a globalizing world. The only compensatory policies that target those experiencing direct losses from globalization are unemployment benefits and trade adjustment assistance which typically consists of job training for those who have lost their jobs from imports.Footnote 81 Unemployment benefits do not compensate completely for lost wages and often end after some period. Job training comes with no assurance of employment and comparable income. These programs do not typically compensate the losers sufficiently to make them indifferent between protection and liberalization.
The most commonly noted form of presumptive compensation comes in the form of a larger public sector. Political scientists and economists point to the correlation between trade exposure and public spending as compensation to the dislocations wrought by globalization.Footnote 82 Such public spending depends on non-linear taxation (and transfers) which are applied on the basis of income and not on the basis of individual characteristics (such as skills).
Yet, the policies that cushion the consequences of economic competition themselves come under attack. A global marketplace creates pressures for the harmonization of economic policy and places redistributive mechanisms under immense strain.Footnote 83 Regulatory policies can be more easily harmonized than redistributive mechanisms because the needs for the latter differ across democracies with different constellations of winners and losers. As adjustment pressures grow so do the needs for redistribution, yet the integrating market that generates the adjustment pressures also makes redistribution more problematic as certain forms of redistribution are not possible (Spector Reference Spector2001) or reduce domestic competitiveness and increase capital flight. Globalization makes redistribution more politically necessary but it is economically and politically problematic. The continued viability of disparate national compensatory policies is at question.Footnote 84
Given the minimal measures to compensate the median voter, and assuming a growing inability to do so, globalization requires either depoliticization or supranationalization, or both. It depends either on insulating state policy from democratic political pressures or taking certain levers out of the domain of the sovereign state. Or at least it does so in some states, those in which the median voter is economically hurt.
If voters vote their pocketbooks in countries in which redistribution from an enriched minority is not possible, globalization will come under attack in democracies with a majority of economic losers. Critical to this conclusion are assumptions about what animates voters and what constrains redistributive policies of economic compensation.
Conclusion
Globalization is an ongoing process. So is democratization. Both are occurring in a system of sovereign states. Yet, globalization, democracy, and sovereignty are incompatible in material terms in their full blown form. The result of the expansion of democracy and continued globalization in a sovereign state system has placed strains and pressures upon all three. Contemporary politics simultaneously evinces forces aimed at constraining globalization, limiting democracy, and reducing sovereignty. This is the inherent result of their incompatibility.Footnote 85
The present, with its pressures on all three, provides a sense of prospective futures. One can imagine a continued backlash to globalization that only increases with continued democratization. In such a scenario, globalization would be the casualty of democracy and sovereignty. Alternatively, there is the vision of world government that has sometimes found adherents. Here, sovereignty becomes the casualty of democracy and globalization. Finally, democracy could be the casualty of continued globalization in a sovereign state system. Voters would still vote, but key aspects of policy necessary for sustained economic integration would be insulated from politics.
Can the three be reconciled? The key entails the political economy of globalization. One possibility is a political ability to change the distributional consequences of globalization, to make it Pareto improving for the median voter in every democracy. Yet, this can only be accomplished through internal economic transfers which themselves come under pressure in a fully integrated global economy.
A second path to reconciliation lies in the role of the non-material as the basis for political choice. The trilemma presumes a particular material basis for political preferences in a democracy. The trilemma is obviated if voters do not vote for redistribution, but also heightened if voters oppose globalization on cultural or social, rather than material, grounds.
A third path to reconciliation is policy in a democracy driven by the superior lobbying ability of the winners from globalization. Were the winners from globalization always more capable of solving the collective action problem and capable of using their resources to capture public policy, there would be no trilemma.
The most likely scenario is an extrapolation of the present. It is a world of contested domains, one in which efforts to constrain globalization, democracy, and sovereignty compete with efforts to expand and maintain them.Footnote 86
Acknowledgments
Earlier versions were presented at the annual meeting of the American Political Science Association, San Francisco, August 30, 2001, at the Colloquium Series on ‘The Globalization Question’, Center for Globalization and Policy Research, School of Public Policy and Social Research, UCLA, October 24, 2001, and at the annual meeting of the International Studies Association, March 25, 2002. The author thanks Jerry Cohen, Amy Davis, Megan Gallagher, Deepak Lal, Michael Mann, Alex Wendt, and anonymous reviewers for comments. The author thanks the UCLA Academic Senate and the Carnegie Corporation of New York for support.
