There are many parallels between U.S. and EU immigration trends and policies despite the clear and important differences in their decision-making institutions. Among the common problems these two entities face are the following: a global economy that promotes (or perhaps requires) a freer migration of capital and labor; fluctuations in the economic cycle that alter the perceived value that immigrants make to national economies at given points in time; greater racial, national and cultural diversity in the immigrants and refugees who seek entry into the US and Western Europe; political backlash from the far right against the growing presence of third country migrants and co-optation of their issues by more mainstream, center right parties; changing patterns of foreign migration from single male sojourners to permanent resident families; the failure of official border controls to stop the flow of illegal immigrants and false asylum applicants; and mounting pressure from cosmopolitans who deplore both the direct and indirect racial discrimination that these policies have spawned and who hold out an ideal of a borderless, prejudice-free international community that directly challenges traditional concepts of national sovereignty and citizenship. But for all these similarities, there are also important differences in how the U.S. and Europe have handled these new challenges, and some of this variation can be traced to institutional factors.

To pose the democratic theory question most abstractly, the issue is whether particular institutional arrangements bias the attainment of fair and efficient immigration policies in predictable ways? Some definitions are needed at this juncture. The term "fair" is here defined as a policy that represents the majority of a policy's constituents' wishes without violating basic, agreed upon fundamental rights as embodied in national or supranational constitutions, or in international treaties. Clearly, this is only one meaning of fair, and a US-centric one at that, but at the least it gives us a useful baseline for analysis. "Efficiency" refers to the proper functioning of the market to provide sufficient labor to meet the needs of an advanced economy. The relevant questions then are: 1. To what degree do U.S. and EU immigration policies meet these criteria? ; 2. Do these policies deviate from these criteria in systematic and predictable ways? ; 3. Finally, do institutional incentives explain the observed variations in their immigration policies?

Posing the institutional question is this way does not presume that institutions by themselves determine immigration policies. Policies are a function of both the preferences of voters and interest groups and the context of institutional structures. Therefore, depending upon the distribution of voter preferences in a country or area, it is possible for very different institutional structures to have very similar immigration policies, and vice versa, for similar structures to produce very different policies. The former (i.e. similarity in policy despite differences in structure) is the case presently with respect to the EU and US policy on undocumented immigrants. Both, for instance, have taken steps in the last fifteen years to restrict the flow of undocumented immigrants and refugees despite very different mechanisms for making such decisions (Cornelius, Martin and Hollifield; 1994). The point is even truer if we acknowledge that EU policy in this area is really determined by national governments that have an even wider variance in institutional structure, but nonetheless exhibit a fairly uniform concern with the problem of false asylum seekers (Fassmann and Munz,1992; Pettigrew,1998; Ucarer and Puchala; 199 ; Papademetriou and Hamilton,1996 ).

While acknowledging that institutions do not completely determine policy, it is the case that they are the source decision-making incentives and constraints that make certain immigration policies more or less likely. This can happen in various ways, such as by fixing the threshold of consent needed to pass legislation, multiplying the number of veto points, or varying the degree of constraint that court-enforced rights play in policy-making. Specifically, my thesis is that the EU decision-making structure biases immigration policy in a direction that will make the expansion of immigrant labor very difficult if and when economic conditions favor expansion. By contrast, the US decision-making structure favors expansion, even under unfavorable economic conditions. Under unfavorable economic conditions, however, the bias of the EU is less evident since it tends to produce an outcome that is both majority-preferred and economically efficient. Indeed, the so-called democratic deficit problem, a topic of great controversy in the EU, really does not apply to EU immigration policy under present conditions in Europe. The democratic deficit in this policy area will likely be realized when economic and labor conditions require more immigration in order to preserve growth and prosperity.

The first section of the paper that follows will examine the democratic deficit argument in some detail and suggest that while parts of it seem correct, other parts do not. In particular, the issue needs to be reformulated more clearly, and some new distinctions need to be made. Secondly, the EU decision-making structure with respect to immigration policy will be compared with the US with an eye towards the relative biases of both. Lastly, returning to the EU democratic deficit problem, I will argue that if and when a true democratic deficit emerges with respect to immigration policy, this might prove to be an important event in the evolution of the EU in so far as the goal is to build upon economic cooperation towards a stronger political union.

The Democratic Deficit and EU Immigration

A common argument in the literature is that the democratic deficit in the EU's current governance structure as it relates to the core issue of citizenship leads to an increasingly restrictionist and discriminatory policy towards non-EU refugees and immigrants (see especially Sorenson, 1996; Papdemetriou; 1996). Since Western Europe has not encouraged or received large numbers of legal economic immigrants since the mid-seventies (Fassmann and Munz, 1992), the most important immediate issues concern refugee policy: the responsibility of European nations for taking legitimate political refugees (i.e. those that meet the stringent definition of the 1951 UN Convention on the Status of Refugees), and the problem of dealing with those who use refugee status for economically motivated reasons. Since the mid-eighties, there has been a dramatic and unprecedented increase in the numbers of persons requesting asylum in EU nations. By the early nineties, there were more than three times the number of asylum seekers into the Member states that there had been in 1985 (Sorenson, 1996;91). There are now also large numbers of illegal immigrants in Europe (Sorenson, 1996: 90). Hence, while legal immigration has dropped considerably, migration still occurs in significant numbers.

In response to the rise in asylum seekers, the EU member states have adopted a number of restrictive measures. Some nations have adopted stricter visa requirements for individuals who come from refugee-producing states. Others have imposed carrier liability on flights and ferries that take persons on board who do not have proper documents. Lists of safe countries from which individuals have no need for asylum have been drawn, and many refugees have been assigned a "temporary status," giving the state the right to expel them when conditions appear to have improved sufficiently. Policies and agreements fashioned at the EU level are primarily concerned with formal control procedures for limiting access and not with harmonizing migration law per se. For instance, the Dublin Convention (1990) represents an attempt by the Immigration Group to deal with the growing problem of asylum abuses. In an effort to put an end to "asylum-shopping," this agreement designated a single Member state to examine a refugee application, depending upon a number of criteria. Similarly, the Convention on the Crossing of External Borders (1991) attempted to tighten controls over third country nationals by more explicitly setting out the definition of external frontiers and the conditions for crossing them.

The harshest critics of current EU refugee policies believe that the EU structure is responsible for an excessively exclusionary and racially biased policy. The argument goes as follows. EU immigration and migration policy has been forged within an intergovernmental framework (Sorenson, 1996; Geddes, 1995; Papademetriou, 199; Butt, 1994; Convey and Kupiszewski, 1995). The Trevi group, the Working Group on Immigration (1986) that grew out of the work of the Trevi Group, the Schengen Agreement (1984), the Dublin convention (1990) and the External Frontiers Convention (1991) are all examples of intergovernmental agreements that exhibit three features First, they cede power on immigration matters to the national governments, and each state has formally equal power in the form of a national veto. Secondly, neither the European Parliament nor the European Court of Justice can over-ride the decision-making or implementation of these intergovernmental decisions. And lastly, the negotiation process in this area tends to be secretive and unaccountable.

Added to the general perception that European "economic integration involves a technocratisation of still more policy areas and issues without direct democratic control" (Sorenson, 1996,122) and that citizens feel locked out of participating in EU decision-making processes, the modus operandi of immigration and refugee policy-making only reinforces the general democratic deficit problem. In the words of one scholar: "Parliamentarians meanwhile have begun to get increasingly anxious about the opaque and uncontrolled decision-making processes by which many of the new decisions at EC level on citizen rights, asylum and refugee policy, and immigration matters are made…The problem of lack of control over the decision-making processes arises from the sensitivity of many of the issues (inviting secrecy of discussion) and the conduct of negotiations on an intergovernmental basis where the negotiators are not accountable to EC institutions, nor frequently to national parliaments. In practice, national parliaments have been presented with deals done by ministers and officials behind closed doors and have been required to approve such deals without amendments because each is a delicate intergovernmental compromise."(Phillip, 1994, 8).

Another related critique argues that the democratic deficit is even greater for the ethnic and immigrant minority groups who live in EU countries. In the words of Andrew Geddes, "a central feature of the Union is that many decisions are taken in intergovernmental forums which often meet in secret, about whose activities remarkably little is known and which are distant from the people of Europe in both a political and geographical sense. This participatory gap-an important aspect of the democratic deficit-is compounded for people from ethnic and immigrant groups in the EU." (Geddes, 1995, 197). In other words, the problem of political response is even greater for ethnic and immigrant groups who would like to counter the anti-immigrant backlash in Europe, but find that they are hindered from doing so by being in the minority and having no few political rights. Even if they were to achieve such rights, inter-governmental modes of decision-making remove the issue one step further from any arena in which they can have impact.

Finally, a third criticism holds that that EU intergovernmentalism leads to a restriction bias that may ultimately prove to be both inefficient and unfair as previously defined. In his book, The Exclusive European Citizenship, Jens Magleby Sorenson argues the intergovernmental structure presented nations that wanted more liberal immigration policies with "a dilemma between being flooded by new immigrants or making their migration and immigrant policies as strict as the strictest member state. Thus when migration issues had to be coordinated between the member states, the tendency of national migration laws has been the erection of stricter and less attractive migration and immigrant policies. A highest common denominator approach will become the standard for the region." (Sorenson, 1996, 146). In other words, the cooperative dilemma that arises when the Member nations separately make decisions about residency and citizenship combined with the unit veto nature of intergovernmental negotiations empowers nations that favor more restrictive policies-they cannot be outvoted by a more moderate majority and forced to compromise.

To unravel this a little more, there are two factors in this immigrant aversion bias. First, there is the problem of the national veto, which empowers the restrictionist outlier in collective decisions, and second, there is the implicit prisoner's dilemma that prevents individual nations from being too generous in their acceptance of immigrants or refugees. With respect to the veto, giving European nations the right to veto or not sign on to an agreement constrains agreements in important ways (Scharpf, 1988). Also, nations are not equally populated, from a purely majoritarian perspective, nation-based negotiations do not necessarily lead to decisions that a referendum of all the individuals in Europe would majority approve. Proposals to make third world immigration or asylum seeking easier across the EU Member states can be vetoed by the nations that prefer highly restrictive policies. This point of course applies to all movements away from the status quo, and not just expansionist outliers. A unit veto or supermajoritarian system requires virtual or total unanimity in order to initiate change. Since the status quo in Europe is highly restrictive to begin with, the inertia of the system makes it hard to move towards more liberalized immigration policies.

The other component of the restriction bias is the implicit prisoner's dilemma between nations that want to have more liberal immigration policies and those that do not. In the integrationist ideal, labor flows freely to the areas that have the most demand. In theory, when demand wanes in a given area, labor will cease to migrate into that area and go elsewhere. But in practice, often labor either cannot go elsewhere (i.e. if there is no free movement across borders) or chooses not to go elsewhere for noneconomic reasons (e.g. because the package of political and social rights are not favorable). Hence even when a nation wants to live by liberal trade ideals, it cannot do so if other EU nations do not cooperate. This then causes the more liberal nation to pursue a more restrictive policy than it might otherwise. Similarly, with respect to refugee policy, a nation that wants to live up to humanitarian ideals will feel more reluctant to do so if the refugees it accepts are prevented from moving elsewhere within the EU or if for want or other alternatives, it must absorb more than its fair share. It is characteristic of these cooperative games that risk averse players often choose outcomes that are suboptimal for fear of being taken advantage of this logic, reinforcing to the restrictionist bias created by the intergovernmental form of agreement.

While this is a powerful and convincing analysis of EU immigration policy, there are a few aspects of the democratic deficit critique that seem a bit muddled. First, with respect to accountability and secretiveness, the question is really one of democratic oversight over the actions of elected and unelected members of the state. This is a problem for all democracies, including the US. The evidence that it is uniquely bad in the area of immigration policy is lacking. Military negotiations and treaties are every bit as sensitive as immigration issues, and they are also often presented to national legislatures as take-it-or-leave it propositions because of the intricacy of the negotiations and the multi-sidedness of contemporary international conflicts. If the democratic deficit is worse in Europe than in the US in this specific sense (and the evidence on that point is lacking), it is not simply an EU deficit per se but also a national deficit by default: that is to say, the fact that the most critical immigration and migration decisions have been ceded by default to national governments means that the deficit lies partly in the national structures themselves. As such, this is really an inherited democratic deficit problem, not a newly created one. Because European nations tend to have stronger state mechanisms and central administrative structures, bureaucratic controls are less directly electoral than the US. From another vantage point, this has also meant that European nations have enjoyed certain advantages in governance as well. As democratic theory so often warns, institutional choices are about trade-offs in advantages and disadvantages.

The restriction bias produced by the national veto and the prisoner's dilemma of cooperation on immigrant policy is a serious structural issue. But ironically, there is actually little evidence that the democratic deficit matters much in this area under current conditions, although it might under different, more favorable conditions. Presently, the EU's restrictive polices, it could be plausibly argued, are fair and efficient as previously defined. They are fair in the sense that they are likely the same policies that would be chosen by European electorate on a referendum vote, or if the matter were given over to the European parliament.

There is no evidence that a majority of Europeans want to admit larger numbers of non-EU refugees except in response to very specific refugee crises (e.g. the British seeking to relieve the plight of the Kosovars). Reviewing data from seven Eurobarometer surveys, Thomas Pettigrew observes that the percentage of Europeans who believed that there were two many "non-EU foreigners in their country" rose dramatically between 1988 and 1991, with clear majorities in Belgium, West Germany, France and Italy. There was also a rapid increase in the numbers of Europeans who wanted to restrict the rights of non-EU nationals (Pettigrew, 1998). Some argue that "in no West European country can politicians or political parties gain votes by favoring new immigration or immigrant voting rights" ( Lahav, 1996; Messina, 1990). While the emergence of anti-immigrant forces has given rise to some transnational political organizing on the part of immigrant groups (Ireland, 1991), the calculus of democracy inevitably favors voting majorities over non-voting minorities. Moreover, it is unlikely that MEPs would chose any differently if it were in their power. Survey evidence indicates that MEP attitudes on immigrants are forged by the same ideological and partisan factors that drive members in the national legislatures (Lahav;1997). Hence, even though national government policy-making and supranational policy-making theoretically could yield very different outcomes, that is not what is happening here.

While there is some concern about the details of particular policies and about non-elected officials controlling elected ones, the simple fact is that there is no deficit in the most important sense that the European majority is deprived of policies that are majority-preferred. To put it another way, there may be a procedural deficit in the sense that the process is not sufficiently democratic (i.e., in the sense of being consultative or connected to elected officials), but there is not much evidence of a substantive policy deficit (i.e., in the sense of immigration policies that lack majority support). One might also argue that in the short run at least the current immigration policies in Europe are probably efficient in the sense that the demand for the labor that is being excluded has been relatively low since the mid-seventies. This, of course, might change over time, and efficient short-run labor policies might not be so efficient in the long run.

Indeed, the true motivation for the democratic deficit argument seems to be an integrationist vision or a cosmopolitan ideal that finds restrictionist policies objectionable on other, broader grounds. To the integrationist, the heavy reliance on intergovernmentalism is a step back from progress towards a politically unified Europe. However, the assertion that restrictive immigration policy derives from the intergovernmental structure is both unproven and disengenuous. One suspects that structure is being blamed for what preference has produced. Without placing a normative value on European preferences, the widespread reluctance to absorb more non-EU populations seems to be what is really driving current policies in Europe. As indicated earlier, structure alone does not determine policy, and structure may not matter at all when preferences are configured in a sufficently high level of consensus. That apparently is the present situation in the EU with respect to immigration.

The cosmopolitans have a slightly different take. They rightly see that national control over citizenship and immigration policy area makes it hard to develop universal social and political rights. This is an admirable goal, but until there is indeed a consensus that these rights should be universal, cosmopolitans have to accept that what the majority in Europe wants is fair by definition. The harsh reality is that the liberal trade and cosmopolitan rights ideals are flawed in application. Labor does not simply move in efficient ways towards areas of demand and away from over-supply. Workers put down roots and labor flows become sticky. This stickiness has external costs to the communities in the form of education costs, social security benefits, medical care expenses and the like. This will mean that immigration and openness towards refugees is not always cost effective, and hence other values have to be traded against economic self-interest. Similarly, beyond the most basic rights to life and freedom, nations and cultures vary significantly in the obligations and opportunities that constituents demand as rights. Pretending that there is harmony on these matters when none exists does not necessarily facilitate their evolution.

Hence, the standard story about the democratic deficit in the EU seems to be right in some parts and wrong in others. There is a restrictive logic to the current structure, but it is not clear that the policies that Europe is currently pursuing are more restrictive than a majority of European voters or a European parliament would prefer. There are problems of secretiveness and unaccountability in the negotiations over immigration and refugee status, but they can be traced partly to oversight problems in the national governments, and are in that sense, inherited from national structures, not simply attributable to the EU. Later, I will argue that a true democratic deficit is more likely under different conditions, and that far from lamenting it, integrationists should welcome it since it will be a more effective motive for greater European integration. But first, in order to lay the groundwork for that argument, I will review the US structure for handling immigration and highlight what I take to be its pluses and minuses.

The Biases of the US Decision-Making Structure.

If the EU structure tends to produce a bias towards restrictiveness, what is the comparable bias of US governmental structures? The sheer numbers of foreign population in the US dwarfs the other OECD countries: in 1991, there were nineteen million in the US as compared to close to 6 million in West Germany, 3.5 million in France, 1.7 million in the UK and 896 thousand in Italy. (Money, 1994). On a per capita basis, however, the US falls more in the middle of the range at 7.9%, lagging well behind Australia at 22.8% and Canada at 16%. Even more significantly, the trend line on US legal immigration is a relatively, steady upward drift. By comparison the trend line in other OECD countries that have accepted large numbers of immigrants like Germany and Australia is much more cyclical (Money, 1994, 4-5). The question I want to raise here is whether that can be attributed to institutional sources?

On the assumption that structural differences may account for differences in immigration policy, it is illuminating to contrast US and EU decision-making structures. More than in the EU, immigration decision-making in the US is multilayered. In the EU, the nation-states control both the borders and the allocation of goods and services to non-EU immigrants/refugees while the EU community facilitates the coordination of those policies. In the US, the federal government controls the borders, but the states and local governments provide the critical social services for immigrants such as emergency health care, schooling and police/fire protection. This explains why immigration policy can sometimes be controversial at the state government level. For example, Prop 187 -the 1994 ballot measure in California that would have denied non-emergency state and local services to undocumented immigrants-was an attempt by a state electorate to mitigate the effects that federal immigration policies were having on California by denying state services to immigrants, and thereby making settlement in California less attractive to prospective immigrants (MacDonald and Cain, 1998; Schockman, 1998; Colino, 1995; Johnson, 1995). As it has turned out, the federal courts have constrained a state's ability to use social policies in this way, but even so, the point is that immigration policy is not simply dictated by the Federal government.

This points to a second difference: the greater role the US courts play in significantly constraining immigration policy in at least two ways. First, they have used the 14th amendment and fundamental rights doctrine to limit the ways that US government can treat immigrants and refugees, and secondly, any attempt to return to a Euro-centric immigration policy would likely be struck down by the courts. In Plyler v Doe (1982), the Court ruled that children of undocumented immigrants could not be denied a free public education, based on a 14th amendment guarantee of equal protection for members of a suspect class (i.e. the children) and the assertion of a fundamental right to education. Similarly, any attempt to revert to pre-1965 standards for legal immigration with quotas favoring white Europeans would run afoul of statutory and constitutional protections against racial discrimination. These legal precedents significantly limit the options available to governments, and in particular, make certain kinds of restrictive immigration policies impossible.

With respect to international law, however, the differences between the US and Europe, may be growing less. Consider, for instance, how the US handled Albanian refugees in the wake of the NATO bombing of Kosovo. Faced with the crisis of a half-million Kosovars fleeing their homes and seeking safety in neighboring Macedonia, Albania and Montenegro, the US along with Germany, Turkey, Norway, Greece and Canada agreed to accept 20, 000 refugees to ease the pressure on Albania. However, it did so in a manner that critics charge was "calculated to deprive the ethnic Albanians of the rights which they would be entitled to under the 1967 Protocol Relating to the Status of Refugees ratified by the United States in 1968." (Musalo, 1999, 2). By proposing to place the refugees on the US Naval Base in Guantanamo Bay, as they had the Haitian refuges following the 1991 coup of Jean Bertrand Aristide, the US could keep the Kosovars off US territory, thereby denying them the right to apply for political asylum in the US.

The reality, say critics, is that while the US is widely regarded by many countries as a leader on refugee issues, it clearly has adopted strategies in recent years to minimize its obligations under international law. This is very similar to the strategies the British (who have a much worse international reputation on refugee matters) have recently adopted to discourage false asylum seekers. Faced with a large and growing backlog of asylum cases and the fear that the benefits of the British welfare state might be a powerful allure to would-be refugees, the British government has restricted benefits (e.g. replacing cash welfare benefits with food vouchers and not giving asylum seekers a choice of where to live), speeding up the claims process and making it harder for refugees to get into Britain (e.g. fining airlines for carrying passengers without proper documents). In short, the US appears to be more similar to Europe with respect to international law than in domestic law.

Thirdly, US decision-making is more majoritarian than the EU system at present. As noted earlier, the EU makes immigration policy largely through intergovernmental negotiation, and the national units are not equally sized. Hence, it is possible (although it is not currently the case) for national negotiations to produce immigration policies that are not majority preferred. That can also happen in the US, but it is harder. With the exception of the US Senate, "one person, one vote" has replaced territorial representation, and the House and the President are strongly influenced by majoritarian pressures.

This majoritarian influence is somewhat greater at the state than the Federal level. In addition to the US Senate (the only violation of the "one person, one vote" principle the Supreme Court allows), the Federal system offers multiple veto points in the organization of the US Congress and supermajoritarian rules (e.g. to break a filibuster, to overturn a Presidential veto) that empower non-median interests and representatives (e.g. Krehbiel, 1998; Brady, 1997). The states mostly mimic the Federal government structure with division of power between three branches (with the exception of Nebraska), but many states supplement their representation with direct democracy mechanisms such as the referendum and the popular initiative. This has both the direct effect of causing majority policies to succeed at the ballot box when they cannot pass through the state legislatures to and the indirect effect of pressuring legislatures to pay closer attention to median voter positions (e.g., Gerber, 1996).

With these comparisons in mind, let us return to how democratic deficits affect US immigration policy. Leaving the court constraint aside for the moment, the prospects for the kind of democratic deficit the EU is vulnerable to are somewhat less in the US case. This is particularly true at the state level where median voter pressures are favored in various ways, as discussed earlier. Moreover, since states do not have unit veto power, any territorial or state-centric to either block departures from the status quo or initiate new policies put forward in the House or Senate must win significant support from representatives from other states. To put it another way, it would be harder for any outlier state that prefers greater restrictions on immigration to prevent the liberalization of national immigration policy unless it had either majority support in the House, or by the logic of pivotal politics, the support of at least a third of the Senate to sustain a filibuster. The same point applies to the liberalizing state outlier.

Secondly, the US structure solves part of the prisoner's dilemma that EU nations face. In the European case, all nations might be better off with a free flowing labor supply across borders, but nations that open their borders when others do not face the prospect that they will be saddled with the social costs of unwanted immigrant labor when economic times turn down. By having one and only one border policy for all in the US, and guaranteeing free movement within the states for immigrants and non-immigrants alike (although the Carter refugee program tried to limit movement for specified periods of time), there is no cooperation dilemma concerning the border. But the multi-layered nature of decision-making in this area still leaves states with a second and related cooperation dilemma in the sense that if one state's social policies are too favorable, they might attract more than their fair share of immigrants and refugees. Clearly, this was an issue that Prop 187 and welfare reform sought to address. By constraining the most restrictive states in terms of what they can refuse to offer immigrants, the courts have de facto limited the effects of this cooperation problem somewhat.

Finally, with respect to the democratic deficit problem that might arise through the secretive and unaccountable implementation of immigration policies by executive agencies, the Congress provides oversight into the administration of immigration policy by the INS. By both casework intervention and congressional hearings, the INS is closely monitored by US elected officials.

The upshot of all these factors is that the US legislative policy is more likely to be based on majority opinion than the EU. Having said this, there are two factors that limit the responsiveness of immigration policy to shifts in majority opinion. First, there is an enormous amount of policy inertia built into the US system. Quite apart from the drag that federalism and the division of power place on policy-making at all levels, the decentralized nature of Congressional decision-making and the multiple veto points that interest groups and individual legislators have in the US system serve to slow down and stop many new initiatives. It takes a great deal of consensus to overcome these inertial pressures.

But the most important democratic deficit in the US with respect to immigration policy-making comes from the courts. Whereas in the EU, it is theoretically possible for nations to create policies through inter-governmental agreements that are not preferred by a majority of EU voters, the US policies can be biased away from the median by court decisions. Whereas in Europe, those who want a more liberalized immigration and refugee policy complain about national control, in the US, those who want to see a more restrictive policy object to the Court's constraints on policymaking. Since the federal courts are more insulated from majoritarian pressures and are less accountable to the people, they create yet another kind of democratic deficit. But this takes us to the critical question of whether a democratic deficit is necessarily a democratic deficiency.

Venue Shopping, Institutional Reform and Democratic Deficits.

The discussion to this point reveals that there are at least three different types of democratic deficits that arise with respect to immigration policy, and that at least two of these help to explain the characteristic biases in US and EU immigration policy. The first type is a deficit caused by the failure of delegation oversight: A democracy (or a union based on democratic principles) delegates the details of immigration policy (i.e. the issuing of visas, the policy for handling of problematic asylum cases, etc.) to administrative officials (especially unelected ones), but then has to hold them accountable. This entails finding out the specifics of what those officials have done and making them responsive to the wishes of the citizens. As suggested previously, this is a problem all democracies struggle with, particularly with respect to technical issues. A growing literature in American political science places the delegation aspects of modern democracies in the framework of principal-agent models (McCuddins, 1985). These situations are often characterized by informational asymmetries in which the agent knows more about the details of a given issue than the principal.

In order to hold agents accountable, it is important to find ways to inform principals about actions that affect them and to design mechanisms (usually electoral) by which principals can bring consequences to agents when they do not like the consequences of the actions agents have chosen (i.e. mechanisms of accountability). In the checks and balances US system, the agency that deals with immigration (i.e. the INS) is ultimately accountable to both an elected President and legislature. The drawn-out saga of Elian Gonzalez nicely demonstrates the degree to which Congressional oversight checks (or interferes with, as the case may be) the INS. Had the Congress not itself been divided over the benefits of giving the boy permanent residency or citizenship, it could have taken the matter out of the hands of the INS in one Congressional act. Or, to take another aspect of this example, the Republican led Congress has taken upon itself the task of investigating the child's seizure to discover whether excessive force might have avoided, demonstrating the downside of accountability can be politicization- i.e., using immigration disputes for partisan political advantage. But one at least cannot say that the INS was operating in an unexamined, unconstrained, policy-making vacuum.

The EU's involvement in immigration and refugee issues is necessarily limited to issues of cooperation in policing borders. The understandable unwillingness of European nations to surrender their control over basic decisions of national membership pushes any democratic deficit with respect to the core issues of immigration and refugee policy back to the national arena. The accountability of national agencies on these issues is a matter of national institutional structure. But as European nations have come to deal with common problems of enforcing borders and trying to reconcile the principle of free labor movement for EU citizens with the national concerns for controlling non-EU migration, immigration has slipped into the least accountable and visible realm. The first type of democratic deficit seems to be most applicable to the EU. However, it is hard to see that this type of deficit necessarily contributes to any particular policy bias per se. In principle, it could serve the interests of liberalizing immigration as much as limiting it.

The second type is judicial and international treaty constraint, limiting the freedom of the US and Europe to close their borders to legitimate political refugees and preventing American states from denying fundamental rights and services to undocumented immigrants. This becomes a democratic deficit when the courts overturn the will of the majority in a particular state or country. The best illustration of this is the court action in preventing the implementation of Prop 187, the California measure that would have denied non-emergency benefits to undocumented immigrants. This deficit type is more prevalent in the US, because of the checking role the courts play in the structure of US government. As noted before, the rights orientation of the US courts contributes towards a liberalizing bias in US immigration policy. In theory, international treaties also protect the rights of political refugees, and in this way, constitute a liberalizing constraint, but both the US and EU nations have devised ways to limit their responsibilities in this area.

The third type of democratic deficit is the mediation of constituent preferences through subnational or formal group representation. This is a deficit in the same sense as the other two types, because the policy that emerges from mediated representation can vary from the majority will. Of course, all republican forms of government as opposed to direct forms of democracy can suffer deficits for any number of reasons-the influence of lobbyists or special interests on representatives, miscalculation on the part of elected officials, party pressures, etc. However, within the context of EU immigration policy, mediation can become problematic when national constituencies make policy for multinational entity. Aside from the fact that the European Member states are unequally populated, the delegation to nations in the EU introduces a prisoner's dilemma cooperation dynamic that favors a more restrictive policy.

The assumption in the literature is that democratic deficits are undesirable. Clearly, they can be, but deficits are not always deficiencies for two reasons. One has already been mentioned-the existence of a deficit does not always mean that the policy chosen is different from what would be chosen if the deficit did not exist. This gets back to my point that policy is both a function of preference as well as institution. There is no evidence that Europe has adopted current policies that are more restrictive than what they would have adopted if the matter had been decided by referendum or even by the European parliament. However, a different configuration of preferences might change things. Assume for the moment that the growth rate in Europe greatly increases and birthrates within the EU countries remain low. Several scholars have observed that natural population increases (i.e. births over deaths) have decreased throughout Western Europe at the same time that the economy has expanded, creating a potential labor shortage in the future.(Munz,1996; Thraenhardt,1996). The amount of free-flowing labor across EU countries has been relatively small., perhaps because the "increasing industrialization in the peripheral countries of the EC" has leveled economic development and pay (Werner, 1994, 160). But insofar as the supply of labor from third countries is motivated by relative low level of pay in the origin country, we can expect it to be high in the future. If the demand for imported third country labor returns to pre-1973 levels, and then the pressure for liberalization could build. Under those conditions, the EU's restrictive bias would become more problematic. Nations that wanted to expand their pool of immigrant labor might want to solve the prisoner's dilemmas associated with liberalized immigration, and to initiate more harmonized labor-expanding policies without the threat of a single member veto.

To play this scenario out further, one might predict that liberalizing forces, led by business elites, might venue shop (i.e. take the issue to the forum where they are most likely to succeed) and ultimately take their case to the EU parliament if they felt that support and the prospects of winning there were greater for any reason. The strategic choices of elites might lead eventually to a general acceptance of the principle that immigration was no longer simply a matter of national interest but was properly a European matter. Policy preferences would end up driving institutional change. This policy-driven scenario is similar to the "benefits of decisional reallocation" scenario of Marks, Hooghe, and Blank who suggested that European nations might surrender control in instances when "the political benefits may outweigh the costs of losing political control or there may be intrinsic benefits having to do with shifting responsibility for unpopular decision or insulating decision-making from domestic pressures." (Marks, Hooghe and Blank, 1996: 349).

Clearly, this scenario of shifting power from the nation to the EU leaves out many of the cultural pre-conditions of closer, future political union. I do not mean to imply that the sense of "we-ness" and trust that develops when people grow accustomed to an arrangement is unimportant. But based on the history of institutional change in the US, it is important to acknowledge that it is often driven by frustrated majority policy preferences. However much it upsets classical notions of constitutionalism, all too frequently the real impetus for institutional change is to overcome policy obstacles by shifting to a more favorable policy venue. A corollary observation is when the shift is in a majoritarian direction, it is often hard to reverse unless there is a massive breakdown in the unity of the larger electorate. When that happens, a minority seeks to change the rules so that it can prevail-hence separation or at least decentralization.

It is hard to say whether the conditions leading to greater union on immigration and citizenship matters will arise in the near future-I suspect not. But there is a second, and even more important, consideration -namely, that democratic deficits may not be deficiencies in the sense that deficits may actually be valuable in some policy areas if the goal is fairness and efficiency as previously defined. The assertion that the unmediated majority should prevail on issues assumes: 1. That the majority will not trample on the basic rights of minorities; and 2., That the majority will choose policies in its best interests. Few, if any, would dispute that these conditions are sometimes violated. Many of the institutions that are perceived to be democratically deficient can also be perceived as checks against the breakdown of these two majoritarian assumptions: The US courts cannot protect minority rights by themselves and administrative officials often make technical choices that are beyond the expertise of the average voter.

Since the situation in the EU is so far from a tyranny of the EU electoral majority (or even the European parliament), this might seem like a purely American problem. That is a fair objection, but my point is only that the concepts that political scientists bring to EU discussions need to build a balanced perspective of democratic deficits or they might put in force democratic expectations that are ultimately destructive. In this sense, the US is an object lesson.

The history of political reform in the US in a nutshell is the story of a populist drift--the gradual but progressive shedding of most forms of indirect and territorial representation in favor of more direct and egalitarian representation. This has put the courts in a more visible role of checking majoritarian excesses. With respect to representation, there is some degree of substitution going on in the US with the courts substituting for the territorial and indirect protections of the original 18th century structures of representation. Perhaps, if the EU follows a similar course, a strengthened EU court system will be called upon to do the same.

To flesh this argument out some, the history of political reform in the US has been towards more openness (i.e. extending the franchise, opening up the workings of government to more external scrutiny) and more unmediated majority control. Hence, political parties have been weakened (i.e. the direct primary system, the end of patronage, etc.), territorial bases of representation have been abandoned to the extent the constitution allows (e.g. the "one person one vote" doctrine), direct democracy has been introduced in 23 states, and indirect representation has ended except in specialized local boards (e.g. the direct election of US Senators). The two major 20th century reform movements in the US have been the Progressive and the Populist. Progressives (both in the pure historical sense of the Progressive movement and the more generic sense of those who follow in that tradition) seek to check the mediation of the popular will by political parties and business interests by supplementing the representative system with direct democracy, campaign finance and lobbying reforms, and political party regulations. Populists increasingly reject any form of mediation, including by representatives themselves, and seek to give more power to the people directly (hence, term limits, the popular initiative, etc). While these two movements diverge on pome important points (see Cain and Miller, 1999; Persily, 1998), the joint product of their influence and efforts has been the removal of Madisonian filters and the handing over of more power to the electorate directly.

This has placed the courts in the role of checking decisions made by electorates that might have been headed off in the past by party leaders or more insulated representatives. This is particularly important in immigrant issues for several reasons. First, nativist fears rise and fall with economic cycles, and decisions taken in panic at the trough of a cycle might not be in the best economic interests of a country (i.e. the efficiency condition). Secondly, because immigration issues often conflate with racial discrimination issues, majority reactions to temporary immigration crises might lead to policies that violate basic rights (i.e. the fairness condition).

Given that the European courts are not yet able to play this role in the EU, the persistence of territorially mediated representation and the democratic deficits caused by unelected administrators may a good thing for the time being. The delegation to nations may produce an overall restrictiveness bias, but nations separately have the freedom to open their borders if the need arises. And secret negotiations by unelected administrative officials might serve to defuse potentially volatile issues to some degree. I would not defend these principles against their logical extremes, but it is important to remember that a democratic deficit may not always be a democratic deficiency, particularly in matters involving rights. The ultimate goal of political union should build in "slack" (i.e. some redundancy of representation) and allow for useful and moderate levels of democratic deficits. This will not guarantee good solutions to vexing problems like immigration, but might facilitate them to some degree.

Archdeacon, Thomas J. "Reflections on Immigration in Light of US Immigration History," International Migration Review, vol 26, no 2, 525

Brochmann, Grete, European Integration and Immigration From Third Countries, Oslo: Scandinavian University Press, 1996

Carchedi, Bruno and Guglielmo Cardechi, "Contradictions of European Integration," Capital and Class #67, Spring 1999, 119

Colino, Stacey, "The Fallout from Proposition 187," Human Rights, vol 22 (Winter), 1995, 16-18

Convey, Andrew and Morek Kupiszewski, "Keeping up with Schengen: Migration and Policy in European Union," International Migration Review, vol 29 (Winter, 1995), 939

Cornelius, Wayne A., Phillip L. Martin and James F. Hollifield, ed. Controlling Immigration: A Global Perspective (Stanford University Press, 1994)

Fassman, Heinz and Rainer Munz, "Patterns and Trends of International Immigration in the Western Europe," Population and Development Review, vol 18, no 3 (Sept. 1992), 457

Geddes, Andrew, "Immigrant and Ethnic Minorities and the EU Democratic Deficit," Journal of Common Market Studies, vol 33, no2, June 1995, 197

Gerber, Elizabeth, "Legislative Responsiveness to the Threat of Popular Initiatives," American Journal of Political Science, vol. 40, 1996, 99.

Ireland, Patrick, "Facing the True (Fortress Europe): Immigrant and Politics in the EC," Journal of Common Market Studies, vol 29, Sept.1991

Johnson, Kevin, "An Essay on Immigration Politics, Popular Democracy and California's Proposition 187," Washington Law Review, vol ?, 1995, 629

Koslowski, Rex, "Intra-EU Migration Citizenship and Political Union," Journal of Common Market Studies vol 32, no 3, Sept. 1994, 369

Kostakopoulou, Theodora, "Why a Community of Europeans Could be a Community of Exclusion: A Reply to Howe," Journal of Common Market Studies, vol 35, no2, June 1997, 302

Lahav, Gallya, "Ideological and Party Constraints on Immigration Attitude on Europe," Journal of Common Market Studies, vol 35, no 3, Sept. 1997, 377

MacDonald, Karin and Bruce E. Cain, "Nativism, Partisanship and Immigration: An Analysis of Prop 187," in Michael B. Preston, Bruce E. Cain, and Sandra Bass, Racial and Ethnic Politics in California , (Berkeley: IGS Press, 1998)

Marks, Gary, Liesbet Hooghe and Kermit Blank, "European Integration from the 1980s: State-Centric V. Multilevel Governance," Journal of Common Market Studies, Vol. 34, no3, Sept. 1996, 341

Martin, Phillip, "Good Intentions Gone Awry: IRCA and the US Agriculture," in The Annals, ed. Mark Miller, vol 534, July 1994, 44

McCubbins, Matthew D., "The Legislative Design of Regulatory Structure," American Journal of Political Science, vol. 29, 1985, 721

Messina, Anthony M. "Immigration as a Political Dilemma in Britain. Implication for Western EuropePolicy Studies Journalvol 23, no 4 (Winter, 1995), 686 "Political Impediments to the Resumption of Labour to Western Europe," Western European Politics, vol. 13, January 1990

Miles, Robert and Diana Kay, "The Politics of Immigration to Britain: East-West Migration in the Twentieth Century," Western European Politics, vol 17, no 2 (April, 1994), 17

Money, Jeannette, Fences and Neighbors: The Political Geography of Immigration Control, Ithaca: Cornell University Press 1994

Munz, R, "A Continent of Migration: European Mass Migration in the 20th Century," New Community 22, 1996, 201

Musalo, Karen, "Fortress American -- The United States' Response to Refugees Poper delivered at Conference on Cosmopolitanism and Nationalism," Stanford University, 1999

Papademetriou, Demetrios G Coming Together or Pulling Apart? The European Union's Struggle With Immigration and Asylum Washington: Carnegie Endowment for International Peace, 1996

Papademetriou, Demetrios G. and Kimberly A. Hamilton, Converging Paths to Restriction: French, Italian and British Responses to Immigration, Washington: Carnegie Endowment for International Pace, 1996

Pettigrew, Thomas F. "Reactions Toward the New Minorities of Western Europe," American Review of Sociology, v 24, no 1, 1998, 77

Philip, Alan Butt, " European Union Immigration Policy: Phantom, Fantasy, or Fact?" European Politics, vol 17, no 2 (April 1994), 168

Scharpf, Fritz M., "The Joint-Decision Trap: Lessons From German Federalism and European Integration," Public Administration, vol. 66, Autumn, 1988, 239

Schockman, H. Eric, "California's Ethnic Experiment and the Unsolvable Immigration Issue: Proposition 187 and Beyond," in Michael B. Preston, Bruce E. Cain and Sandra Bass, Racial and Ethnic Politics in California, vol 2, (Berkeley: IGS Press, 1998)

Thraenhardt, D. "European Migration From East to West: Present Patterns and Future Directions," New Community 22, 1996, 227

Ucarer, Emek M. and Donald J. Puchala, ed. Immigrations Into Western Societies: Problems and Policies, London: Pinter, 1997

Weintraub, Sidney, "North American Free Trade and the European Situation Compared," International Migration Review, vol 26, no 2, 506