Dear Workshop participants:
This paper, which I co-authored with Dwayne Woods of Purdue University, grew out of our attempt to understand the effects that travel liberalization will have on Cuba’s political future. The Cuban government last year announced a set of changes to its travel regulations that are the most sweeping in over five decades. As you will see, there is a connection in this story to political protest through the Exit, Voice, and Loyalty paradigm of Albert O. Hirschman. As always, I (and also my co-author) thank you for the opportunity to present at the Workshop and for any feedback you provide.
NO WAY OUT: TRAVEL RESTRICTIONS AND AUTHORITARIAN STABILITY
Why do autocrats differ in their propensity to allow their citizens to travel abroad?This article posits that there is an inverse relationship between the freedom of foreign movement and authoritarian stability. Authoritarian leaders recognize this and hence try to control foreign travel by their citizens. The result isthat autocracies that impose restrictions on travel last significantly longer than those that do not, for two reasons. First, travel restrictions decrease the costs of repression for regimes. Secondly, countries that allow international travel integrate themselves significantly more in the global society than regimes that restrict it.There is an authoritarian dilemma inherent in this choice then, since the very human capital that governments desire to develop economically is based on openness to and integration with the outside world.
Keywords: migration, democracy, globalization, travel restrictions.
Word count: 8,510
The relationship between international migration and domestic politics has drawn increased interest in comparative politics in recent years.[1] Despite this extensive treatment, scholars have not explicitly addressed why autocrats differ in their propensity to allow their citizens to travel and move to other countries, and what effects these restrictions have on the stability of their governments.[2] This article posits that there is an inverse relationship between the freedom of foreign movement and authoritarian stability. Authoritarian leaders recognize this and hence try to control the foreign movement of their populations, fearing that exit will result in increased internationalization of domestic dissent and less domestic control.
There is, moreover, an authoritarian dilemma inherent in this choice. The very human capital that governments desire to develop economically is based on openness to and integration with the outside world. The deeper an authoritarian regime pursues global integration through travel, monetary transfers, and communication flows, however, the more difficult and costly it is for that regime to impose restrictions on travel.[3]
Our findings reveal that regimes that impose restrictions are significantly more durable than those that do not, for two related reasons. First, travel restrictionsdecrease the costs of repression for regimes because the freedom of movement, like other human rights, is a public good that allows citizens to coordinate their information, resources, and activities (Bueno de Mesquita and Smith 2009, 2010). In addition, countries where foreign travel is permitted integrate themselves significantly more in the global society than regimes that have restrictive travel policies. This suggests the social linkages and networks that develop between émigrés and foreigners on the one hand and local citizens on the other significantly increase the democratic prospects of less democratic nations.
The findings in this study highlight then the need to move beyond simplifiedaccounts ofthe relationship between migration and democratic political reforms (Hirschman 1970, 1993). In these accounts, the two ways to express dissatisfaction with an authoritarian government – foreign travel and anti-government protest – are mutually at odds, that is, the use of one impedes the other. Hirschman rightly anticipated that in the presence of politically motivated emigration, it would be difficult for anti-government dissent to develop. But he did not envision how social globalization, motivated to some extent by emigration,could increase citizen political involvement at a later stage.[4] As countries become more socially integrated with the outside, particularly through ties that bind émigrés with their compatriots back home, citizens are able to pressure their governments more effectively and elites are in a better position to identify with and benefit from democratic reforms.
The following two sections point to gaps in our understanding of the relationship between emigration and democratic reforms and articulate an alternative approach that factors globalization into the behavior of both masses and elites in authoritarian regimes. The third section details the empirical approach we have developed for testing these relationships. A fourth section presents the results of our analysis. We conclude by drawing implications and policy lessons from our findings.
Exit and Voice: An Inverse Relationship?
In an award-winning article on the collapse of the German Democratic Republic, Albert O. Hirschman (1993) tried to understand why the GDR had been a stable authoritarian regime for so long but had collapsed so suddenly.[5] In the original formulation of his exit, voice, and loyalty approach (hereafter EVL), Hirschman (1970) implied that the blocking or limiting of exit should have a seesaw effect on voice. More specifically, if citizens cannot express their grievances with the regime in power by leaving, they may be more likely to speak up, making protest the dominant means of redressing grievances in the political arena. In what amounted to a modification of his earlier EVL framework, Hirschman claimed that the sudden collapse of the GDR could be explained by a change in the relationship between emigration and protest: where a regime is widely unpopular and authorities attempt to block exit, exit may actually go along with belief in the efficacy of voice.[6]
Two decades have passed since the publication of Hirschman’s account, and his interpretation of East Germany’s collapse has been challenged, among others for underestimating the role mass migration can play in failing to produce collective voice(Pfaff 2006; Pfaff and Kim 2003). But while these works have continued to refine the EVL framework initially introduced by Hirschman more than four decades ago, an increasingly accumulating literature has called into the question the utility of the model for understanding broader trends in the relationship between globalization, in particular the movement of people across national borders, and domestic democratic reforms (e.g., Hoffmann 2008; Koslowski 2005).[7]
Hirschman intended his approach to be useful in explaining the likelihood of democratic political reforms in authoritarian regimes. He did so by trying to understand the relationship between emigration and protest. While the EVL framework provides some good intuitive insights, however, it raises a number of important questions that it cannot satisfactorily answer. This failure stems from its inability to fully take conditioning factors into account and to travel to cases outside those for which it was originally conceived, namely, state-socialist regimesthat simultaneously restrict the exercise of exit and voice.
Consequently, we set out not to test a more refined theory of exit, voice and loyalty dynamics. Instead, we use the original model to ask three relatively open-ended questions. First, is the exit-voice dynamic conditioned by globalization? In other words, is the calculus on travel shaped by the degree to which a regime is integrated into the global society? A second and related question is whether regimes that restrict travel more also integrate themselves less with the outside world. Finally, is it possible to identify a causal relationship between travel restrictions and regime stability? That is, do travel restrictions, directly or indirectly, increase the ability of autocracies to hang on to power?
The extent to which exit results in voice, and the ability of loyalists nationals to rely on voice to demand political reforms, varies greatly. North Korea, for example, has one of the most restrictive exit policies in the world, going as far as executing anyone caught trying to flee the country. The regime rightly views exit as a signal to others of the existence of domestic dissent. In Cuba, on the other hand, exit has been used as a valve to diffuse domestic grievances by ridding the country of those who might be more propense to engage in protest (Pedraza 2007).
In China, exit is now to some extent a function of global economic integration. The communist party permits its citizens to travel abroad for study, business, and tourism, but carefully controls travel visas for dissident intellectuals, academics, and members of ethnic and religious minorities.[8] Moreover, China attempts through various channels to control the way that Chinese citizens experience travel abroad. China’s Approved Destination Status (ADS) policy, for example, allows Chinese citizens to travel in organized group tours to countries the government has approved (Arita at al.2011). The objective of the policy, adopted in 1995, is to accommodate the growing demand by Chinese to visit other countries and at the same time allow the government to have some control over which countries they visit and the duration and nature of those visits. The Chinese case suggests that the context in which autocracies respond to the threat of voice has changed in last four decades: a significant number of governments have become more democratic (Doorenspleet 2005). The demise of superpower rivalry in particular created a world that is lot less hospitable to the policies of regimes that suppress human rights and restrict political competition (Boix 2011; Jansson, Lindenfors, and Sandberg 2013, 2).
Figure 1 presents two plots of travel restrictions from 1981, the first year for which data is available, to the most recent year for which we have data, 2011. The source of this information is the CIRI Human Rights Data Project (Cingranelli and Richards 2008), the only repository of standardized comparative data on the freedom of travel. The plot displays both the distribution of values as well as the average for a particular year, allowing a trend over time to be discerned. The plots thus capture global trends in the regulation of foreign travel, first for all regimes, and then for countries where authoritarian tendencies prevail over democratic ones judging by the polity scores of country-year observations. We have used a polity score of 0 as our threshold dividing autocratic observations from democratic ones.[9]
Figure 1. Travel Restrictions in the Era of Globalization, 1981-2011
Note: A score of 0 indicates that this freedom was severely restricted, a score of 1 indicates the freedom was somewhat restricted, and a score of 2 indicates unrestricted freedom of foreign movement.
Figure 1 very clearly establishes that whereas no global trend exists in travel regulation for the last three decades, a more restrictive trend over time is evident when the comparison is limited to autocracies – despite the reality that the world has become increasingly more democratic since the onset of the Third Wave of democratization in 1974.[10] Similar to the displays in Figure 1, Figure 2 plots correlograms of economic, social and political globalization for all countries over time for which there is data. The information comes from the KOF Globalization project (Dreher 2006; Dreher, Gaston, and Martens 2008), a quite extensive collection of standardized cross-national time-series data on globalization in its various manifestations.
Figure 2. Global Progress in Economic, Social, and Political Globalization, 1981-2010
Note: Economic globalization refers to data on flows of trade, stocks of foreign direct investment, portfolio investment, and income payments to foreign nationals, all as a percentage of GDP. Social globalization refers to data on telephone, letters, and information flows into and out of countries, flows of “goods, services, income, or financial items without a quid pro quo”, international tourist arrivals and departures, and the foreign or foreign-born population in a country. Political globalization refers to embassies in a country, membership in international organization, participation in U.N. Security Council Missions, and international treaties ratified.
As Figure 2 indicates, the world is becoming more integrated whether the measure of globalization used is economic, social or political.
Figure 2 raises the possibility that autocratic elites, while attracted to the opportunities presented by globalization, also see increasing contact with and dependence on the outside as a threat to their power and legitimacy. If this is so, a secular trend towards more interdependence may be specific to democracies, not the world as a whole. Figure 3 reveals however that autocracies are also partaking in the challenges and opportunities presented by globalization.
Figure 3. Progress in Economic, Social, and Political Globalization for Autocracies, 1981-2010.
If autocracies tend to restrict travel compared to democracies (Breunig, Cao, and Luedtke 2012), they seem to be doing so because, not despite globalization. The question then is whether these restrictions are having their intended effect on the stability of authoritarian regimes. A positive correlation exists between restrictions on foreign travel and authoritarian stability: autocracies that impose restrictions on outbound travel last significantly longer than those that do not. This relationship is present whether authoritarian stability is understood in terms of how many days a regime is in existence (r=0.19; p<0.000)[11] or in relation to movement (or lack thereof) along a continuous one-dimensional indicator of the extent of ‘democrativeness’ of a political system (r=-0.14; p<0.000).[12] The following section explains the reason why this relationship is present.
The Conceptual Model
Hirschman claimed that in the presence of politically motivated emigration, it would be difficult for anti-government dissent to develop. But he did not envision how social globalization, motivated to some extent by exit,could increase voice at a later stage. His conceptualization of exit as an individual act and voice as a collective one involving significant risks, moreover, can severely circumscribe the domain of cases to which the terms may apply (e. g., communist one-party autocracies). We argue instead that democratization and social globalization require us to relax the scope conditions of these two terms and define exit and voice as not simply politically motivated exile and costly anti-government opposition, respectively. Instead, exit can take the form of temporary exile and even economically motivated emigration since the latter, by responding to the government’s inability to meet the economic needs of its population, is an expression of disapproval. Voice, moreover, can take less demanding forms such as voting for the opposition party in party-dominated autocracies (Pfutze 2012).
Travel restrictions are likely to affect the type and extent of integration an autocracy pursues with the outside world.Contrary to the EVL model, however, we are not interested in the relationship between exit and voice per se, but in how the two produce (or fail to lead to) democratic reforms. We also subsume the category of loyalty into the realm of voice (as hinted by Hirschman in his distinction between insurgent voice and reformist voice) and ask instead how the regulation of exit affects the probability that a country will engage in reforms that increase its level of democracy. In so doing, we adapt the classical Dahlsian (1971) argument about democratization as a function of the costs of tolerating or suppressing a political opposition to the relationship between travel restrictions and authoritarian stability.
In our view, restricting travel limits the capacity of citizens not only to emigrate, but also to return, and this explains why travel restrictions lower repression costs for governments: by preventing citizens from expressing their dissent in the first place, or severing the link between dissenters (those who have already left) and their societies, governments that severely restrict travel make it difficult for citizens to organize and mobilize around oppositional grievances. The variable we employ to measure this right, the CIRI Human Rights Data Project’s “freedom of foreign movement and travel” indicator, embodies the notion that the freedom to leave one’s country of origin and to return to that country whenever one pleases are fundamental human rights enshrined in the United Nations’ Universal Declaration of these rights. As a result, the Project codes this right as severely restricted when countriesrestrict all or nearly all the foreign travel of its citizens (as in most communist countries), and when they“do not respect the right of citizen refugees outside of the country’s international borders to return to their homes.” (Cingranelli and Richards 2008, 47).[13]
A country may be more socially globalized because of its own policies, those of its neighbors, or both. For our purposes, what matters is the extent to which social integration helps countries promote democratic reforms, controlling for other factors that may affect this integration such as travel policies and the level of development.[14]
We believe that democracies pursue policies that are more amenable to global integration along societal dimensions, but we are not claiming we can isolate the precise mechanisms in the external environment that influence a country’s political system (Brinks and Coppedge 2006, 467). Indeed, external forces can influence a country’s level of democracy through at last five different channels (Levitsky and Way 2010, 38-39).[15] The indicator of social globalization we use captures some of those channels – such as income or financial transfersa country receives without a quid pro quo, namely remittances and foreign aid. Other variables we employ such as the average level of democracy in a country’s immediate neighborhood help control for other important channels of diffusion.
Diffusion variables in previous studies have not explicitly captured processes that reflect how a nation’s citizens respond to the external environment. Many of the variables used, even those that proxy for spatial diffusion, mostly reflect the effect of external factors on a country’s elite.[16] Our study is the first to include a variable that primarily reflects how the external environment affects citizens in the recipient nation, in addition to more commonly used indicators of political and economic diffusion.