Key themes

Much research examining the implications of regime type is contested. With the exception of the democratic peace theory—which posits that democracies do not go to war with other democracies (and that we discuss at greater length)—research on the consequences of regime type has yielded conflicting results. Some of the inconsistency in the empirical analysis stems from differences in research design and the measurement of key variables across studies. This variation makes it hard to evaluate competing claims and has contributed to a good deal of confusion with respect to how regime type affects a host of outcomes of interest. In the last ten years, however, research on the consequences of regime type has evolved and there are two key takeaways we wish to highlight.

First, studies that use a continuous approach to disaggregating political systems show that hybrid regimes, or those countries that sit in the middle of the autocracy–democracy spectrum, perform less well than either their fully democratic or fully authoritarian counterparts in a number of areas. The countries in the ‘messy middle’ are particularly vulnerable to insecurity, including from repression, terrorism, and prospects for internal conflict (Boswell and Dixon, 1990; Muller and Weede, 1990; Hegre, 2001; Fearon and Laitin, 2003). They are also the least durable (Gurr, 1974; Gates et al., 2006; Knutsen and Nygard, 2015) and accordingly most volatile regime type (Epstein et al., 2006). Full autocracies and consolidated democracies have internally consistent institutional arrangements, which create a self-enforcing equilibrium. The inconsistency and incoherence created by hybridism, in contrast, creates challenges in terms of security. Hybrid systems lack the benefits of democracy that encourage citizen satisfaction and therefore stability, such as avenues for citizens to participate and articulate their preferences. At the same time, they do not have the concentrated power and repressive capacity that facilitates control and stability in full autocracies.

Some scholars and policymakers have interpreted the ‘messy middle’ argument to mean that political liberalization—or movements from autocracy towards greater democracy—will unleash political violence and deteriorate government performance. p. 73↵Political liberalization, the argument goes, disrupts the institutional procedures of a transitioning state. Samuel Huntington (2006), for example, argued that the process of democratization is associated with mass mobilization, which can trigger violence if the political institutions are not prepared to accommodate this level of participation. The instability surrounding the Arab Spring has given this view greater credence in the minds of many. Protesters’’ efforts to topple long-standing dictators in Egypt, Libya, Yemen, and Syria unleashed instability and violence that continue to plague the Middle East and North Africa. Indeed, full autocracies rarely become mature consolidated democracies overnight. Transitioning countries typically pass through the messy middle where mass politics mixes with authoritarian elite politics—generating a view that democratization produces instability.

We offer a note of caution when drawing such sweeping conclusions. Making the argument that political liberalization triggers instability requires disentangling it from hybridism (Hegre, 2001). Our discussion in Chapter 4 underscored that not all countries that are categorized in the middle range of democracy indicators should be considered ‘in transition’, as many become entrenched in a more permanent state of hybridism. Likewise, some countries in the messy middle may be backsliding from democracy to autocracy. In other words, countries that lie in the mid-range of the democracy–autocracy spectrum include hybrid regimes, liberalizing states, and backsliding states. We can reasonably expect the behaviour of these types of states to differ. Recent studies seeking to disentangle these dynamics find little empirical support for the notion that democratization raises a country’s risk of civil war. Empirical studies show that democratizing countries have only a marginally higher share of conflict onset during the transition compared to years without transitions (Gleditsch and Ruggeri, 2010). Before making broad conclusions about the risks associated with political liberalization, more research is needed. For example, are some liberalization processes—perhaps those that occur ‘organically’ and without international intervention or that occur gradually—less likely to lead to instability?

Second, studies that use a categorical approach to disaggregate political systems (such that democracies lie in one category and dictatorships in the other) find that democracies outperform dictatorship on almost every indicator examined. Democracies are less likely to fight inter-state wars against other democracies (Levy, 1988). They are less likely to employ repression against their citizens than autocracies, and civil wars in democracies are comparatively less lethal (Davenport, 2007; Gleditsch, Hegre, and Strand, 2009). Most recent research indicates that democracies grow their economies at a rate that is at least on par with dictatorships and that the growth they produce is of higher quality—both less volatile and more likely to benefit the people they govern. This suggests that societies that feature free and fair elections are safer and more prosperous, and their citizens enjoy a higher quality of life.

The superior performance of democracy relative to dictatorship raises a number of critical questions and implications. How should policymakers approach full dictatorships, such as those in the Middle East or Central Asia? If, as some research suggests, movement towards hybridism ups the chance of internal conflict and insecurity, should political liberalization be avoided? Should we reinforce the dictators in office to lessen p. 74↵the chance of turmoil? Or is this view short-sighted? If democracy delivers better outcomes for the people living under these regimes and enhances a country’s prospects for prosperity over the long-term, should we work to encourage democratic development to avoid repeated cycles of instability?

These issues are neither easy to address nor straightforward in their response. A host of additional factors, including national security interests, culture, and past (in)experience with democracy often further complicate the picture. However, here we present the cutting-edge research that should shape our thinking on these important questions. We provide a more in-depth discussion of the research on the effects of regime type on a state’s conflict propensity, levels of terrorism, economic and human development, corruption, and human rights.

Regime type and conflict

Regime type and terrorism

Since 9/11 one of the primary arguments of Western democracy advocates has been that democratization will stem a country’s production of terrorists and terrorist groups. As democracy spreads, the thinking goes, discontented citizens will have more constructive outlets for political grievances and opportunities to express their views through the p. 79↵political process, alleviating their need to use violence to achieve their goals (Crenshaw, 1981; Eubank and Weinberg, 1994; 1998; 2001; Eyerman, 1998; Li, 2005). U.S. President George W. Bush made this logic explicit in a speech in 2005. ‘When a dictatorship controls the political life of a country, responsible opposition cannot develop, and dissent is driven underground and toward the extreme. And to draw attention away from their social and economic failures, dictators place blame on other countries and other races, and stir the hatred that leads to violence.’ U.S. President Barak Obama shared these views. His administration’s National Strategy for Counterterrorism stated in 2011 that ‘promoting representative and accountable governance is a core tenet of U.S. foreign policy and directly contributes to our counterterrorism goals’ (White House 2011, pp. 4–5).

Does greater democraticness make a country less likely to produce or experience terrorism? The prevailing wisdom from academic literature indicates no. In fact, a majority of empirical studies published in the last fifteen years generally shows that democracies experience more terrorism than autocracies (Eubank and Weinberg, 1994, 2001; Weinberg and Eubank, 1998; Pape, 2003; Chenoweth, 2010). However, a number of recent studies suggest that the relationship between regime type and terrorism is somewhat more nuanced. In keeping with one of the key themes in this chapter, there is some evidence that the relationship between regime type and terrorism is non-linear, with hybrid regimes and new democracies experiencing more terrorism than full autocracies or full democracies. We discuss these findings in greater detail.

According to most scholars, terrorism is ‘the threatened or actual use of illegal force and violence by a non-state actor to attain a political, economic, religious, or social goal through fear, coercion, or intimidation’ (Global Terrorism Database).⁴ Potential terrorists hold policy preferences that differ from the state and are commonly more extreme than those of the general population (Lake, 2002). Because they cannot mobilize broad support for their positions and are weak relative to states they oppose, terrorists seek political change by targeting the civilian population with violence. In many ways, terrorism may be viewed as one part of a larger repertoire of political contention (Chenoweth, 2013; Ash, 2016).

There are a number of explanations for why democracies are more prone to terrorist activity than autocracies. The first argument suggests that democratic norms and institutions create a more permissive environment for terrorism than closed authoritarian systems. Democracies are likely to experience more frequent terrorist attacks because greater executive constraints and preservation of individual rights provide a more hospitable environment for terrorism than is found in more repressive environments (Li, 2005; Pape, 2003). Democratic freedoms, such as freedoms of movement, association, and expression, provide opportunities for groups to form, operate, recruit, and coordinate terrorist activities without fear of intrusion from the governments they oppose. Moreover, expansive and secure civil liberties hinder democratic governments’ ability to suppress discontented groups. The political and civil liberties that democracies extend to their citizens mean that there are greater restrictions on policing, as well as legal frameworks that make it more difficult to convict terrorists, lowering the costs p. 80↵of conducting terrorist activities. Autocrats, in contrast, have nearly unlimited discretion to target and repress potential terrorists.

Likewise, freedom of the press—a core feature of democracies—creates incentives for terrorists to target such states. This is because market-driven media companies are the most enthusiastic in reporting about violent events, providing free publicity to terrorist groups and exacerbating the fear these groups intend to create (Crenshaw, 1981; Hoffman, 2006; Gadarian, 2010). Democracies place fewer restrictions on the media and often produce more sensationalist news coverage, which might provide more extensive attention to terrorist events than would occur in closed authoritarian settings.

Not only do the institutional features of democracies make them more vulnerable to terrorist activity, but terrorists may also be more likely to target them because democracies are more likely to make concessions. Democracies are more accountable and responsive to domestic pressures for policy change, making them more likely to modify policy as a result of violence (Ezrow, 2017). Moreover, democratic publics are unlikely to tolerate the costs imposed by suicide terrorist attacks and more likely to push for changes in state policy that will lessen the likelihood of future attacks (Pape, 2003). Although voters may not share the policy preferences of terrorist groups, terrorist attacks can lead to changes in government policy consistent with terrorist goals. For example, one of the stated intentions of Al Qaeda prior to the 2001 attacks was removal of U.S. troops from Saudi Arabia. Nearly all U.S. troops were removed by 2003.

Research on regime type and terrorism has recently moved beyond the dichotomous classification of regime type. A number of studies now suggest that the relationship between democracy and terrorism is nonlinear (Kurrild-Klitgaard et al., 2006; Abadie, 2006; Chenoweth, 2010). Consistent with the ‘messy middle’ argument we discussed, these studies show that it is neither the freest nor the most repressive states that experience the most terrorism, but rather those in the middle of this spectrum. This implies that countries experiencing political transitions from autocracy to democracy are particularly prone to terrorism (Chenoweth, 2010). Political transitions may heighten opportunity, unleash long-standing grievances, and trigger competing terrorist organizations to compete with one another for influence in a newly democratic environment (Chenoweth, 2010). Similarly, there is evidence that new democracies are especially vulnerable to terrorism (Eyerman, 1998; Piazza, 2013). Terrorist organizations view new democracies as fragile and less capable of countering their actions (Eyerman, 1998). Also, in the earliest days of democracy, groups with grievances may maintain their previous approach of using violence to achieve their objectives. It takes time for such groups to recognize the utility of eschewing armed struggle in favour of nonviolent political engagement through democratic institutions. Advanced democracies, in contrast, generally do not suffer from high levels of terrorism unless they interfered in other countries’ affairs through military intervention or occupations, in which case such countries often become targets of transnational terrorism (Chenoweth, 2013; Pape, 2003; Savun and Phillips, 2009).

It is important to note in closing this section that the types of countries most likely to experience terrorism may be changing (Chenoweth, 2013). Although consolidated democracies still experience several hundred attacks per year, this figure is lower than p. 81↵in past decades, indicating that robust democracies are indeed becoming more immune to terrorism over time. Only a very few advanced democracies (such as Israel) continue to experience high numbers of terrorist acts. Instead, factionalized democracies (such as India), partial autocracies (such as Algeria and Pakistan), and occupied countries (such as Iraq and Afghanistan) have become the most frequent targets. This change is an interesting and important puzzle for scholars to explore. Future research will need to examine what explains the shift from terrorism in democracies to terrorism in semi-democratic and non-democratic contexts, especially after 9/11.

Regime type and economic performance

Regime type and quality of life

Economic growth is not the only way to evaluate development and is in many ways an inadequate indicator of socio-economic progress. How does regime type affect a government’s capacity to satisfy people’s basic needs, such as access to food and clean water, the way economic benefits are distributed throughout society, and the quality of government services like health care and education? While conclusions about regime type’s effect on economic growth are contentious, the research on democracy and development is more conclusive: most studies find that democracies provide a better quality of life for their citizens than autocracies.

First, democracies spend more on public goods such as education, health care, and social security than non-democracies. One study, for example, finds that democratically-elected politicians in Africa are more responsive than their authoritarian counterparts to the demands of rural groups that form the majority of citizens in most African countries (Stasavage, 2005). More specifically, contested elections led leaders to spend more on education and prioritize primary schools over universities. The focus on primary relative to university spending benefits the rural majority as this tends to be the only formal schooling they receive. Additional research has found similar patterns in Latin America where democracies have also spent more on education and health than autocracies (Segura-Ubiergo, 2007). Likewise using an innovative data set of night lights visible from satellites, scholars have even found that democratization is associated with a substantial increase in electrification (Min, 2008).

Democracies’ focus on the needs of the majority translates into substantive differences in the welfare of the people living under them. Citizens in democracies live longer, healthier, and more productive lives, on average, than those in autocracies (Przeworski et al., 2000; Halperin et al., 2009). Consider, for example, differences in infant mortality rates across regime type. Social scientists consider the infant mortality rate to be a useful indicator of societal welfare because it serves as a window into how governments provide for the economic and social well-being of their citizens. Democracies significantly outperform autocracies on this metric at every level of income (Navia and Zweifel, 2003). One innovative study, for example, examined the survival probabilities of infants born to the same mothers before and after their country underwent a democratic transition (Kudamatsu, 2012). It found that infant mortality fell by 1.2 percentage points, or 12 per cent of the sample mean, after democratization. The study concluded that infant mortality rates declined after democratization because of improvements in public health service delivery, not an increase in affluence.

It is important to note that some scholars question democracy’s positive effect on infant mortality rates. Ross (2006), for example, contends that previous studies produced overly optimistic conclusions about the effects of democracy on infant mortality p. 87↵rates as a result of selection bias arising due to missing data from high-performing autocracies. Once missing data problems are accounted for, he argues, there is no significant relationship between democracy and infant mortality rates. Democracies do maintain the edge over autocracies in education and health spending, but some studies suggest that the benefits of these policies do not seem to translate into lower infant mortality rates.

In a variety of ways, democracy appears particularly beneficial for the poor and women. Democracies pay higher manufacturing wages than autocracies; growth in democracies, therefore, allows workers—who tend to be poor—to capture a larger percentage of economic expansion (Rodrik, 1999). This is especially advantageous for women, who are disproportionately employed in manufacturing sectors. According to Przeworski et al. (2000), women living under autocracy, ‘participate in gainful activities at the same rate as they do in democracies, and as workers, get lower wages … they also have more children, see more of them die, and are themselves more likely to die in childbirth’ (p. 271). The higher wages and services that democracies deliver even translate into greater calorie consumption among the poorest segments of society (Blaydes and Kayser, 2011).

Why might democracies perform better on these dimensions? First, in democracies voters can hold their governments accountable for their choices and actions. In his influential work on the relationship between governance and famine, Amartya Sen (1999) finds that famine has never occurred in any independent and democratic country with a relatively free press. This is, in part, because the democratic electoral process allows citizens to penalize governments that allow famines to occur; political leaders, in turn, are strategic and seek to avert famines. According to Sen:

Famines kill millions of people in different countries in the world, but they don’t kill the rulers … if there are no elections, no opposition parties, no scope for uncensored public criticism, then those in authority don’t have to suffer the political consequences of their failure to prevent famines. Democracy, on the other hand, spreads the penalty of famines to the ruling groups and political leaders. (p. 180)

The political and civil liberties that democracy affords largely determine the extent of accountability in a political system and greater respect for these principles benefits development (Sen, 1999). Freedom of the press, for example, enables the media to report on policy disasters, like famines, drawing these issues into the public discourse. Where citizens lack information about government performance, in contrast, they are unable to hold politicians to account (Keefer and Khemani, 2004).

Overall, democracy entails greater levels of accountability than dictatorship, leading them to produce superior outcomes. Even in authoritarian settings, however, more accountability can enhance policy outcomes. In China, for example, informal accountability structures increased public goods provision (and thus human development), even in the absence of democracy (Tsai, 2007). In those villages where solidarity groups existed—groups based on shared moral obligations or shared interest—the provision of public goods improved. These groups served to enhance public goods provision by spreading information about the performance of local officials. Because local officials p. 88↵are sensitive to this information given that it affects their moral standing in the community, this form of accountability creates incentives similar to elections. Likewise, there is also evidence that local elections in China benefit local citizens. Local elections may shift the tax burden from households to enterprises (Zhang et al., 2004), and direct elections of a village’s leader may increase public goods investment in the village (Luo et al., 2007).

The last explanation for the observed variation in developmental outcomes across democracies and dictatorships focuses on the incentive structures created by each type of system (Bueno de Mesquita et al., 2005). According to this line of thinking, all leaders seek to maintain power and will pursue a set of policies designed to sustain support among the group of backers whose support they need to retain office (their ‘winning coalition’). Leaders can use private goods, such as cash payoffs, or public goods, such as access to health and education, to satisfy their key constituents. Those who depend on a relatively narrow set of backers to keep them in office, as in autocracies, engender loyalty through patronage and other private payoffs. Democratic leaders, in contrast, (in theory) require the support of 51 per cent of the public, making private payoffs a prohibitively expensive strategy to pursue. Democratic leaders, therefore, shift their strategy to the provision of public goods, which benefit everyone in society. In other words, as the number of supporters a leader must satisfy to remain in power grows, the more likely it is that he or she will use public versus private payoffs.

Despite the large number of studies suggesting that democracies enhance education, public goods, health care, and other development outcomes of interest, research is ambiguous about democracy’s effect on inequality. Although the theoretical motivations we outlined in this section suggest that democracy should improve inequality, the empirical evidence is mixed. While there are a number of studies that show that democracy reduces inequality (Acemoglu and Robinson, 2000; Muller, 1988; Moon, 1991; Rodrik, 1998), there are also several compelling studies that find no statistically significant effect (Sirowy and Inkeles, 1990; Bollen and Jackman, 1985; Deininger and Squire, 1998; Gasiorowski, 1997).

Why might democracy fail to mitigate inequality?

Democracy may not enhance the distribution of wealth in those countries if the middle class is able to disproportionately capture the benefits of the democracy relative to the poor. Ross (2006), for example, shows that although democracies spend more money on education and health than non-democracies, these perks seem to accrue to middle- and upper-income groups. Ross argues that in some settings, political competition binds rulers to constituencies that are relatively well-off, subsidizing access to health and education services for those who could otherwise use their own resources. Likewise Nelson (2007) explains, ‘in democracies, pro-poor reallocation of resources across levels of service and among regions may be particularly difficult politically … middle class demand for more and better education and health services is virtually unlimited’ (p. 27). In sum, although democracies re-allocate income and invest in their citizens, the more politically active middle class tend to benefit more than the poor, which can reinforce inequality in these countries.p. 89↵

Regime type and corruption

Corruption is the misuse of office for private gain. The ‘gains’ may accrue to the individual official or to the groups or parties he or she belongs to. Corruption can take the form of bribery, embezzlement, the intentional misappropriation of assets, or nepotism. Corruption can be further disaggregated into grand corruption and petty corruption. Grand corruption refers to major abuses of power that take place at the national level of government, such as an official who rigs the bidding process for state contracts in order to gain kickbacks for personal gain. Petty corruption is the abuse of power at the local level, such as an official requiring a relatively small bribe or favour in return for some good, a basic service that the official should provide for free, or other opportunity.

The scale and cost of corruption is difficult to assess, but the World Economic Forum in 2016 estimated that each year the cost of corruption amounts to roughly 5 per cent of the world’s GDP, including about $1 trillion in bribes. Moreover, empirical studies have consistently demonstrated that the poor are disproportionately burdened by corruption, because they pay the highest percentage of their income in bribes. For example, in Paraguay, the poor pay 12.6 per cent of their income to bribes while high-income households pay 6.4 per cent (The World Bank, 2018). Corruption also discourages these citizens from accessing health services and education if they cannot afford to pay bribes that are requested. And embezzlement and the diversion of public funds for personal gain reduce the government’s resources available for development and poverty reduction spending.

In addition to economic and quality of life considerations, corruption has important security ramifications. Corruption is a tool that countries such as Russia and China use to gain international influence. Russia, for example, leverages corrupt relationships with politicians and prominent economic actors in the Balkans and parts of Eastern Europe to acquire political influence in those countries. These relationships also harm Western businesses, which are unable to compete in some environments because of their inability to pay bribes and engage in corrupt practices. Finally, corruption can facilitate crime and terrorism. Corruption among security forces and border guards can help the movement of drugs and illicit weapons across borders. Likewise, security sector corruption can compromise the capacity of these services to manage threats posed by insurgents and terrorist organizations.

Given the significant costs of corruption, what can we say about its relationship to regime type? Corruption directly benefits those groups or individuals participating in it, including through personal enrichment and the ability to gain political support. These potential benefits are attractive to many politicians and people, regardless of whether they live in a democracy or autocracy (Drury et al., 2006). Is there anything systematic we can say about the prevalence of corruption?

Before delving into this, it is important to briefly discuss how scholars measure corruption. The broad scope of corruption makes it difficult to quantify. Researchers struggle with what to include in such a measure and face challenges gathering objective data. Corruption is carried out, in most cases, away from the public eye and cannot be observed in public records. Two of the most widely used measures of corruption, p. 90↵Transparency International Corruption Perceptions Index and the World Bank’s corruption index, therefore, rely instead on perceptions of corruption. Both of these indices aggregate information from a number of sources that include country risk ratings produced by business consultancies, surveys of international or domestic business people, and polls of country inhabitants. Both groups aim to reduce measurement error by averaging different sources and use similar (and overlapping) sets of inputs. In sum, the data do not measure corruption itself but rather opinions about its prevalence. Because measures of corruption are imperfect and indirect, we must therefore use caution when we draw inferences from most research on the subject (Treisman, 2007).

So how does regime type affect corruption? The current consensus indicates that democracy has a non-linear effect on corruption—democratization may increase corruption in the short run, but corruption declines over time as democracy deepens (Montinola and Jackman, 2002; Treisman, 2000, 2007). This is particularly true for countries with a free press (Adsera et al., 2003; Brunetti and Weder, 2003; Chowdhury, 2004). Press freedom exposes corruption to voters, who, equipped with this information, can punish corrupt politicians in the polls. The consolidation of democratic institutions—including independent judiciaries and social attitudes and norms—enable citizens to expose corruption, make it an issue of political importance, and facilitate recourse ranging from public hearings to voting corrupt officials out of office.

In addition to a country’s time under democracy, levels of economic development also influence democracy’s ability to mitigate corruption. Democracy may reduce corruption but only in economies that have already crossed a GDP per capita level of approximately U.S.$2,000 (Jetter et al., 2015). For poorer nations, democratization may actually increase corruption. Consistent with this, democratic India has done no better than China at checking corruption and may have fared worse given its modest level of economic development (Sun and Johnston, 2009). Underdevelopment, the argument goes, reflects a scarcity of legitimate alternatives for gain, and pervasive vulnerability to exploitation. Just establishing elections and other institutional hardware is not enough to create the accountability required to curb corruption.

A body of case study evidence lends credence to the notion that democracy has a non-linear effect on corruption. In Latin America, for example, the wave of democratization that swept across the continent starting in the late 1970s initially led perceptions of corruption to rise (Weyland, 1998). Democratization dispersed political power and incorporated more institutions into decision-making processes, effectively increasing the number of actors demanding bribes. Under authoritarian rule citizens could get by bribing officials in the federal government; under democracy, however, state and municipal governments and legislators in Congress also demanded their share. A study notes, for example, that in Brazil the ‘commissions’ that public officials extorted rose from about 8 to 12 per cent of a contract’s value under the military regime to between 15 and 20 per cent under democracy (Weyland, 1998). A number of other scholars have noted similar dynamics in countries ranging from Russia (Mohtadi and Roe, 2003), Turkey (Mohtadi and Roe, 2003), Indonesia (McLeod, 2005; Robison and Hadiz, 2004) and Thailand (Ammar, 1997; Case, 2002; Hicken, 2001; Pasuk and Baker, 1998; Rock 2000).

Relative to established democracies, however, authoritarian regimes tend to be more corrupt. Autocrats are far more reliant than their democratically-elected counterparts on the distribution of patronage in order to secure loyalty (Bueno de Mesquita et al., 2005). This is particularly true for personalist dictatorships (Chang and Golden, 2010). In a study of highly personalized dictatorships in Africa, ‘neo-patrimonial regimes’ differ fundamentally from other types of autocracies in their extraordinary reliance on the exchange of material rewards for political support (Bratton and van de Walle, 1997). Likewise, corruption in those autocracies with leaders who are uncertain about their longevity in office is likely to be especially severe. Autocrats with short time horizons will be the most predatory and seek to maximize the personal benefits of holding power p. 91↵while their access to such opportunities last (Wright, 2008). To summarize, while transitions to democracy can lead corruption to proliferate, fully consolidated democracies are less corrupt than autocracies.

Regime type and repression

Repression is defined as a form of socio-political control used by authorities against those within their territorial jurisdiction to deter specific activities and beliefs perceived as threatening to political order (Goldstein, 1978). The particular magnitude or combination of repressive activities that governments use varies across time, space, and context, but all governments use some form of repression (Davenport, 2007). Although all governments use it, a robust body of research shows that repression is far more prevalent—in both severity and incidence—in dictatorships than in democracies (Poe and Tate, 1994; Poe, Tate, and Keith, 1999; Davenport and Armstrong, 2004; Vreeland, 2008). Scholars have referred to the robust body of empirical evidence showing democracy’s pacifying effect as the ‘domestic democratic peace’, mirroring the international relations research on interstate war. In autocracies, repressive leaders cannot be voted out of office, meaning that repressive leaders in these systems often go unpunished. Moreover, because they are not popularly elected and exclude large swaths of the population from decision-making, dictators must substitute force for legitimacy as a means of obtaining public acquiescence to regime decisions.

Repression comes in many forms, each of which serves a distinct purpose for the regime (Frantz and Kendall-Taylor, 2014; Fein, 1995; Davenport, 2007a). Broadly speaking there are two categories of repression: civil liberty or empowerment rights repression (i.e. censorship, restrictions on assembly) and personal integrity rights repression (i.e. torture, disappearances, political imprisonment). Dictators use civil liberties restrictions to make it harder for the political opposition and public to mobilize anti-regime activity. Physical integrity rights violations, in contrast, are more focused and tend to target specific individuals or groups that threaten the regime. Most dictators use a mixture of both. Moreover, in addition to state-sponsored repression, some authoritarian regimes permit (or even direct) non-state actors, including thugs, vigilantes, mercenaries, and secret death squads to perpetrate violence against regime dissidents. These irregular actors enable regimes to further stifle dissent while giving the regime plausible deniability about their involvement (Rudbeck et al., 2016).

In addition to the fact that democracies repress less because repressive leaders can be voted out of office, there are additional reasons that democratic leaders rely less on repression than their authoritarian counterparts. First, democratic institutions—including elections, a free and open media, and checks and balances—raise the cost of using repression to stifle dissent (Davenport, 2007; Linz, 2000). Of all aspects of democracy, scholars have argued that executive constraints have the greatest pacifying influence on democratic leaders (Bueno de Mesquita et al., 2005). Autonomous courts, for example, can constrain state behaviour by resisting the state’s attempts to p. 93↵suppress dissent, including by raising the likelihood of litigation against abusive leaders (Mitchell et al., 2013; Powell and Stanton, 2009). Independent media, for its part, raises the likelihood that human rights violations will be exposed.

Second, democracies are likely to have less need to use repression because they possess alternative pathways of control, such as participation in the political system, that are absent or constrained in autocracies. Finally, some scholars have emphasized how democratic norms of compromise and negotiation reduce repression because they provide a ‘realistic way to accommodate demands with a minimum of conflict’ (Henderson, 1991, pp. 123–4). Democratic processes and decision-making, in other words, provide a non-coercive way to manage conflict before grievances mount and invite repression.

While full democracies clearly repress less than autocracies, some research argues that the relationship between regime type and repression is non-linear (Fein, 1995; Gartner and Regan, 1996). Some scholars have argued that there is ‘more murder in the middle’, positing that hybrid and transitioning systems are more repressive than either full autocracies or consolidated democracies (Fein, 1995; Regan and Henderson, 2002). This argument is based on a widely held view that repression is largely a response to dissent. The more dissent a government faces, including protests, riots, terrorism, and other challenges to state authority, the higher state repression is likely to be (Davenport, 1995; 2000; Gartner and Regan, 1996; Poe and Tate, 1994). Hybrid regimes, the argument goes, face more regime-threatening dissent than either full democracies (which channel dissent in less threatening ways) or full autocracies (which have effectively quashed mobilization). A recent study, however, disputes this argument. This study argues that dissent is endogenous to repression; the state acts to prevent dissent from ever occurring, and groups often self-censor in the very anticipation of repression. Once this strategic interaction is accounted for, the study finds that dissent has no effect on levels of repression (Ritter and Conrad, 2016).

Rather than an inverted U-shape between regime type and repression, a growing body of research suggests instead that there is a threshold effect (Bueno de Mesquita et al., 2005; Davenport and Armstrong, 2004). At low levels of democracy, including in new or weakly institutionalized democracies, democratic institutions are unlikely to have much effect on repression. Above a certain level of democratic development, however, democracy reduces repression. In other words, movement along the autocracy–democracy spectrum does not lower levels of repression until a certain level of democracy has been reached.

Although it is difficult to pinpoint the democratic threshold above which human rights improve, it is clear that full democracies do a better job of protecting the rights of their citizens than autocracies. Moreover, research shows that regime type is among the most important determinants of repression levels (Hill and Jones, 2014; Davenport and Appel, 2014). Among all the variables scholars have examined—including economic development, population size, and international constraints arising from human rights treaties and NGO activity—democracy has one of the substantively largest effects on repression (Hill and Jones, 2014). This relationship is significant because it underscores the importance of democracy for protecting human rights.p. 94↵

Conclusion

In the seventy years since the end of World War II, the United States and the West have fostered a global order dominated by states that are liberal, capitalist, and democratic. They have promoted the spread of democracy to strengthen global norms and rules that constitute the foundation of our current international system. However, despite the steady rise of democracy since the end of the Cold War, over the last decade or so we have seen dramatic reversals in respect for democratic principles across the globe. A 2015 Freedom House report stated that the ‘acceptance of democracy as the world’s dominant form of government—and of an international system built on democratic ideals—is under greater threat than at any point in the last 25 years’.

In this chapter, we examined why democratic decline matters. We examined the effects of democracy on a range of outcomes, including the propensity for countries to fight wars, internal conflict, terrorism, economic growth, quality of life, corruption, and repression. Two key themes were woven throughout this chapter. First, hybrid regimes, or those countries that sit in the middle of the autocracy-democracy spectrum, perform less well than either their fully democratic or fully authoritarian counterparts in a number of areas. The countries in the ‘messy middle’ are particularly vulnerable to insecurity, including from repression, terrorism, and prospects for internal conflict. Second, research suggests that democracies outperform dictatorship on almost every indicator we examined. Democracies are less likely to employ repression against their citizens than autocracies and experience less terrorism. Most recent research indicates that democracies grow their economies at a rate that is at least on par with dictatorships and that the growth they produce is of higher quality—both less volatile and more likely to benefit the people they govern.

In conclusion, we believe that the academic record demonstrates that even after we set democracy’s intrinsic value aside, government is better when it is more democratic. Although democratic decision-making can be slower, this process is more likely to weigh risks, thereby avoiding volatile and ruinous policies. When something is going wrong, democratic leaders hear about it and have incentives to take action. If democracy recedes in the face of growing pressure from authoritarian powers like China and Russia, it would have important effects on the outcomes we discussed in this chapter.