Back on Top

Chavez hails Venezuela poll win

[Granted, my attempts to create some revolutionary blog while in Norway was a flop--I'm a terrible journalist, which is apparently true also of blogging.  Notwithstanding, this is a great place for me to keep notes as I work on my dissertation.  To that end, I'm going to bite the bullet and turn the page...]

This last week I reread two seminal books in the literature on democratization, and another important work which applies some of their insights to explaining authoritarian longevity.  This post serves to summarize and contrast them before adding more material to the story:

Acemoglu and Robinson (2005) present a formal model to explain why some countries are democratic while others are not.  Their model is based on three "building blocks":
1) individual economic incentives determine political attitudes, such that people act strategically to optimize their welfare;
2) different groups have opposing interests in society, which provides a source of conflict; 
3) political institutions solve commitment problems by affecting the future distribution of dejure power.

The authors argue that democracy is more likely to be created when there is sufficient social unrest in a non-democratic regime that cannot be defused by limited concessions.  Societal actors are distinguishable by inequalities in resource and the defacto power by which to change distributions of the former.  When the disadvantaged are provided the temporary opportunity to change their status, the advantaged may attempt one of three strategies to maintain their relative positions: repression, concessions, and democratization.  Repression may work, but it is a costly maneuver, especially in a capital-intensive society.  Elites' use of force can dampen productivity and create losses that are harmful.  Concessions may work, but the commitment problem remains--that is, even if temporary payments induce loyalty, it does not promise to hold indefinitely.  Where citizens use their defacto power to force democratization, however, it enables the creation of more durable, long-term agreements, such as parties and legislature, which ensure that redistribution will continue in the future.  

In considering the authors' explanation, one can think of types of governance as a continuum.  At one end there is rule by one person; at the other, all citizens have a say in government.  To formalize their argument the authors imagine a society composed of elites vs. citizens (two groups).  The bargaining between them occurs over one issue--redistribution.  Elites prefer to not redistribute, and citizens (the poor) prefer to redistribute.  Primarily, the comparative statics center on these assumptions.  They authors do consider how the presence of a middle class or alternative identities affects when and whether elites prefer repression, concession, or democratization.  Primarily, however, the workhouse model is based on the assumptions that there are relatively few, cohesive groups bargaining over one issue, and that each group is a unitary actor with nearly perfect information.

The insights of Acemoglu and Robinson (2005) are many.  It is therefore not surprising that it has been hailed as a major contemporary work on democratization.  The authors provide an explanation for when and why limited concessions may not be enough, and how expectations are moderated by relative inequality.  They also add in the role of social stratification and alternative identities.  Generally, the model demonstrates that given complete information, societal actors can act in their best interest and generate a number of different forms of government, the explanation for which is rooted in inequality, defacto power, and expectations about the future.

Bueno de Mesquita et al. (2004) ask why autocrats last longer in office.  They argue that survival is based on how leaders manage their supporters (called the winning coalition, W) relative to the set of people with the qualities required to choose/benefit from leadership (the selectorate, S).  Management of supporters and the selectorate is based on redistribution of goods, which is the result of taxation and spending.  A smaller winning coalition, relative to the selectorate, encourages loyalty by limited access to goods, which are more expedient to deliver privately.  As the number of supporters that the leader needs increases, he/she must switch from providing private goods to providing public goods.  Increasing the size of the winning coalition thereby affects of number of economic decisions, and ultimately makes the leader more susceptible to defeat in office. 

The authors claim to endogenize the interaction described by Acemoglu and Robinson (2005).  Institutions affect coalition size, which in turn, affects the choice of future institutions.  Their explanation for leader tenure maps nicely onto Acemoglu and Robinson's explanation of democratization.  If we think of elites as W and the poor as S, we can imagine how changes in the relative sizes of each group would affect the incentives to redistribute.  What Acemoglu and Robinson call revolution Bueno de Mesquita et al. call leader defeat; repression equal purges, and concessions equal small-scale inclusion.  

Together, the stories suggest that autocracy is a situation where a small number of elites support the leader, relative to the rest of society.  When rest of society (the selectorate) become powerful enough, those in the winning coalition  may support repression (discouraging a widening of the coalition).  The leader may also be able to widen the coalition just enough to co-opt revolution.  Otherwise, elites must adapt to increasing the political arena to accommodate new demands, which entails a shift from private to public goods provision.  The two explanations are fairly complementary, and add to each other.  For one, Acemoglu and Robinson do not consider how the distribution of elite vs. citizen membership affects the dynamic.  This is precisely what Bueno de Mesquita et al. account for by considering the relative size of W to S in society.  Bueno de Mesquita et al. also highlight the ability of citizens to move from one segment to another.  The authors do not explain when and how a challenger might arise from the selectorate, however, which Acemoglu and Robinson explain as a shift in the defacto power of the disenfranchised.  Bueno de Mesquita also suggest that sources of conflict may differ by the group that initiates them, which Acemoglu and Robinson do not address (they assume that each group acts in tandem).  

For their compelling nature and complementarity, the two stories are incomplete.  For one,  some important concepts are not clearly defined, such as the term revolution.  It is also not clear who are elites, who are citizens, who is in the winning coalition, at any given point.  For any given issue, who is in W/S/R?  How does that differ when one considers multiple bargaining issues?  Secondly, one thing that both Bueno de Mesquita et al. and Acemoglu and Robinson ignore is the incentive for someone in the winning coalition--or an elite--to challenge the incumbent authoritarian.  Is it necessarily true that elites prefer to avoid a revolution, or might we expect entrepreneurial elites to create tension so as to capitalize on it?  Thirdly, although Acemoglu and Robinson touch on the issue, it remains unclear how exactly cross-cutting cleavages and multiple identities complicate "affinity" and create coordination problems.  Finally, Bueno de Mesquita et al. go beyond Acemoglu and Robinson in considering how membership in each group affects the source of conflict that might emerge.  They do not, however, go into detail as to how leaders' response to different forms of conflict might affect future politics.

The two arguments are convincing, but perhaps too abstract.  One must seriously consider what are the concepts to get at the need to strategically adapt.  Different actors have different (assumed) payoffs from, sources of, and vulnerabilities to defacto power.  The actors can also be expected to differ with regard to political capital and its convertability.  The arguments also do not adequately deal with how political differentiation and the bargaining between groups are affected by multiple issues.  What is more, both Acemoglu and Robinson and Bueno de Mesquita et al. overlook the role of order in determining outcomes.  The order in which grievances, opportunity, concessions, and institutions occur suggests that there should be more variability than the authors suggest.

Back to the broader issue--why do non-democracies use democratic institutions if they are not on their way to becoming full-fledged democracies?  Gandhi (2007) argues that dictators must thwart challenges to rule and solicit cooperation, but that the severity of problems differ by societies.  As Acemoglu and Robinson suggest, sometimes concession is not enough, for which institutions are necessary to co-opt opposition members.  Gandhi uses a similar explanation for seemingly democratic institutions in non-democracies as would Acemoglu and Robinson and Bueno de Mesquita  et al. The need for cooperation, societal polarization, and the observation of opposition strength induces dictators to offer monetary concessions and institutions that would create policy concessions as well.  Opposition members then decide whether or not to revolt.  Institutions are used by dictators who cannot avert threats to their rule otherwise, and the use of institutions by dictators affect specific liberties and policies.  They do not necessarily lead to democracy, which is the main point highlighted by Gandhi.  This point is largely overlooked by Acemoglu et al., who see democratization as any movement towards democracy.  Gandhi's empirics are not particularly convincing, nor does she address the variety of ways in which concessions can be made, institutions can be used, and threats can be averted.  The book is nevertheless an important contrast to the two major works, however, in that it introduces the expectation that political co-optation--also conceived of as organizational proliferation (see Haber 2006)--is a unique response to commitment problems that arise in society over the issue of redistribution which does not necessarily produce democracy.  Indeed, Gandhi shows that it produces particularly stable forms of non-democracy, though they may be more peaceful and moderate than more conventional forms of authoritarianism.

Economic Origins of Dictatorship and Democracy and the Logic of Political Survival are major cornerstones on which a stronger theory of democracy can be built.  They pose politics as bargaining between actors--who are distinguishable on the basis of entitlement--over different aspects of redistribution.  To mitigate societal conflict, the leader incentivizes their loyalties through the use of repression, concession, and institution-building.  Whether someone is entitled (in the winning coalition) or not, and whether he/she is enfranchized or not (relatively better off) shapes the particular threat that he/she poses to the leader.  The uniqueness of threats posed by the distribution of citizens demands unique combinations of repression, concession, and institutions to thwart a truly formidable challenge.  From these works, we can expect various actors in society to fluctuate with regard to relative enfranchisement and entitlement.  We can also expect them to decide, based on position and likeliness, whether to threaten the incumbent regime.  To the extent that the leader values office and can foresee the threat, he/she should manipulate the conditions to stay in office.  In so doing, the probability of regime change is altered, as are the actors that contribute to its survival.  These expectations nevertheless go beyond the aforementioned writings, however, in their emphasis on the non-stationarity of actors and different types of threats and threat response.  In combination with Gandhi, a theory of democratization must also account for the use of democratic institutions to respond to particularly threats, stabilize authoritarianism, and prevent democracy.