Bruce Mesquita - The Logic of Political Survival

The Replacement or Deposition Rule

Since our theory focuses on the actions leaders take to stay in office, we need to define the circumstances that result in their being deposed. We identify two different deposition rules, but focus only on the first. The second produces similar analytic results. The rules emphasize different thresholds that challengers must pass to come to power. Two plausible deposition thresholds are what we label (1) the constructive vote of no confidence and (2) simple deposition.

A constructive vote of no confidence is the more demanding of the two illustrative deposition rules. 12It gives the incumbent an advantage. Under this rule, the challenger must attract enough supporters away from the incumbent so that the incumbent no longer retains backing from W members of the selectorate. In addition, the challenger must have the endorsement of a coalition at least of size W to take over. That is, the constructive vote of no confidence indicates that if no one currently controls a coalition of size W , the incumbent remains in office. In a parliamentary democracy this is equivalent to the survival of a prime minister who only retains a minority coalition so long as no opponent assembles a majority coalition (Strom 1990).

The simple-deposition rule relaxes one restriction; specifically, the challenger simply has to pull off sufficient supporters from the incumbent’s chosen set of backers. If fewer than W members remain loyal to the incumbent, she is deposed and the challenger enters office. Under this rule if the incumbent loses control of a coalition of size W , the default favors the challenger. Although other deposition rules are plausible, for the sake of continuity and clarity, here we will use only the constructive vote of no confidence rule.

Political Systems: Analogies But Not Equivalence

The size of the winning coalition and the selectorate can be readily related to conventional labels for describing different political systems. As we have already intimated, in modern presidential democracies with nearly universal suffrage, S is about equal to N and W is typically a simple majority of S . In such systems, citizenship automatically holds out the prospect of benefits, both in terms of public policies and in terms of private benefits. In our framework, leaders, as agents of the selectorate, choose policy positions that they believe will satisfy those who keep them in power. They do not necessarily choose the policies most desired by the citizenry or even by their backers. Leaders may have their own policy preferences, after all. Still, since any citizen in a democracy can be a member of the winning coalition, the policy preferences of all citizens must be considered by leaders as they formulate the public policies they pursue and the private goods they allocate. In addition to the consumption of public policies by all citizens, any enfranchised resident in a democratic system with direct election of a president as chief executive has about a 50 percent chance of obtaining some, probably very small, private benefits through membership in the winning coalition.

In nondemocratic political systems, the picture is different. In a single-party dictatorship, for instance, S may not be equal to N . S may be only 10 or so percent of the total citizenry, as was common in some communist and fascist states. S might be much smaller than that. It might consist of a tiny minority of citizens. Authoritarian regimes run the gamut in terms of the size of the selectorate. When the selectorate is small, this means that the policy preferences of the vast majority of residents ( N - S ) can be ignored as a part of daily, routine politics. 13Only the preferences of the citizens in S need attention. For the citizens in the selectorate, then, there is a chance of gaining access to private benefits in the future and there is a current influence on policy choices. For the disenfranchised set of people, N - S , neither of these benefits exists. For the members of W there is influence on current policy and there are immediate private gains to be had. In universal-suffrage democracies, W is large. In single-party dictatorships, even if S is universal, W is small. Communist Party membership in Vietnam, for example, is only about 3 percent of the population and only a fraction of that 3 percent are needed to keep the leadership in office.

Monarchies have still different characteristics. In a typical monarchy, only a very small number of people have a routine prospect of becoming members of the winning coalition. Recall that King John’s England had 236 barons who comprised the entire selectorate. Of course, they had their own followers who were likely to be rewarded by them for their good service, but only supporters among these barons were really vital to the king’s election and survival in the face of rivals. In traditional monarchies, the selectorate is small. The winning coalition, W , is naturally even smaller because it is a subset of S . In some monarchies the value of W is determined by a very limited-franchise majoritarianism and in others it is not. We have already mentioned examples of elective monarchies. Some had even smaller selectorates than those we have mentioned.The Holy Roman emperor, for instance, was chosen by a majority vote among seven electors, a very tiny group indeed (Eulau 1941). 14These few constituted S .

Monarchies were sometimes only nominally majoritarian; still, many did function at least in a quasi-majoritarian manner. Without the support of a majority of the royal family it was often difficult for an individual to become a monarch or to remain as the leader. During the Merovingian, Carolingian, and Capetian dynasties in France, for instance, it was not automatically the case that the king’s eldest son succeeded him. Rather, there was a quasi-electoral process and a clear requirement that the future king have support among a significant proportion of the royal family and other influential individuals. Indeed, this was a central concern in the early years of the reign of Philip Augustus (1179-1223) in France, who, unlike many of his predecessors, was designated by his father to succeed to the throne (Baldwin 1986). His father did so out of fear that Philip would otherwise not be selected by the nobles. Even with his father’s endorsement, there was controversy surrounding Philip’s ascent to the throne. Similar patterns of election and competition for kingship can be found in the history of the monarchies of medieval Castile, León, England, and elsewhere. Indeed, recognition of this essential fact is at the heart of William Shakespeare’s Hamlet .

Sometimes majoritarian monarchies expanded the selectorate beyond the royal family. At least since 1295, the king of England depended partially on majority votes by the House of Lords in Parliament to retain many of the privileges of office. These lords were drawn from a somewhat broader set of people than the few barons on whom earlier English kings depended. Chapter 8 explains the specific conditions under which kings had incentives to increase the size of the selectorate and to increase (or at least to accept an increase in) the size of the winning coalition.

Like monarchies, military juntas include very few people. Typically, a military junta depends on a handful of colonels or generals to form the selectorate, and some small fraction of those are in the winning coalition. Juntas, like monarchies, need not be majoritarian within their very limited selectorate.

Though it is simple enough to relate W and S to well-known regime types, we make a conscious effort to move away from categorical discussions of political systems. The links between the size of W and S and regime types like democracy, autocracy, monarchy, and junta are useful as heuristic devices, but no two democracies are alike, nor are any two autocracies, monarchies, or juntas. Inevitably, categorical discussions of regime types lead to the construction of arbitrarily drawn boundaries. That, of course, is why there are so many different ways that people define democracy. 15