Political Contingency: Studying the Unexpected, the Accidental, and the Unforeseen

6. Region, Contingency, and Democratization

Chapter 6 Region, Contingency, and Democratization Susan Stokes Recently,1 Robert Dahl noted that the challenges facing the world’s roughly 200 countries vary, from the transition to democracy in non-democracies, to the strengthening or consolidation of democracy in newly democratized countries, to the deepening of democracy in older democracies.2 As we grapple toward an understanding of transitions to democracy and of its consolidation and deepening, we frequently try to discover general laws of cause and effect, ones that operate in the same way over time and space. With a few exceptions, we have ignored the role of contingency in encouraging or impeding democratization. Yet mounting evidence points toward spatial unevenness in democratization and in the consolidation of democracy . And this spatial unevenness can be the result of contingent choices and chance events. The dictionary definition of contingency is something that is likely but not certain to happen, happens by chance, or dependent on something else. In this chapter I define a contingent cause or outcome as one that depends on choices or events whose probability of occurring is low. After discussing contingency in general and offering some examples of political phenomena that are subject to contingent causation, I focus on contingencies related to spatial location. I examine the role of region in democratization . In so doing, I hope to clarify what we mean by regional effects in politics, and to specify whether, and which kinds of, regional effects represent a form of contingency. What Is Contingency? Political outcomes may be contingent in at least three senses. A potential cause may have its effect only in the presence of some background 171 Shapiro_pp097-202 7/17/07 3:12 PM Page 171 condition or additional cause, and the fact that an interaction is required makes the event unlikely (interactions). Or an outcome may appear ex ante unlikely because it depended on two or more factors that had to occur in a particular temporal sequence (sequence). Or it may rely on a choice by an actor who might have made another choice (choice). All of these senses of contingency have been discussed at length in recent scholarship.3 Interactions In interactions, A and B cause C; with only A or only B, C does not occur . Unless both A and B are certain to occur, then, according to probability theory, the fact that they both have to occur makes the outcome C less likely than it would be had only A or only B been required. Almond and Verba’s The Civic Culture contains an example.4 They find that, in advanced democracies, people who trust government also participate more in politics. But in developing countries trust does not cause participation. If trust is A and development—the necessary background condition—is B, and if participation is C, then A and B together cause C, but A alone or B alone does not. Kalyvas offers an example in which the chance simultaneity of two causally unrelated events has an effect that would have been absent had the two events not coincided.5 In England in the early 1970s and France and early 1980s, leftist governments came to power and instituted policies of nationalization. In both countries major economic crises followed these nationalizations, crises that were caused by events that had little to do with the nationalizations. The chance near-simultaneity led, in the minds of the mass publics of both countries, to “the association of nationalization with economic crisis and the subsequent rejection of nationalization by public opinion.”6 If nationalization is A and economic crisis B, then had either A or B not occurred, the discrediting of nationalization (C)—and, indeed, according to Kalyvas, a broader collapse of a Keynesian economic hegemony —would also not have occurred. Indeed, had these events not unfolded in the particular order that they did—for instance, had economic crises preceded nationalizations—then C would not have followed. Kalyvas ’s example leads us nicely, then, into a consideration of sequence and contingency. 172 s u s a n s t o k e s Shapiro_pp097-202 7/17/07 3:12 PM Page 172 Sequence A must happen before B in order for C to occur; if B happens before A, C does not occur. In addition to Kalyvas’s example, here are two more. First, where the habit of political contestation among opposing elites preceded mass electoral participation, Dahl explains...