I think of institutions that are robust over time, and lend themselves to compar- isons across settings, as structured. They persist in roughly the same form from year to year, and their similarities to and diVerences from objects sharing their label in other places also persist.7 Thus, the US Congress, or the New York Assembly, or the Irish Dail are structured in this sense. So, too, is a parliamentary cabinet, a judicial court, an administrative bureau, a regulatory agency, a central bank, an electoral regime, even a political party, a royal court, or an army. Rational choice institu- tionalism has explored many of these. There is surely variation among the myriad instances of any one of these structured institutions; but there are also powerful central tendencies. This is what induces us to group them together and to think it sensible to compare them.
Other institutions are less structured. Like structured institutions, they may be described as practices and recognized by the patterns they induce, but they are more amorphous and implicit rather than formalized. Norms, coordination activity, cooperative arrangements, and collective action are instances of what I have in mind. Senatorial courtesy, for example, is a norm of the US Senate eVectively giving a senator a veto on judicial appointments in his or her state (Binder and Maltzman 2005; Jacobi 2005). Seniority was a norm of both chambers of the US Congress for most of the twentieth century, establishing queues or ladders in congressional committees on which basis privileged positions—committee and subcommittee chairs, the order of speaking and questioning in hearings, access to staV, etc.—were assigned.8 Neither of these norms is a formal rule of the institutions.
Various forms of patterned informal interaction, including coordinated agreements like which side of the road to travel, sharing rules like ‘‘split the diVerence,’’ and understandings like ‘‘tit for tat’’ (Axelrod 1984) and ‘‘taking turns’’ (Ward 1998), also constitute unstructured institutions. These patterns emerge informally and often are not actually written down as formal rules; they simply come to be known as ‘‘the way things are done around here.’’ They are, in short, equilibrium patterns.
Collective action—the capacity of a group of individuals to coordinate for mutual advantage—sits close to the boundary between structured and unstruc- tured institutions. Sometimes it takes the form of well-organized and formalized arrangements; other times it looks spontaneous and idiosyncratic. Interest- group political organizations described by Olson (1965) constitute instances of the former, while intergroup ethnic relations, sometimes peaceful sometimes not, are often patterned but unstructured and implicit (Fearon and Laitin 1996).
- Structured Institutions
Probably the single biggest success of the rational choice institutionalism program is the analysis of structured institutions. There are several factors that facilitate rigorous analysis and thus account for this success.
First, politicians in these settings are selected in a relatively well-deWned way— election to legislatures or party oYces, appointment to courts, regulatory agencies, or higher executive posts. Politicians may thus be thought of as agents of (s)electors (Bueno de Mesquita, Smith, Siverson, and Morrow 2003). Their activities while in oYce will be motivated in part by the objectives of the (s)electorate—see below.
Second, politician objectives can be speciWed with some precision, due in part to selection eVects. In the literature these objectives are often grouped into oYce preferences and policy preferences.9 Ideal-types holding preferences of the former category care primarily (only?) about oYce and the perquisites that come
with incumbency—salary, inXuence, control of staV, generalized prestige. More recent work, under the rubric of career concerns, places special emphasis on selection eVects.10 The policy preferences ideal-type cares about policy outcomes. In the spirit of Downs (1957), oYce-oriented politicians make policy in
order to win elections whereas policy-oriented politicians win elections in order to make policy.11
Third, politician behavioral repertoires are delineated by institutional rules and processes. A legislator on the Xoor of the chamber, for example, may seek recog- nition from the presiding oYcer or not. If he does, he may oVer a substantive motion, a second to a motion, an amendment to an existing motion, a procedural motion (to table, to recommit, to adjourn, etc.), a point of order or information, and so on—some of which are permitted by the rules (‘‘in order’’) and some of which are not (‘‘out of order’’). If a vote is called, he may vote yea, vote nay, or abstain (in whatever manner of vote expression is required). That is, the ‘‘legislation game’’ may be written down and the strategies available to the politicians speciWed.12 In other structured institutional settings, the repertoires of judges and bureaucrats may be portrayed in clear-cut ways.
Fourth, outcomes are clearly implied by the conWguration of rules in a struc- tured institution. These rules prescribe the mechanism for aggregating behaviors into a Wnal result. Thus, any combination of behavioral repertoires by institutional politicians maps into a speciWc outcome.
Fifth, payoVs may be inferred from the objective functions of politicians. Policy- oriented players will prefer the combination of behavioral repertoires that map into more desirable outcomes. OYce-motivated politicians will prefer those repertoire combinations that improve their prospects with their (s)electorate. If the selection mechanism chooses politicians with policy preferences closely aligned to those of their (s)electorate, then we may not be able to distinguish between the two preference types empirically. The strategic choices of oYce types and policy types will be observationally equivalent.
Finally, there is the matter of (s)electorate preferences. The (s)electorate is the collective principal that chooses an institutional politician to act as its agent. With their preferences in hand, we complete the circle. (S)electorates are vulnerable to two kinds of ‘‘agency problems’’—adverse selection and moral hazard. The Wrst problem is associated with hidden information—characteristics of the prospective agent that cannot be known in advance by the principal. Is the politician of ‘‘high quality?’’ Does he or she share policy preferences with the (s)electorate? The second problem is associated with hidden action—strategic agent behavior that
may not be discernible by the principal. Does the politician support the preferences of the (s)electorate in arenas where his or her behavior cannot be directly observed (an unrecorded vote, a secret committee meeting or party caucus, a meeting with a lobbyist)? The connection between (s)electorate and politician entails some form of delegation from principal to agent and is characterized by more or less accountability by the agent to the wishes of the principal. The rational choice literature on each of these facets of institutions is vast.13
- Unstructured Institutions
The Archimedian lever of rational choice institutionalism is provided by the structure of structured institutions. This structure embeds the logic of optimization in a strategic context. The context of unstructured institutions is more Xuid, providing a less Wrm foundation for analysis. Many more things are possible; many more contingencies need to be accounted for. However, considerable pro- gress has been made.
The great success story in this region of the rational choice institutionalism program is the logic of collective action (Olson 1965). The foundational basis for this work is the analysis of public goods, dating back to the early work of Samuelson (1954). Collective action for a group is a public good, an outcome desired by its members but diYcult to elicit costly contributions for its production. Members, according to this logic, are attracted to the free-riding option since non-contribu- tion is a dominant strategy in the collective action game. Mancur Olson took this insight and demolished prevailing pluralist and Marxist views on groups by arguing that they will not of necessity form around common interests and objectives (as these more sociological arguments had taken for granted) precisely because of the logic of free-riding. Individual contributions are both personally costly and often only trivially important in achieving a group goal, especially in large groups. So individuals are tempted to abstain from contributing. This temptation is reinforced by the realization that everyone else will be tempted to free-ride.
Groups do form and not everyone free-rides all the time. Why? Answering this question has constituted something of a light industry. Olson argued that since success in inducing an individual to contribute does not come from the prospect of realizing group objectives (which will be enjoyed if the group succeeds whether she contributes or not, and whose contribution is negligible in any event), then it must
come from some other source. Groups must be able to oVer things of value to contributors and only to contributors—selective beneWts, not collective beneWts. The group objective is Wnanced, therefore, as a byproduct of bribing individuals to contribute with private compensation.
One of the earliest responses to Olson’s classic was a book review by Wagner (1966). There he pointed out a glaring omission in the byproduct logic of Olson’s theory of collective action—namely, the role of leadership. (Also see Frohlich, Oppenheimer, and Young 1971.) Wagner suggested that even Olson’s byproduct logic must have some source of implementation. Inventing the term political entrepreneur, he argued that particular individuals may make unusually large contributions of time and energy and Wnancial and (especially) logistical resources not (only) because they care passionately about the group’s objective but (also) because they see an opportunity to parlay this investment into something personally (read: selectively) rewarding. It is no surprise, for example, when a congressman from south Florida (home to many retirees) provides political lead- ership on issues beneWting the elderly—the electoral connection supplies the ex- planation (whether the congressman is personally passionate about these issues or not). Likewise, it is surely not entirely explained by ‘‘generosity of spirit’’ when a young lawyer takes on a cause—say, the lead-poisoning of inner city infants—even though there may be no immediate remuneration. Applying the career concerns logic just suggested about the congressman, this political entrepreneur takes lead- ership of an issue in order to advance a personal agenda (of which Wnding a solution to the issue at hand may be part, but only part), possibly parlaying his public spirit into a political career, a network of contacts, future remuneration for his legal practice, etc. The leadership explanation is not entirely compelling in all settings. But it invites us to scrutinize some of the less obvious motives of those who assume the mantle of leadership. (On the rational choice analysis of leadership more generally, see Fiorina and Shepsle 1989; and Shepsle and Bonchek 1997, ch. 14.)
A feature of all collective action from a purely rational perspective is that outcomes are not Pareto optimal. Everyone would be better oV if there were some way to coerce contributions. Selective beneWts and political entrepreneurs are two of the most important contributions of rational choice institutionalism to an appre- ciation of solutions to collective action phenomena. Leadership, in fact, may be interpreted as giving some agent the authority to wield carrots and sticks—that is, provide selective incentives—to induce contributions to group objectives and thus move the collectivity onto the Pareto surface. (Indeed, this is a rough approximation of arguments made centuries ago by Hobbes and Hume to justify the existence of the state. Generally, see Buchanan and Tullock 1962; Hardin 1982; Sandler 1992.)
A third ‘‘solution’’ to the problem of collective action is best understood in the problem writ small—the problem of cooperation. Axelrod (1984) paved the way to understanding how to get individuals to seize a cooperation dividend, rather than leaving it on the table, by examining repeated prisoners’ dilemma (PD)
situations.14 In the PD an individual can cooperate with another and capture a beneWt, exploit the cooperative inclinations of the other by non-cooperating and do even better while the other suVers a loss, or join his opposite number in non-cooperation and get nothing. A dominant strategy in the one-shot PD is for both individuals not to cooperate, producing a zero payoV and something left on the table. (What is left on the table is a positive payoV had both cooperated.) The idea exploited by Axelrod, and I count this as the third important solution to collective action problems (along with selective beneWts and leadership), is repeat play. Axelrod noticed what game theorists had discovered even earlier—that repeat play allows for ‘‘history contingent’’ strategies. Thus, in the play of a PD game at any time t, each player may take into account the way the game was played in earlier periods, and make his or her behavior in the current interaction contingent on previous play. Today’s play, therefore, determines not only today’s payoV but will inXuence the behavioral choices of others tomorrow. This may, depending upon how much the players value tomorrow’s payoV relative to today’s, induce them to eschew their dominant strategies in the one-shot play of the PD and choose to cooperate instead. Indeed, unlike leadership and selective beneWt solutions to collective action, repeat play is more like an invisible hand.
I have oversimpliWed this discussion, but it allows me to observe that history dependent behaviors in equilibrium—‘‘tit for tat,’’ ‘‘take turns,’’ ‘‘split the diVerence’’—come very close to the ordinary language meaning of norms and conventions.15 The program of rational choice institutionalism thus provides analytical handles on the collective action problem writ large and writ small.
