Increasing Returns,Path Dependence,and the Study of Politics June 2000 you work for a firm with an annoying boss and bad pay. depends heavily on the actions of others.This is less You have a clear option:Seek work elsewhere,either true of some aspects of politics-such as answering an at one of a large number of other firms or by setting up opinion poll question or voting-than others.Even in business on your own.Your ability to move depends on voting,however,the lumpiness of election outcomes the state of the labor market,but the existence of (in the absence of a pure system of proportional competitive options sets clear limits on how annoying representation)means that the actions of a person who your boss can afford to be and on how bad the pay can does not want to "waste"her vote may well turn on get. what she expects others to do Or suppose you invent a great new product.Assum- A crucial feature of most collective action in politics ing that you can get financial backing (which you is the absence of a linear relationship between effort should be able to do-it is a great idea,and the market and effect.Instead,collective action frequently involves generates a ready supply of venture capitalists),your many of the qualities conducive to positive feedback prospects are good.Nothing stops you from going into (Marwell and Oliver 1993).A central reason is the business or selling the idea to someone.Either way,the prevalence of adaptive expectations.When picking the new,superior product gets to see the light of day,and wrong horse may have very high costs,actors must you reap considerable benefits from your innovation. constantly adjust their behavior in the light of how they The setting of consumers,at least in the textbook expect others to act.Whether you put energy into case,is similarly atomistic.My decisions are essentially developing a new party,or join a potential coalition,or independent of my expectations regarding the choices provide resources to an interest group may depend to of other consumers.12 There is no need for explicit a considerable degree on your confidence that a large attempts to coordinate behavior;the market simply number of other people will do the same.In addition, aggregates the isolated decisions of individuals. many types of collective action involve high start-up These highly stylized examples illustrate the flexibil- costs,which reflects the fact that considerable re- ity,fluidity,and atomization of economic markets.In sources (material or cultural)need to be expended on contrast,political "markets"are generally far from organizing before the group becomes self-financing. flexible and fluid.In politics,the consequences of my That collective action processes in politics are very actions are highly dependent upon the actions of often subject to increasing returns explains why social others.What I get depends not just on what I do,but scientists are often struck by the considerable stability (mostly)on what others do.Following Olson's (1965) of patterns of political mobilization.over time.Lipset path-breaking work,students of politics have long and Rokkan's (1967)work on political parties in Eu- recognized the "logic of collective action."Most of the rope exemplifies this dynamic:Key historical junctures "goods"produced in politics are public goods;it is produced major political cleavages.These political difficult to limit their consumption to those who helped divisions became organized into political parties.Once provide them.As a result,individuals will have a strong they have surmounted initial start-up costs and fueled tendency to free-ride.Creating conditions favorable to processes of adaptive expectations,these parties are collective action is a principal issue in political life. reproduced through time,which generates "frozen" The problem is not limited to the fact that the public party systems. sector produces public goods.Given the reliance of Recent work by Skocpol (1999)on extensive volun- politics on mechanisms of collective decision backed by tary associations in the United States provides addi- authority,laws themselves have the character of public tional strong evidence of the organizational persistence goods for those who benefit from them.In the words of that can result from positive feedback.Skocpol identi- Marwell and Oliver (1993,42),"influencing govern- fied all such organizations that had ever enrolled more ment policy almost always has very high jointness of than 1%of the American population (or half that supply."These circumstances generate major collective amount for single-gender groups)and tracked them action problems. over time.The results,which cover 58 groups since the There is another reason political action frequently 1830s,reveal striking organizational continuities.Al- requires coordination.Many of the goals pursued by though some crossed the 1%threshold only for a political actors have a“lumpy'or“winner-take-all” relatively short period,26 remain above it today.Of quality (politicians seeking reelection,coup plotters, these,16 had reached the 1%mark by the 1940s,and a and lobbyists either win or lose;legislation either number of them stretch back much farther.A large passes or is rejected).Unlike economic markets,in number have fallen from the list,but they managed to which there usually is room for many firms,in politics stay on it for many decades.Among the 40 organiza- finishing second may not count for much.Indeed-the tions founded before 1900,19 stayed above the 1% Menshiviks in 1917 come to mind-it can be extremely mark for at least five decades.Ten of the 40 are still problematic.Again,the effectiveness of my actions above that threshold,a century or more after their founding.13 In short,despite massive social,economic, This represents a critical difference between economics and poli- tics,but a number of important qualifications need to be made.For instance,consumption often involves significant externalities,which 1 It should be emphasized that this evidence understates organiza make consumer choices interdependent.As already noted,condi- tional persistence,since many groups that fall below the demanding tions of independent consumption often do not apply to high-tech 1%threshold nonetheless continue to have a very large membership. products,which frequently involve network externalities.For a good Also,they may have existed as quite large organizations for long discussion of some of these complications,see Hirsch 1977. periods before initially crossing the threshold. 258