Journal of Theoretical Politics

A spatial theory of the comparative consequences of unified and divided government clarifies the relationship between institutions, partisanship and gridlock. Under formally specified conditions, executive and legislative institutions are shown to inhibit but not prohibit the convergence of public policies to those most preferred by the legislative median voter. Extensions illustrate the theoretical robustness of the primary empirical expectation. Under all but extreme conditions, gridlock is a predictable consequence of super-majoritarian procedures. Gridlock occurs in divided and unified government alike, and in the presence of any degree of party strength. Concluding discussions speculate on the effects of alternative assumptions and on applications of the theoretical framework to settings other than US national government.

In any society characterized by diversity and specialization, each citizen - the butcher, the baker, the candlestick-maker - is a special interest on some dimensions of public policy and a member of the general public on many other dimensions. Each citizen enjoys being at the receiving end of special interest handouts even at the expense of inefficiencies imposed on general public. But relative to the status quo involving inefficient redistribution on most dimensions of public policy, including many dimensions on which the citizen is a member of the general public, the vast majority of citizens would surely be better off if the government did not cater to special interests at all. With this common interest in place, why do democratic societies have such a hard time ridding themselves of special interest politics?

This article argues that the driving force underlying special interest politics is an information asymmetry: each citizen has a precise idea about the special interest handouts she is receiving and a vague idea about the price she is paying, as a member of the general public, for inefficient redistribution. I explain why special interest politics are ultimately impervious to reform (though we can tame them at the margin); I examine the trade-offs inherent in reform efforts (what it costs to tame them); and I spell out why special interest politics can be a Good Thing (why we should not tame them all the way even if we could).

A common mental model sees lobbies pushing policy in a certain direction from the status quo with a certain force, the outcome being a function of how hard lobbies push and in what direction, according to the `parallelogram of political forces' at work. When this model is formalized and lobbies are seen as playing a game in which they are attempting to maximize their payoffs, net of lobbying costs, the prediction is that lobbies seldom if ever will push in the direction of what they ideally want, for their lobbying will partly aim to offset the efforts of others, leading to waste of resource. There will often appear to be more conflict of interest between two lobbies than actually is the case. This calls into question methodologies for studying and for measuring power that assume that we can infer groups' preferences from their lobbying behaviour. From observing their behaviour we can infer things about their power if we know their preferences or we can infer things about their preferences if we know their power. But it is not possible to make both inferences.