Using techniques from the experimental and computational social sciences, this dissertation identifies three central flaws in one of our most important mechanism designs – elections – and identifies institutional methods to solve these problems.

Mechanism design is the “engineering” side of economic theory. Where game theory “takes the rules of the game as given, and [] makes predictions about the behavior of strategic players,” mechanism design “is about the optimal choice of the rules of the game.” For mechanism designers, identifying the optimal choice of rules typically depends on predicting how individuals will behave in certain environments. Both mechanism design and game theory, therefore, rely on identifying equilibriums – that is, joint strategic solutions where no one involved has an incentive to deviate from the strategy that they are playing.

Economists argue that electoral accountability can be seen a special version of mechanism design. The logic is this. Electoral institutions establish processes for selecting candidates for public office, in such a way as to incentivize them to act in the public interest. To achieve this result, electoral institutions lead to a cascade of mutually-reinforcing incentives. These institutions (a) incentivize voters to retrospectively evaluate officials' performances such that (b) victorious candidates will tend to act in the public interest. Therefore, electoral accountability is a set of institutions, and the incentives that predictably flow from it. The promise of mechanism designers is that, if the institutions are created correctly, officials will be accountable to the public. And this accountability will occur as a matter of course: mechanically.

In fact, however, this mechanism is broken. This dissertation solves its problems. Chapters 2 and 3 represent three efforts at testing rational choice-based theories of electoral accountability through randomized control trials. These studies take aim at concepts that are crucial to our understanding of a central feature driving retrospective evaluation of policy. Chapter 4 operates at a more theoretical level, but nonetheless retains its behavioral approach. It offers an approach to legal chance that highlights the strategic alternatives facing firms. Chapter 5 offers institutional fixes to the problems identified in Chapters 2, 3, and 4.