The debate about the cost-effectiveness of randomized field experimentation ignores one of the most important potential uses of experimental data. This article defines and illustrates “downstream” experimental analysis—that is, analysis of the indirect effects of experimental interventions. We argue that downstream analysis may be as valuable as conventional analysis, perhaps even more so in the case of laboratory experimentation.

The purpose of this Comment is to put the current debate regarding testing necessary conditions into perspective and to point out a particularly troubling aspect of the “all cases” research design (Seawright 2002).

Prior to the recent spate of books and articles in the social sciences (Ragin 1987; Dion 1988; Braumoeller and Goertz 2000), the debate over the testing or the confirmation of necessary conditions took place in the philosophical literature, mainly in terms of Hempel's (1945) paradox of the ravens. In what follows, I briefly review Hempel's paradox and the Bayesian solution to it. I argue that Seawright's account, while Bayesian in nature, relies on an assumption that no Bayesian would be willing to make.

Few better ways of checking and improving statistical methods exist than having other researchers go over your results, and so I especially appreciate the efforts in Anselin and Cho (2002), hereinafter AC. In this note, I make two main points.

In this study, we unite two experimental traditions to examine the impact of discursive processes on political decision making. We directly manipulate the presence and timing of discussion in the “divide-the-dollar” game to assess the effects of discussion on participants' allocations and perceptions of the game's legitimacy. To investigate the influence of structure, we also manipulate the presence of a majority/minority cleavage among participants. The dependent measures in all instances are the players' allocations, the outcome of the game, and psychometric indicators of legitimacy perceptions. Results indicate that the presence of discussion can generate outcomes that are perceived as more equitable and fair in some circumstances—namely, when a cleavage is present. These findings establish the utility of this paradigm, as well as an important baseline for assessing the probable impacts of proposals to integrate deliberation into political decision making.

In “Testing for Necessary and/or Sufficient Causation,” Jason Seawright (2002) presents an intriguing new Bayesian technique for testing theories that posit necessary conditions. His technique differs in substantial ways from those presented in separate articles by Dion (1998), Ragin (1987), and ourselves (2000) and constitutes a fundamental critique of those procedures. Having criticized quite a few procedures and theorists in our own article, we can hardly complain; rather, we recognize the constructive spirit of the endeavor and, in that same spirit, would like to offer a rejoinder.

We read with interest the sundry points raised by King in his comment (2002) on our article (Anselin and Cho 2002). Given the space and time constraints in this forum, it is impossible for us to address adequately all of the issues here. It is sufficient to say that we strongly disagree with King's claims that we are “missing the point.” We will pursue a more thorough and formal rebuttal elsewhere, but would like to set the record straight on a few issues in the limited space provided here.

The complex question of what counts as appropriate evidence for necessary and/or sufficient causation merits careful, ongoing discussion. I thank Kevin Clarke (2002) and Bear Braumoeller and Gary Goertz (2002) for their thoughtful responses to my (Seawright 2002) article, and I discuss their comments in turn.

We show that temporal, spatial, and dyadic dependencies among observations complicate the estimation of covariance structures in panel databases. Ignoring these dependencies results in covariance estimates that are often too small and inferences that may be more confident about empirical patterns than is justified by the data. In this article, we detail the development of a nonparametric approach, window subseries empirical variance estimators (WSEV), that can more fully capture the impact of these dependencies on the covariance structure. We illustrate this approach in a simulation as well as with a statistical model of international conflict similar to many applications in the international relations literature.