We wish to thank Justin Buchler for his kind words and trenchant insights about our book. Comparing our Intuitionist Model to Zaller’s RAS model gives us big shoes to fill. We would like to think that we offer a contribution in the spirit of Lakatos’s “positive problemshift” (Imre Lakatos, “Falsification and the Methodology of Scientific Research Programmes,” in Imre Lakatos and Alan Musgrave, eds., Criticism and the Growth of Knowledge: Proceedings of the International Colloquium in the Philosophy of Science, London, 1965, 1970). We believe that our model preserves what the RAS model predicts and adds to it by making additional predictions. In particular, our model offers an explanation for why some highly aware partisans do not always act in the sophisticated ways predicted by spatial theory. If people are not motivated to second-guess themselves, they are more likely to choose party loyalty over policy congruence.

We also appreciate Buchler’s two critiques. First, it is true that the act of measuring people’s attitudes and behaviors could alter the way in which they generate those attitudes and behaviors. What we observe on a survey or in the lab may not travel to the real world. It is worth noting here that this is a general problem for empirical science. Even physicists face this problem: The act of observing what subatomic particles do may affect what they do. In our case, we are somewhat heartened by the results reported in Chapter 5 with respect to economic performance and presidential evaluations. The survey data come from a panel of individuals interviewed every month from 2011 to 2015, and economic performance was measured using the actual monthly change in unemployment. Our indicators of reflection came from one of the panel surveys, while the evaluations of President Barack Obama came from many panel surveys over a four-year period. Consequently, we do not think there is much to worry about when it comes to testing interference. Nonetheless, the mere fact that people are taking a survey about politics means that some people could be induced to think more than they normally would, and Buchler is quite right that we cannot definitively dismiss this possibility.

Moreover, we agree that some day-to-day decisions may not be sufficiently consequential to warrant reflection among even those with the highest propensity for reflection. This, we argue, is an outcome consistent with our model, insofar as it provides for the possibility that some decisions will simply not engage System 2 (i.e., higher order, cognitively intensive) processes. It is true that our tests are designed to engage issues that people care about, and for which reflection is a possibility. To that end, we may observe, on balance, more reflection in the lab than we would on most mundane decisions. The important point, from our perspective, is that we observe systematic variation in reflection across individuals as a function of the way they process information.

Second, we wholeheartedly agree that more work needs to be done to investigate whether reflection influences electoral outcomes in ways that improve democratic accountability. Our speculative test in the conclusion was just that—speculative. We hope that our work spurs more research on this question. Beyond the empirical question, we would also love to see more formal theorists, like Buchler, incorporate behavioral models of decision making into their formal models. We see a great deal to be gained from collaborations at the intersection of political psychology and formal theory, and we are glad that scholars like him are pointing the way.