Trust and Reciprocity: Interdisciplinary Lessons for Experimental Research

14. The Transformation of a Skeptic: What Nonexperimentalists Can Learn from Experimentalists

373 Chapter 14 The Transformation of a Skeptic: What Nonexperimentalists Can Learn from Experimentalists Margaret Levi R eading through this collection of papers for the purpose of writing a concluding note on “what social scientists can learn” has proved a daunting task. My own work is in historical and comparative political economy. My role in the scholarly division of labor is to reveal the micro foundations of macro events and outcomes that are often unique and then to provide the link between the micro and the macro. Even in the finest comparative research, the capacity to control for the diversity of independent variables and multiple motivations is at best limited and, more often, hopeless; it is certainly just about impossible to proceed as systematically and rigorously as the experimentalists do. At first glance, then, nothing could be further from what I do than the very micro, carefully controlled experiments documented in this book. Moreover, mine is the view of someone who is skeptical as well as ignorant about experiments. I am in awe of the rigor of experimentalists , and I am persuaded by many of their findings. At the same time, like other social scientists who deal with data derived from the events and behaviors associated with real polities and economies, I often have had difficulty understanding how to translate the insights of experimental research into the ways the people I observe regularly interact with one another and the conditions under which they inter- 374 Trust and Reciprocity act. Second, much of the experimental research of the past conflates trust and cooperation or trust and reciprocity. I am concerned about the extent to which experiments can actually get at the question of trust as opposed to the behavior and choices that trust elicits. Third, despite the popularity of evolutionary theory among some of the social scientists I most respect and admire, I remain dubious of the usefulness of expending so much intellectual effort in forging links with this research program. My ignorance is also multifaceted. I am not conversant with the nuances of experimental research, of how experimenters ensure that they actually get at what they want and in an uncontaminated way. Only by attending the workshop that produced these papers did I finally begin to grasp the distinctions among the disciplinary approaches to experimental research. I am just beginning to get a good feel for the complexity of the relation between game theory and experimental design. Reading these papers has eliminated some of my most egregious forms of ignorance and thus alleviated at least part of my skepticism. Moreover, I am now far more aware than before of how experimental research can inform other kinds of social science scholarship. In particular , I believe experimentalists have something to teach the rest of us about the methodological issues in studying the relations among trust, cooperation, and reciprocity; the limits of game theory and of the standard model of rationality; the value added of other theoretical approaches, most especially those derived from cognitive science and evolutionary biology; and, of course, the substantive issues of the inquiry : trust, trustworthiness, cooperation, and reciprocity. Methodological Issues In chapter 3 of this volume, Russell Hardin, a political philosopher, emphasizes conceptual distinctions, and his kind of work is sometimes faulted for overconcern with the fine points of linguistic differentiation . However, useful experiments demonstrate why such distinctions matter. Precision regarding the rules, the decision to participate, the order and direction of moves, and the scenario facilitate distinction among various motivations and behaviors. It is not that all agree on the terms. For example, when referring to the belief that emerges from thick, ongoing relationships, Hardin uses the term “trust” for what Toshio Yamagishi, in chapter 13 of this volume, labels “assurance.” Karen Cook and Robin Cooper (chapter 8, this volume) think Hardin may be right that risk is what is at issue to the first mover in the trust-honor game, but Kevin McCabe, Vernon Smith, Catherine Eckel, and Rick Wilson (in chapters 6, 9, and 10 of this The Transformation of a Skeptic 375 volume) consider trust to be in play. The usage of the terms may differ, but what is important here is that the analysts are clear in what they mean and can justify their inference that the experimental design tells us something about what they claim to be investigating. This pluralism about terms does not obviate the fact that experimentalists often conflate...