Using Elites to Study Elite Decision-making: Opportunities and Pitfalls

What are the opportunities and pitfalls of quantitative research in which elites are used to study elite decision making? Decision making by elites is a central research focus of political science, yet studying elites in their natural environment can be a challenging task. Two main issues confront researchers interested in elite behavior: (1) the difficulty of obtaining access to elites, and (2) the strategic importance for elites to uphold and project a specific public image. Elites have busy agendas, and are shielded by gatekeepers. Their political careers likely depend on the image they communicate to the outside world. Researchers will have to find a way to deal with these issues. To gain access and confidence, they can invest in their relations with elites. This personal approach fits well with, for example, elite interviews. Such qualitative methods make optimal use of all available context information and generally require smaller sample sizes. Such methods can be very useful to examine the specifics of a case and/or arrive at a better understanding of decision making processes (e.g., Kingdon, 1984; Weyland, 2009, 2014). Quantitative approaches might add to such studies of elites, as they can explore general patterns. The anonymity of numbers that these studies provide might be appealing to elites. Still, convincing them to participate requires much time, effort, and good-will. Especially, as they require larger sample sizes to obtain valid findings. In this paper, we present a compromise between access and sample size: experiments involving mid-level elites, specifically local politicians. Experiments can benefit the study of elite decision making by testing causal processes in elite decision making directly, rather than inferring them from, for instance, self-reports or secondary data. Random assignment to different conditions can help to deal with strategic self-presentation by elites. Research shows that elite decision making processes have both similarities and differences compared to non-elites (Hafner-Burton, Hughes, & Victor, 2013; Haigh & List, 2005; Linde & Vis, 2017). It appears that selection processes surrounding their appointment, as well as experience of elites, might make them behave differently than non-elites. Using mid-level elites as experimental participants could enable researchers to explore the effects that experience might have on decision making, while allowing large sample sizes. By comparing these mid-level elites to a student sample, the effect of selection processes can be explored. In this paper we will review the literature on the use of heuristics in elite decision making, and evaluate the pitfalls and benefits of using mid-level elites as experimental subjects. To arrive at recommendations for further research, we will complement these insights with our experience in conducting a pilot study on the use of the availability and representativeness heuristics (Gilovich, 2002). Our study will use well-known decision situation, with a solid basis in prior, non-elite, research. We will then compare responses of Dutch local politicians from big cities against findings in prior research. To validate the experimental set-up, we will first test it on a student sample.