Graphical Illustration of Interaction Effects in Binary Choice Models: A Note Jason R.V. Franken, Joost M.E. Pennings and Philip Garcia 1 (Original submitted December 2016, revision received August 2017, accepted September 2017.) Abstract Graphing procedures for evaluating power or interaction terms in binary logit and probit models are illustrated in an application to hog producers’ decisions based on transaction cost economics’ hypothesised positive effect of the interaction of uncer- tainty and asset specificity on contract use. Results support the hypothesis, particu- larly for producers that are otherwise on the cusp (near the 50/50 probability) of choosing either contract or spot transactions based on their responses for other variables. Such insights may not be drawn without use of the demonstrated graph- ing procedures. Keywords: Asset specificity; binary choice models; contracts; transaction costs economics; uncertainty. JEL classifications: D86, C25, L14, Q13. 1. Introduction Agricultural economists often evaluate discrete choices of a binary nature, e.g. tech- nology adoption, conventional or organic production, contractual or spot exchange. The challenge for agricultural economists is that in the binary logit and probit models used to examine these discrete choices, conventionally reported marginal effects of power or product terms may not be statistically significant, while the variable in fact has potentially important and economically relevant impacts. The primary purpose of this note is to demonstrate procedures for graphically assessing the marginal effects of product and power terms in binary probit and logit models, and to raise awareness of these procedures among agricultural economists. With volatile prices and growing contract use, the capital intensive hog industry offers an interesting research context 1 Jason Franken is in the School of Agriculture, Western Illinois University, USA. E-mail: jr- franken@wiu.edu for correspondence. Joost Pennings is with the Department of Finance and the Department of Marketing, Maastricht University, the Netherlands, and with the Depart- ment of Marketing and Consumer Behavior, Wageningen University, the Netherlands. Philip Garcia is in the Department of Agricultural and Consumer Economics, University of Illinois at Urbana-Champaign, USA. Journal of Agricultural Economics doi: 10.1111/1477-9552.12257 Ó 2017 The Agricultural Economics Society