Money and Markets: Essays in Honor of Leland B. Yeager
Money and Markets: Essays in Honor of Leland B. Yeager
Money and Markets: Essays in Honor of Leland B. Yeager
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
6 Good ideas <strong>and</strong> bad regressionsThe sad state <strong>of</strong> empirical work <strong>in</strong>public choiceSteven B. Caudill *It was the late 1990s. As I walked down the hallway my young colleague, JohnWells, a times-series econometrician, was st<strong>and</strong><strong>in</strong>g <strong>in</strong> his <strong>of</strong>fice doorway mutter<strong>in</strong>gsometh<strong>in</strong>g about an “<strong>in</strong>tervention analysis.” I asked, “John, isn’t an <strong>in</strong>terventionwhen your family <strong>and</strong> close friends get you <strong>in</strong>to a room <strong>and</strong> confront you aboutestimat<strong>in</strong>g bad regressions?”(S.B. Caudill)ForewordWe face an epidemic <strong>of</strong> bad empirical work <strong>in</strong> economics <strong>and</strong> much <strong>of</strong> it can befound <strong>in</strong> the public choice area. My goal is to characterize <strong>and</strong> criticize a cavalierapproach to empirical work that I first noticed <strong>in</strong> papers <strong>in</strong> the areas <strong>of</strong> publicchoice <strong>and</strong> the sociology <strong>of</strong> economics. I <strong>in</strong>clude the latter because much <strong>of</strong> thework <strong>in</strong> the sociology <strong>of</strong> economics is an extension <strong>of</strong> public choice <strong>and</strong> many <strong>of</strong> thesame researchers are <strong>in</strong>volved. If my characterization <strong>of</strong> public choice is “goodideas <strong>and</strong> bad regressions,” then my characterization <strong>of</strong> the sociology <strong>of</strong> economicsis either “worse ideas <strong>and</strong> bad regressions” or “no ideas <strong>and</strong> bad regressions.”There are no theories <strong>in</strong> the sociology <strong>of</strong> economics. Both public choice <strong>and</strong> thesociology <strong>of</strong> economics employ a common approach to empirical research that Ihenceforth refer to as the PC approach.Despite the suggestions <strong>of</strong> my colleagues, I do not plan to “name names” so that,<strong>in</strong> the words <strong>of</strong> my market<strong>in</strong>g colleague, Herbert Rotfeld, “more people will feelguilty.” The lone exception I make is to exam<strong>in</strong>e, <strong>in</strong> detail, the bad regression <strong>in</strong> thepaper by Lab<strong>and</strong> <strong>and</strong> Taylor published <strong>in</strong> Economic Inquiry <strong>in</strong> 1992 entitled, “TheImpact <strong>of</strong> Bad Writ<strong>in</strong>g <strong>in</strong> Economics.” I use this paper <strong>and</strong> identify the authorsbecause the econometrics there<strong>in</strong> has previously been criticized <strong>in</strong> pr<strong>in</strong>t by McCloskey(1992). I merely echo McCloskey’s criticisms to evaluate my characterization <strong>of</strong> abad regression.In this Foreword I wish to <strong>in</strong>dicate why I, <strong>in</strong> particular, am bothered by the PCapproach to empirical work. The explanation is found <strong>in</strong> my background <strong>and</strong> tra<strong>in</strong><strong>in</strong>gas an economist, the general nature <strong>of</strong> PC empirical work, <strong>and</strong> my experienceteach<strong>in</strong>g econometrics courses to PhD students at Auburn University.