Marc F. Bellemare Posts

I unfortunately have too little time for a proper post this week, but I wanted to make time for a quick post. A grad-school friend and colleague sent a link to an interesting new(-ish) paper by Kahn and Whited (2017) that has been making the rounds in finance, and which is forthcoming in the Review of Corporate Finance Studies.

The title of the article is “Identification Is Not Causality, and Vice Versa.” Here is the abstract:

I was in Helsinki last week for the UNU-WIDER workshop on the Vietnam Access to Resources Household Survey (VARHS) data, presenting work that my coauthors and I have been doing using these data.

One thing that I saw a few instances of during the workshop was the following. A researcher wants to a variable x in a regression, but that variables needs to be logged. Because there are many zero-valued observations of x, and because log(0) is undefined, the author simply uses log(x +1), or log(x + 0.001), or log(x + 0.00001), and so on.

This post is about what to do in such cases. There are many instances in development where you’d like to include a financial variable–say, the value of chemical fertilizer used on a given plot, for example–where many observations will have a zero-valued observation–in the chemical fertilizer example, not everyone in the data will use chemical instead of organic fertilizer, and so they will report a zero when you ask them what was the value of chemical fertilizer used on any of their plots.

When you want to log a variable x but that x has many zero-valued observations, there are three things you can do in principle:

(Update: Thanks to my colleague Jason Kerwin for writing, late in the day, that there was no link to the paper. The post now includes a link.)

That is the title of a new working paper I have with Stacie Bosley, from Hamline University here in St. Paul, and two undergraduate coauthors, who are also from Hamline.

Here is the abstract:

Consumer financial fraud is costly to individuals and communities yet academic research on the subject is scarce, in part due to how difficult it is to find reliable data. Using a lab-in-the-field artefactual experiment, we study judgment and decision-making as well as the correlates of victimization in a prototypical pyramid scheme fraud. We record demographic, psychological, cognitive, and behavioral characteristics for 452 subjects at the 2017 Minnesota State Fair, and we estimate the impact of an information treatment—specifically, a reminder to pay attention to the odds of winning or losing—on our subjects’ behavior in relation to pyramid scheme fraud. Our results indicate that this straightforward, simple treatment reduces fraud uptake, but only for subjects with a post-secondary education. Our findings show correlates of victimization beyond cognitive ability, including impulsivity, risk preferences, religiosity, and prior exposure to pyramid scheme fraud. Subject reliance on probabilities in decision-making and the accuracy of subjective expectations are the most statistically significant predictors of the decision to invest in a fraudulent pyramid scheme. Our results can help inform the targeting of consumer protection interventions as well as the potential content of those interventions.

My interest in the topic comes from my interest in behavior in the face of risk and uncertainty as well as in lab-in-the-field experiments, which I have previously run in Peru to study the behavior of farmers in the face of output price risk. And given that we got access to the University of Minnesota’s Driven to Discover building at the Minnesota State Fair, it was nice to run experiments by recruiting from a pool of subjects that is more representative than that of the usual lab experiment, which typically consists of college students.