Comments on the AEA code of conduct
I write to express my general support for the Draft AEA Code of Professional Conduct and to add to it. While the inclusion of age, gender, race, and ethnicity seem obvious and understandable, I find the omission of “political ideology” to the list of attributes troubling. While one’s sexual orientation or disability status or genetic information are unlikely to have a bearing on one’s views of say, the minimum wage on employment, one’s political views are likely to have a bearing on that question. One has to only take a casual look at the stream of research on minimum wages published by the Institute for Research on Labor and Employment (IRLE) and UMass Amherst to admit of the possibility that the ways those scholars view the world influence their findings or, at a very minimum, put them at odds with other scholars who also work in the area (Fn. 1).
This discussion by the AEA on a professional code of conduct for our profession also seems to be taking place simultaneously with discussions of how diversity is important for deciding on who we interview and who we bring in for campus visits. However, such discussions seem to be dominated with an emphasis on gender and ethnic diversity, which to me, seems awfully limited. After all, at a university, the kind of diversity we ought to care most about is the diversity of thought and as Nicholas Kristof put it in a recent piece in the NY Times, “Universities should be a hubbub of the full range of political perspectives from A to Z, not just from V to Z” (Fn. 2).
Along those lines, let me offer a personal anecdote. Before pursuing a PhD, I worked at a consulting firm for two years. At the very start of my job, we went through a training and had colleagues from all over the world in the same room and I was struck by how similar they sounded, regardless of where they were from and how they looked. That same experience was repeated over and over again including at a 2-week workshop where, as before, we had a diverse group from all over the world (including someone who identified as gay, an African-American, and a close-to-even split between males and females). Yet at the very end of the workshop someone remarked – “Now I know I have a group of 25 friends who think just like me.” I could offer other examples from my experience but I hope that they convey the point I am trying to make – diversity has got to be more than skin deep.
I finally want to touch on another issue that’s less directly tied to the code of conduct per se and more tied to the issue of integrity in research and how we as a profession go about practicing it. Everyone reading this is likely to have come across the well-known study “Are Emily and Greg More Employable than Lakisha and Jamal?” While it is one of my favorite papers and I have nothing but the highest respect for both authors, it is an open question that I lay out there – would this study have received the attention it did if it had reported a null set of results as opposed to the results it did, holding constant everything else about the paper? While there is no magic bullet that I can think of which would address all my concerns related to how research is conducted, I think pre-registering studies would help. (Fn. 3) The requirement that all studies published in an AEA journal provide their data and code helps too (I have downloaded such data/ code more than once and used it for examining my own work) but the AEA could encourage all journals (including non-AEA ones) to adopt such a policy. Finally, encouraging studies that come up with null results to be also submitted is desirable. After all, if the requirement for every audit study examining discrimination to be published is that it find effects in the expected direction, then a true understanding of the extent and degree of discrimination will elude us.
Fn. 1: See Isaac Sorkin “Are there long-run effects of the minimum wage?,” (RED, 2015) or Ekaterina Jardim et al. “Minimum Wage Increases, Wages, and Low-Wage Employment: Evidence from Seattle” (NBER Working Paper No. 23532) for recent examples of research that run counter to the stream of research published by the IRLE.
Fn. 2: https://tinyurl.com/hae5bwu
Fn. 3: https://tinyurl.com/jdf3u94
This discussion by the AEA on a professional code of conduct for our profession also seems to be taking place simultaneously with discussions of how diversity is important for deciding on who we interview and who we bring in for campus visits. However, such discussions seem to be dominated with an emphasis on gender and ethnic diversity, which to me, seems awfully limited. After all, at a university, the kind of diversity we ought to care most about is the diversity of thought and as Nicholas Kristof put it in a recent piece in the NY Times, “Universities should be a hubbub of the full range of political perspectives from A to Z, not just from V to Z” (Fn. 2).
Along those lines, let me offer a personal anecdote. Before pursuing a PhD, I worked at a consulting firm for two years. At the very start of my job, we went through a training and had colleagues from all over the world in the same room and I was struck by how similar they sounded, regardless of where they were from and how they looked. That same experience was repeated over and over again including at a 2-week workshop where, as before, we had a diverse group from all over the world (including someone who identified as gay, an African-American, and a close-to-even split between males and females). Yet at the very end of the workshop someone remarked – “Now I know I have a group of 25 friends who think just like me.” I could offer other examples from my experience but I hope that they convey the point I am trying to make – diversity has got to be more than skin deep.
I finally want to touch on another issue that’s less directly tied to the code of conduct per se and more tied to the issue of integrity in research and how we as a profession go about practicing it. Everyone reading this is likely to have come across the well-known study “Are Emily and Greg More Employable than Lakisha and Jamal?” While it is one of my favorite papers and I have nothing but the highest respect for both authors, it is an open question that I lay out there – would this study have received the attention it did if it had reported a null set of results as opposed to the results it did, holding constant everything else about the paper? While there is no magic bullet that I can think of which would address all my concerns related to how research is conducted, I think pre-registering studies would help. (Fn. 3) The requirement that all studies published in an AEA journal provide their data and code helps too (I have downloaded such data/ code more than once and used it for examining my own work) but the AEA could encourage all journals (including non-AEA ones) to adopt such a policy. Finally, encouraging studies that come up with null results to be also submitted is desirable. After all, if the requirement for every audit study examining discrimination to be published is that it find effects in the expected direction, then a true understanding of the extent and degree of discrimination will elude us.
Fn. 1: See Isaac Sorkin “Are there long-run effects of the minimum wage?,” (RED, 2015) or Ekaterina Jardim et al. “Minimum Wage Increases, Wages, and Low-Wage Employment: Evidence from Seattle” (NBER Working Paper No. 23532) for recent examples of research that run counter to the stream of research published by the IRLE.
Fn. 2: https://tinyurl.com/hae5bwu
Fn. 3: https://tinyurl.com/jdf3u94
posted by sutirthabagchi | 2:31 PM
|
0 Comments
