Skip to main content

Enjoying sunny San Diego while observing how economics and the AAEA are changing

David Zilberman, professor, agriculture and resource economics | January 8, 2020

San Diego boatsThe annual meeting of social scientists and economists (ASSA Meetings) was in San Diego, which was quite an improvement over previous years, where we met in the cold northeast. This was especially enjoyable, as I was remembering the meetings in Philadelphia, which were drastically underprepared for the snow that happened to occur during those meetings. The main attraction of the ASSA meetings is the job market. The ASSA meeting becomes a global job market for the social sciences, finance, and even political science. You can recognize job candidates by their impeccable dress and nervous look. Thank goodness that I don’t have to look for a job, or interview for that matter. There are some very entertaining talks with leading economists and policymakers, many technical sessions, and reunions of various academic associations. I was there to discuss a paper, meet friends, and meet with the board of Agricultural and Applied Economics Association (AAEA). I have done it now for three years, and it has been quite enjoyable. That being said, I am quite relieved that that was the last time.

Open access and the future of journals

The big change from an Association perspective is that academic literature is swiftly moving towards a world of open access journals. This would suggest that subscription revenues will decline drastically, and many journals will have to rely on government/philanthropic donation as well as charging publication fees directly to author institutions. Although publisher income will likely decline substantially, the move towards open access comes with significant social benefit as it transforms papers into public goods. I suspect that we may also see a reduction in the number of journals, as many will fail to raise enough revenue in an open-access world, though I’m not sure. All that being said, it is clear that AAEA income will decrease since AAEA receives a hefty payment from our publisher. So what do we do about it? We need to raise funds. One of the things we are doing is establishing appreciation funds honoring old members. Last year, we established five and this year, we received a significant contribution that will allow us to establish several more. We also look to obtain other resources for supporting our activities in the future, as we would like to enhance what we provide to our members, improve mentoring programs, and maintain high-quality journals.

At the conference, I learned that top economic journals are beginning to adopt a very rigorous process to verify results. In the future, most authors will have to submit data and follow specific procedures to allow readers to duplicate results. In principle, it is a positive strategic decision since people are concerned with irreproducible results and society may currently be less trusting of science. However, I think we have to go about this carefully, and think deeply about the consequences. More rigor, for example, may lead to exclusion. Left to its own devices, this policy will bias accepted literature towards a few universities with the most resources, so the procedure must be designed in such a way to prevent concentration and increase the breadth of publishing universities. On a related note, I personally worry about economics becoming homogenous. Our field has become quite enamored with big data and less concerned with ideas, modeling, and speculation. I personally believe that economics is about stories that must be supported, in part, by rigorous data. However, sometimes the pursuit of extra rigor leads people to forget about history, context and encourages people to focus too much on statistics rather than the story of economics.

Machine Learning

Speaking about data, one of the most interesting events in this meeting was an excellent talk by Susan Athey on machine learning and economics. She distinguishes that while computer scientists developed algorithms aimed to minimize prediction error, economists rely on theory, develop estimates that minimize bias, and use data to present hypotheses. With traditional economic models that previously relied on small sample sizes, economists were not able to differentiate between groups of individuals. This led to policies that were appropriate for the average person. However, averages rarely exist. Populations are heterogeneous. There are big gains when we are able to have differentiated policies that adjust to heterogeneity. The big trend that I applaud is that economists have access to big data sets that allow us to use different techniques to identify these sources of heterogeneity in the population, group people accordingly, and change policies that adjust for variability and evolve over time. This type of approach is consistent with my work on adoption, which posits that people don’t adopt technology at the same time because of this heterogeneity. Dynamic processes (e.g., learning) push adoption and changes in behavior all the time. Machine learning has particularly incredible potential in agriculture. A variety of remote and on-field data sources (e.g., satellites, GIS, field sensors) provides a huge amount of information that allows us to monitor crop response to varying conditions. We can use this information to change the application of inputs in varying conditions, which may reduce waste and allow us to get more with less. Of course, this vision will take a long time to implement completely, and we have to make sure that technologies will be accessible and affordable.

Machine learning can be the main source of social benefit, as long as we continue to keep the applications centered around the story of economics. To that end, we must recognize that data is more than numbers. And when we want to understand what is happening, we need to also study narrative. If we want to understand behavior, it is critical to interview people about their decisionmaking, as well as study what they have done previously (revealed preference). It is true that people will not always tell you the truth, but you can infer much of it in the same way that we use our numerical data to eliminate bias, errors, and deception. In the end, economics is about understanding human bboatsehavior and improving decision rules for consumers and for society.

Interdisciplinary collaboration

The AAEA started as an association of farm economists, and now encapsulates a much broader field of applied economists. As we continue to evolve, our challenge is to collaborate with other sister associations (environmental, energy, and health). Eventually, we could have larger joint meetings and overcome some of the silos to improve knowledge sharing across similar fields and techniques. Furthermore, I believe that as applied economists it would serve our discipline to be more multi-disciplinary and to understand deeply what is going on in related sciences (agriculture, social science, etc.). This is appearing to be a challenge for AAEA, and if we are able to provide support for members who want to reach out to other collaborators outside our field, we will be stronger and more relevant in the face of changes in policy and technology.