,
9 tweets,
3 min read
Read on Twitter
Little THREAD on academic research in economics and GAFAs.
Last night I asked myself: how much EMPIRICAL research has been published in ECONOMICS using data from GAFAs? The answer is: almost NONE.
Last night I asked myself: how much EMPIRICAL research has been published in ECONOMICS using data from GAFAs? The answer is: almost NONE.
Why GAFAs? Well, we talk a LOT about them in competition circles. They are among the very top for market capitalisation. They have TONS of data. This SHOULD attract the attention of economists! 
Selection criteria? Papers published in top 5 econ journals over the past 5 years (May 14-June 19) using microdata from GAFAs. A few THOUSAND articles. Why top 5? Because that’s what you need for an academic career. Why past 5 years? Because I had only 1 hour free...
Findings:
American Economic Review (excluding P&P): 1 (using data from G trends). Not addressing IO questions;
Econometrica: 0;
Journal of Political Economy: 2 (1 with FB data, 1 with G trends). Not addressing IO questions;
Quarterly Journal of Economics: 1 (G patents);
American Economic Review (excluding P&P): 1 (using data from G trends). Not addressing IO questions;
Econometrica: 0;
Journal of Political Economy: 2 (1 with FB data, 1 with G trends). Not addressing IO questions;
Quarterly Journal of Economics: 1 (G patents);
Review of Economic Studies: 1 (on prediction markets) plus 1 using Amazon Mechanical Turk. Not addressing IO questions.
There is a small bunch of (great) papers using eBay data, or Yahoo!, or Expedia – but these are not GAFAs.
What does this all mean? Various possibilities:
There is a small bunch of (great) papers using eBay data, or Yahoo!, or Expedia – but these are not GAFAs.
What does this all mean? Various possibilities:
1.I’m totally wrong, I missed tons – apologies then, but which ones?
2.Papers with GAFAs data are plentiful in other fields. Possible, I’m not an expert: I am referring to econ top 5.
3. No interesting questions about GAFAs that attract the attention of economists: implausible.
2.Papers with GAFAs data are plentiful in other fields. Possible, I’m not an expert: I am referring to econ top 5.
3. No interesting questions about GAFAs that attract the attention of economists: implausible.
4. Data are simply not made available for academic research. I think this is true and the source of the problem.
Mind you: these guys do give $ millions to some academics, but no data. Rather they sponsor policy agendas. E.g. this:
Mind you: these guys do give $ millions to some academics, but no data. Rather they sponsor policy agendas. E.g. this:
Or this:
assets.documentcloud.org/documents/2995…
(what bothers me is not the funding, but the lack of disclosure)
assets.documentcloud.org/documents/2995…
(what bothers me is not the funding, but the lack of disclosure)
Bottom line: it's very hard for a young researcher to have an academic career in economics studying these guys directly. That’s a big loss, since the smartest are then forced to look elsewhere.
How can we change this BAD equilibrium?
Ciao /FIN
How can we change this BAD equilibrium?
Ciao /FIN
