The Economist Magazine has changed its tone since its new ownership took over. Another example is presented here in a piece that argues that applied microeconomics is a mess. The article makes some reasonable points as it highlights that teasing out cause and effect is difficult. The piece throws a strange punch at Gary Becker as the author enjoys playing the role of the critic.
Sherwin Rosen said many things to me. One thing he said was "Don't be an assassin". For the field of economics to make progress, we must focus on constructive ideas that improve our existing machinery. We are not going to throw away our machinery.
I want to make several points;
1. Applied micro researchers have access to more data than ever and have access to more computing power and more easy to use and sophisticated econometric methods than ever before. Improved canned software in Stata allows applied researchers to relax many of the statistical assumptions that researchers made in previous years. Such "robust" estimates allow us to march towards learning the truth.
2. Due to "natural experiments", discontinuities, and explicit randomizations, we now have more variation in "cause variables" (the X's) than ever before.
3. The advent of Google and the rise of Economics in Europe and outside of Western nations means that the current set of applied micro researchers are aware in "real time" about what findings are emerging in the top 5 journals and NBER and IZA and CEPR working papers. Now that there are so many applied micro economists working around the world, this competition fosters innovation and progress.
4. Replication is rising as an important piece of our advance as a "science".
5. Leading firms such as Amazon highly value quantitative training. Undergraduates are aware of this and they are investing in the math/computer programming and economics and stats training to have the option to pursue this. Some of these young people will opt into doing a PHD in economics and applied micro grows stronger due to this influx of talent.
6. Thanks to scholars such as Raj Chetty, the power of using administrative data (such as IRS tax data) are now more clearly seen all over the world. I expect that more government officials who "know that they do not know" the answers for unlocking economic development will increasingly partner with the J-PAL and other economists to help them to experiment and learn. This is Hayek as applied microeconomist at its best.
All of these points make me highly optimistic about our future.
My Concerns;
1. Applied microeconomists are shrewd about marketing. Before we write a paper, we know how it likely will be interpreted by the media and the bloggers. A tenured professor who has certain political leanings can step "towards the line" by writing the paper in the way to raise the probability that his/her work triggers a media response. We see this all the time in environmental economics. As the line between research and activism blurs, economists must be very careful here.
2. The war on the perfect competition model. Many of our predictions such as the effects of a minimum wage are based on assuming perfect competition in the labor market. Many economists such as Joe Stiglitz argue that there is pervasive market power in labor markets and output markets across our economy. The presence of market power (that firms have price setting ability) has key implications for applied micro studies. Such market power means that we can only make progress in applied work if there is a much closer connection between game theory and strategic play and applied estimation research (no more "taking prices as given" implicit assumptions).
3. Some economists have "career concerns". In our 30s, we focus on research excellence while later in life many of us begin consulting or seeking government jobs. Do such expected career trajectories have implications for the applied micro research that scholars conduct while young?
The good news with respect to #3 is that in an age of replication and in an age where young economists can make a name for themselves if they successfully challenge an earlier study that this dynamic possibility disciplines "bad behavior" by the current generation.
4. The Lucas Critique continues to be ignored. Economics has no physics constants. The economy is always evolving. The Austrians were right about this. In an economy featuring forward looking optimizing decision makers, what do "reduced form" coefficient estimates mean? Which structural estimates are "structural"? Research in climate economics is plagued by this. Past correlations between extreme heat and economic growth tell me little about the future relationship between these variables and I have argued that the past correlation plays a causal role in leading to a smaller relationship between these variables in the future. Why? If we have learned from the past experience that extreme heat hurts our economy, and if we expect climate change to cause greater heat in the future, we will invest to attenuate the future effect of heat on economic growth.