In the Year 2030, How Will Applied Microeconomists Study the Consequences of the 2020 COVID Shock?

An active research field in climate economics uses natural disasters as "natural experiments" to learn about the economic effects of such shocks.  Amine Ouazad and I study how major hurricanes impact bank lending and securitization patterns in this study.   Gary Lin, Rhiannon Jerch and I study local public finance dynamics in the aftermath of disasters in this study.    In both papers, our estimation strategy relies on the fact that we have a "treatment group" and a "control group".   Intuitively, there are places that are roughly similar that have and have not been recently hit with a shock.  This allows us to use standard statistical methods to estimate the impact of the shock on those places that did get hit.  An assumption in such work is that "general equilibrium effects" can be ignored so that the set of places not hit by the shock are not affected by the responses in the places that were hit by shock. Intuitively, if all of the people in the places hit with a hurricane move to the areas not hit by the hurricane, then this is an example of an enormous general equilibrium effect!   

The 2020 COVID shock is not a localized natural disaster.  All of us have been hit and hit on multiple dimensions.  Think of yourself in the year 2030.  How will the 2020 shock's effects persist in affecting your future quality of life?   Ask the counter-factual, who will you become given that we have experienced this shock? Who would you be if this shock had never occurred?  Applied economists try to recreate these two paths to study causality.  In this case, how will we proceed with this important research?

I will turn 65 in 2030 so the 2020 shock (assuming I am still alive) will have little impact on my earnings prospects but my taxes are likely to be higher at that point.    My son will turn 30 in 2030.  How will his life be affected by the shock of 2020?  How much less human capital will he have because of his schooling disruption?  How will the labor market be affected in the medium term because of the restructuring of firms that may occur? How will the new pattern of COVID induced transfers and taxes affect his career investments?  

How will structural labor economists model how the COVID shock influences human capital decisions, savings decisions and work decisions?  

To be specific, suppose that the marginal tax rate on high income earners rises over the next few years.  Will researchers who seek to study the labor supply implications of this incentive change pool data from 2010 through 2030 to study how this policy change affects labor supply?  What is the right control group here?   If a researcher has access to individual level IRS income data, she can include a person fixed effect in an earnings regression and study how the average person's earnings varies with changes in the taxes (that rose in the post COVID economy) but that is assuming that the error term is uncorrelated with the change in taxes and that doesn't sound kosher.  The dynamic error term in such a regression will feature many key contemporaneous time trends about the local macro economy.

COVID 2020 both changes us as people (in terms of health capital and maybe even aspirations and imagination), changes our portfolio's worth and changes the tax rules we face going forward.  Given that economists are greatly interested in persistence, how will we model the persistence of the COVID 2020 shock when we don't have a control group?  

We can follow Doug Almond's 1918 flu methodology and estimate dummy variables such as "were you at a critical age in 2020" and make relative comparisons for people of different ages in the year 2030 about their well being.  For example, if you have a 5 year old child in 2020 and this kid has spent more time at home with her WFH parents --- does this kid at age 15 in 2030 now achieve more relative to siblings who were 10 in 2020 and lost schooling time?   

FINALLY,  a technical point.  I asked Dora whether in the year 2030 the QJE will receive reduced form papers that break the sample into "BEFORE 2019" and "AFTER 2022" and then do a Chow Test to test for coefficient stability.

If the Null Hypothesis of coefficient stability is not rejected, will such QJE paper authors be brave enough to claim that COVID2020 doesn't impact their study's dynamics and they will proceed to pool their data?

When we reject this null hypothesis, do all papers using data from before 2020 become Economic History as we are in a "new economy" because a major non-stationarity that we can't really model has occurred?

In this case, the standard errors in our papers will grow larger at a time when due to p-hacking --- the critical value in studies is getting smaller. This will mean that very few scholars will have statistically significant findings to report.   

Note that structural applied micro economists will face an awkward issue at RESTUD and JPE explaining why their "deep" time invariant structural parameters are changing over time (before and after COVID).  A sober referee may ask the researcher to provide a microfoundation for such dynamics.