Step 05

Analysing the results and discussing the mechanisms

The numbers, what they mean, and why.

I regress on different populations and state subgroups to better understand what's happening.

Effects of climate shocks on casual labor force participation

+/- indicates an associated increase/decrease in informality following a shock, all lags included. The number of signs indicates the effect's strength as a combination of significance and magnitude. Conventional levels are used (∗p<0.1; ∗∗p<0.05; ∗∗∗p<0.01). For detailed scientific results, refer to the manuscript.

SubgroupDroughtsFloods
General population
+++non significant
High-income states
−−−
Low-income states
+++
Agricultural states
−non-significant
Less agricultural states
+++−
Male workers
++non significant
Female workers
+++non significant
Rural areas
++non significant
Urban areas
++not very significant (+)

I also run regression to look at the effects of climate shocks on a handful of labor market measures (unemployment, employment, labor force participation) for each of those subgroups, and at the relative size of each industry in relation to climate shocks.

Those analyses serve as a reminder that informality is not the whole picture and motivate the emphasis on the role of opportunity in the discussion. I use those insights to interpret the results in the Conclusion. 

As always, full results are in the manuscript.