Unmitigated Disasters JIE 2024

But it is the uninsured losses that drive the macroeconomic cost; insured losses are less consequential in the aggregate, and can even stimulate growth. By helping to finance the recovery, insurance mitigates the macroeconomic cost of disasters. Many countries lack the capacity to (re)insure themselves and would stand to benefit from international risk sharing.

The Stata code referenced in this replication package produced the original findings of the article cited. They mainly consist of estimated growth responses to scaled disaster impulses in a dynamic panel setting, allowing for four types of natural disasters and distinguishing between insured and uninsured losses. The code enables the exact replication of our results when using the Munich Re NatCat statistics presented in the JIE paper.

For reasons of confidentiality, this replication package contains no data from Munich Re. The code can be run using public data, such as the data included in this replication package based on the EM-DAT database (publicly available at public.emdat.be). The datasets included in this package are sufficient for running the code, but the quantitative results do not replicate the estimates in the paper and should not be relied upon for other purposes.

For background and detailed guidance, see the READ_ME file and other materials included in this replication package.