Andrew B. Martinez
My research interests span applied macroeconomics, econometrics, international economics, and time series econometrics.
I will be available for interviews at the EEA meetings in Naples, the RES meetings in London, and the ASSA meetings in Atlanta.
Can forecasts of natural disasters alter their destructiveness? Poor forecasts increase damages when individuals do not mitigate risks based on the false belief that they will be unaffected. We test this hypothesis by examining the impact of 12-hour-ahead forecasts on hurricane damages and find that larger errors in the storm's predicted landfall location lead to higher damages. The cumulative reduction in damages from forecast improvements since 1970 is about $82 billion. This exceeds the U.S. government's spending on these forecasts and private willingness to pay for them. The benefits from forecast improvements are underestimated and individual adaptation decisions matter.
The trajectory and path of future outcomes play crucial roles in policy decisions. However, analyses of forecast accuracy typically focus on the point forecasts. Examining the path forecasts provides additional insight into the forecast dynamics. I propose a test for differences in path forecast accuracy using the link between path forecast evaluation metrics and the joint predictive density. This path test nests and extends existing joint multi-horizon forecast accuracy tests. Simulations highlight that there are trade-offs when using path forecast accuracy tests to detect a broad range of differences in the forecasts. I compare the Federal Reserve's Greenbook path forecasts against four DSGE model forecasts. The results show that differences in the forecast dynamics play an important role in the assessment of path forecast accuracy.