Saturday, October 29, 2016

Risk and Adaptation: Evidence from Global Hurricane Damages and Fatalities

We examine whether countries adapt to hurricanes. A spatially refined global tropical cyclone data set is created to test for adaptation. We find evidence of adaptation in most of the world by examining the effects of income, population density, and storm frequency on damage and fatalities. In contrast, there is no evidence of adaptation to damage in the United States, leading to a damage function which is 14 times higher than other developed (OECD) countries.
Over the last decade, the average annual global damage from tropical cyclones (hurricanes) has reached $26 billion dollars with 19,000 lives lost (Mendelsohn et al., 2012; CRED, 2012).1 These losses measure the remaining damage and fatalities given existing adaptation. But what would the damage and life lost have been if no adaptation had occurred? What evidence exists that there have been any adaptations? This paper seeks to quantify the answers to these questions from an empirical analysis of the damage and fatalities caused by individual tropical cyclones from around the world
For the empirical analysis, we build an original dataset of more than 1,400 storm landfalls around the Earth from 1960 to 2010. The cumulative damage from these storms totals almost $0.75 trillion and approximately 400,000 lives lost. We drop observations from before 1960 because global observations of earlier storms were more erratic and estimates of damage and lives lost were unreliable (HRD, 2014). Hsiang and Jina (2014) correctly note that 6,700 storms have been recorded by humans since 1950, but many of these storms do not make landfall and fewer still can be linked with direct economic damage or human fatalities. Thus, our dataset represents the full record of storms between 1960 and 2010 that can be matched with publicly available damages and fatalities.
We ... calculate what the American damage would be if we used the damage coefficients for the other-OECD regression. The damage per storm would fall from $2.0 billion to $166 million, over an order of magnitude drop.  The aggregate annual damage from tropical cyclones in the United States would fall from $15.3 to $1.2 billion. We also estimate the damage in the other-OECD countries if they had the damage function of the United States. The damage per storm in the other OECD countries would increase from $231 million per storm to $9.4 billion, a forty fold increase. There is no question that the United States is far more vulnerable to tropical cyclones than other similar countries, controlling for the frequency and intensity of storms and for income and population density....

This paper develops a theory of adaptation to tropical cyclone damage and fatalities in order to test for the presence of adaptation. The theory provides several new ways to test for adaptation by comparing empirical parameters of damage and fatality regressions against hypothetical parameter values one might expect if there was no adaptation. We also test hypotheses raised in the literature concerning the adaptation response by areas that are more frequently threatened.

We then conduct an empirical analysis based on a new spatially-explicit historical dataset of more than 1,400 storms that have caused serious damage or lives lost over the last 50 years. We match the damage and fatalities of each storm with the cyclone characteristics at landfall and the characteristics of the impacted local area struck. A set of multiple regressions are then estimated with this new dataset to test for adaptation. Several tests are undertaken. First, we look at economic damage and explore whether the elasticity of income and to a lesser extent the elasticity of population density is unitary. Second, we look at fatalities and explore whether the elasticity of income is negative and whether the elasticity of population density is less than unitary. Third, we test if there is a negative relationship between impacts and the underlying storm frequency. We decompose these main findings into adaptation across levels of development and urbanization. The primary evidence of adaptation lies in the income and population density results. There is also evidence of a small adaptation effect with respect to frequency. The actual damage and fatalities are much less than potential damage and fatalities as if no adaptation measures were being undertaken.

There are useful policy insights from these results. Most of the world appears to have taken effective precautions to reduce fatalities. Overall, fatalities from tropical cyclones have fallen over time, suggesting an increasingly effective adaptation. As incomes rise, fatalities fall. As population density increases fatalities do not increase proportionally. Places with more frequent high intensity storms manage to reduce fatalities. Unfortunately, places with more frequent low intensity storms have slightly higher fatalities per storm.
The results strongly suggest that higher income increase adaptation. As incomes rise, damage does not rise proportionally. The income elasticity of damage is less than one in every country except the United States. As countries develop, the damage from tropical cyclones will be a smaller component of their income.
One final insight concerns the damage function of the United States versus the rest of the world. Since 1990, the United States has been responsible for 60% of global tropical cyclone damage even though it is struck by only 4% of global storms. The reason appears to be the damage function of the United States. The income elasticity in the United States is unitary (the income elasticity in all other countries is less than one). This result implies there is no adaptation in the United States. The constant term and the elasticity with respect to intensity are also significantly higher than in other countries....

by Laura A. Bakkensen and Robert O. Mendelsohn
Journal of the Association of Environmental and Resource Economists
Volume 3, Number 3; September, 2016; Online July 12, 2016
Keywords: Adaptation, Economic damages, Hurricanes, Natural disasters, Risk, Tropical cyclones.

No comments:

Post a Comment