«

»

Sep 26

Print this Post

Using Fatalities to Predict Humanitarian Aid

Last Tuesday, an earthquake of magnitude 7.7 shook with such violence that it created a new island off the coast of Pakistan, as well as tragically killing at least 515 people and affecting at least 300,000 more. The international community has attempted to provide relief and humanitarian assistance. A 2013 social science paper indicates that number of fatalities is the key consideration when predicting the amount of aid that is provided in the aftermath of natural disasters. The paper employs an advanced econometric technique to arrive at its conclusion: for every additional fatality reported, the international community responds with about $9,000 of additional aid. The number of people affected, in contrast, was found to be irrelevant in predicting aid flows. Based on this finding and the number of fatalities reported, we would expect donations to total roughly $463,500.

The estimate above is in part based on the assertion that the paper’s model considers all of the factors that are correlated with donations and fatalities. If some of these factors are excluded, the model is unreliable because it suffers from what econometricians call “omitted variable bias.” For this very reason, the model does include several important items such as the disaster type and location. However, missing from the analysis are variables like the recipient country’s GDP, population, and population density. Without performing the analysis, we don’t know for sure whether GDP for example is correlated with either donations or fatalities. I can think of plausible arguments why richer countries might receive different (i.e. systematically greater or lesser) donation amounts. Put simply, further analysis may be appropriate before accepting this paper’s conclusions.

What’s unfortunate with respect to the Pakistani earthquake is that even if we take the study’s conclusions as given, the above estimate of total donations probably overstates what ultimately is received. We know this because relief efforts have been stymied by attacks on security officials and aid groups. Such obstructions will almost certainly reduce the amount of aid delivered.

About the author

Benjamin Bohr

Benjamin is a manager at Fulcrum Financial Inquiry. He specializes in statistics and data analysis for use in high-stakes litigation, business valuations, and other financial matters.

Permanent link to this article: http://betweenthenumbers.net/2013/09/using-fatalities-to-predict-humanitarian-aid/

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>


*