Statistician Analyzes Global Financial Integrity's Illicit Financial Flow Study, Finds It Is Not Fake
January 18th, 2012
January 18th, 2012
Benford’s law states that in many data sets the overall distribution of the significant digits tends to be logarithmic so that the occurrence of numbers beginning with smaller first significant digits is more often than those with larger ones. We investigate here recent data on illicit financial flows from developing countries and reveal that the data does submit to Benford’s law. Further, the general improvement in the statistical accuracy which we observed supports the applicability of the normalization process used to limit the inclusion of the countries in the database for which the illicit financial flows are not substantial.
Readers of U.S. political blogs may remember Benford’s Law from when statistician Nate Silver called into question whether or not one polling firm was fabricating polling results. Tariq Ahmad Mir took Global Financial Integrity’s reports on illicit financial flows from developing countries and put them to the statistical test. The result?
“The failure of a data set to follow Benford’s Law arouses suspicion not only about its authenticity but also the process involved in its generation which in turn necessitates further research for ascertaining the quality of the data… tendency of the data to follow the law might be an indication of its truthfulness… We investigated the validity of Benford’s law for the recent data on illicit financial flows from developing countries and found the observed frequencies of the significant digits to be in accordance with the predictions of the law.”
Global Financial Integrity checks out! I wonder how the books of a random selection of shell corporations based in the Cayman Islands would do?