Wednesday, September 26, 2012

Statistical evidence that elections in Russia and Uganda were fixed

Wired alerts us to a paper by a team of Austrian statisticians suggesting that recent elections in Russia and Uganda were riddled with fraudulent voting practices.

The study focused on hunting down anomalies in regional voting patterns, rather than on larger pools of data as is the usual trend. By focusing on regional activity, patterns clearly emerged indicating that when a high voter turnout in a specific area was combined with a high consensus for a particular candidate, more than a little ballot-fiddling had gone on - namely, the introduction of large numbers of false votes and the destruction of real ones. Or ballot stuffing, as it is more plainly called.

The paper's authors say: "We show that reported irregularities in recent Russian elections are, indeed, well-explained by systematic ballot stuffing. Extreme fraud corresponds to reporting a complete turnout and almost all votes for a single party."

Statistical graphs compared voter turnout in specific geographic regions with the percentage of individuals in that region that voted for the winning candidate. Plotted out, these results appear as an 'election fingerprint', where anomalies in the data are immediately evident. Russia and Uganda's fingerprints are both smudged rather than arranged in an orderly cluster - the hotspots represent areas where nearly everyone voted, and all for the winner.

Read the full paper in the Proceedings of the National Academy of Sciences.

No comments: