Election Recount Unlikely To Reveal Evidence Of Vote System Hacking

Putting aside the near impossibility of a successful hack in elections run by localities, the statistical variances that so concern Prof. Halderman are easily explained by the demographics of the areas involved, as has been pointed out by the staff of political polling aggregation website FiveThirtyEight.

This group, started in 2008 by polling expert Nate Silverman, explained how the statistical differences disappear when demographics are taken into account in a Twitter post.

At this point, Wisconsin state board of elections will have to decide whether to honor Stein’s request for a recount. They can decline, and in addition the winning candidate Donald Trump, can object to the recount.

Unlikely event

The problem is that there’s not really any way to recount the election results on the electronic voting machines used by Wisconsin, and if the paper ballots don’t show any evidence of tampering, then there’s not much more that can be done.

The statistical analysis in Michigan is even more problematic. The reason is that Michigan doesn’t use electronic voting machines. All voting in Michigan is done with paper ballots.

The only way for the Russians to tamper with the votes there would be to send thousands of Russian imposters into the state to pretend to be someone else and then vote. Chances are, an invasion of Russian imposters would be noticed.

Thus it appears that the recount that’s being demanded by Stein is an exercise in futility. While it’s conceivable that it might uncover some tampering or other vote fraud, there’s virtually no chance that it will change the outcome of the election.

First, the elections in all three states would have to be overturned, which isn’t likely at all, for Trump to lose the electoral vote. But in the meantime it serves to add at least some doubt about the validity of the election during the transition and is sure to provide fodder for the most credulous conspiracy theorists for years to come.

Second, it depends on the Green Party being able to afford to pay for the recount. The cost is considerable and the Green Party has yet to raise the necessary funds. Furthermore the recount can be challenged by the Trump campaign and any changes in the count are just as likely to go in the direction of Trump as Clinton.

What’s being demonstrated here is that statistics can be a powerful tool for finding the truth, but if you don’t know how to apply that tool, then you won’t find the truth. In the case of the current recount effort, it appears that a knowledge of statistics isn’t enough.

The truth also lies in the application of the statistics and that can make all the difference. Big data after all isn’t magic—the magic is in the knowledge of how to use big data, not in the data itself.

Originally published on eWeek