Tuesday, November 08, 2016

The Smell Test for a Rigged Election

By Marcus Ahn

In a normal election year, questions about rigging the vote would be dismissed as a matter of tinfoil hat and conspiracy mongering. But this, of course, is not a normal election, and per a recent Reuters report, nearly 70% of Republicans believe a Clinton victory would be because of illegal voting or vote rigging. So, will the vote be “rigged” in favor of Secretary Clinton? Is there even a way to assess the prospect of foul play?
 Yes, there is. While certainly not foolproof, a decent proxy of whether or not the vote is indeed rigged will be Trump’s relative performance in Michigan vs. Pennsylvania and New Hampshire vs. North Carolina. To understand why, it’s important to first review some background information.
Integral to claims of vote rigging are electronic voting systems, or Direct Recording Electronic (DRE) Systems. Some swing states like Pennsylvania and swaths of North Carolina have these DRE systems, while other states like Michigan and New Hampshire have traditional paper ballots. Rigging the vote in a state with paper ballots would be difficult given the physical medium -- that isn’t to say that chicanery can’t take place, but it’s a lot harder. With a DRE system, especially in states like Pennsylvania where there is no paper trail or physical record of the votes cast, the prospect of tampering is far more viable and, more importantly, scalable. With a paper ballot, a corrupt official could stuff a few extra ballots, but with a DRE system, a motivated or well-informed hacker could stuff a few hundred thousand. 
Michigan vs. Pennsylvania
Michigan has a traditional paper ballot while Pennsylvania has an electronic system with no paper trail. Neither state allows early voting. Trump has consistently polled better in Pennsylvania than Michigan—the Real Clear Politics poll of polls average has Trump down by just 2.4% in Pennsylvania compared to 4.7% in Michigan, while Nate Silver’s 538 blog pegs Trump’s chances of winning Pennsylvania at 23.2% compared to a 21.5% chance of Trump winning Michigan. Basically, all things being equal, Trump should perform better in Pennsylvania than Michigan. The key difference, of course, is the discrepancy in voting equipment.
If Trump outperforms in Michigan and underperforms in Pennsylvania, it should cause some concern with respect to the smell test.
New Hampshire vs. North Carolina
New Hampshire has a traditional paper ballot while swaths of North Carolina use an electronic system, or DRE. The comparison is not as straightforward as that of Michigan and Pennsylvania, because North Carolina permits early voting while New Hampshire does not. But nonetheless, all things being equal, Trump appears to have a substantially better chance of winning North Carolina than New Hampshire. Trump has consistently polled better in North Carolina; the Real Clear Politics poll of polls average has Trump up by 1.4% in North Carolina, while he trails by 0.6% in New Hampshire. Nate Silver’s 538 blog pegs Trump’s chances of winning North Carolina at 48.3% compared to a 33.9% chance of Trump winning New Hampshire. Again, the key difference is paper ballots versus the more cheater-friendly electronic system.
If Trump outperforms in New Hampshire relative to North Carolina, it should likewise cause some concern with respect to the smell test. 
If both occur and he outperforms in Michigan vs. Pennsylvania and outperforms in New Hampshire vs. North Carolina, a closer look at the propriety of the election might be warranted.
Let’s all hope that this isn’t the case. There might not be enough tin foil hats to go around.


No comments: