Sweet Carolina?

By Kraken Wood December 1, 2020

The single most accurate predictor of an election at county or state level is voter registration by party. Not all states register voters by party, but those that do are easier to predict than those that do not. The metric, depending on state, is usually over 90% accurate in determining which party will gain in a gain, and consequently, which will lose.

The exceptions typically come in very large counties like Mecklenburg, which only trended Republican in registration thanks to a higher concentration of new voters registering as independent. If they vote like most urban voters, it can trend the county in the opposite direction in actual voting. Barack Obama won North Carolina by 0.32% in 2008, with the Democrats holding a 13.8% registration advantage over Republicans. The GOP cut that gap by 0.8% (down to 13.0%), which accurately projected the state returning to the Republican column. Mitt Romney won it by 2.08%. The state trended 3.7% more Republican, down to 9.3%, which foretold Donald Trump's 3.66% victory.

North Carolina trended another 3.8% toward the GOP, reducing the Democrat advantage down to 5.5% in the four years since Trump's election. That suggests a dominant victory of over 5 points based on historic registration trends. Despite 94 of 100 counties trending Republican in registration, Trump's margin was reduced substantially, down by over 2.2% from 2016. North Carolina appears to be part of a "Southern Steal" strategy involving Georgia, and was won over a substantial amount of fraudulent votes that may have also helped to re-elect Governor Roy Cooper.