Will Trump or Biden carry Arizona?

Multiple press outlets called the state for Biden. However, many have complained that this call was too early and that Trump still has a chance.

After looking at some basic numbers, it seems that Trump indeed has not only a shot of winning the state but will likely win it.

Our analysis is based off of the number of registered voters by county. We looked at the net difference between registered Republicans and Democrats, with the idea that Trump would win in those counties with more registered Republicans.

This logic is extremely simple but effective in Arizona as we will see.

Then, we compared this net difference in party registration to the current difference in votes between Trump and Biden, by county. So, we would compare a county’s net registration (the number of voters registered as Republicans minus those registered as Democrats) to its net votes (the number of current votes in favor of Trump minus those in favor of Biden).

In theory, these net figures should be more or less in-line, of course assuming that independent voters do not heavily favor one party and that turnout by party does not significantly vary.

We found that our basic analysis worked in every county in Arizona, except in Maricopa where there are currently a few hundred thousand ballots that have yet to be counted. Our analysis infers that these ballots should slant heavily towards Trump. Again, this is an extremely basic and simple analysis, but insightful.

Table 1: Arizona General Election 2020, Actual Net Registration Figures by County, Provisional Net Votes by County

Source: ZettaCap, Arizona Secretary of State

Can you spot the anomaly?

Maricopa stands out as the only county that has significantly more registered Republicans while having significantly less votes for Trump. Additionally, we should keep in mind that this county voted for Trump in 2016.

The fact that (apparently) most of the outstanding (uncounted) votes come from this county makes us wonder if this anomaly will normalize.  In other words, once the votes are counted, it would look a lot less awkward if Maricopa’s Net Votes were somewhere above zero.

These are back of the envelope type calculations, and done very quickly. So, take them with a grain of salt.

The way these numbers stand now, they just don’t look right. Perhaps something extraordinary could have occurred specifically in this county that made it not only different from its voting pattern from 2016 but also from the rest of the state in 2020. It is possible. But, it does not seem likely.

The most likely scenario is that once these outstanding ballots are counted in Maricopa, its Net Votes will no longer stand out so much and more or less fall into an expected range based on its Net Registrations.

If this expected range is hit after counting the outstanding ballots, Trump would gain more than enough to overcome the current 46,000 deficit.

For those interested in these numbers, go to the Arizona Secretary of State site where they provide registration figures for each county and for each election.

For those who just want the data on the number of registered voters:

G.E. stands for General Election. This data is the official Arizona registration data provided by the state. The total does not match up as it includes other parties and independents.

We should also mention that we forecasted Trump would win Arizona and were initially surprised by the early call in favor of Biden.

Also note that betting markets still are in favor of Arizona going for Biden, with an approximate 75% probability as of publication of this post.