This is a controversial topic. I feel compelled to write about it because I don't see most American's ears perking up about it.

First off I want to draw the lines around this essay. My dissertation advisor always taught me to advance and take only the exact intellectual territory you wish to seize and no more. I am only writing about the 2022 Arizona midterm elections. I am not suggesting that this election was "stolen" or that there was foul play involved. What I hope to do in this essay is to illustrate that the tabulation phase of this election is strange. Furthermore I hope to shed light on the way in which this viewpoint is being preemptively delegitimized.

State of the World

I voted in-person on November 8th like the crazy conspiracy theorist I probably am. I had to wait about 30 minutes in line at a "Vote Center". Upon entering, I typed my address into a computer and a ballot was printed, which was strange to me. I've voted in plenty of elections before but this is only my second regular-cycle election of voting in Arizona. When I lived in Ohio, I would show up to a "Polling Location", my ID would be checked, and I was given a pre-printed ballot. After asking around I discovered that this is so any Arizonan from any area of town (maybe any county?) can vote at any "Vote Center".

Vote counting has been arduously slow in AZ. It is now Sunday November 13th, 5 days after the election and we still don't know who the next governor will be. There seems to be no clear answer on how many votes remain to be counted, instead percentages are given (percentages of what?). Democratic candidate Katie Hobbs has a narrow lead over Republican candidate Kari Lake.

Consulting the Oracle

I don't remember state elections taking this long until the 2020 presidential election (notable exception Gore v. Bush). My parents would vote in person at their local Polling Location, usually a church or school - they would sometimes take me and the whole process would only take about 10 minutes by the way. That evening, sometime after dinner, but before bedtime, the local news would announce the winners.

I consulted Google. I searched for "how long did it take to count votes in the 1800s". I was surprised but I should have been - Google has been deep-curating their search results on hot topics to control conversations since COVID and likely before. That query yielded pages and pages of articles written in the past week about how it's "totally normal" and "to be expected" for vote counting to take this long.

Unsatisfied, I searched DuckDuckGo for the same query. DuckDuckGo had articles long it took to count votes in the 1800s. Turns out weeks. Ok, I accept that - it was a pre-car era.

Diving deeper, I started to look for specific examples. I looked up the election date for a few elections, then searched the New York Times archives for an article announcing the winner to determine when the election had been decided. Here's what I found.

Nixon/McGovern 1972

Election date: Nov 7 Election decided: Nov 9

Reagan/Carter 1980

Election date: Nov 4 Election decided: Nov 7

JFK/Nixon 1960

This is an interesting example because JFK squeezed by with just a .17% advantage in the popular vote over Nixon. It was close.

Election date: Nov 8 Election decided: Nov 10

As a nation we could count enough votes countrywide in just two days to determine the winner of a presidential election seven decades ago. It is an absolute certainty that average societal IQ decreases about three points every decade. The average person is stupider than the average person 60 years ago. This is no excuse when it comes to calculating ballots. Our modern-day-IQ-equivalent-peers in Florida were able to count enough ballots to determine winners the night of the election which goes against my "stupider people" hypothesis. So what is the cause?

Premature Optimization

A thought error I see all the time in more junior software developers is premature optimization. The temptation is great when one begins coding to solve for every hypothetical problem that could arise in a system regardless of whether or not the system will ever experience the hypothesized conditions. This makes for a lot of "feel good" work but progress is slowed and software is bloated by these extraneous features that nobody asked for and that the system doesn't require to carry out its job.

After interacting more with the business side of software, developers learn that software isn't an intellectual exercise about programming patterns and "best practices". It is a tool that accomplishes a purpose, usually one that generates money or interest in some way. Keeping a clean-enough codebase is important but tradeoffs must be made so that the software is kept lean and accomplishes what it set out to do efficiently. This means avoiding lengthy philosophical discussions about design patterns or "how Google does it". It means executing precisely and quickly. It requires knowing what code moves the ball down the field and letting anything that doesn't slide.

I see this everywhere outside of software development now. Self proclaimed "innovators" lack the humility to execute effectively. Our inflated sense of self worth says "nobody wanted a better PDA until Steve Jobs invented the iPhone, and I'm like Steve Jobs". These "leaders" fail to accomplish the basics instead opting to achieve "greatness".

The election process in Arizona is a perfect example of premature optimization. Maybe not premature, there was a forcing function, I assume COVID lockdowns encouraged a lot of rapid change in the way in which elections are handled in the state. At any rate, it's time to say that this "solution" is garbage and that the legacy way of doing things just worked better in the sense that it accomplished the goal (determining a winner fairly and quickly) more efficiently.

Future Elections

Here is how I'd like to see AZ elections carried out in the future (at least the in-person part, I'm not touching mail-in voting since that is a sacred cow for half of the country).

There are a fuckton of polling places. They're in churches, schools, your neighbor's garage. You can't walk two blocks without seeing one on election day. They're open 9-7. Election day is a state holiday, employers must give the day off or pay time and a half. Paid time off must be given for part of the day to go vote (AZ already has this). You vote at your local precinct no exceptions. You show your ID when you show up to vote (to this date nobody has been able to properly explain to me why this is controversial other than we're not supposed to think that way shame on you).

Ballots are preprinted and counted by the polling place staff the evening of the election after polls close. If it takes more than a couple of hours they brew coffee and count them until the wee hours. Why count at the polling places? Technical and practical reasons. Technically you have a lot of simple tasks being done, i.e. adding votes. Splitting this among a lot of workers is generally a good idea. A lot of people count a few votes instead of what we have now which is a few people counting a lot of votes. The more practical reason is that a distributed system is harder to compromise. Even if you sabotage a few nodes, it's unlikely that it will severely disrupt the election results. In a centralized system only one vulnerability risks compromising the integrity of the election.

Wishful Thinking

As of this writing, Katie Hobbs has about a 1.4% lead over Kari Lake. It is looking like Hobbs is going to be our next governor so we must turn to her with our hopes for doing away with the bloated and inefficient system we have now.

Unfortunately Hobbs seems unlikely to implement any of these changes, it's against her own interests. As of writing, Hobbs is our Secretary of State, which means she is our chief election official in AZ. Regardless of whatever part she may have played in shifting towards this new system of voting, she has owned it as our chief election official and had the power to change it but didn't after seeing it crumble during the 2020 elections.

Hobbs chose to not recuse herself as chief election official during...her own election...which is totally not weird...

Going down the rabbit hole as usual I was able to turn up one example of this exact situation happening before. In the 2018 Georgia gubernatorial election, Stacey Abrams (D) ran against Brian Kemp (R). Brian Kemp was at that time the Secretary of State and thus Georgia's chief election official. Kemp won the election. Stacey Abrams eventually accepted Kemp as the winner of the election but refused to say the election was legitimate, and even went so far as to suggest foul play was involved. Was Stacey Abrams called an election denier and conspiracy theorist? Hardly, she ran again in 2022. Will Kari Lake be treated the same way if she is defeated? We'll have to wait and see - at this rate we might not know until the San Francisco peaks are green again.

I recently removed social share links from my site. If you enjoyed what you read, please take the time to tell one person directly about this blog.