Anyone who has ridden three to a back seat on a road trip knows the rule: you take turns. Window seat for a few hours, then the dreaded middle seat for a few hours. Simple. Fair. What could possibly go wrong?

The Lodi City Council is considering a change to how it picks its Mayor and Mayor Pro Tempore. Instead of a vote each December, the seats would rotate automatically by district. The Mayor Pro Tem moves up to Mayor, the next district in line becomes Mayor Pro Tem, and everyone gets a turn. Five districts, five years, one tidy loop.

I liked the idea in principle, then I started doing asking what-if...

The calendar has opinions

Here's the fun with the numbers. Lodi has five council districts, so the rotation repeats every five years. But council seats are elected every four years: Districts 1, 2 and 3 in one cycle, Districts 4 and 5 two years later. Five and four do not get along. They are the Ben and Jerry of numbers.

Because of that mismatch, about once every 20 years a district's turn as Mayor lands right after its own election. If the incumbent loses, the new member has been on the council for all of fifteen minutes. The draft resolution sensibly requires 12 months of service before becoming Mayor, so that district would get skipped.

Fair enough. Skip 'em and move on.

Then the skips took over the group chat

Except once one district is skipped, the remaining four rotate the next four years. That is, you'll notice, exactly the election cycle. So the skipped district comes up again right after its next election. If voters pick someone new again, it gets skipped again.

I ran the scenario forward. In the worst case, District 3 gets skipped all 23 times between 2035 and 2123. Uh-OH!

To be fair, that's the extreme. So I got fancier and had AI simulate 10,000 possible futures where incumbents win 75% of the time. Under that scenario, districts lost an average of about 7.5 Mayor turns over a century. In the unluckiest futures, one district would wait more than 30 years between terms.

Fixing it with one sentence

The fun part of policy work is finding the one sentence that does most of the lifting. Here is the line that solved this problem: nobody can serve as Mayor Pro Tem during a year their seat is on the ballot.

Since the Mayor Pro Tem becomes Mayor the next year, that one rule means the person next in line for the gavel never faces an election first. In the same 10,000 simulated futures, lost Mayor turns dropped from about 7.5 to zero and no district waited more than nine years. Which may never happen with some luck.

Add a "you keep your place in line" rule for any district that gets passed over, plus a "fewest terms goes first" tiebreaker, and the rotation more or less balances itself.

And then the what-ifs arrived

Of course, a policy isn't complete until it survives the what-ifs.

What if the Mayor resigns in March? The most recent former Mayor still on the council steps in, so the Mayor Pro Tem stays on track. What if the Mayor Pro Tem moves away? The next in line serves out the term and becomes Mayor the following year. What if someone simply doesn't want the job? What if a seat is vacant? What if redistricting renumbers everything?

Each answer to these questions required a sentence or two in the resolution. This turned a four-line resolution about the mayoral rotation into a seven-section policy. That's what "simple" looks like after you've asked "but what if?" enough times.

Why easy policy can be hard

Every business owner knows this feeling. You write a simple return policy, and then a customer shows up with a un-legible receipt and a product you haven't sold in 10 years. Good rules work on the normal days and the weird days.

This lesson applies far beyond City Hall. Nobody did anything wrong in drafting the rotation. The idea of giving every district a fair turn is a good one. Things just started to break down once I ran the clock forward a few decades looking for unintended consequences or abnormalities.

Taking turns should be easy. It just took a little fun with numbers to see that it is harder than it appears.