Prop 50 and Game Theory: Tit-for-tat

Today, California voters are deciding whether to temporarily abandon the independent redistricting commission they created fifteen years ago. The reason? Texas redistricted mid-decade to gain about five Republican House seats, and Democrats want to punch back.

Welcome to democracy in 2025, where the path to saving fair representation might require a detour through unfair representation. If that sounds contradictory, buckle up. We're about to take a tour through iterated game theory and a strategy called "tit for tat with cooperation" (or "generous tit for tat"). It turns out the math has some uncomfortable things to say about noble principles meeting brutal reality.

The Setup: Two Players, One Map, Infinite Rounds

Here's the game in its simplest form. Two players (let's call them Red State and Blue State) each get to draw electoral maps. They have two choices: cooperate or defect.

Cooperate means drawing fair maps that create competitive districts and accurately represent voters. Defect means gerrymandering the hell out of your state to maximize your party's seats.

If both cooperate, democracy works. Competitive elections. Representation that matches actual voter preferences. The whole Schoolhouse Rock fantasy.

If one defects while the other cooperates, the defector wins big. They pack their opponents into a few districts and crack the rest, gaining maybe five or ten extra House seats. The cooperator? They're Charlie Brown, and the football just got yanked away again.

If both defect, both parties gerrymander, and democracy takes a beating. But here's the thing: neither side loses ground relative to the other. It's mutually assured distortion.

Sound familiar? It should. Political scientists call this the Prisoner's Dilemma, and it's been the subject of game theory research since the 1950s. The twist is that redistricting isn't a one-time choice. States keep redrawing maps every decade (or in Texas's case, whenever they feel like it). This makes it an iterated game, where the same players face the same choice over and over.

And iterated games? They change everything.

Enter Tit-for-Tat: The Strategy That Won by Losing

In the 1980s, political scientist Robert Axelrod ran a series of tournaments where game theorists submitted strategies for playing iterated Prisoner's Dilemma. Complex algorithms competed against each other in round after round of choices: cooperate or defect?

The winner shocked everyone. It wasn't some sophisticated strategy with machine learning or probabilistic decision trees. It was tit-for-tat, submitted by mathematician Anatol Rapoport. Four simple rules:

1. Start nice (cooperate first)
2. Be provocable (if opponent defects, defect back immediately)
3. Be forgiving (return to cooperation after one retaliation)
4. Be clear (make your strategy obvious)

Tit-for-tat didn't win by crushing opponents. It won by teaching them that cooperation pays better than exploitation. Defect against tit-for-tat, and you get punished. But cooperate, and tit-for-tat cooperates right back. The strategy creates an environment where mutual cooperation becomes the rational choice.

Now look at California's Proposition 50.

Started nice? California created an independent redistricting commission in 2008 for state legislative districts, expanded it to congressional districts in 2010. While other states were gerrymandering themselves into pretzel shapes, California took itself out of the game.

Provocable? After Texas redistricted mid-decade and Donald Trump bragged that "Texas just picked up five seats" at a January rally, Governor Gavin Newsom announced California would gerrymander right back. "Two can play that game," he said.

Forgiving? Prop 50 includes a built-in sunset clause. After the 2030 census, control returns to the independent commission. This is retaliation, not a permanent strategy shift.

Clear? You cannot get more transparent than putting the question directly to voters: "Should we temporarily gerrymander to counter Republican gerrymandering?"

By the textbook definition, this is tit-for-tat.

The Provocation: When Always-Cooperate Meets Always-Defect

The context matters here. Texas didn't just redistrict after the census, which is normal. They redistricted in 2024, mid-decade, after the Supreme Court's 2023 decision gave states more leeway on redistricting. Missouri, North Carolina, and Ohio followed suit. Republicans gained somewhere around 15-20 House seats total from these maneuvers, expanding their already existing control of the House.

Democrats, meanwhile, had been playing the role of the eternal cooperator. Besides California's independent commission, states like Michigan and Virginia had also moved toward fairer maps. New York's Democratic legislature actually had their gerrymander struck down by courts and replaced with a fair map.

In game theory terms, Democrats were playing "always cooperate" while Republicans were playing "always defect." And here's what the math says about that matchup: the always-cooperator gets destroyed.

Always-cooperate only works when both players use it. Against a defector, it's just unilateral disarmament. You lose seats, you lose power, and eventually you lose the ability to protect the very democratic norms you're trying to defend.

California had cooperated for 15 years. Texas's response was to redistrict mid-decade specifically to gain House seats. The signal could not have been clearer: cooperation will not be reciprocated.

Prop 50 is the retaliation move. It suspends the Citizens Redistricting Commission for one cycle and hands redistricting to the legislature, with the explicit goal of gaining about five seats to counter Texas's five-seat gain. Then, after 2030, the commission returns.

The Voter's Dilemma: Principles Meet Payoffs

Here's where it gets uncomfortable. Polling shows about 60% of California voters support Prop 50. These are the same voters who created the independent commission in the first place. They like fair maps. They believe in the principle of non-partisan redistricting.

But they also watched Texas redistrict mid-decade with zero consequences. They watched the Supreme Court effectively give up on policing partisan gerrymandering. They watched Republicans expand their House majority through maps alone.

Charles Munger Jr., who funded the original commission initiative, opposes Prop 50. His argument: California loses the moral high ground. If Democrats gerrymander too, they can't credibly argue for federal legislation banning the practice.

It's a valid point. There's something genuinely troubling about voting to make your democracy temporarily less democratic.

But game theory has a response: moral high ground doesn't matter if you're not in power to use it. Unilateral disarmament is only noble if it changes your opponent's behavior. When it doesn't, it's just surrender with better PR.

The mathematical reality is harsh. In an iterated game against a defector, always-cooperate is the losing strategy. You have to retaliate. Not because retaliation is good, but because it's the only thing that creates incentive for the opponent to cooperate.

The Outcome: Two Paths Forward

If Prop 50 passes, California Democrats gerrymander for one cycle. They gain somewhere around five seats, offsetting Texas's mid-decade redistricting. The commission returns in 2031, and California signals loud and clear: we prefer cooperation, but we will retaliate against defection.

The question then becomes: do Republicans get the message? Game theory suggests they might. Tit-for-tat's power lies in its clarity. When retaliation is swift, proportional, and followed by an offer to return to cooperation, it teaches the opponent that defection doesn't pay.

Maybe Republican-controlled states decide mid-decade redistricting isn't worth the retaliation. Maybe Congress finally passes federal redistricting standards. Maybe the Supreme Court reverses course. Or maybe Republicans escalate, and we get a race to the bottom where every state with unified party control gerrymanders every chance it gets.

If Prop 50 fails, California maintains the moral high ground but loses five House seats. Democrats continue playing always-cooperate while Republicans play always-defect. The payoff matrix suggests this ends with Republicans controlling redistricting in their states while Democrats unilaterally disarm in theirs.

From a pure game theory perspective, that's the irrational outcome. You can call it principled, but the math calls it losing.

Beyond the Game: What Happens Next?

The danger with tit-for-tat is that it can spiral. If both players get stuck in a defect-retaliate-defect cycle, cooperation never returns. That's why Prop 50's sunset clause matters. It's not California saying "gerrymandering is fine now." It's California saying "we will match your defection for exactly one round, then offer cooperation again."

Will Republicans hear that signal? That's the multi-billion-dollar question.

Other states with independent commissions are watching closely. Michigan, Arizona, Colorado. They may respond based on what they observe here.

Congress could pass redistricting standards. The Supreme Court could change its mind about partisan gerrymandering being a "political question." Until either happens, we're probably fated to a race to the bottom where every state with unified party control gerrymanders every chance it gets.

The game theory lesson here is simple: cooperation emerges when both players know they'll face each other again and when defection carries consequences. Prop 50 attempts to create those consequences.

The Verdict: Sometimes You Have to Defect to Defend Cooperation

From a pure game theory perspective, Prop 50 is rational. It's textbook tit-for-tat, and tit-for-tat is the mathematically optimal strategy for iterated Prisoner's Dilemma.

That doesn't make it feel good. There's something uncomfortable about democracy voting itself into a less democratic state, even temporarily. The irony is almost too on-the-nose: using democratic means to approve undemocratic gerrymandering in order to preserve democratic representation.

But the truth is that defending cooperation sometimes requires punishing defection. If one player always cooperates while the other always defects, the cooperator doesn't change the game. They just lose it.

Will it work? That depends entirely on whether Republicans recognize the signal. If they see California's retaliation and decide mid-decade redistricting isn't worth the cost, cooperation might return in 2031. If they escalate instead, we get the race to the bottom.

The voters decide today. Either way, we'll find out whether democracy can play game theory better than game theory predicts.