📈 economics

Misrepresentation in District-Based Elections

This paper proposes a misrepresentation-minimizing cutoff rule that balances local district majorities with statewide vote proportions, demonstrating how adjusting the weight of statewide representation can reduce gerrymandering and offering a novel metric to evaluate the representativeness of U.S. House district boundaries.

Original authors: Yunus C. Aybas, Oguzhan Celebi, Surabhi Dutt

Published 2026-02-16

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Yunus C. Aybas, Oguzhan Celebi, Surabhi Dutt

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a state is a giant pizza, and the voters are the people who want to eat it. The state is divided into slices (districts), and each slice gets to pick one topping (a representative) for the whole pizza.

The big problem this paper tackles is a tug-of-war between two ways of deciding who gets the toppings:

The "Local Majority" Rule (First-Past-The-Post): In every single slice, whoever gets the most votes wins that slice's topping. If 51% of people in a slice want pepperoni, that slice gets pepperoni, even if only 49% wanted it.
The "Statewide Flavor" Rule (Proportional Representation): If 60% of the entire state wants pepperoni, then 60% of the total toppings on the pizza should be pepperoni, even if that means some slices where pepperoni lost locally get pepperoni anyway.

The authors ask: How do we balance these two? If we only listen to the local slices, the final pizza might taste nothing like what the whole state actually wanted. If we only listen to the whole state, we might ignore the specific tastes of individual neighborhoods.

The "Misrepresentation" Meter

The authors invent a "Misrepresentation Meter" to measure how unhappy the voters are. This meter has two dials:

Dial A (Local Unhappiness): How many people in a specific slice are stuck with a topping they hate because their neighbor won? (e.g., You wanted cheese, but pepperoni won your slice).
Dial B (Statewide Unhappiness): How far off is the final pizza from the state's average taste? (e.g., The state wanted 60% pepperoni, but the pizza ended up with 80% pepperoni).

The paper suggests we can't just ignore one dial. Instead, we need a Weight. This weight is like a volume knob that decides how much we care about Dial B compared to Dial A.

Volume Knob at 0: We only care about local slices. This is the current system (First-Past-The-Post).
Volume Knob at Infinity: We only care about the statewide average. This is Proportional Representation.
Volume Knob in the Middle: We care about both, but we have to make a trade-off.

The Magic "Cutoff" Switch

The paper's biggest discovery is that the best way to balance these two isn't a complicated algorithm. It's a simple Cutoff Rule.

Imagine a line drawn across a graph of voter support.

If a slice's support for a party is above the line, that party wins the slice.
If it's below the line, they lose.

The genius of the paper is showing that where you draw that line changes based on your Volume Knob.

If you don't care about statewide fairness (Knob = 0), the line is drawn at 50%. (You need a simple majority to win).
If you care a little bit about statewide fairness, the line moves slightly below 50%. This helps the party that is slightly underrepresented statewide to win a few extra close races.
If you care a lot about statewide fairness, the line moves up or down significantly to ensure the total number of seats matches the total votes perfectly.

The Analogy: Think of it like a thermostat. The current system (FPTP) is stuck at a fixed temperature. This paper says, "We can install a smart thermostat." If the house (the state) feels too cold (underrepresented), the thermostat automatically lowers the bar to let in more heat (seats). If the house is too hot (overrepresented), it raises the bar to keep the heat out.

Gerrymandering: The "Shuffling" Game

Gerrymandering is when politicians draw the pizza slices in weird shapes to make sure their party wins more slices than they deserve.

The Old Way: Under the current rules, if a party is good at shuffling their voters into "safe" slices, they can win a huge number of seats even with fewer total votes.
The New Insight: The paper shows that as you turn up the "Statewide Fairness" knob (moving away from the strict 50% rule), it becomes much harder to cheat.
- If you try to shuffle voters to win extra seats, the "smart thermostat" (the new rule) notices that the statewide balance is off and adjusts the cutoff line to cancel out your trick.
- It's like trying to cheat at a game where the referee (the rule) is watching the entire scoreboard, not just the individual plays.

What the Data Says

The authors looked at real US election data (like in Texas, California, and North Carolina) to see what "Volume Knob" setting would make the current results look "optimal."

In some states (like California): The current results only make sense if we care almost zero about statewide fairness. Even a tiny bit of concern for fairness would suggest a different number of seats. This means the current map is very sensitive to how votes are distributed.
In other states (like Texas): The current results hold up even if we care quite a bit about statewide fairness. The maps here are more "robust" or resistant to the idea of changing the rules.

The Takeaway

This paper doesn't just say "Proportional Representation is better" or "Districts are better." It says: "There is a whole spectrum of rules in between."

It gives us a tool to measure exactly how much a state values local control versus statewide fairness. It suggests that we don't have to choose between a broken system and a radical overhaul. We can simply turn the knob to find a rule that minimizes voter unhappiness, makes it harder for politicians to rig the game, and ensures the final delegation actually tastes like the state that elected it.

1. Problem Statement

The paper addresses a fundamental tension in representative democracies that utilize single-member district (SMD) elections: the conflict between local district majorities and statewide proportionality.

The Conflict: First-Past-The-Post (FPTP) rules respect the majority in each district but often produce a statewide delegation seat share that deviates significantly from the statewide popular vote. Conversely, Proportional Representation (PR) aligns the delegation with the statewide vote but may require awarding seats to parties that lost the local district majority.
The Gap: Existing literature often treats these as separate issues (measuring partisan bias or axiomatic properties of specific rules) without explicitly modeling the tradeoff or providing a unified framework to balance local representation against statewide proportionality.
Objective: The authors aim to characterize the optimal seat-allocation rule that minimizes a weighted sum of "misrepresentation" from both local and statewide perspectives, and to analyze how this framework interacts with geographic concentration of support and gerrymandering.

2. Methodology and Model

A. Theoretical Framework

The authors model a two-party election (Parties A and B) across $N$ single-member districts of equal population.

Inputs: A profile of district vote shares $p = (p_1, \dots, p_N)$ for Party A. The statewide vote share is $\bar{p} = \frac{1}{N}\sum p_i$ .
Distortion Measures:
1. District Misrepresentation ($Dist$): The mass of voters represented by the non-preferred party. If Party A wins district $d$ , the misrepresentation is $1-p_d$ (B voters); if B wins, it is $p_d$ (A voters).
  $Dist(x) = \sum_{d=1}^N [x_d(1-p_d) + (1-x_d)p_d]$
2. Statewide Misrepresentation ($Agg$): The absolute deviation between the party's seat share and its statewide vote share, scaled by $N$ .
  $Agg(x) = |N\bar{p} - S|$
  where $S$ is the total seats won by Party A.
Objective Function: The total misrepresentation is a weighted sum:
$\Phi(x; w_A) = Dist(x) + w_A \cdot Agg(x)$
Here, $w_A \in [0, \infty)$ represents the normative weight placed on statewide proportionality relative to local majorities. $w_A=0$ corresponds to pure FPTP logic; $w_A \to \infty$ corresponds to pure PR.

B. Optimization Strategy

The problem is solved in two steps:

Conditional Optimization: For a fixed seat total $S$ , the rule that minimizes $Dist(x)$ is to award the $S$ seats to the districts where Party A has the highest vote shares (the "top- $S$ " allocation).
Global Optimization: The optimal seat total $S^*$ is chosen to minimize the total objective function $\Phi(S; w_A)$ , given the top- $S$ allocation.

3. Key Contributions and Results

A. The Misrepresentation-Minimizing Rule (Structural Characterization)

The central result (Theorem 1) is that for any weight $w_A$ , the optimal rule is a uniform vote-share cutoff rule.

Mechanism: Party A wins every district where its vote share $p_d$ exceeds a threshold $t(w_A)$ .
The Cutoff Dynamics:
- At $w_A = 0$ , the cutoff is $t = 0.5$ (standard FPTP).
- As $w_A$ $w_{A}$ increases, the cutoff shifts continuously.
  - If Party A is underrepresented (seats < proportional share), the cutoff drops below 50%, converting narrow losses into wins to increase seat count.
  - If Party A is overrepresented, the cutoff rises above 50%, converting narrow wins into losses to decrease seat count.
- At the limit ( $w_A \to \infty$ ), the rule converges to Proportional Representation (PR).
Significance: This defines a one-parameter family of rules that smoothly interpolates between FPTP and PR, tracing the Pareto frontier of feasible tradeoffs between local and statewide misrepresentation.

B. Geographic Concentration and Majorization

The authors analyze how the geographic concentration of support affects misrepresentation using majorization theory.

Definition: A profile $p$ majorizes $q$ if $p$ is more concentrated (cumulative support in the strongest districts is higher for all $k$ ) while maintaining the same statewide mean.
Result (Proposition 2): Under the misrepresentation-minimizing rule, a more concentrated support profile yields weakly lower total misrepresentation.
Crucial Insight: This monotonicity holds only if the rule is optimized for the specific weight $w_A$ used to evaluate the outcome. Fixed-threshold rules (like standard FPTP) do not necessarily satisfy this property because the induced seat total can jump discontinuously, potentially increasing statewide distortion.

C. Axiomatic Characterization of Endpoints

The paper provides axiomatic characterizations for the two endpoints of the family:

FPTP ( $w_A=0$ ): Characterized by Strong Monotonicity. Increasing support in a district never causes a party to lose a seat it previously won. This holds only when districts are treated independently (no statewide coupling).
Proportional Representation ( $w_A \to \infty$ ): Characterized by Gerrymandering-Proofness. No party can increase its seat total by reshuffling votes across districts while keeping the statewide total constant. This holds only when the seat total is determined solely by the statewide mean.

D. Gerrymandering Cost

The authors quantify the vulnerability to gerrymandering by defining the gerrymandering cost: the minimal change in district vote shares required to achieve a target seat total.

Result (Proposition 5): For an overrepresented party, the cost of engineering additional seats strictly increases as the weight $w_A$ moves away from 0 (FPTP) toward PR.
Implication: Moving away from pure FPTP toward a misrepresentation-minimizing rule with a positive $w_A$ acts as a complementary lever to limit gerrymandering, making it harder to manipulate maps for partisan gain without abandoning district-based representation entirely.

4. Empirical Analysis

The authors apply the model to U.S. House elections (1990–2024) to infer the implicit weights rationalizing observed outcomes.

Method: They invert the model to find the range of $w_A$ for which the observed FPTP outcome is actually the misrepresentation-minimizing solution.
Findings:
- State Heterogeneity: There is significant variation across states. In states like California and Washington, the implied weight for FPTP to remain optimal is very low; even a small concern for proportionality would shift the optimal rule away from FPTP. In states like Texas and Georgia, FPTP remains optimal even under high weights for proportionality.
- Redistricting Impact: The analysis tracks how changes in district maps (e.g., in North Carolina) shift the implied weights. When maps are drawn to favor a party, the implied weight required to rationalize the FPTP outcome drops, indicating a deliberate decoupling of local majorities from statewide representation.
- Texas vs. California: Despite both having Republican overrepresentation, California's district map is less robust to proportionality concerns (lower cutoff) than Texas's, suggesting California's current map creates a larger distortion between local and statewide representation.

5. Significance and Conclusion

Theoretical Contribution: The paper unifies the FPTP and PR debate by showing they are endpoints of a continuous, optimal family of rules defined by a single parameter (the weight on proportionality). It provides a rigorous metric for evaluating the "representativeness" of any district-based system.
Policy Implication: It offers a novel alternative to traditional anti-gerrymandering approaches (which focus solely on map geometry). By adjusting the election rule (the cutoff threshold) to reflect a societal preference for proportionality, states can make gerrymandering more costly and difficult to execute without abandoning single-member districts.
Empirical Tool: The "implied weight" serves as a diagnostic tool for policymakers and courts to evaluate whether current district boundaries and election rules align with the state's actual representational goals, providing a data-driven basis for reform proposals.

In summary, the paper demonstrates that the tradeoff between local majorities and statewide proportionality is not a binary choice but a continuous spectrum. By optimizing the cutoff threshold based on a societal weight, one can minimize total misrepresentation and simultaneously increase the resilience of the electoral system against partisan manipulation.