My regular readers will know that I’m fascinated by systems with “emergent properties” — systems where a few simple rules that individuals operate by result in complex and surprisingly organized behavior by the group — group properties that don’t seem to have any obvious direct relationship to the simple rules.
Birds flocking is a great example. Each bird has very simple algorithms — “If I’m on the outside of the flock and the bird next to me gets farther away, move toward that bird.” The result is the organized flock movements we’re all familiar with.
This came to mind when reading this Economist leader on gerrymandering of congressional districts in America.
Time to bury Governor Gerry
AMERICAN exceptionalism comes in many forms, but one of the odder ones is the way it sets its electoral boundaries. In every other democracy worthy of the name, independent commissions perform the sensitive and vital task of adjusting boundaries to take account of shifts in population. But in no fewer than 44 of America’s 50 states, it is state legislatures, composed as they are of party politicians, who decide where the lines should be drawn for seats in the House of Representatives in Washington, DC.
Now obviously other democracies don’t have wonderfully perfect political systems. But still, it made me wonder: could that single fact — that one rule-of-the-system — be largely responsible for many of America’s political (hence economic) failings? Could it inevitably — as an emergent property of the system — render our national politics and economic policy-making dysfunctional or destructive?
Maybe it’s just me grasping for simple, single-answer, one-handed explanations and solutions, but could this fact be responsible for many of the ills that plague our economy — the emergence of ever-longer “jobless recoveries” since the 80s, stagnation of middle-class earnings, increasing inequality, etc.? It’s not hard to imagine such a fact resulting in systematically distorted legislative approach to economic issues.
Could the proximate causes that we keep claiming for these problems — globalization, technology improvements, education issues, etc. — be red herrings, with the ultimate cause lying utterly elsewhere? I’m nowhere near being able to answer that question (if it’s even a useful question), but I did turn up a couple of recent papers that illuminate the subject. Both consider whether, and to what degree, federal courts can and should impose rules and restrictions on state-level gerrymandering of congressional districts.
An Interstate Process Perspective on Political Gerrymandering — an anonymous “Note” from the March 2006 Harvard Law Review — outlines four ways of looking at gerrymandering:
It points out that almost all the analysis of gerrymandering to date has been about intrastate stuff. It cites one article on Interstate Effects (“Partisan Gerrymandering and Disaggregated Redistricting” by Adam B. Cox, Supreme Court Review, 2004) and discusses it at length (emphasis mine):
…as Professor Adam Cox has recently pointed out, both academic commentators and the Justices in Vieth have failed to account for crucial conceptual differences between gerrymandering for state legislative districts and gerrymandering for congressional delegations. Unlike the harms from gerrymandered state legislative districts, the full harms from a state’s gerrymander of its congressional delegation can only be determined by looking simultaneously at all other states’ districting plans, as its delegation is only a subset of a larger legislature. With each state controlling the process of drawing its own congressional district lines,a state’s political gerrymandering might impose costs well beyond its borders.
Some extracts from Cox’s paper may make this clearer:
…evaluating the potential political gerrymander of a single congressional districting plan in isolation prevents a court from identifying the harms, if any, that stem from the manipulation of the composition of Congress as a whole.
To the extent that the harm of congressional gerrymanders can only be identified from a legislature-wide institutional perspective, the Court will inevitably fail if it tries to pin the injury on individual congressional redistricting plans.
He proposes that federal courts adopt a risk/probability-based approach to policing of state congressional redistricting:
So long as courts reduce the level of bias in each state congressional plan in some relatively uniform fashion,they will decrease the probability that the state-level biases will accumulate into an unacceptable level of national bias.
…Judicial intervention can in theory lower the probability that state redistricting schemes combine to produce Congress-wide injuries. But if that probability is low in the first place, judicial intervention becomes less appealing.
Returning to the HLR article, it then moves to its central topic — what it calls Interstate Process defects:
…gerrymanders should be alarming because they result from “interstate process” defects. …the legislative process that produces it fails to account adequately for outsiders’ preferences — outsiders who suffer very real harms from gerrymandering. These harms include being governed by potentially unresponsive politicians, having a House of Representatives with an “unfair” partisan balance, and feeling compelled to alter the political processes of one’s own state.
The interstate process perspective thus echoes the familiar economic concept of externalities.
…the concept of an interstate process problem helps justify rigorous judicial intervention in these constitutional areas.
…one state’s political gerrymandering can impose significant costs on out-of- state interests. Yet those out-of-state individuals are not participants in the political process that produces the gerrymandering state’s districting plan.
…Since those harms are uniquely visited upon outsider states, in-state residents likely will have no reason to prevent those harms.
Thus political gerrymandering may actually be a more worrisome species of interstate discrimination than the kinds of discrimination remedied by existing law.
…the people of the entire nation, not just the people of an individual state, have an interest in each state’s selection of federal representatives. Political gerrymandering ignores that interest…
I’ve saved the unfortunate news — from Cox — for the end:
[Whether] courts should continue to police congressional partisan gerrymanders turns importantly on the answer to empirical questions—about how likely it is that injuries will aggregate substantially, and how likely it is that courts will improve the situation. These questions are underexplored. …the legal literature rarely asks whether state-by-state manipulation regularly leads to substantial levels of national partisan bias. While the political science literature has engaged the question a bit more directly, it is somewhat divided on the answer.
He cites two articles from the political-science literature that do address the question — both co-authored by…Cox.
How does your state determine congressional districts?