The 50/50 gerrymander

by Ted on December 11, 2003

Gary Farber has a thought experiment posted about a mandate requiring Congressional districts to be drawn to create districts that are as competitive as possible.

That is, the goal in drawing district lines would be that all districts be as evenly divided between likely Republicans and Democrats as is predictable. You know, the opposite of the way gerrymandering has been functioning, overall, since the days of Eldridge.

It’s obviously not going to happen, but it does set one a’thinkin.

The House is filled with representatives in completely safe seats- they represent districts whose residents are so overshelmingly Republican or Democratic that the representative can operate without fear of being defeated by the other party. To an unpleasant extent, they don’t really have to answer to the public as long as they stay in the good graces of their party. The obvious advantage of Gary’s proposal is that it would sharply reduce the safety of these seats. When every seat is competitive, party loyalty would presumably weaken, as representatives would be significantly less likely to stick with the party in the face of public opinion. (The Senate tends to look less like kindergarden, because Senators are less likely to be totally safe from a challenge from the other party.)

Sometimes I worry that the House is losing the ability to even try to solve the problems that face the United States. This kind of radical shake-up would change our current system, forcing compromise as a default mode. Tom DeLay’s unbendingly partisan and conservative strategy can be described in many colorful ways, but one of them is “re-election strategy”. He can keep his seat forever without offering an inch to moderates, Democrats or liberals. Gary’s plan would change that; DeLay have to offer something to moderates or lose his seat. So would Nancy Pelosi. Right now, they just have to win, by hook or by crook. This plan would mean that they’d have to negotiate, and hopefully produce outcomes that would be in the interest of the most people.

However, the increase in moderation would be paid for by some awfully unrepresentative districts. Quite a few communities in the United States genuinely swing right or left without any help from gerrymandering. Before I started thinking about this, I would think that a Platonic ideal of a districting map would be blind to party voting patterns; it would be built around keeping communities together. If we achieved that, there would be a lot of natural safe seats. (I can’t imagine how to achieve competitive gerrymanders in Utah, for example.)

Furthermore, it’s not as if liberal and conservative districts are laid out in a checkerboard. On the contrary, Democratic voters are more likely to live in densely-populated cities, and Republican voters are more likely to live in less-populated rural areas. Truly competitive districts would have to be created with some awfully wiggly gerrymanders exploding out of cities.

I wouldn’t envy the job of a representative of one of these 50/50 districts. In Houston, let’s say that you’re turning Tom DeLay’s safe suburban Sugarland district into a 50/50 one. So you lop most of Sugarland off and make a snake into a poorer Hispanic district in southern Houston. The partisan gerrymanders do have a saving grace to them; a representative knows what the interests of his district are. How do you represent this new district?

Flash forward into the future 20 years. Let’s say that the gerrymanders are a fabulous success, and the House is filled with moderates. Wouldn’t that seriously distort the meaning of a history of a district voting Republican or Democrat? A popular long-serving representative would gradually force himself out of the district he was representing, as the new gerrymanders cut out his base and grabbed more territory from the other party.

Then there’s the old “who watches the watchmen” angle; who do we trust to create these gerrymanders?

I’ve probably said five dumb things already, but it’s late. On balance, I’d have to say that I’d be opposed to this plan, but the right commentator might be able to shake me out of it.



Katherine 12.11.03 at 5:36 am

Opposed, though it’s probably slightly better than the current mess.

1. How on earth do you this and keep adjacent districts? Parcel out Manhattan to Schenectady a few blocks at a time? I just doubt it’s even possible.

2. Party registration does not mean the same thing in all parts of the country. New England Republicans are more liberal than some southern Democrats. If that’s true of politicians it’s doubly true of voters.

3. There’s nothing to be done with one district states, but then they’re really not the problem. The large states that vote overwhelmingly one way ARE a problem. Say Texas is 60% R, 40% D. Does this mean you make 30 districts that are 60% R, 40% D? That would be just fine with DeLay; it’s not so far from what his latest plan tried to do–splitting Austin into teeny bite sized pieces. Or do you try to have as many districts 50-50 as possible and have a few lopsided ones to make the difference? But then where do you draw the lines? You’ve got the same untrustworthy people doing it.

4. If you’re actively avoiding having districts represent likeminded communities, you might as well switch to proportional representation.

5. This might well reduce the number of minorities in the House to close to zero. I don’t know if that would violate the Voting Rights Act, but I think it’s a bad thing in any case.

How does England do Parliamentary districts?


Chris Clarke 12.11.03 at 5:42 am

We just had a mayoral election in San Francisco that was pretty evenly split between Dem and Green, with no GOP candidate. How would San Francisco fit into this scheme?


Brian Weatherson 12.11.03 at 5:54 am

I don’t think we should aim for 50/50 in each seat. You want to have lots of competitive seats to make the House reflect the country, but if a region has a political tendency, that should be easily reflected in who it sends to Congress. As long as there are lots of competitive seats, then even if the leadership are in safe seats, they won’t stay as leader very long if they put several of their colleagues at risk, so Ted’s worries can be addressed without creating 435 marginal seats.

I think we should aim for making each district as compact as possible, with perhaps some preference for having boundaries reflect ‘communities of interest’. (So it’s a good idea to have prominent geographic features, like rivers or highways, divide districts, especially in urban areas.)

Here’s the principles the Australian Electoral Commission uses:

* community interests within the division, including economic, social and regional interests
* means of communication and travel within a division
* physical features and area
* existing boundaries of divisions
* enrolment (both current enrolment and projected enrolment three and a half years after the redistribution).

Note that potential or real political implications are not considered in any way by the Committee.

The link is to a fairly wide-ranging FAQ on how boundaries are drawn in Australia. I think it works fairly well. It certainly produces a reasonable share of safe seats.

It might not look very meaningful to those outside the area, but here’s a map of the electoral districts for Victoria. Lots of marginal seats, few silly looking electorates, lots of seats where you can quite naturally take about the character of the electorate, just as we’d want. Lots of safe seats too, but only where the character in question is filthy rich or fairly poor. (Note that most of the squiggly boundaries are so either (a) the boundaries can follow rivers or (b) to make the electorates the same size.)


Joshua W. Burton 12.11.03 at 6:02 am

Stop making districts entirely; move the problem one category up the abstraction ladder. Have the state legislature vote on districting _algorithms_, nine years ahead of time, and then let the bits fall as they may at the next census, by the blind application of open source code to standard-format data. Any partisan Trojan horses in the code will (1) probably backfire due to demographic surprises, (2) be exposed and sit out there as objects of ridicule for years before they kick in, and (3) lack the political immediacy of post-census districting, and thus not be worth their prompt political cost.

Second best would be to let the legislatures do exactly as they like. The Dems gave away the House in the two decades leading up to 1994, by robbing a lot of swing seats to build a few minority-solid ones. With freedom to bungle, I have confidence that Tom Delay will in time return the favor, probably by misunderstanding that collar counties around US cities are swinging left as they age.


Matt McIrvin 12.11.03 at 6:06 am

You don’t want to *force* a ratio of 50/50; I doubt it’s even possible, for the reasons others have stated.

I tend to think that as much of the districting as possible should be done by an algorithm requiring little human input. Of course people could and would still try to game things by legislating changes to the algorithm, but at least everything would be on the table.


Dan Western 12.11.03 at 6:45 am

I think there’s some wisdom in one of the ideas above. Perhaps we need a compromise: voters should vote as “natural communities” as often as possible.

Where a “natural community” isn’t possible, a degree of defense should be necessarily proportionate to the bizarreness of the district, and should be required to be done in order to make the district more, not less, evenly spaced.

Thus: “Distict 12: Brooklyn.” No problem. “District 17: Bits of Manhattan, right along the Jersey border, than a u-turn out to Schenectady.” Problem, needs to be defended.


Chris Bertram 12.11.03 at 8:21 am

I’m for some version of PR myself. Anyway, a big drawback of the Farber proposal (if I’ve understood it correctly) is that a small but uniform swing in the national vote would produce a landslide for one side – highly undesirable.


Brian Weatherson 12.11.03 at 8:41 am

Yeah, for the US House it’s hard to see a good reason not to use PR. The main problem with using PR for, say, the House of Commons is that it leads to too much instability. (We could debate for ages whether this kind of instability is a good thing or bad thing in the Commons. I think it’s a bad thing, but that’s a different debate.) But this simply isn’t an issue for the US House. The only way the House can lead to real instability is if one side has firm control and resolves to use it as forcefully as possible. (See Gingrich, Newt.) A PR system would lead to exactly the kind of messy horse-trading kind of system that characterised the House for most of the last century, but probably without some of the most destructive features.

But when it comes to reforms to the US political system, “too good to be true” is too often literally true, so don’t hold your breath on that one.


Doug 12.11.03 at 9:52 am

vox diaboli: I thought that over here on the left side of the isle, party discipline as seen in European parliaments was often praised as something we needed more of in the US?

Also, take a moment to consider the large parliaments in Ukraine and Russia in the early years of the transition – almost no partisan attachment. But the vast swathe of non-partisan deputies was known as The Swamp for a good reason, and proved far more malleable in the hands of a strong president than more partisan bodies in other transition countries.


Greg Hunter 12.11.03 at 12:21 pm

I would like to return to the original intent of the electoral process and have my congressman/senator cast my vote for President. The process would put added pressure on the elected officials “to pick the correct person”. If they chose poorly, the voters would more likely throw them out of office.

The splitting of communities in the re districting process is inherently wrong and inefficient from a community standpoint. It exacerbates the splits that are already manifested in the community and requires the community to get 2 or more legislators on board.


Jeremy Osner 12.11.03 at 2:10 pm

What about this idea, which has been nagging at me for a while: make the house of reps about 4 or 5 times larger than it is now, and each district correspondingly smaller. This would have two principle benefits to my way of thinking: it would reduce the degree by which sparsely-populated states are overrepresented in the house; and it would make each representative’s constituency a smaller, more easily identifiable group.


Alan Peakall 12.11.03 at 2:17 pm

Isn’t the argument that ‘the splitting of communities in the re-districting process is inherently wrong’ the rationale for the minority-majority districts that are some of the most geographically absurd gerrymanders? Or by community do you mean ‘residential community’ and not (for example) ‘black community’?


sennoma 12.11.03 at 4:46 pm

I like Jeremy Osner’s suggestion, which seems to mesh nicely with the “natural communities” idea as well. In the case of relatively large “natural communities” like, say, Manhattan, these could be split geographically or simply allocated extra reps according to population.


ucblockhead 12.11.03 at 4:53 pm

Isn’t this antithetical to the idea of a “representative” democracy? I mean, if a district is created that is 80% Republican, then that representative represents the views (roughly speaking) of 80% of his constituents. On the other hand, if you attempt to make all districts 50-50, then most representatives will only represent the views of a little over half of their constituents. Taken over all of congress, this means a less representative body. If you are not going to try to make congressmen representative of physical areas, I think you’d be better off giving up the idea of physical districts entirely.

It seems to me that having people in congress who represent the views of their constituents is more important then “competitive” elections. If we don’t like the views of certain congressmen who stick around forever without having to even really campaign, maybe we need to admit that what we really don’t like are the views of the people in the district that congressman represents.

Personally, I think a good argument could be made for drawing districts in the exact opposite manner, so as to magnify the number of mostly liberal and mostly conservative districts. (Which is not at all the way political parties want to gerrymander, as a 60% majority is nearly as safe as a 90% majority.) This would increase the number of people overall who would vote the way their congressman would vote on any particular issue.

That’s the idea behind representative democracy, isn’t it?


Katherine 12.11.03 at 4:58 pm

“I would like to return to the original intent of the electoral process and have my congressman/senator cast my vote for President.”

This momentarily sounded crazy to me, and then I realized it’s a lot like a parliamentary system, which most of the world has. (Though with more scheduled elections.) I don’t think it’s the intent of the U.S. electoral process though–separation of powers and all that.

Nor do I think we want a 2000 member house of reps. It’s just too big.

The recent New Yorker article I read on this topic said that it was once thought that gerrymandering would sort itself out, but now the software is just too good. (They did note the Democrats’ strategic blunder on majority minority districts.)

I do like PR, either nationally or by state; and either as a supplement or substitute for first-past-the-post. But it requires a Constitutional amendment, and everyone will speak of virtues of your own local district (and Congresspeople are often good at constituent services), and it will never happen. I don’t think the Supreme Court will actually step in either. I think the only hope of change is a Congressional statute or a bunch of state statutes switching to the Iowa method (which sounds similar to the Australian one). An algorithm might be ok too; I don’t know enough about the technical side of that one. I’d prefer this happen on the national level, because if it’s on a state level it punishes the party who does the right thing in the legislatures they control.


Matt Weiner 12.11.03 at 5:04 pm

Ted–I agree with you about keeping communities together–no way could you get a competitive district in Pittsburgh–but Utah currently has one Democratic Representative, Jim Matheson, to two Republicans. I wasn’t here when that happened, but according to this site he was first elected in 2000 and the Leg. tried to gerrymander him out by including more conservative areas and less Salt Lake City. Obviously it didn’t work.


Matt Weiner 12.11.03 at 5:09 pm

I don’t know if the Dems can be completely blamed for building minority districts. States are legally obliged to create them under many circumstances.
I don’t much like proportional representation, because I think that it could leave large swathes of states unrepresented. Wouldn’t the parties load up the list with candidates from swing areas? I agree with Matt McI.–go for a nonpartisan process like they have in Iowa.


Greg Hunter 12.11.03 at 5:41 pm

The electoral process was the intent of the original framers of the US system and they did not have any “separation of powers” issues. I cannot foresee any problem with returning to the original intent (senator/congress vote for president). This process would make the electorate focus ever more intently on the person placed in office and the relationships that he/she formed.

Increasing the representatives would help have more diversity, but I think the gerrymander splits should be “hard and fast” so that a community is not split.

For example if a borough of NY has enough population for a Rep. then draw the line around the borough. If the 5 boroughs have enough for four Reps., then put two together and have one Rep. I know there are problems, but there would be many more benefits, especially a cohesive voice from the local pols to the DC pols. The same could be done for Counties.


Pouncer 12.11.03 at 6:24 pm

I think it would be interesting to step back to requiring state legislatures to choose U.S. Senators. A strong majority party in the state would choose a party member for the federal level, of course, but a split legislature would have to come up with a more popular, moderate, conciliatory representative. And shifts in local interests, over the course of three or so local elections, would allow gradual change over SIX years for their Senator. It might easily be in a divided state that two similar politicians trade a Senate seat back and forth as one or the other party gets a majority in the home legislature …

But for the Federal House, I’m about inclined to demand that Representatives be selected the same way states select jurors — more or less randomly from the voter’s registration lists. “Representative” is the ideal, after all.


Katherine 12.11.03 at 6:28 pm

Hang on a minute. The Constitution never provided for Congress to vote for President except to break a tie or when no one has a majority of the electoral college. State legislatures were to choose the electors (or rather, to determine how the electors were chosen), who were to choose the President. U.S. Congressmen and Senators could not be electors. If no one got a majority, it went to the House, but with one vote per state.

From Article II, section I:
“Each State shall appoint, in such Manner as the Legislature thereof may direct, a Number of Electors, equal to the whole Number of Senators and Representatives to which the State may be entitled in the Congress: but no Senator or Representative, or Person holding an Office of Trust or Profit under the United States, shall be appointed an Elector.

The Electors shall meet in their respective States, and vote by Ballot for two Persons, of whom one at least shall not be an Inhabitant of the same State with themselves. And they shall make a List of all the Persons voted for, and of the Number of Votes for each; which List they shall sign and certify, and transmit sealed to the Seat of the Government of the United States, directed to the President of the Senate. The President of the Senate shall, in the Presence of the Senate and House of Representatives, open all the Certificates, and the Votes shall then be counted. The Person having the greatest Number of Votes shall be the President, if such Number be a Majority of the whole Number of Electors appointed; and if there be more than one who have such Majority, and have an equal Number of Votes, then the House of Representatives shall immediately chuse by Ballot one of them for President; and if no Person have a Majority, then from the five highest on the List the said House shall in like Manner chuse the President.”

It goes from there about the procedure if it goes to Congress. (This was partly superseded by the 12th Amendment, but if we’re talking about the original text of the Constitution here it is.)


russ e 12.11.03 at 7:28 pm

Congressional districts, and state legislative districts, should be based on historical communities and geographical areas. And the maps should be drawn by citizens commissions like they are in Iowa.

In 2002, 4 of 5 Iowa districts were competitive, 20% of the total competitive races in the country. Their map looks like a vertical line and a horizontal line through the middle of the state with a bullseye on Des Moines. Compare that with the gerrymandered nonsense in Illinois.

Rural districts should not split counties, suburban districts should not break up municipalities and urban districts should respect community boundaries.

The problem with this is a dogmatic, literal interpretation of “one man, one vote.” Districts should be allowed to vary by a few percentage points in order to allow boundaries that respect historical communities. Otherwise you’re constantly hiving off a few precincts or wards here and there to reflect minute changes in population distribution. We have tolerated one representative for every 900,000 Montanans and one for every 500,000 Illinoisans for years.

The current system has turned the House of Representatives from the most changeable chamber, reflecting the changing moods of the people as the Framers intended, to a hidebound nest of entrenched incumbants.


Brian Weatherson 12.11.03 at 7:41 pm

I don’t get Matt’s comment about loading up on candidates from swing districts. The thing about PR, at least once you get to a large enough system, is that every area is a swing area. It really matters whether the Democrats get 85% or 88% of the vote in Manhattan. Indeed, it even matters whether the Republicans get 63% or 66% of the vote throughout Texas.

I suppose there’s an empirical assumption here that I don’t really believe – that some geographic regions are significantly more electorally volatile than others within a state. Maybe there’s some evidence for that, but I find it hard to believe except right at the fringes.


Gary Farber 12.11.03 at 7:46 pm

This is a side issue, but Chris Bertram says: “I’m for some version of PR myself. Anyway, a big drawback of the Farber proposal (if I’ve understood it correctly) is that a small but uniform swing in the national vote would produce a landslide for one side – highly undesirable.”

I’m not sure it’s axiomatic that this is highly undesirable. The argument for that outcome is that it allows a clear policy to be established for a time and for a government to have the strength to take strong action. Inevitably there will be a return pendulum swing of opposing ideology and corrective measures will be taken to deal with any excesses.

On Chris’s other point, while I have the greatest respect for the reasoning of the many fine minds who argue for PR, I’ve always seen reason to view it with something close to horror, based upon the vivid examples we have of it working in practice, namely, Italy and Israel, where anyone familiar with their Parliamentary workings would want to run screaming from the room at the idea of taking up such a system.

PR demands a highly unstable government creating by sewing together a coalition of disparate minority bits, which pretty much mandates an inherently corrupt process that rewards extremist views that hold a balance of power, and yet tends to fall apart with great rapidity in any event. It incorporates practically every flaw there is: instability, corruption, stasis, paralysis, self-contradiction, incoherence, and extremism, for the bargain of being able to say that it is “more representative.” It makes a very good case for the value of a certain degree of “small r” republicanism over pure democracy.

On the question of shapes of districts under my proposal: have any of the questioners any familiarity with the American system of the last fifty years? Districts that look like hyper-space squids with threads that may run over a hundred miles but be one hundred feet wide, twisting and turning in a Rorschach blot beyond imagining are quite the norm. That’s what we have. So objections to that under a system using said mapping to achieve competitiveness can’t possibly say it would be moving to a more flawed system for that reason.

If I didn’t say that my system would include mandating adherence to the Voting Rights Act, I meant to.

From the comments here, my proposal doesn’t seem to meet with anyone’s favor, or even favour. Does no one see any virtue to forcing races to be competitive (note: that the races will be competitive doesn’t mean that the results will necessarily come out any particular way), and thus encouraging genuine consideration of competing idea, as well as encouraging moderation and compromise? Right now, only about 35 out of 435 seats in the House are at all competitive, and a number have become hereditary. Doesn’t that bother anyone?

I thank very much Ted for the link. I do think you over-estimate the effect of this on party, however, Ted. As it is, candidates are comparatively independent of party precisely because of the safety of most seats and the overwhelming power of incumbency. Once you are in office you have great — though not total, to be sure — power to raise money on your own, and to distance yourself from your party to a large degree, and thus perpetuate yourself with little help from your party, in many districts. This is one of many huge differences between the American and parliamentary systems.

“To an unpleasant extent, they don’t really have to answer to the public as long as they stay in the good graces of their party.”

It’s not necessarily at all so much the party, as staying in the graces of the interest groups/corporations/sources-of-money they choose to pander to/represent.

Needless to say,it would be preferable if they were Representatives of the people of their district rather than of sources of money.


epist 12.11.03 at 8:44 pm

It seems to me that the problem with district-centered national votes is that they encourage a pork-barrel mentality. When you vote for ‘your’ rep, where the pronoun denotes a neighbourhood, you should consider what the effects on the neighbourhood would be for the different candidates. But the effects on your neighbourhood can be very small, and at odds with the national interest.

This is precisely the dynamic that underpins pork-barreling. If a rep is successfull in bringing government money into his district, he will increase his chances of re-election, since his constitutents will be pleased by the extra wealth. But the nation is not best served by such a policy (for a variety of, I hope, obvious reasons).

I understand the appeal of district representation; i.e. neighbourhoods and communities are important democratic blocks, and should have voices. But it seems to me that the best expression for local voices is in the state legislatures. Let the people vote for state candidates that will represent the local interests of their districts, and reshuffle the way the Feds dole out money to let the state governments have a voice. That way we can clear the federal elections for reps that serve the interests of much broader sections of society (classes, genders, race and ethnicities, occupations, age cohorts, etc.). These groups would still nominally be bounded by state borders, but since the parochial interests of individual communities would no longer be (at least) the dominant factor, voters could cast their lots with issue-candidates without feeling like they betrayed the interests of those in their communities.

And as a result, the House would look less like a horse-trading parlour, and more like the pointy end of the war of ideas and ideologies that demoacratic government is supposed to be (at least by my lights).


Kilroy Was Here 12.11.03 at 9:43 pm

The simplest answer here is to abandon the one district-one representative model. Instead, have ‘super-districts’ that have 5 or 10 representatives, and then use preference voting (or ‘choice voting’) to determine the representatives. For more info, take a look a the Center for Voting and Democracy’s website.

Simple and easy. How can we get it done?

Kilroy Was Here


Jim Flannery 12.11.03 at 11:25 pm

Regarding the San Francisco mayoral race: The results of the vote really say nothing about this proposal, which addresses party registration: the vote was 53% Dem/47% Green, but the registration split is roughly 80% Dem/3% Green. What’s interesting about this result is the fact that likely more than half the democrats who voted jumped party lines (precinct-level results aren’t out yet but it’s hard to see those pockets of republicans voting for Gonzalez) because they saw the Green as the Democrat in the race.

This goes to what strikes me as bizarre about this proposal: it’s intended to produce “more compromise” … I certainly don’t think we’re in danger of running out of compromising Democrats, and haven’t been for the last 8 or 9 years. Democrats seem motivated enough to pander to the right, even in the “safest” districts, without any structural help.


ahem 12.12.03 at 12:50 am

On the other hand, if you attempt to make all districts 50-50, then most representatives will only represent the views of a little over half of their constituents. Taken over all of congress, this means a less representative body.

Um, no. The basic principle in parliamentary systems, at least, is that you represent every member on an equal basis when in comes to constituency matters. So if you’re a Tory in Sedgefield or a socialist in Folkstone, you can write to Blair or Howard and have him deal with the wonky pavement that the council won’t fix. The with the US federal system is that congressional districts are too big for Representatives to ‘represent’ individuals, so they end up beholden to lobby groups.

Anyway, experience suggests that representatives in 50-50 districts are actually much more representative of their voters, simply because they can’t assume that they’ll get in next time and have to prove that they’re not simply toeing the party line if that line ignores the demographics of the electorate.

But: one big problem with Congressional districting is that it lowers the barrier to entry, and thus produces a House that is almost entirely made up of pork-seeking numbskull. The notion of there being so many uncontested House elections would horrify anyone from a country with a parliamentary system.

Let me explain: in Britain, any prospective MP has to fight one or two losing battles — it’s known as ‘blooding’ after the initiation rite at a fox-hunt. You get selected for a constituency in which your party finished third, and have to prove yourself on the stump by either keeping your deposit or getting the runner-up spot. The ones who come out of that experience most promisingly get to stand for safer seats next time. (Tony Blair, for instance, lost in true blue Beaconsfield before getting Sedgefield, where the populace would vote for a pig with a red rosette.)

By contrast, most Congresscritters have never fought a tough election in their life. (I give you Rep. Katherine Harris, Jeb Bush’s fiddler in more ways than one.)

This helps account for the fact that some of the doziest constituency MPs in Britain outshine 90% of U.S. Representatives. Of course, that sometimes causes problems: lots of ‘bloodees’ in 1997 got elected thanks to huge swings against the Tories, and a lot of them aren’t very good.

So, I’d say that House elections could at least have RON as an option. If the other main party won’t compete, then give the voters the chance to chuck out the incumbent.


GFW 12.12.03 at 1:58 am

If you think it’s bad now, wait till the rest of the country catches up with California.

Why not have the House membership made up of:

One half: Iowa-like compact districts.

One half: proportional representation with a minimum threshold for parties to get into parliament (5% is the standard one in Europe)

Just like the Germans in their great stable system, you’d vote twice. Once for a representative, and once for the party you like most. The House is still weighted heavily to the two party system because of the single member districts (and of course the Senate is still the same), so the system remains stable.

But, lo and behold: viable third parties. Party lists that look like America (minorities and women included!!!). And you don’t have to Gerrymander the districts. Parties discipline would be strengthened (since you have toe the party line to get on the list), platforms become more important (since you have to attract voters on the second ballot mainly based on party attractiveness). Since every vote counts toward a party list (not like the Democrat’s vote in DeLay’s single member district), voter turnout reaches European levels. All sorts of wonderful things happen when we switch to the system that every industrialized democracy has except for Britain, the US, and Canada.

See Robert Dahl for more.


Jason McCullough 12.12.03 at 5:35 am

This “solution” suffers the same outcome-based reasoning as the problem. Congressional districts were created so you’d have a single representative for a (mostly) homogenous group of geographical people to represent their views. The various district compactness solutions seem a lot more compatible with this.

I really like what Dan said about natural communities.


John 12.12.03 at 6:20 am

(I give you Rep. Katherine Harris, Jeb Bush’s fiddler in more ways than one.)

Was Florida Secretary of State an elective office? If so, that must have been a competitive race.


russ e 12.12.03 at 2:45 pm

The solution to gerrymandering is not to try and change a bedrock of the Constitution. It is about as likely as getting rid of the electoral college. The way to end gerrymandering and get a more representative House is to–end gerrymandering, the original meaning of which is to draw a district to favor a party or individual.

Natural communities are the only way to district congressional seats. In Chicago the wards are gerrymandered every census every which way and most citizens are hard pressed to tell you what ward they live in. They do know which community they live in, however. The disconnect between natural community and gerrymandered government boundaries confuses and essentially disenfranchises many citizens.

Our statehouse districts and congressional districts are a morass as well.

If we divided Illinois’s 18 House seats in a sensible manner, with some allowance for differences, Chicago would share its 4.2 seats with suburban Cook County’s 3.6 seats [7.8] and share 8 reps. Lake County’s .9 [North burbs] would get 1, McHenry and Kane share 1 [NW burbs], Dupage and Will share 2 [W and S burbs], and five continguous counties around Rockford with .75 would be the core of another district. That takes care of 13 seats and corresponds to real communities and regions. 1 seat would be centered on the two Illinois counties across from St. Louis [.8] and 2 seats broadly in the Peoria, Springfield, Champaign triangle [1.65]. That would leave all other counties in Illinois to fill out some of the other districts and share 2 seats between the Mississippi-Illinois valleys and the Ohio-Wabash valleys.

As far gaining locks on districts, parties change, issues change, communities change demographically. Every ten years some states will have to redistrict simply because they’ve lost or gained a seat.

Making districts comprehensible eases the citizen’s task of figuring out who’s accountable.


Matt Weiner 12.12.03 at 7:10 pm

Brian–The thought was this:
Suppose 1/3 of the state’s voters will always vote Democratic, 1/3 of the state’s voters will always vote Republican, and 1/3 of the state’s voters will base their vote on the candidate list. The parties would want to craft their list so as to appeal to the swinging 1/3. If the swinging 1/3 are unequally geographically distributed, then the regions with high concentrations will be unequally represented. For instance, my guess is that in Pennsylvania the lists would have lots of suburban candidates, because the GOP can count on most of the rural voters and the Democrats can count on most of the urban voters.
My model is obviously oversimplified–you have to get people to show up. But I also worry about how primaries would work under PR. latest proposal seems like it might solve that; I’d have to think it through.

Comments on this entry are closed.