Ace of Spades HQ

« Selective Declassification To Undermine the Bush Administration | Main | Sullivan Vs. Sullivan »

November 09, 2005

In Defense of Political Gerrymandering

Everyone seems to be in favor of ending the terrible practice of political gerrymandering, except for actual voters, who soundly defeated initiatives to end the technique in Ohio and California. So, let me defy the pundit class and argue in favor of the voters:

1. There Is No Such Thing As A Non-Political Gerrymander. Arnold's "solution" in California was to take the power of drawing electoral maps out of the hands of "politicians" and put it into the hands of "non-political" retired judges.

Non-political? Judges? Retired or not, judges are political. Not only are judges in many states elected through expressly partisan contests, but even in states where judges are appointed, they're appointed by a political actor (usually a governor, or a panel selected by the governor). Let's just say you have a much better chance of being appointed to be a judge, and promoted thereafter, if your politics agree with the dominant politics of your state.

Further, who would have appointed this panel of judges? Presumably the Governor, or a commission with members appointed by him. Call me crazy, but a Republican governor would tend to appoint Republican-leaning judges to this panel (or appoint Republican-leaning officials to the commission which would then pick which judges would draw the electoral map).

This isn't taking politics out of the process. It's simply making the politics slightly more labrythine but almost as predictable.

2. If the Sentiment of a State Is Strongly In Favor Of One Party Or the Other, Why Shouldn't That Party Get to Draw the Electoral Map?

In both California and Ohio, the voters didn't like the supposed anti-gerrymandering initiatives partly because Californians, being strongly Democratic, wanted to preserve the advantage the Democrats had in their state, and Ohioans, being less strongly Republican but still Republican-leaning, wanted to preserve the Republicans' advantage there.

It's not just "politicians conspiring to protect incumbents." It's the voters "conspiring" to protect incumbents. And when a Democratic state has Democrats drawing its electoral map, it elects more Democrats to Congress-- precisely what the voters want. Same for Republican states.

Get this-- very-blue-state California likes putting Republicans at a political disadvantage, and likes electing as many liberal Democrats to Congress as is mathematically feasible. That's what the voters want. Why shouldn't they have that?

3. Gerrymandering Actually Results In Non-Homogenous Pools of Voters Voting In One District... For At Least Most Districts.

Isn't this what the good-government types want? "Diverse" populations in the same electoral district?

It's a myth that, for example, gerrymandering results in nothing but monolithic districts, where 90% or more of voters share the same basic politics and therefore can be expected to reliably elect the most ideological of partisans year after year. They may do the latter, but not because the district itself is a 90% ideologically-pure monolithic block.

Why? Because to do so -- to pack a district with 90% of like-minded voters -- would be to waste 39% of those votes.

The purpose of gerrymandering isn't to create districts of 90% Republicans or 90% Democrats. If you did that, you'd actually have less reliable Congressional districts than you would from even say a purely random gerrymander. What you want when you gerrymander is to create districts which have a comfortable partisan advantage -- say 55-60% or so -- and not a single additional partisan voter in that district. Because those "excess" partisan voters can be used in another district, to try to make that district one in which a comfortable 55-60% margin of voters will reliably elect your guy.

Actually, when you think about it, the more natural way to draw electoral lines is to make districts of overwhelmingly liberal sentiment, overwhelmingly conservative sentiment, and overwhelmingly moderate sentiment. Shouldn't like-minded people be grouped together to elect their own? In such a regime, one could be pretty sure that any elected official from such a district was truly representing around 90% of his constituents -- because they almost all basically share his views, as well the views of their fellow voters.

Gerrymandering eschews this "natural" method of grouping like with like and instead attempts to use a comfortable majority of, say, liberal voters to beat-down an uncomfortable minority of conservative voters. Gerrymandered districts are mixed districts -- because it makes partisan sense to do it that way.

Now, here's the caveat-- that's what you do when you're seeking to maximize the number of districts in which you have a comfortable-but-not-overwhelming majority (which will lead, in most years, to winning the maximum number of possible Congressional seats).

You deploy your own troops -- your own reliable voters -- strategically, trying to maximize the number of districts in which they'll form a comfortable majority.

But for the other guy's troops -- well, in many cases, yes, gerrymandering does result in districts that are nearly 100% partisan. Because with regard to the other guy's voters, you want to have all those "wasted" votes. Let some districs vote 99% to 1% for a Republican -- they're still only winning a single seat out of all those votes. All those votes above the 51% mark are essentially useless.

But those sorts of arrangements can't be contrived very often. Where possible, the disfavored party will have as many of its voters crammed into as few districts as possible, giving them guaranteed victories in those few districts, and not a chace in hell of winning anywhere else. But geography and mathematics dictate that such super-crammed districts will be fairly rare. More common will be disctricts of a safe majority of 55% for the favored party and 45% for the disfavored party, doomed to perpetually lose by 8-10 points.

4. Highly-Gerrymandered Districts Do Result In Predictable Majorities... Until Catastrophe Hits.

Suppose you build a city whose codes require buildings to be resistant to earthquakes measuring 6 or less on the Richter scale. You can expect that your city will suffer fairly minor damage for any earthquake of magnitude 6 or less.

At magnitude 7, you'll lose about half of your buildings.

But at magnitude 8 or more, you'll lose much of your city.

Gerrymandering is very similar. You try to make districts that will be safe for your own candidates-- under normal circumstances. You attempt to build in a safe 55% advantage for such candidates in as many districts as possible, for that will maximixe the number of districts you carry -- again, under normal circumstances.

You don't try to make districts in which your candidate can expect 80% of the votes, because those excess votes are "wasted" and could be better used to pad the advantage in another district. You simply don't need such a huge partisan advantage to carry that district -- once again, under normal circumstances.

But what happens when public sentiment shifts sharply, and 55% is no longer enough? What happens when a party has so alienated the country that 5% of its leaning-voters -- voters it was counting on to make that crucial 5% margin of victory -- suddenly shift in sentiment and vote the other way out of disgust or in protest?

What happens, in other words, when a political earthquake of magnitude 8 or more on the Richter scale strikes? Then the carefully-constructed bastions, built to resist earthquakes of magnitude 6 or less and not an order of magnitude more, suddenly begin to wobble and then fall like dominoes.

That's what happened in 1994, pretty much. Gerrymandering wasn't quite as advanced then as it is now -- now we have powerful computers and sophisticated demographic software to aid in the drawing of political lines for maximum electoral gain -- but it was still pretty advanced, the art being 200 plus years old. But when the perfect political storm hit -- Jim Wright, the Post Office scandal, kited checks, gays in the military, Hillarycare, the Contract with America -- the sturdy-but-not-invulnerable electoral city of Democratic Congressional dominance came crashing down.

Less gerrymandered districts can be expected to have some amount of turnover year by year. More gerrymandered districts will have less turnover, and occasionally zero turnover. But in a political earthquake, a highly gerrymandered map will result in catastrophe for the disfavored party.

Which is not necessarily a bad thing. Congress will tend to remain in the same party's hands for a period of time, with few seats changing hands, until suddenly, a lot of seats change hands -- 20, 30, 40 or even 60 seats being captured by the other party, almost always resulting in a change of power. And sending an unmistakable signal to politicans that change is now officially demanded. Remember how President Clinton had to assert he was still "relevant" after 1994, and some began to speak of a return to late 1800's governance, when the Speaker of the House held the real power?

That was all hot air, of course, but the 1994 elections unmistakably changed politics in the next decade.

What sends more of a signal-- a fairly routine changing of the guard in minor ways every two years, or an enormous shift in power every ten or fifteen years or so?

And, on that point, I have to say: Republicans should not be quite as confident as they seem to be about how nicely they've managed to gerrymander districts in a majority of the states. If the party should come into as great disfavor as the Democrats did in 1994, they'll find that they're losing most of the carefully-drawn 55% majority districts they drew up in 2000 and 2001, and thereby losing the House as well.

A more sophisticated gerrymander is a gerrymander that will produce more victories under normal circumstances and many more losses under abnormal circumstances. Because the margins-of-error are smaller, by design. The expected margin of victory is designed to be smaller -- not quite razor-thin, but pretty thin -- so as to create another small expected margin of victory in another district.

If old-school gerrymanders resisted earthquakes of magnitude 6, or new, computer-assisted ones can only resist earthquakes of 5.5. And there may be tremors indicating such a strike is on the horizon.

Likely to happen? I don't think so. But it's not as unlikely as many Republicans seem to think. There is a potential perfect storm brewing -- high gas prices, dissatisfaction about the war in Iraq, citizens who still don't grasp that this country is growing at a healthy 3.8-4.0% clip, and out-of-control runaway spending -- that might make a lot of Repubican-leaning voters pull the lever for Democrats, and make a lot of strong Republicans simply sit the 2006 midterms out.

A lot of these factors are out of Bush's and the Congressional Republicans' control, but one -- the budget -- most decidedly is not. I would strongly suggest they take care of the one problem they have the most control over.

As for the rest-- well, there's always is the power of prayer.

posted by Ace at 01:14 PM

Comments

"Further, who would have appointed this panel of judges? Presumably the Governor, or a commission with members appointed by him"

It's worse than that. I'm pretty sure they would have been appointed by the legislature itself.

Posted by: Knemon on November 9, 2005 02:23 PM

Yup.

"Requires panel of three retired judges, selected by legislative leaders, to adopt new redistricting plan"

http://www.voterguide.ss.ca.gov/prop77/title_summary.shtml

Oh well, it's moot now.

I was 2 for 8 yesterday - 79 and 80 went down (they actually did even worse than Arnold's props).

Hey, it's better than Bob Shrum's record ...

Posted by: Knemon on November 9, 2005 02:25 PM

I'd rather have term limits.

Posted by: V the K on November 9, 2005 02:27 PM

"overwhelmingly moderate sentiment"

Overwhelmingly Moderate ... like Rudy. I like that.

Okay, I'll shut up for a while.

Posted by: Knemon on November 9, 2005 02:28 PM

Oh, brother!

Any redistricting plan drawn up would then be subject to a vote of approval by the voters. Is there a reason why you omitted that little factoid, Ace?

Posted by: on November 9, 2005 03:06 PM

If you want a non-partisan, deterministic, flexible method of automatically redistricting, they should be looking towards a mathematical solution.

Finite element analysis on the contour map from the Census department should break any state (or county, or whatever) into whatever number of desired districts are desired.

Of course, then everyone can argue over the math was done correctly, but at least then we can let our eyes glaze over. ;)

Posted by: Al on November 9, 2005 03:10 PM

For years now, there has been no reason to vote in local elections because the outcome has been predetermined.

There is no competiton among republicans, no competition among democrats, and no competition between demcrats and republicans. God forbid if politicians had to come up with ideas that worked. Now they just sit on their fat asses and their election is guaranteed.

We get stuck with shit republicans that are no better than the shit democrats. They sing one song to the voters and a different song in Washington. Our state is half way down the toilet because of illegal immigration and both parties are to blame. Nothing will change. But hey, thanks for your endorsement of the status quo. Go ahead, you can flush now.

Posted by: on November 9, 2005 03:12 PM

Any redistricting plan drawn up would then be subject to a vote of approval by the voters.

huh? not here. Nobody voted on the last plan that (finally) got pushed through.

And in some states (like here), judges are elected officials. The only appointees in Texas I'm aware of are State Supreme Court justices.

You're right on about the thinner margins of victory too - our district (the old Texas 11th, Chet Edwards - D) was split to try to take Chet down. Didn't happen.

Posted by: Dave in Texas on November 9, 2005 03:13 PM

Ace, I admire your attempt to be contrarian, which is especially endearing if you actually believe everything you just wrote.

Point 1: True, but there is such a thing as *less* political gerrymanders. You may disagree with the need to remove or mitigate the presence of partisan politics in the drawing of districts-- fine. But if the goal was to minimize the presence of partisan politics in this process, what process would your recommend? Who can play an impartial party here? Or is that such a misguided goal as to be ignored completely?

Point 2: You wrote:
It's not just "politicians conspiring to protect incumbents." It's the voters "conspiring" to protect incumbents. And when a Democratic state has Democrats drawing its electoral map, it elects more Democrats to Congress-- precisely what the voters want. Same for Republican states.
Uhhh, you want to rethink that point? We are not a parliamentary democracy, we are a representative democracy, with representation based on districts, not states. Allowing-- nay, encouraging-- a party to exercise its power to gerrymander an entire state to secure that party's dominance is anti-democratic.

I'd be curious to see whether or not all those 'fangled computers can draw a "perfect" partisan map in, say, California, and what that map would likely produce. My guess is it would be impossible to produce a 100% partisan state, but you could probably come close. How is a state with nothing but Democrats at all representative of voters if, say, 49% of the voters are Republicans? It's not.

Of course, this is what does happen all the time-- I'm not arguing it's wrong on technicalities, just wrong on principle.

Point 3: That's an excellent, oft-neglected point, you earn a cookie.

Point 4: Your argument is basically in favor of institutionalizing "temporal super-majorities," which is contrary to the Founder's intent. After all, we have House elections every 2 years not just for shits and giggles, but to allow for the real chance that seats can turnover.

You-- and the politicians in charge, albeit for more personally selfish reasons-- may be able to see merit in political stability of one-party control of Congress for longer than 2 years, but fundamentally, such interests are in contradiction with the electoral process.

Ultimately, this point is kind of moot anyway-- you're in effect trying to excuse the "sins" of gerrymandering by telling us that "It's not really that bad anyway, a big political shift can render gerrymandering ineffective." Well, duh.

I guess in the end the most important point to remember is that the Founders did not envision creating a representative democracy that would rely anywhere near to the extent that it does on political parties, let alone a two-party system. Thus, the present day interest in having a stable political system would probaly ever have occurred to them, or if it did, would have been anathema to their intentions of small-d democracy.

Besides, given the majority votes needed in Congress for most things anyway, I think even if we had a more "unstable" political system, with changes in party every two years, you'd still see less practical instability than one might think. After all, simply consider the presence of the filibuster in the Senate, or the need to get veto-proof majorities in Congress. You still need a super-majority to be effective in Congress.

Hell, remind me, which party is in control of Congress and the White House again? Does it really matter-- on a strategic level-- that we have 51 or 59 Republican Senators in Congress?

----
Anyway, Ace? Good topic to write about, love your stuff.

Cheers,
Dave at Garfield Ridge

Posted by: Dave at Garfield Ridge on November 9, 2005 03:23 PM

Any redistricting plan drawn up would then be subject to a vote of approval by the voters.

huh? not here. Nobody voted on the last plan that (finally) got pushed through.

Huh is right. We are talking about California:

Approval Process. In developing a plan, the panel would have to hold public hearings and could receive suggested plans from the public and the Legislature. Once the panel unanimously approves a redistricting plan, the plan would be used for the next primary and general elections. The Secretary of State would place the plan on the general election ballot for the voters to consider. If the voters approve the plan, it would be used until the next redistricting is required. If the voters reject the plan, another panel would be appointed to prepare a new plan for the next primary and general elections.

Posted by: on November 9, 2005 03:32 PM

For years now, there has been no reason to vote in local elections because the outcome has been predetermined.

I'm not sure how this would change. Under any redistricting scheme, local elections could become uncompetitive, unless you could somehow create districts that divide 50-50, (or 33-33-33, if you include independents) which would be impossible.

I live in a Democratic district. I can vote and hope for the best, but I short of moving there's nothing I can do about it. I don't see how this will change no matter how they redistrict it.

The way I see it, districts can be made that reflect voting patterns, or made somewhat randomly. Each seems to have strengths and weaknesses.

Posted by: Jason on November 9, 2005 03:37 PM

We are talking about California

lighten up Francis. Ohio is also mentioned. And gerrymandering certainly isn't limited to those two states. My "huh" was head scratching over the concept of voter approval of the maps, not a challenge to you personally, whoever you are.

Posted by: Dave in Texas on November 9, 2005 03:39 PM

Panel of Retired Judges. This measure requires that a three-member panel of retired federal and/or state judges (“special masters”) develop redistricting plans. The measure requires that the judges meet a number of criteria, including that they have never held partisan political office.

1. Judicial Council (an administrative body of the court system) collects list of retired judges willing to serve on a panel. The judges must not have:

Held partisan political office.
Changed their party affiliation since their judicial appointment.
Received income over the past year from specified political sources.

2. Judicial Council randomly selects a pool of 24 judges from the list of volunteers. The two largest political parties must have equal representation.

3. The four legislative leaders (two each from the majority and minority parties) nominate a total of 12 judges from the pool. The leaders each nominate three judges with party affiliations different than their own. Each leader is then able to eliminate one of the nominated judges.

4. From the nominated judges remaining on the list, three judges are selected at random to serve as the panel. Each of the two largest political parties must have at least one representative.

5. The selected judges pledge, in writing, to not run for offices affected by the districts they draw or accept public jobs (other than judicial or teaching) for the next five years.

This could include employment of legal and other experts in the field of redistricting and computer technology. Funding for the panel would be limited to a maximum of one-half of the amount spent by the Legislature on redistricting in 2001 (adjusted for inflation beginning after the 2010 federal census). For the first redistricting plan under the measure (to be developed for use at the next primary and general elections following the measure’s approval), the funding would be provided from the state General Fund.

Posted by: on November 9, 2005 03:42 PM

Dude, sorry, but if you'll allow me to threadjack for a minute, this is a hot news tip the board simply must check out:

John Derbyshire over at NRO just posted that Melissa Theuriau, who is apparently a French news vixen, has -- mercifully -- been left unscathed by the recent violence outside Paris.

All I can say is "Thank God."

http://www.fresh99.com/news-anchor-melissa-theuriau.htm

We now return to our regularly scheduled programming.

Posted by: Blacksheep on November 9, 2005 04:29 PM

Blacksheep, dude, you're *just* discovering Theuriau? Man, she's the best thing to come out of France since fries.

I'd love to snack on her frog legs, that's for sure.

Cheers,
Dave at Garfield Ridge

Posted by: Dave at Garfield Ridge on November 9, 2005 04:46 PM

Sorry, but this is the first time I've ever heard of her. But then again, everything is apparently old to you, Dave. Sure you aren't living backward in time like Merlin or something?

Posted by: Blacksheep on November 9, 2005 04:55 PM

Nah, not everything is old to me. But a face like her's, holy schneikes, it gives me faith in the greatness that once was France*

*You know, like before the Revolution.

Cheers,
Dave at Garfield Ridge

Posted by: Dave at Garfield Ridge on November 9, 2005 05:08 PM

Interesting ideas, Ace! I still voted yes on 77, though.

Posted by: SJKevin on November 9, 2005 05:17 PM

Way to stick your neck out AFTER the vote came in. At least I committed before the election... and took some grief for it as well.

But better late than never....

Posted by: steve sturm on November 9, 2005 05:58 PM

Great explanation Ace. I really hadn't thought the issue that far through as I don't have a problem with the practice. I am glad I understand it much better now.

Posted by: Silk on November 9, 2005 06:52 PM

I see one overarching flaw in the whole premise. Dave hit on this a couple times in his post but I would argue that its central to the whole discussion.

The framers never based their whole representative democracy thingy on a strict and exclusionary 2 party system where congressional districts would be redrawn to serve the interests of the two parties, not the interests of the people. But thats exactly what we are assuming here. Which ever party happens to be in power gets to redraw the district to best serve their needs, not the needs of the people, and punish the other party.

Districts were supposed to be drawn to incorporate groups of people with similar interests and then candiates would vie to be the one to represent them. We now have parties cherry picking people to be in certain districts (they have figured out that 55-60% rule in many areas) and building guaranteed wins. Worse, they are guaranteeing that there will only be two effective parties. In the worst case, in states with stable or shrinking populations, where tidal waves and earth quakes are rare, like the northest, you rarely see turnover. Look at MA. See a lot of anything but democrats in Congress....ever? Does this mean that everyone supports them? Hardly. Just means that the opposition has been effectively taken out of the process. As was stated up above, a great many republicans don't bother to run for congress anymore let alone vote . It may be a myth that gerrymandering results in monolithic districts but its a fact that it results in guaranteed results.

Yet there is a republican governor and has been for quite some time. Constitutional officers as well. How can this happen? Gerrymandering. If we are going to assume that we only have two parties, and for all intents in purposes we do, I would just as soon we divided up the number of seats in a state based on party registration. Let them be contested on a state wide basis. If there are 5 democratic seats, top 5 democrats get the seats.

Does this fundementally change our gov't and give us a body of essentially junior senators? Yep. And its worse than the silly gerrymandering games we play now with districts based on voting patterns rather than constituent needs, how? More people would feel they have a say in the process and vote and run. Parochialism would go down. Hell you might even encourage real 3rd party participation if you could get a grass roots campaign to register enough voters for a 3rd party seat. Money might actually have to play on a more even if still uneven field.

Radical concept but certainly no more radical than some of the ridiculous congressional districts being cooked up in back rooms of party offices.

Posted by: JackStraw on November 9, 2005 08:48 PM

The only appointees in Texas I'm aware of are State Supreme Court justices.

Actually, that's not true. I have an uncle who recently retired as a family court judge in East Texas. He had been originally appointed to the bench by Mark White because of his years of work for the Democratic Party in East Texas.

Point 1: True, but there is such a thing as *less* political gerrymanders.

Dave, Dave, Dave, Dave, Dave. The entire point of gerrymandering is political. The defniniton of "gerrymander:"

To divide (a geographic area) into voting districts so as to give unfair advantage to one party in elections.
- The American Heritage® Dictionary of the English Language, Fourth Edition
Copyright © 2000 by Houghton Mifflin Company.

Now, if you mean that it is possible to redraw districts without it being political, you are correct.

Posted by: Steve L. on November 10, 2005 08:27 AM

Excellent analysis. Too many people on both sides are trying to take the politics out of what is essentially a political matter.

Voters arn't machines and it's doubtful that any sort of non-political mechanism can be built that would be truly free of political bias.

Posted by: Carl W. Goss on November 10, 2005 01:36 PM

Ace:

in the 2004 California elctions (and, BTW, CA has strict term limits for its state legislature) not one seat changed party hands.

Not in the state legislature, and not in Congress.

It would be impossible for any plan to be more Gerrymandered than the current situation.

If you want a government that is responsive to the people, then you want a government where the elected officials feel like screwing up can get them kicked out of office. California does not have such a government. And it shows.

Posted by: Greg D on November 13, 2005 04:39 PM