CVD homepage
What's new?
Online library
Order materials
Get involved!
About CVD

The Economist

How to Rig an Election: In a normal democracy, voters choose their representatives. In America, it is rapidly becoming the other way around
April 25, 2002

Imagine a state with five congressional seats and only 25 voters in each. That makes 125 voters. Sixty-five are Republicans, 60 are Democrats. You might think a fair election in such a state would produce, say, three Republican representatives and two Democrats.

Now imagine you can draw the district boundaries any way you like. The only condition is that you must keep 25 voters in each one. If you were a Republican, you could carve up the state so there were 13 Republicans and 12 Democrats per district. Your party would win every seat narrowly. Republicans, five-nil.

Now imagine you were a Democrat. If you put 15 Republicans in one district, you could then divide the rest of the state so that each district had 13 Democrats and 12 Republicans. Democrats, four-one. Same state, same number of districts, same party affiliation: completely different results. All you need is the power to draw district lines. And that is what America provides: a process, called redistricting, which, through back-room negotiations too boring for most voters to think about, can distort the democratic system itself.

All countries, in the interests of equal representation, adjust their electoral boundaries to reflect population changes. Most democracies hand over this job to independent commissions, which content themselves with tinkering with existing boundaries. In America, in all but a few states, that idea sounds elitist and undemocratic. So every ten years, after the census, politicians in state legislatures meet to draw new voting maps which are approved by the state governor. Since America's population is both faster-growing and more mobile than that of other old democracies, and since the Voting Rights Act actually requires minorities to have special ìmajority-minority districtsî in order to get an equal chance to elect candidates of their choice (ie, their race), redistricters end up doing a lot more than tinker.

The results are as bizarre as you would expect. Florida's 22nd District is 90 miles long and never more than 3 miles wide. It consists of every beach house lining Route A1A along Florida's Gold Coast from West Palm Beach to Miami Beach. You could say about this district, as used to be said of the old Texas 6th (which was a road from Houston to Dallas), that you could kill most of the constituents by driving down the road with the car doors open. Other districts look like donuts, embryos or Rorschach tests.

But the champion gerrymandering comes from Illinois. Chicago has two Hispanic areas. They are in different parts of the city, but that has not discouraged the good politicians of Illinois from creating a constituency consisting of these two areas only. They lie on either side of a black part of the city like the bread of a sandwich. Worst of all is the state's extraordinary 17th District, which is a crab (see chart). Though most of it lies in the western part of the state, two claws stretch out towards the eastern part to grab Democratic cities in order to make the surrounding 18th and 19th districts more reliably Republican.

Weirdly shaped districts like these are signs that a crime has been committed. Again, start with Florida. This year, the Republican-controlled legislature has proposed a map with 18 Republican-leaning seats and seven Democratic ones. But as the 2000 presidential vote showed, Florida's electorate is split perfectly down the middle. The map has been rigged outrageously to favour the Republicans.

Florida is gaining population and seats. But it is just as easy to rig elections if your population is falling. Michigan, for example, will lose a seat this time. There, the Republican-dominated state assembly has managed to arrange matters so that six Democratic incumbent congressmen will have to slug it out among themselves for only three Democratic-leaning districts. Democrats will probably lose three seats in a state that Al Gore won.

Michigan also provides an extreme example of what clever redistricting can do for an individual. Mike Rogers represents the 8th District around the state capital, Lansing. He squeezed into office by a mere 160 votes in 2000, and had to wait even longer than George Bush for confirmation of his victory. The new redistricting plan tacks on a lot of Republican suburbs to his seat. So, after only two years, the man who won by the narrowest of margins in 2000 finds himself in such a safe Republican seat that no Democrat is bothering to challenge him in 2002.

Needless to say, Democrats are equally partisan. In Georgia they have drawn a map which means they will probably pick upómirabile dictuóboth of the state's new districts. And in North Carolina, long notorious for outrageous reapportionment, the chairman of the state redistricting committee is running for a new congressional seat that he himself mapped out.

And now technology makes it worse

Such things have long been staples of American political life. It would be too much to claim that redistricting has fundamentally altered any nationwide election result. But this year is slightly different, and in some ways worse, for two reasons. First, new software has made it easier to draw more ìreliableî electoral mapsóie, to be more exact in your partisanship. Until the 1990s, legislators had to draw districts using coloured pens on acetate sheets spread out on big maps on the floor. Computers appeared in the 1990s, but only big, sophisticated ones could handle the demographic data, putting the cost beyond all but a few states.

Now the Census Bureau puts out digitised maps, called TIGER/Line files. New geographic information systems for mapping and analysing demographic data cost only a few thousand dollars, work on ordinary Windows operating systems, and can draw up partisan maps automatically. This has turned gerrymanderingósorry, redistrictingófrom an art into a science.

Second, the 50-50 split in the 2000 election has changed what the parties want from redistricting. Under the old plans, you maximised your seats by drawing up districts which you would win narrowly. That was risky, because it gave your opponents a chance. Now the parties have adopted a policy of safety first. Because the House of Representatives is so closely balanced, legislatures try to maximise the number of safe seats for each side, drawing competitive districts only if they cannot avoid it.

In California, the Democrats in the legislature passed up a chance of grabbing risky seats from Republicans, and approved a map with only one competitive district out of 53 seats in Congress. That district is the disgraced Gary Condit's. ìIf the average Californian doesn't like his congressman,î says a Republican adviser, Dan Schnur, ìthe only option is to call the moving vans.î It is a similar story in the other big states that have issued their maps so far.

All in all, reckons Charlie Cook, a political analyst, with four-fifths of the states having issued their new district plans, there will be fewer than 50 competitive races this time (meaning races in which the candidates are only a few points apart) compared with 121 ten years ago. Of those 50, only half will really be toss-ups. This is worsening existing trends. In 1998 and 2000, nine out of ten winning candidates in the House of Representatives won with 55% of the vote or more. That was the lowest percentage of close races of any election year since 1946, save one. In other words, redistricting is becoming a glorified incumbent-protection racket. And that is having all sorts of odd effects.

For one thing, it means the Democrats probably cannot take over the House this year unless a miracle occurs. The House will be decided by the toss-up seats. Roughly half of them are Democratic, half Republican. To overcome their current six-seat deficit, therefore, Democrats will have to take three-quarters of the closest seatsósomething they cannot do unless there is a dramatic change in the national mood.

The 2002 redistricting plans are making an already change-resistant Congress even more immutable. Only six sitting congressmen were defeated in the general election in 2000, a re-election rate of 98%. Such a result, which would hardly shame North Korea, is becoming the norm: the re-election rate has averaged more than 90% since 1952. Not surprisingly, congressmen are reluctant to leave their warm nests. Only 28 have announced their retirements so far, compared with 64 in 1992.

The combination of larger numbers of safe seats and increasingly expensive election campaigns is undermining the quality of American politics. There are now two categories of House races: the overwhelming majority, where the incumbent is a shoo-in and which national parties ignore, and a tiny number of competitive races into which the parties pour all their money and energy. Of course ìall politics is localî. But in the current political arrangement, the local concerns of a handful of seats are inflated by a vast amount of national attention and end up deciding the balance of Congress.

Redistricting is also reinforcing a self-perpetuating quality in American politics. Incumbents anyway find it easier to raise money than challengers (House incumbents outspend challengers by five to one.) If they can make their seats safer by redrawing boundaries, they discourage challengers even more. And that in turn must depress voter turnout. The connection is not direct, since turnout usually depends on the races at the top of the ticketófor president or governor. But it is hard to believe there is no link between America's astoundingly high re-election rates and its astoundingly low voter turnout.

Putting it into cleaner hands

So what, if anything, can be done? Some states already use alternative systems that could be copied. Iowa lets civil servants draw new lines without reference to incumbents or regional voting patterns (rather as in Europe). Five other states hand redistricting authority over to bipartisan commissions, sometimes with a neutral tie-breaker approved by both parties.

Neither system works perfectly. But either would be better than the existing one. Both would limit partisan gerrymandering by removing debates about redistricting from legislatures, leaving them free to get on undistracted with their proper business, such as crafting budgets. Best of all, they do seem to work quite well. Washington and Iowaówhich use alternative systemsósaw more competitive House races in the 1990s, in proportion to their population, than other states.

Extending such practices would not be easy: politicians would naturally be reluctant to cede power. But even this barrier is not insuperable, at least in states which allow people to sponsor referendums. Citizens in Arizona, for instance, demanded a referendum to approve a redistricting commission in 2000, and, to the surprise of most experts, the measure passed. As the campaign-finance battle has shown, it is possible to reform America's electoral system, even if it takes years. And there are still years to go before the 2010 round of redistricting arrives.


top of page

Copyright © 2002 The Center for Voting and Democracy
6930 Carroll Ave. Suite 610 Takoma Park, MD 20912
(301) 270-4616 ____ [email protected]