More than 90 percent of the nation’s voters will go (or not go) to the polls on November 5 knowing that, as far as the House of Representatives is concerned, the elections in their districts will be largely a symbolic exercise. The main reason is that the winners in most states have been predetermined by the state officials and party operatives who drew the congressional district lines.

Political gerrymandering-usually done to protect incumbents-has become so pervasive that it’s easy to drift into the incorrect assumption that both its causes and its poisonous effects are irresistible historical trends. Among the effects:

• A steady decline in the number of competitive House races, from about 151 a decade ago to 44 this year, according to my colleague Charlie Cook. (California, with 53 districts, has only one competitive race; Iowa, with five districts, has four, thanks to a depoliticized redistricting process.)

• Difficulty finding good candidates in the more than 90 percent of House districts in which the minority party’s nominee appears to have little chance.

• A perception among voters that their House ballots do not really count.

• A House of Representatives that is unrepresentative of the nation’s largely centrist population, with most Democratic members well to the left and most Republicans well to the right of center, because they fear challenges only in the primaries, which are dominated by liberal Democratic and conservative Republican voters.

• A commensurate increase in partisan rancor, resulting from the clash of liberal and conservative ideologues.

• Ever-fewer moderate centrists, who do best in truly competitive districts.

• Waning interest in the compromises necessary to make Congress work.

The causes of these destructive trends are less familiar. But at least to some extent, the trends are the unintended consequences of a succession of well-meaning but clumsy interventions into our electoral system over the past four decades by the Supreme Court.

First, since the 1960s the justices have carried their "one-person, one-vote" rule to mindless extremes by making a fetish of strict numerical parity among House districts. One unfortunate effect has been to increase the number of opportunities for political manipulation by requiring that every House district in the nation be redrawn after every decennial census to reflect even the most minor of population shifts. Another effect has been to legitimize disregard for the traditional districting standards-integrity of political subdivisions, natural and historical boundaries, compactness, and contiguity-that had been the political system’s only built-in restraints on incumbents’ efforts to manipulate district lines to their advantage.

Second, in 1986 the Court’s liberal majority announced an absolutist and unbalanced interpretation of the 1982 Voting Rights Act amendments that virtually required states to trash what was left of the traditional standards by creating as many majority-black and majority-Hispanic districts as possible.

Third, since 1993 a more conservative Court, appalled by the racial balkanization of the electorate that liberal justices had engineered, has ruled that overtly race-based districting violates the Constitution. But in doing so, the Court has encouraged even more gerrymandering of safe Democratic and Republican seats-with much the same balkanizing effect, given that the vast majority of black voters vote Democratic.

None of these excesses was required by the Constitution or by statute. None was necessary to end gross malapportionment or to make sure that black and Hispanic votes count. Analysis of the Court’s missteps shows how social engineers on the bench can make messes by going to hubristic extremes, and points to a way of undoing some of the damage.

None of this is to say that the House would be free of gerrymandering and full of moderates had the Supreme Court been more restrained. Gerrymandering has been a familiar practice at least since the word was inspired by Massachusetts Gov. Elbridge Gerry’s role in creating a salamander-shaped district in 1811 to help his party. And sometimes gerrymandering increases the number of competitive districts, as when a dominant party moves some of its incumbents’ favorite precincts to help candidates in adjacent districts.

But the modern norm has been gerrymandering based on bipartisan deals to protect incumbents. And by fostering more such political deals, "the courts have played a destructive and unnecessary role," in the words of congressional scholar Norman Ornstein of the American Enterprise Institute for Public Policy Research.

The Court was right in the 1960s to attack the egregious malapportionment of legislative seats that had left urban voters grossly underrepresented in much of the country, and to set a target of one person, one vote as a remedy. But it was wrong-wrong, stupid, and arrogant-to enforce that slogan with absolutist zeal by requiring that congressional districts come as close as possible to precise mathematical equality.

The justices carried this to absurd extremes in decisions such as Karcher v. Daggett, in 1983, which struck down a New Jersey redistricting plan on the grounds that a variation of less than 1 percent between the most- and least-populous districts was too great. This in the face of dissenting Justice Lewis F. Powell’s warning, which echoed previous dissents: "An uncompromising emphasis on numerical equality would serve to encourage and legitimate even the most outrageously partisan gerrymandering." Or, I would add, outrageous bipartisan gerrymandering.

The Court gave still more impetus to such safe-seat gerrymandering in the 1986 decision Thornburg v. Gingles. Congress had intended to require the creation of some majority-minority districts to remedy the unwillingness of whites in some areas to elect black or Hispanic candidates. But the justices once again took a good idea to destructive extremes by laying down rules widely seen as requiring states to gerrymander the maximum possible number of safe seats for blacks and Hispanics.

By further shredding the traditional standards, this also further legitimated all forms of gerrymandering. The ruling aggravated the ideological polarization of the House by ensuring the election of black and Hispanic representatives who held views well to the left of center, and by removing black and Hispanic voters from suburban districts-thus helping conservative Republicans in those districts to beat moderate Democrats. The packing of minorities into a few districts also had the indirect effect of helping Republicans score net gains in the House. And that explains the first Bush administration’s strange alliance with liberal civil-rights groups in pushing for race-based districting.

Ironically, on the same day that Gingles was decided, the Court declared in another case that partisan gerrymandering was unconstitutional if it was so severe as to "consistently degrade" the influence of either major party (or any other political group). But that decision, in Davis v. Bandemer, was no obstacle to bipartisan gerrymandering of safe seats. In any event, the ruling has been virtually a dead letter.

In a line of 5-4 decisions since 1993, the conservative justices have put the brakes on race-based districting, holding that the Constitution bars the creation of "bizarre" district lines with the predominant purpose of "balkanizing" the electorate into racial "enclaves." These decisions have left states guessing whether the swing justice, Sandra Day O’Connor, would find that their plans had violated the Constitution by trying too hard to create more black and Hispanic districts-or had violated the Voting Rights Act by not trying hard enough.

The Court suggested a way out of this box last year in Hunt v. Cromartie. It ruled that packing mostly Democratic black voters into weirdly shaped districts was OK as long as the state’s primary goal was to create not a safe black district-but rather a safe Democratic district!

This was no great solution for those who prefer competitive districts. Indeed, no great judicial solution exists. The justices are not about to clean up the mess by suddenly inventing a new "constitutional" right to more-competitive districts. And we have seen enough judicial invention already. The best solution to safe-seat gerrymandering is for states to adopt Iowa’s depoliticized model of redistricting.

The Court could usefully do some uninventing, however, starting with its extreme extensions of the one-person, one-vote principle and its misguided interpretation of the Voting Rights Act. The justices should make it clear that nothing in the Constitution or the law requires states to chop up municipalities or jump over rivers in pursuit of precisely equipopulous districts, to draw Rorschach-test-like districts in pursuit of racial proportionality, to help create safe seats for incumbents, or to make moderates an endangered species in the House of Representatives.