The ridiculous-looking, 50-percent black congressional district in Dallas County, Texas, depicted on this page, was drawn in 1991 for the explicit purpose of creating a safe black seat (for one Eddie Bernice Johnson) in Congress.

Its geometry makes it one of the most irregularly shaped congressional districts in the nation, as was pointed out on Dec. 5, when the Supreme Court heard oral arguments in this term’s two big racial-gerrymandering cases.

(They are Bush v. Vera, in which a lower court struck down the Dallas district, along with one majority-black and one majority-Hispanic district in the Houston area, and Shaw v. Hunt, in which a lower court upheld two majority-black districts in North Carolina.)

Justice Sandra Day O’Connor-the swing voter who will almost surely determine the outcome of both cases-sounded like she was itching to strike down most or all of these five districts. But in the case of the Dallas district pictured here (District 30), that would be a mistake.

I don’t like uncouthly shaped, race-based districts either. Nor do I like it when (as here) the weirdness of the shape is largely attributable to the self-serving efforts of politicians to draw safe districts for themselves. But the Dallas district should nonetheless stand.

This is not to say that the Court should give states carte blanche to engage in unrestrained racial gerrymandering to increase minority representation. Nor is it to fault the Court’s June 29,1995, decision (in Miller v. Johnson) striking down as a racial gerrymander a majority-black district that had been improperly forced on the state of Georgia by the Justice Department, in the department’s headlong pursuit of proportional representation at all costs.

But the Texas case presents the Court with an opportunity, at long last, to draw a principled and intelligible line between unconstitutional racial gerrymandering, on the one hand, and permissible efforts to enable once-disenfranchised minority voters to elect representatives of their choice (as states are required to do by the 1982 amendments to the Voting Rights Act), on the other.

A good case can be made for striking down the two Houston-area districts before the Court. These districts, which border one another, were contorted into extreme and confusing shapes with dozens of split precincts for the purpose of separating black from Hispanic voters, so as to create one safe black and one safe Hispanic district. The alternative would have been two districts in which candidates would have to build cross-racial coalitions.

In contrast, the Dallas district appears to be on the constitutional side of the line. The process by which it was created-during the decennial reapportionment of 1991, in which Texas gained three new seats in Congress-shows why.

It was clear after the 1990 census that black voters in and around South Dallas were sufficiently numerous and geographically concentrated that a reasonably compact new congressional district could have been drawn with a narrow black majority. It was also clear that these black voters had been victims of a long history of race-based disenfranchisement, that they were politically cohesive, and that-given racial bloc voting-they could probably elect a representative of their choice only if such a district were created.

These circumstances suggested to all those involved in die 1991 Texas redistricting that a strong case could be made that creation of a district with a black majority (or near-majority) in the Dallas area was required by §2 of the Voting Rights Act, as construed by the Supreme Court in Thornburg v. Gingles (1986).

And, in fact, the near-majority-black district that was initially proposed by Eddie Bernice Johnson-who then headed the state senate’s Subcommittee on Congressional Districts, and whose goal was to create a safe black seal (for herself)-was reasonably compact.

But Johnson’s plan also provoked a pitched battle between her and two white Democratic incumbents from the Dallas area, Martin Frost and John Bryant. They engaged in a tug of war for various pockets of Democratic voters-conspicuously including black voters-and for areas like the white neighborhood near South Dallas in which Frost lived.

The result was a geographically tortured compromise that put large numbers of black voters who lived near the black South Dallas core of the new District 30 into the Frost and Bryant districts. Meanwhile, District 30 was drawn with long, crooked tentacles shooting out to the north and west, mainly to pick up enough voters (a majority of them white) to create a 50-percent black district with a total population large enough to satisfy the one-person-one-vote requirement.

The proper legal inferences from this history were aptly suggested by Justice Stephen Breyer in his questioning at the Dec. 5 argument:

The state’s goal of creating a majority (or near-majority) black district was warranted by its compelling interest in complying with the Voting Rights Act. Contrary to the lower court’s opinion, the new District 30 was "narrowly tailored" to achieve this goal, despite its bizarre shape, because the predominant purpose of its extreme deviation from compactness was not to maximize black representation, but to protect the neighboring white Democratic incumbents. Such gerrymandering to protect incumbents-however distasteful-violates neither the Constitution nor the Court’s precedents.

In other words, while the Dallas district was surely race-based, and while it was surely gerrymandered, it was not a racial gerrymander. The gerrymandering was, rather, of the partisan variety.

To be sure, it can be said that (as the lower court stressed) "racial gerrymandering was an essential part of incumbency protection," in the sense that white and black Democrats vied to pull black neighborhoods into their districts, based on the (entirely accurate) assumption that blacks would overwhelmingly vote Democratic. But this was emphatically not racial gerrymandering in the sense of drawing weird shapes in order to "segregate" minorities from whites, or to isolate them in one district, or to enhance minority voting power.

And while some of the justices appear to be shocked-shocked!-whenever they see racial "stereotyping" going on, in this context that’s just a pejorative label for the statistically accurate inference of probable political orientation from ethnic identity-whether it be Irish, Polish, Hispanic, or black. It’s the sort of practice that has long been so universally employed and so easily camouflaged in the redistricting process that any effort to eradicate it would only drive it underground, and thus would be an exercise in futility, ending in hypocrisy.

None of this is to suggest that there is anything admirable about the forces that drove the Texas districting process, the most potent of which was the partisan and self interested gerrymandering engaged in both by Johnson (who went on in 1992 to win the safe black seat she had created) and by the incumbent congressmen of both parties.

But this practice (like the use of racial and ethnic "stereotypes") has pervaded the redistricting process all over the country throughout for history. In Texas especially, the tradition for decades has been to give protection of incumbents priority over compactness in drawing district lines.

And any judicial effort to bar politicians from manipulating the redistricting process to serve their own personal or partisan interests would be a novel exercise in judicial activism with little basis in the Constitution. It would also be as futile as ordering the bears to stop defecating in the woods.

That may help explain why the three-judge lower court, in an opinion by Judge Edith Jones of the U.S. Court of Appeals for the 5th Circuit, made no effort to strike down any of the majority-white districts in Texas-some of which were no less bizarre in shape than District 30, and no less gerrymandered to suit the personal and partisan self-interest of incumbent politicians. The only districts the court struck down were the 50 percent black district in Dallas and the majority-black and majority-Hispanic districts in the Houston area.

For this apparent double standard, Judge Jones and others of the "colorblind Constitution" school of thought have been plausibly accused of engaging in a peculiarly color-conscious and result-oriented gerrymander of their own: attacking the constitutionality of politically motivated gerrymanders only when the beneficiary politicians and groups happen to be black or Hispanic.

Beyond that, the rules laid down by Judge Jones-a Reagan-appointed darling of the Republican right-would make it even more difficult to create majority-minority districts, as is sometimes required by the Voting Rights Act, without sacrificing incumbent white Democrats. How convenient-for the Republicans.

Justice O’Connor-whose comments at oral argument seemed to echo Judge Jones’ insistence on compactness in the drawing of majority-minority districts (and only those districts)-should heed the brief filed in the Texas case by Penda Hair of the NAACP Legal Defense and Educational Fund.

"By imposing a special, regular-shape-maximization requirement only on majority-minority districts," the brief contends, the Jones opinion "mandates that die stale discriminate against racial minorities, compared to other groups seeking recognition of their voting strength," and thus "heaps an additional disadvantage on minority groups which already suffer barriers to participation in the political process."