Bartels, who is 44, teaches political science at Princeton University. Redistricting is not his specialty. He focuses on voter behavior, looking at questions such as how uninformed voters affect elections, or how partisan loyalties bias people's assessments of political events. That is why he doesn't vote: "I think I can do all that with a better perspective," he says, "if I don't have an active rooting interest in one side or the other."
It may well be Bartels' reputation for scrupulous neutrality that got him his messy job. New Jersey has an unusual system for redrawing its legislative map every decade. It creates a commission made up of five Democrats and five Republicans, who meet to see if they can agree on a plan. If they can't--and, since the system was put in place in 1966, that's usually been the case--the chief justice of the state supreme court appoints a non-partisan 11th member to cast the deciding vote. In 1981 and 1991, the tiebreaker was Donald Stokes, who was dean of Princeton's Woodrow Wilson School of Public and International Affairs. Stokes died a couple of years ago; Bartels was his replacement.
He began by examining each party's proposals in light of several criteria that courts and legislators have deemed important. These included everything from a district's racial makeup to population equality and physical shape. In addition, Bartels applied two other standards: The first, developed by Stokes, sought to produce a state legislature reflective of the statewide vote for each party; the second, Bartels' own, sought to give as many voters as possible a chance to vote for incumbents they knew. Soon, though, it became apparent that the two parties were adopting different approaches to racial representation. "When it became clear that this would be the point of contention," Bartels says, "expensive lawyers from D.C. appeared, rapidly and prominently."
This is because New Jersey, which holds legislative elections this year, offers the first chance for Democrats and Republicans to test the U.S. Supreme Court's ongoing redefinition of the Voting Rights Act. In the 1990s, the Republican Party benefited greatly from the theory that the only way to ensure fairness for minorities was to create so-called "majority-minority" districts, in which African Americans or Hispanics might need as much as 65 percent of the population to elect one of their own. Although this did, indeed, produce black and Hispanic officeholders, it also concentrated minority voters in fewer districts and hurt Democrats overall.
This year, Democrats are eager to see whether "minority opportunity" districts--that is, districts in which African Americans, Hispanics and Asians might not be a majority, but still have enough numbers to be politically potent--can pass constitutional muster. Spreading minority voters over more districts in this way should give Democratic candidates a chance to win more seats. That is what they proposed in New Jersey. Republicans, understandably, didn't like it, and Bartels was called in to settle the issue.
In the end, he sided with the Democrats' plan because he felt it better handled the other criteria he wanted to apply. The GOP took the issue to court, and in early May a three-judge federal panel upheld Bartels. The case now goes to the U.S. Supreme Court. The stakes are high, and the issue is likely to be followed closely all over the country. "Wherever it arises," says Maurice Cunningham, chairman of the political science department at the University of Massachusetts- Boston, "this question will continue to be litigated. This is very important, and potentially very damaging to Republicans' prospects."
