AS WE approach the completion of the 1990 census and state officials begin to position themselves to redraw congressional and state legislative districts in 1991, we hear the usual cries of alarm about the evils of gerrymandering. I believe the problem is not nearly as bad as it is portrayed.
Gerrymandering can be defined as the drawing of legislative districts to obtain partisan or factional advantage. The term was first used in 1812 when the Jeffersonians in the Massachusetts legislaDavid C.Saffellture split a county to dilute the strength of the Federalists. The new, oddly-shaped district resembled a salamander and it was dubbed a "gerrymander" in honor of Gov. Elbridge Gerry, who approved the plan. Some would say that all legislative districting since 1812 has been gerrymandering.
As populations grew and shifted in the first half of this century, gerrymandering produced serious problems of unfair representation. In Vermont, which had not reapportioned its legislature since 1793, one representative served 24 people while another served 35,531. The disparity between the most and least populous state senate districts in California was 422 to 1. In the South it was commonplace to draw political districts to shut out blacks.
Following a series of Supreme Court decisions in the 1960s, problems of population inequality and the purposeful dilution of black and Hispanic representation were eliminated in relatively short order. Since then critics have focused much of their attention on partisan gerrymandering -- drawing districts that are essentially equal in population to aid a political party -- and on efforts to devise a set of neutral standards that will produce fair representation. I think it's time to declare victory in this war and let state officials practice the grand old art of gerrymandering if they wish.
Worry about partisan gerrymandering seems largely unnecessary because states haven't done it well even when they've tried. With the exception of California's 1980 plan for congressional districting, in which the districts were oddly shaped and did not follow community boundaries, most gerrymanders have been unsuccessful. Recent failures include an obvious attempt by Republicans in Indiana to expand control of their state legislature in the 1980s and a 1982 congressional map in Massachusetts drawn by the opponents of Rep. Barney Frank. At the time Frank noted: "If you asked legislators to draw a map in which Barney Frank would never be a member of Congress again, this is it."
Most partisan gerrymanders fail because we can't predict accurately how voters will behave. Voting behavior is becoming more fickle, and successful gerrymandering is possible only if there is a pattern of consistent voter turnout and partisanship.
Likewise, we should stop trying to devise "neutral" districting standards. Some of these standards are impossible to obtain and others, taken to extremes, have encouraged gerrymandering. The idea that no part of one district should be completely separated from any other part of the same district is universally accepted. Problems occur with the other standards.
As a practical matter, perfect compactness (circles or squares) is impossible to obtain. Moreover, district shapes may need to be extended in order to follow existing political boundaries (cities and counties) or to protect the interests of racial or ethnic groups.
Concern about population equality also has been taken to extremes by federal courts. For example, in 1983 the Supreme Court overturned a congressional districting plan in New Jersey even though the maximum deviation in population difference among the districts was less then 0.7 percent. Although most model districting standards allow some flexibility in the size of districts (deviations up to 5 percent), even this relaxed guideline may lead to a disregard for honoring political boundaries, such as cities and counties. In some cases streets and even apartment buildings have been split down the middle.
Population variation occurs even under the strictest guidelines. During the 1980s, South Dakota had one U.S. representative, and it was the nation's largest district with over 700,000 people. Montana was the smallest state with two congressmen -- each representing about 400,000 people. And, of course, congressional districts can change dramatically because of population shifts during a decade. In the 1980s nine Florida districts and eight California districts grew by over 20 percent, while a district in Detroit declined by over 15 percent.
Then there are serious questions about the accuracy of the 1990 census. In 1980 the Census Bureau estimated that it counted about 99 percent of the white population but only about 94 percent of blacks. This year's count would be even less accurate; many are calling for an adjustment to compensate for the minority undercount.