Bushmanders and Bullwinkles: How Politicians Manipulate Electronic Maps and Census Data to Win Elections / Edition 1 available in Hardcover
- Pub. Date:
- University of Chicago Press
For years Mark Monmonier, "a prose stylist of no mean ability or charm" according to the Washington Post, has delighted readers with his insightful understanding of cartography as an art and technology that is both deceptive and revealing. Now he turns his focus to the story of political cartography and the redrawing of congressional districts. His title Bushmanders and Bullwinkles combines gerrymander with the surname of the president who actively tolerated racial gerrymandering and draws attention to the ridiculously shaped congressional districts that evoke the antlers of the moose who shared the cartoon spotlight with Rocky the Flying Squirrel.
Written from the perspective of a cartographer rather than a political scientist, Bushmanders and Bullwinkles examines the political tales maps tell when votes and power are at stake. Monmonier shows how redistricting committees carve out favorable election districts for themselves and their allies; how disgruntled politicians use shape to challenge alleged racial gerrymanders; and how geographic information systems can make reapportionment a controversial process with outrageous products. He also explores controversies over the proper roles of natural boundaries, media maps, census enumeration, and ethnic identity. Raising important questions about Supreme Court decisions in regulating redistricting, Monmonier asks if the focus on form rather than function may be little more than a distraction from larger issues like election reform.
Characterized by the same wit and clarity as Monmonier's previous books, Bushmanders and Bullwinkles is essential background for understanding what might prove the most contentious political debate of the new decade.
|Publisher:||University of Chicago Press|
|Product dimensions:||5.50(w) x 8.50(h) x 0.90(d)|
About the Author
A professor of geography in the Maxwell School of Citizenship and Public Affairs at Syracuse University, Mark Monmonier is the author or coauthor of eleven books, including How to Lie with Maps, Cartographies of Danger: Mapping Hazards in America, and Air Apparent: How Meteorologists Learned to Map, Predict, and Dramatize Weather, all published by the University of Chicago Press.
Read an Excerpt
Bushmanders and Bullwinkles
How Politicians Manipulate Electronic Maps and Census Data to Win Elections
By Mark Monmonier
The University of Chicago PressCopyright © 2001 The University of Chicago
All rights reserved.
Twist and Clout
GEORGE HERBERT WALKER BUSH, forty-first president of the United States, shares a unique political legacy with Elbridge Gerry, our fifth vice president, under James Madison. In 1812, while Gerry was governor of Massachusetts, his party, Thomas Jefferson's Democratic-Republicans, controlled the state legislature. In redrawing senatorial district boundaries after the census of 1810, the Jeffersonians hoped to win more seats by packing Federalist voters into a few strongholds while carving out a long, thin Republican district along the northern, western, and southwestern edges of Essex County. Gerry disliked the plan but signed the remap into law anyway — a veto, he thought, would be improper. The Federalist press was not amused. When a reporter pointed out the new district's lizardlike appearance, his editor exclaimed, "Salamander! Call it a Gerrymander!" Artist-cartoonist Elkanah Tisdale added the wings, teeth, and claws (fig. 1.1) that enshrined Gerry's name in the language as a political pejorative, with its hard g mispronounced like the j in Jerry. Ironically, the sinuous district crafted by the governor's cronies is far less troublesome in form, if not intent, than the cartographic manipulations encouraged by the Department of Justice under the Bush administration. I call them "bushmanders."
This new species is also more ragged around the edge than its nineteenth-century ancestors. In figure 1.2, for instance, New York State's Twelfth Congressional District, crafted as a Hispanic-majority district in 1992, looks more intricate and fragile than the famed Essex County senatorial district of 1812. Although the unadorned Massachusetts prototype provoked cynical slurs, Tisdale's sinister enhancements probably account for its longevity as a political icon: widely reproduced in books and articles on electoral manipulation, the classic gerrymander is rarely rendered as an unembellished contour. By contrast, New York's Twelfth District needs no adornments to explain journalists' delight in labeling it the "Bullwinkle District," after the loquacious moose who shared a Saturday morning spotlight with his cartoon-show sidekick Rocky the flying squirrel. Although the narrow rows of comblike prongs and wider blobs awkwardly connected by thin corridors only faintly resemble antlers, once some clever wag linked the district's contour to the beloved talking moose, the name stuck. Nicknames are rare, though: most bushmanders inspire descriptions like "the 'Z' with drips" (Louisiana's Fourth District), a "spitting amoeba" (Maryland's Third District), and "a pair of earmuffs" (Illinois's Fourth District).
The emergence of these new political critters in the early 1990s is partly a consequence of the Voting Rights Act, passed in 1965 and modified several times. In addition to banning racial discrimination in voter registration, the law defends the right of minority voters to elect candidates of their choice and demands federal scrutiny where past abuse has been especially flagrant. Among other provisions, the Department of Justice must approve the postcensus remap in several states, mostly in the South, as well as the three New York City boroughs containing parts of the Bullwinkle District. Interpreted broadly, the Voting Rights Act also prohibits political cartographers from splitting a district in which members of a minority group constitute a majority. This stricture raises a thorny question: Can the Justice Department deny preclearance when a state chooses not to form a thinly stretched minority-majority district? In rejecting plans submitted by Georgia and North Carolina, George Bush's map editors answered with an assertive "Yes!" Intimidated by earlier rulings as well as eager to accommodate black and Hispanic leaders, New York's mapmakers won preclearance on their first try.
Bushmanders would be difficult, if not impossible, without computers. In Gerry's day, and for more than a century thereafter, the basic building block for congressional districts was the county. Although New England mapmakers eagerly split counties along town lines, political cartographers elsewhere preferred to combine whole counties wherever possible and to split cities, if needed, only along existing precinct boundaries. To explore different configurations, they spread out their maps on a large floor and tallied district populations by hand or by adding machine. Redistricting became more troublesome after the mid-1960s, when the Supreme Court insisted that states not only reconfigure congressional and legislative districts every ten years, as the Constitution intends, but minimize variation among districts in population size. In the early 1990s, with the Civil Rights Division of the Justice Department poised to reject plans that ignored possible minority-majority districts, states turned to interactive computers, electronic maps, and detailed census data, which made it easy to accumulate blocks inhabited by African Americans or Spanish-speaking Americans and link dispersed minority neighborhoods with thin corridors inhabited by few, if any, nonminority "filler people." To ignore this technology was to invite federal judges to draw the lines themselves. After all, Justice officials in Washington had similar tools, as did African American and Hispanic interest groups eager to sue for apparent violations of the Voting Rights Act.
The result typically was a district difficult to describe with maps or words. New York City's Bullwinkle District, for instance, is a polygon with no fewer than 813 sides. In a bill approved by the state legislature and signed into law by the governor in June 1992, District 12's perimeter requires 217 lines of verbal description, which read like the itinerary of a taxi driver trying desperately to run up the meter. Figure 1.3 shows how part of the boundary twisting across Brooklyn helped elect a Hispanic to the House of Representatives by capturing blocks rich in Spanish surnames while avoiding blocks where Hispanics are a minority. The line became law as
to Linwood street, to Glenmore avenue, to Cleveland street, to Pitkin avenue, to Warwick street, to Glenmore avenue, to Jerome street, to Pitkin avenue, to Warwick street, to Glenmore avenue, to Jerome street, to Pitkin avenue, to Barbey street, to Glenmore avenue, to Schenck avenue, to Liberty avenue, to Barbey street, to Atlantic avenue, to Van Siclen avenue, to Liberty avenue, to Miller avenue, to Glenmore avenue, to Bradford street, to Liberty avenue, to Wyona street, to Glenmore avenue, to Pennsylvania avenue, to Liberty avenue, to Vermont avenue, to Atlantic avenue, to New Jersey avenue, to Jamaica avenue, to Vermont avenue, to Fulton street, to Wyona street....
Lawmakers are word people, and before they vote on a redistricting bill, boundaries composed on a computer screen are converted to verbose lists of street segments, watercourses, and other fixed features.
Because politicians and election officials need to see where voters live, redistricting officials convert the lists back into maps. To supplement the electoral maps of individual states, the Bureau of the Census publishes the Congressional District Atlas, a standardized cartographic reference for the fifty states. Because the Atlas's letter-size pages are too small to show on one map the intricate details of computer- crafted gerrymanders, a single district can extend across a dozen pages or more. In the edition for the 103rd Congress, published in 1993, fragments of New York City's Bullwinkle District appear on twenty-two pages. Identified vaguely on the separate, single-page county maps for Brooklyn, Manhattan, and Queens, District 12 crops up in greater detail on eighteen partial-county inset maps, mostly printed one to a page, at various scales. Figure 1.3, extracted from one of the Brooklyn (Kings County) insets, illustrates the symbols and level of detail. Especially complex portions of the boundary in parts of Queens required eight additional subinset maps — insets of insets — focused on small areas at even larger scales. The cartography of bushmanders is, to coin a word, insetuous.
The Congressional District Atlas affords a useful, if idiosyncratic, indicator of the bushmander's radical geometry: the ratio of map pages to House members. Consider New York, which lost three seats after the 1990 census. In 1983 the state's thirty-four congressional districts occupied seventeen pages of maps. Ten years later thirty-one districts required seventy-seven map pages — two and a half pages of maps for every seat. According to the national maps in figure 1.4, for most states the index jumped markedly between the 1983 and 1993 editions. For the earlier year, the highest ratio is an even two for Oregon, where five districts required ten map pages. Only three other states had a ratio over one, and seven states registered equal numbers of map pages and House members. By contrast, the map for 1993 shows twenty-five states with a ratio over one, and twelve of these have an index greater than two. Texas, which required 177 map pages to describe thirty districts, had the highest ratio (5.90). Six states were ahead of New York (2.48): Florida (4.52), North Carolina (4.50), Louisiana (4.00), Georgia (3.18), Arizona (2.50), and South Carolina (2.50). Only one of these six, Arizona, had had more map pages than congressional districts ten years earlier.
There's a pattern here: most states with high ratios for 1993 are in the South, most have high relatively large percentages of African American or Spanish-speaking residents, and as figure 1.5 shows, most required preclearance by the Department of Justice. Of these factors, the need for preclearance yields the strongest correlation, however imperfect. Five of the nine states requiring statewide preclearance have ratios of two or higher, as do three of the seven states requiring preclearance for specific counties and only four of the remaining thirty-four states. Among the latter, Illinois (2.30) and Maryland (2.12) designed noncompact districts to elect minority representatives, Washington (2.44) used several complex boundaries in partitioning the Seattle area, and Nevada (2.00), which centered one of its two districts on Las Vegas, threaded the boundary in and out of the city limits.
Close examination accounts for several anomalies. Preclearance is irrelevant to Alaska (1.00) and South Dakota (1.00), each with only one district. California, with fifty-two seats and only four preclearance counties, scored a low ratio (0.94) despite several districts stretching across four or more map pages. More intricate boundaries are also apparent for the South's three full preclearance states with map-seat ratios less than two: between its 1983 and 1993 editions, the Atlas registered substantially increased ratios for Alabama (0.43 to 1.71) and Virginia (0.60 to 1.45), which in 1992 elected their first black congressmen since Reconstruction, and for Mississippi (0.40 to 1.40), which redrew its sole minority-majority district to ensure a safer seat for Mike Espy. Although preclearance and racial intent cannot account for all bushmanders, the species flourished in the South, where dispersed populations of rural African Americans thwarted the easy construction of compact minority-majority districts.
Why would a Republican administration favor African American and Latino candidates, almost certain to be Democrats? For the same reason that Elbridge Gerry's Jeffersionian Republicans packed Federalist voters into Essex County's inner district: by creating safe districts in which minority candidates were likely to win, the Bush Republicans added white voters to formerly Democratic districts, which responded, as hoped, by electing Republicans. There was another advantage, though. A widely shared resentment of minority-majority districts, often perceived as yet another affirmative action strike at the prerogatives and values of the white middle-class majority, fueled white dislike for the Democrats' policies and politicians. Because the GOP did not openly advocate minority districts, Republican candidates were free to rail against the bushmanders' flagrantly contorted shapes. Reinforcing the perception of an antiwhite conspiracy were lawsuits filed by aggrieved "filler people" and promptly challenged by pro–civil rights Democrats. Adding to the irony, federal judges appointed by Presidents Reagan and Bush won the public approval of white Republicans by condemning the racial gerrymanders that helped their party take over the House in 1994.
A hypothetical example explains how racial gerrymandering works. The fifty-four squares in figure 1.6 represent counties in a fictitious state accorded three congressional districts in the post-1990 reapportionment. Vying for these seats are two equally fictitious political parties: the Traditionalists and the Modernists. As their name suggests, the Traditionalists salute the flag, love Mom and apple pie, and prefer rural environments, specifically the fifty-one nonshaded cells, representing rural counties, where they outnumber Modernist voters by a slight majority, 81,600 to 71,400. (To simplify the arithmetic, I kept the numbers small, much smaller than the population now required for three House seats.) By contrast, the Modernists form a substantial majority in the three urban counties, where they outnumber Traditionalist voters 24,000 to 3,000. If all candidates ran at large and every voter had three votes, all three Modernist candidates would win a seat, 95,400 to 84,600. And if the state was partitioned among three equal-sized, equally urban districts, the Modernists would still take every seat.
Numbers and maps explain what's happening. The three comparatively compact districts in the left-hand map yield identical tallies. Within each district, the Modernists reap 1,400 votes from each of seventeen rural counties and 8,000 from the single urban county, for a total of 31,800 votes:
23,800 votes in the seventeen rural counties (17 × 1,400)
+ 8,000 votes in the single urban county
= 31,800 votes for each Modernist candidate
With their slight edge in the countryside more than offset by a weak showing in the city, the Traditionalists can muster only 28,200 votes:
27,200 votes in the seventeen rural counties (17 × 1,600)
+ 1,000 votes in the single urban county
= 28,200 votes for each Traditionalist candidate
With a reverence for compact districts and winner-take-all victories, the Traditionalists might find some solace in an electoral system that reinforces traditional (small t) values.
Suppose, though, that their Traditionalist brethren in Washington intervened on behalf of a racial minority of Modernists concentrated in the three urban counties. Suppose that Modernists in the state legislature were encouraged, if not coerced, into a redistricting plan that gave minority-group members the opportunity to elect a "candidate of their choice" by placing all three urban counties in a single district, as in the right-hand map in figure 1.6. Confident that they enjoyed a statewide majority — 95,400 to 84,600 would impress most politicians — the Modernists approve the plan and hope for the best. After all, they've courted minority voters over the years, and fair is fair.
But look again at the numbers. In the district linking the three cities, the Modernist candidate — presumably a member of the minority — wins easily. With eleven rural and three urban counties, District 2 is very safe Modernist territory.
15,400 votes in the eleven rural counties (11 × 1,400)
+ 24,000 votes in the three urban counties (3 × 8,000)
= 39,400 votes for the Modernist candidate
The opposing candidate is thoroughly trounced.
17,600 votes in the eleven rural counties (11 × 1,600)
+ 3,000 votes in the three urban counties (3 × 1,000)
= 20,600 votes for the Traditionalist candidate
The other two districts produce different outcomes. Packing Modernist voters into a long, thin district akin in shape and spirit to the Essex County gerrymander yields a pair of highly rural districts in which a Traditionalist candidate enjoys a smaller but nonetheless comfortable majority. With twenty rural counties and no urban counties, each district has 32,000 Traditionalist voters (20 × 1,600) but only 28,000 Modernist voters (20 × 1,400). Able to control the congressional remap, the Traditionalists capture two of the state's three seats and proclaim themselves champions of individual and minority rights.
Meandering boundaries on the left-hand map serve the Traditionalists in two ways: by wasting Modernist votes in District 2 and by diluting Modernist strength in Districts 1 and 3. Racial gerrymanders crafted under the guise of the Voting Rights Act do this too, and so do purely partisan gerrymanders, which are often more subtle in geometry. By adding here and dropping there a clever politician can craft boundaries that confer advantage with little hint of cartographic manipulation. The real test, it seems, is whether the outcome at the polls is substantially better than a party's inherent strength as measured by voter registration or previous elections. In this sense the 1812 manipulation of Massachusetts senatorial districts worked extraordinarily well: although Jeffersonian candidates throughout the state received only 50,164 votes, in contrast to 51,766 votes for their Federalist rivals, Gerry's party won twenty-nine of the Senate's forty seats. Although other districts were drawn to waste Federalist votes, only the outer Essex district was flagrantly misshapen.
Whether a gerrymander is suspicious or subtle often hinges on geography. In major cities like New York and Chicago, where less affluent African Americans inhabit large, generally contiguous but segregated neighborhoods, redistricting officials can usually corral a comfortable majority of black voters without resorting to bizarre boundaries. But even here complex shapes can arise when political cartographers try to maximize the number of black-majority districts or insist on a 60 or 65 percent "supermajority," more certain to elect a minority candidate. By contrast, Hispanic voters, who are less numerous and assimilate more easily, tend to be distributed more widely across the city, in discrete neighborhoods requiring thin corridors like those of New York's Bullwinkle District (fig. 1.2). Carving out even one Latino-majority district can be a delicate task. And in the Deep South, dispersed settlement of African Americans in small towns and rural areas makes black-majority districts difficult if not impossible without long corridors and amorphous appendages. Without computers, a broad interpretation of the Voting Rights Act, and a Justice Department eager to offer minority candidates safe districts, bushmanders would not have appeared.
Excerpted from Bushmanders and Bullwinkles by Mark Monmonier. Copyright © 2001 The University of Chicago. Excerpted by permission of The University of Chicago Press.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Table of Contents
1. Twist and Clout
2. Gerry's Legacy
3. Thing Majorities
4. Redrawing the Lines
5. Gauging Compactness
6. Props and Propaganda
7. Immunizing Incumbents
8. What a Friend We Have in GIS
9. A Tale of Two Censuses
10. Beyond Boundaries