Karcher v. Daggett

KARCHER, SPEAKER, NEW ASSEMBLY, JERSEY et DAGGETT

al. v. et al. Argued No. 81-2057. March 1983 Decided June Court, Marshall, Bbennan, J., of the in which opinion delivered the Stevens, Blackmun, Stevens, O’Connor, JJ., J., joined. filed a and White, J., dissenting opinion, in post, p. 744. filed concurring opinion, JJ., J., Rehnquist, joined, Burger, post, and which C. Powell p. p. Powell, J., dissenting opinion, post, 765. filed a 784. Guido, the cause for Jr., argued appellants.

Kenneth J. Sachse, R. E. him on Harry the briefs were With Loftus Becker, Cummis, Charles Jr., Donald Clive S. Simon, J. D. Sokol, D. Leon J. Michael Walsh, Baranoff, Jerald J. A. Farkas. Solomon, Marinari, Lawrence T. and Robert Hellring argued appellees. Bernard the cause for With him on the brief were Jonathan L. Robert Goldstein, S. * Raymar, Stephen Dreyfuss. L. opinion delivered the of the Court.

Justice Brennan question presented by appeal appor- this an whether tionment districts satisfies I, §2, Art. justification of the Constitution without need for further if largest percent district is than less one greater than the of the smallest district. A three-judge Jersey’s District Court declared New 1982 re- apportionment plan authority unconstitutional on the of Kirkpatrick (1969), Preisler, U. S. v.White (1973), population Weiser, 412 U. S. 783 because the devi- among although ations districts, small, were not the result good-faith population equality. of a effort to achieve We affirm.

After the of results the 1980decennial census had been tab- Represent- ulated, the Clerk of the United States House of Jersey atives the notified Governor of New the number Representatives to which the was entitled had de- State Accordingly, Jersey Legis- creased from to 14. the New required reapportion lature was the State’s Legislature passed reappor- districts. The State’s 199th two by tionment bills. One the Governor, was vetoed the signed although significant second, law, into occasioned dis- among minority voting satisfaction those it who felt diluted strength city App. 83-84, of Newark. 86-90. response, Legislature problem the 200th returned to the apportioning congressional when it in Jan- districts convened (S-711) uary swiftly by passed bill introduced 1982, and it pro Senate, Feldman, tem of the State Senator President Moore, Foelber, A. Hess filed a *Roger Allan Richard P. and Michael amicus curiae Republican urging brief for the National Committee as affirmance. apportionment at issue this case.

which created the January by signed on be- The bill was Governor (hereinafter Plan). coming A ch. 1 Feldman L. 1982, Pub. resulting apportionment appended map of the infra. by every plan legislature, Like considered the Feldman average population per districts, an Plan contained with (as census) by 526,059.1 determined the 1980 district population. same On the Each district did have the figure average, from the district differed “ideal” each largest people. district, or about 726 The 0.1384%, population District, includes had a Trenton, Fourth which embracing District, 527,472, smallest, and the Sixth most population' County, a The dif- 523,798. of Middlesex people, them or 0.6984% of the 3,674 ference between was average populations The of the other districts also district. including Bergen County, most of District, varied. The Ninth population in the had a State, northeastern corner along District, while the of the Third 527,349, only App. 524,825. 124. shore, Atlantic was legislature plans appreciably it had before other largest deviations between and small- smaller receiving the most attention est districts. The one by po- designed Reock, Jr., Dr. a District Court was Ernest Rutgers University professor Director of litical science and A the the Research. version of Reock Bureau Government 1 early 1981, this In the Three sets of census data are relevant case. the preliminary figures showing that total Bureau of the Census released Jersey 7,364,158. 1981 it released population New was October (and Orange the data, population the of East corrected which increased whole) 3, 1. All calcula by Appellants n. people. Brief for State a is, legislature the opinion refer the data available to tions in this —that concluded, the figures. proceedings below had the After the October population correction the Bureau of the Census made an additional population the Orange, people, bringing and total adding another 188 East was not Ibid. Because 7,365,011. this last correction State issue, need it the we legislature at the time enacted available to the not consider it. by Legislature Assemblyman 200th

Plan introduced of 2,375, Hardwick had a maximum difference figure. average Id., 0.4514%of the at 133. immediately

Almost after the Feldman Plan became a law, group varying including of individuals with in- interests, all Republican Congress Jersey, cumbent from Members New sought apportionment plan declaration that violated injunction against pro- I,Art. of the Constitution2 §2, and an ceeding primary Repre- with the election for United States plan. three-judge sentatives under the A District was Court 2284(a). § pursuant convened to 28 S.U. C. The District hearing February par- Court held a on at which the depositions ties submitted a affidavits, number moved summary judgment, right and waived their to introduce summary judg- further evidence in the event the motions for ment were denied.

Shortly opinion thereafter, District Court issued an declaring and order Plan Feldman unconstitutional. De- nying summary judgment resolving the motions for whole, case on the record as a the District Court held that the population variances in the Feldman Plan were not “unavoid- despite good-faith equality,” able effort to achieve absolute Kirkpatrick, rejected supra, appel- see at 531. The court argument lants’ that a im- deviation lower than the statistical precision equiva- of the decennial census was “the functional equality.” Daggett lent of Kimmelman, mathematical 1982). (NJ Supp. appellants F. It also held that justi- had failed to show that variances were by legislature’s preserving purported goals fied minor- part: Representatives composed In relevant “The House of shall be every People of Members chosen second of the several Year States ....

“Representatives apportioned among . . . be the several States shall may respective which Union, according be included this to their within . Numbers . . .”

730 population. strength anticipating ity voting shifts enjoined appellants from conduct- The District Court Ibid. ing general primary Plan, under Feldman but or elections stayed pending appeal 455 Court, to this U. S. that order was (1982) chambers), prob- J., noted and we 1303 (Brennan, (1982). jurisdiction, 457 U. S. able

I—I hH “high justice § a standard of I, 2, Article establishes congressional apportionment dis- common sense” for “equal representation equal people.” numbers tricts: (1964). Wesberry Sanders, 376 S. Precise math- v. U. impossible may equality, however, be to achieve ematical representation” “equal imperfect world; an therefore the only requiring to the extent of standard is enforced apportioned population equality to achieve “as districts nearly practicable.” 7-8, 18. As we ex- id., as See plained Kirkpatrick further in v. Preisler: nearly practicable’ requires

“[T]he ‘as standard good-faith precise make a effort to achieve State Reynolds equality. mathematical Sims, U. S. Unless variances among congressional districts are to have shown resulted despite justify effort, such the State must each variance, S., no matter how small.” 394 U. at 530-531. “permits only popula- I, §2,

Article therefore, limited despite good-faith tion variances are which unavoidable equality, justification effort to achieve absolute or for which Id., Accord, is shown.” at 531. Weiser, White U. at 790. shape litigation questions

Thus two basic over legislation apportioning deviations state dis- popula- First, tricts. the court must whether the consider among tion differences districts could have been reduced altogether good-faith eliminated effort to draw districts population. equal challenging apportionment leg- Parties proof islation must bear the burden of on if issue, this they fail to show that the differences couldhave been avoided apportionment upheld. scheme must be If, however, the plaintiffs population can establish that the differences were good-faith equality, result of a effort to achieve proving significant State must bear the burden of that each necessary variance between districts was to achieve some legitimate goal. Kirkpatrick, S., 532; 394 U. at cf. Swann 440, 443-444 Adams, U. S.

h-1 r-H Appellants principal argument in this case is addressed to question They the first described above. contend that regarded per product Feldman Plan should be se as the of a good-faith population equality effort to achieve because the population among maximum deviation districts smaller predictable than the undercount census available data.

A Kirkpatrick squarely rejected nearly argument. identical nearly practicable’ approach “The whole thrust of the ‘as adoption is inconsistent with of fixed numerical standards regard which excuse variances to the without particular circumstances of each S., 530; case.” 394 U. at supra, see White Weiser, at n. 8, 792-793. Adopting any using population equality, standard other than the best census data see 394 available, U. would subtly representation. equal erode the Constitution’s ideal legislators If state minimis knew that a certain de level of acceptable, they differences was would doubtless equality.3 strive to achieve that rather Id., level than 3There is some evidence in the from which one infer that record could precisely Karcher, this is happened Jersey. Speaker what in New Alan Assembly, one-percent testified that he had set maximum deviation as upper any plans seriously by legislature, limit for to be considered (Karcher Record pp. deposition), Doc. No. but there is no evi- im- different standard would Furthermore, choosing 531. the into of review- process of arbitrariness degree port high Ibid. case, In this appellants ing plans. apportionment of 0.7% deviation should approximately that a maximum argue minimis. de argument, If we how accept be considered 0.8%, 1%, of or 1.1%? 0.95%, deviations regard are we to involves a cer- absolute standard, including equality, Any even the census point out, As artificiality. appellants tain are not well-known restlessness perfect, data counts for American means that population particular people are before are Yet long they completed. localities outdated at to any with the data hand problems apply equally pop- standard we could choose.4 As between two ulation-based or less than something equality only equality — standards — § of 2. I, the former reflects the Art. aspirations To of small though accept legitimacy unjustified, popu- lation in this case would mean to the basic reject deviations Kirkpatrick Wesberry. of We decline premise appel- lants’ to that far. The unusual of their go rigor invitation of standard has been noted several times. Because we have that absolute rigor, population equality required in only the case paramount objective apportionment any attempt improvements one-percent to dence serious seek below level. problems certainly concededly apply Justice arbi Such White’s post, trary solution, solely five-percent 782, apparently see at selected in Kirkpatrick discarding avoid the embarrassment the actual result No de minimis line tied to actual along reasoning. with its post, any way 771-772, mitigates differences identified at between the any eligible, registered, number of adults or or actual voters in two dis below, infra, 736-738, systematic tricts. As discussed see at some unless effort made correct the distortions inherent census counts total population, population equality deviations from the norm of are far more likely If to exacerbate the differences between districts. a State does attempt to use a measure other than total or to “correct” the figures, may inconsistent, conjectural haphazard, census it not so in a or do Kirkpatrick, infra, S., 534-535; manner. 740-741. 394 U. see at § congressional for districts, I, 2, which command of Art. regards Legislature outweighs National local inter may apportioning ests that a State deem relevant districts representatives legislatures, for to state and local but we questioned population equality have standard for con gressional g., See, Weiser, districts. e. White v. 412 U. (1973); Regester, 412 793; U. S. Mahan White principle Howell, 410 U. S. population equality proved districts has not unjust economically socially experience. or harmful (1924) Washington Cf. Co., 219, v. Dawson & 264 U. S. 237 (Brandeis, dissenting); B. J., The Nature of the Cardozo, (1921). anything, Judicial If Process 150 this standard difficulty legislatures should cause less now for state than it adopted Wesberry. rapid did in when we it in advances computer technology during and education the last two dec relatively simple contiguous ades make it to draw districts of equal population and at the same time to further whatever secondary goals Finally, the State has.5 to abandon unnec essarily interpreta a clear and oft-confirmed constitutional impair authority Dept. cases, tion would our other Florida Nursing Health v. Florida Home Assn., 147, 450 U. S. (1981) concurring); 153-154 Pollock v. Farm J., (Stevens, (1895) (White, ers’ Loan Trust Co., 429, J., & 157 U. S. 652 dissenting), implicitly open plethora would to door of re quests may that we reexamine other rules that some consider many problems Jersey

5 Note that Legislature that the New en in drawing equal population countered districts with stemmed from the de cision, appellees any challenged, municipalities which never not to divide Jersey between two districts. The entire of New is State municipalities, populations 329,248 divided into 567 with ranging from (Newark) (Tavistock 36, Borough), Appellants to 9 n. See Brief for 38. Preserving intact, however, permissi perfectly subdivisions while goal, a secondary failing ble as is not a sufficient excuse to achieve population equality infra, specific without showing at described Kirkpatrick Preisler, supra, Weiser, White 533-534; 740-741. See at 783, 412 U. S. 791 734 supra, 149-150, and would preju- at Cardozo,

burdensome, an seeking the rule of law relied upon those who have dice Jersey, New congressional apportionment equipopulous supra, Nursing Assn., Home Florida at see (Stevens, de there are no thus reaffirm We J., concurring). minimis be variations, which could practicably I, standard Art. nonetheless meet the avoided, but which §2, justification.6 without perfectly equality objects that “the rule absolute Justice White Rockefeller, v. Wells sort,” compatible ‘gerrymandering’ of the worst Post, (1969) (Harlan, J., 776. That dissenting). U. S. justify certainly beyond requiring States may true to some extent: be reasons, explicit, precise might which be ex

population deviations with effect, Kirkpatrick inhibitory prevent little to pected some does have Robins, Backstrom, gerrymandering. generally & is known as what Eller, Gerrymandering: Exploratory An of Partisan Measure Issues Minnesota, Applied Gerrymandering Minn. L. Rev. Kirkpatrick’s object, achieving (1978); popula n. 4. cf. 394 U. equality, required than what would to address tion is far less ambitious gerrymandering level. on constitutional Kirkpatrick actually promotes event, any the additional claim that (as it) completely

gerrymandering opposed merely failing stop to to is any empty. principle population equality prevent A federal does long taking steps gerrymandering, good-faith from to inhibit so a State as g., e. population equality See, effort is made to achieve as well. Colo. V, § (guidelines compactness, contiguity, Art. as to Const. boundaries political subdivisions, interest); Const., and communities of Mass. 4-100(2) CI, § (boundaries); (McKinney § Art. Y. Amended N. Elec. Law boundaries). 1978) (compactness and lack of a de minimis rule encour- argues

Justice White further that Post, litigation ages and intrusion federal courts into state affairs. gainsaid the de minimis rule proposes It cannot be that he 777-778. would litigation proves experience have made this case unattractive. But that upon cases in which a court is existing appor- federal called to invalidate an tionment, plan stead, and sometimes substitute court-ordered in its frequently newly arise apportionment plan not because a enacted fails to of Kirkpatrick, partisan politics meet the test because but frustrate the legislature efforts of a state to enact new after a recent census has existing plan g., e. grossly malapportioned. See, shown (Colo. Lamm, 1982); Shayer Carstens Kirkpatrick, Supp. 543 F.

B appellants’ theory The sole difference between and argument rejected Kirkpatrick appellants we is that have proposed gives a de minimis line that the illusion of rational- ity predictability: imprecision the “inevitable statistical They argue: “Where, here, census.” the deviation imprecision from ideal district size is less than known figures, equivalent the census is the variation functional Appellants problems of zero.” Brief for 18. There are two approach. appellants with this on the First, concentrate systematically extent to which the census undercounts actual figure precisely which is not which, known —a if it even were known, would not relevant to this case. imprecision Second, the mere existence of statistical does not among equiva- small make deviations districts functional equality. lent of appellants rely

In the Court, District Court and before this exclusively on an Dr. Trussell, affidavit of James a Princeton University demographer. App. 97-104. Dr. Trussell’s carefully worded statement reviews various studies 1950, 1960, undercounts and 1970 censuses, decennial (1) important and it draws three conclusions: “the undercount (2) likely percent”; in the 1980census is one be above “all places evidence date indicates that all are not under- counted extent, to the same since the undercount rate has depend age, been shown to on race, sex, income, and educa- (3) “[t]he tion”; and distribution of the undercount New Jersey is . . unknown, . and see no reason to believe that it uniformly spread municipalities.” would be over all Id., Assuming purposes argument 103-104. that each of (WD Mo.), aff’d, Supp. summarily 541 F. 922 (1982); 456 U. 966 S. (Kan. Brier, Donnelly Meskill, O’Sullivan Supp. 1982); v. 540 F. 1200 v. (Conn. 1972); Cahill, (NJ David Supp. 345 F. 342 Supp. F. 1972); Illinois, Skolnick (ND State Electoral Board Supp. 336 F. 1971). Ill. support appellants’ they correct, do is these statements argument. one-percent appellants’ is little more benchmark essence, attempt present de minimis line with an attractive to

than an authority. Dr. Trussell’s state- patina Neither a scientific specifies precise any appellants’ a other evidence ment nor Jersey, Dr. in New Trussell’s level for the undercount impossible it is to makes clear that discussion of census anything develop of the undercount on but reliable estimates His at 98-101. conclusion id., scale. See a nationwide “likely percent” one seems to be above the 1980undercount previous in censuses and on the undercounts to be based procedures adopted guess to a as to how well new if Therefore, would we ac- reduce the undercount work. theory cepted appellants’ that the national undercount level ability tell the differ- a limit on our to use census data to sets populations we districts, ence between percent might far one well be forced to set that level above completed.7 analyses are when final 1980census admits, id., Dr. of a As Trussell the existence one- percent would be irrelevant to devi- undercount among if ations districts the undercount were distributed among evenly districts. The undercount the census af- only accuracy fects the of the deviations between districts the extent that the undercount varies from district to dis- one-percent explain one-percent trict. For undercount populations the census districts, deviation between two approxi- in the smaller district undercount would have to be mately large larger three times as as the undercount in the Passel, Robinson, & J. generally Siegel, Coverage J.

7 See J. Na Census, by Sex, Population Age, Preliminary tional in the 1980 and Race: (Nov. 1981) (Record 31) by Demographic Analysis Estimates Doc. No. (hereinafter Passel). previous Estimates the national undercount g., e. range See, censuses from 2.5% 3.3%. Panel on Decennial Census (Nat. Plans, Counting People Appraisal An 1980: of Census Plans 2 1978). Acad. Sciences district.8 It is of highly unlikely, course, that this condition true, holds since especially appellants have utterly failed to introduce evidence that showing the districts were designed for the compensate probable undercount. Dr. Trussell’s affidavit states that the rate of may from undercounting vary municipality but it does municipality, not discuss by how much it or to what extent may vary, those variations would be reflected at the district level, with many municipalities combined. Nor does the affidavit indicate that the factors associated with the rate sex, of undercounting race, age, — (more from district vary district, or importantly) etc. — that in the smaller districts populations reflect the rele- vant factors more than the in the populations dis- larger tricts.9 Dr. admits, As Trussell the distribution undercount Jersey New is completely unknown. Only by bizarre coincidence could the systematic undercount example, hypothetical As an assume in a congres State with two sional population 502,500, districts District A has a B and District has a 497,500. population 5,000, per The deviation them is or one between cent of the mean. If percent, the statewide undercount is also one and it is evenly districts, distributed A between the two District will have a “real” 507,525, population B population and District will have a “real” 502,475. Only The percent. deviation between them will remain one if people three-fourths of the uncounted in the live in B State District will the populations. two have If equal districts three-fourths of the uncounted A, people happen to live in District the deviation districts between the two will increase to 1.98%. instance, accepted For it is the rate of undercount in the census for significantly than the higher black on nationwide basis is rate population. generally of undercount for white 9-20. Yet the Passel census of the districts in the Feldman Plan is unrelated to District, instance, percentage of blacks each district. Fourth mean; largest it population, district in terms 0.268% above the has black, id,., at population, App. 17.3% black 94. The First District is 14.6% exactly population. The undercount average and it is almost in overall any particular predicted only percentage district cannot be from counted, district, blacks in the but to the extent that blacks are in the rela in the Fourth District than undercount would more severe tively populous less First District. relationship

census bear some statistical to the districts by the Plan. drawn Feldman may systematically population, census undercount may vary place. undercounting place to

the rate of from meaningless facts, however, Those do render the dif- population congressional districts, as ferences between by contrary, counts. To the determined uncorrected census only provide less than the census data reliable—albeit population perfect of the districts’ “real” relative —indication say certainty Even if cannot with that one dis- levels. one larger merely higher it trict than another because has is say certainty that the district count, census one can likely larger larger census count is more to be than the with a than it to be smaller or the same size. That other district certainty decisionmaking. City is sufficient for Cf. New- 1978). (DC Supp. Further- Blumenthal, ark 457 F. represents popula- the “best more, because the census count Kirkpatrick, S., 394 U. at it is available,” tion data see only good-faith attempts population to achieve basis explain population equality. Attempts deviations on the to preci- supported with a of flaws in census data must basis id., at 535. sion not achieved here. See

C in the deviations that the census-based Given among it districts, Plan differences Feldman reflect real they significantly re- have been avoided or is clear that could equality. good-faith duced with a effort achieve inappropriate accept it would be alone, For that reason equivalent” “functionally to a with dis- Feldman Plan as equal population. tricts plans intro- found that several other

The District Court Legislature devi- had smaller maximum in the 200th duced Supp., F. at 982. Feldman Plan. ations than the Appellants and n. 9. 412 U. Weiser, Cf. White object plans considered the District that the alternative comparable Plan the Feldman because Court were not political profoundly. g., their characters differed e. See, (affidavit Jr.) App. (arguing Woodson, S. H. that al- plans protect ternative failed to the interests of black voters areas). in the Trenton and Camden We have never denied apportionment political process, legisla- is a or that state pursue legitimate secondary objectives long tures could as objectives good-faith those were consistent with a effort to population equality achieve same time. Neverthe- require population less, the claimthat considerations among congressional belongs differences districts more properly judicial inquiry to the second level of in these cases, see at in infra, 740-741, which the State bears the burden of

justifying particularity. the differences with any unnecessary event, it was for the District Court to finding plans rest its on the existence of alternative with rad- ically political Kirkpatrick, different As in effects. “resort simple transferring the to device of entire subdi- population contiguous visions of known between districts produced would have districts much closer to numerical equality.” Starting 394 U. at 532. the S., Feldman legislature Plan itself and the census data available to the at App. the time it enacted, 23-34, was see one can reduce the population by plan merely shifting maximum deviation of the municipalities a handful of from one district to another.10 10 According population figures Reock, following to the used Dr. the 1982, adjustments in to the Feldman Plan as enacted Pub. L. ch. would 0.449%, population reduce its maximum than variance somewhat lower legislature: the version of the Reock Plan introduced To Fifth (from District, Eighth District), add Oakland and Franklin Lakes and (from District). Lake, Hillsdale, and the Ninth To the Woodcliff Norwood (from District). District, Sixth add North Brunswick the Seventh To the (from District, District), the Fourth and South Seventh add Roosevelt District). (from District, Eighth and Helmett'a To the Plainfield Sixth (from District). To Ninth Town the Fifth add Montville Boonton (from District). District, Edge River the Fifth add Oradell changes Shifting particularly of these are obvious. the small town Some brings apprecia- from the Fourth to the Seventh District both Roosevelt mean, already nearly by the bly and the town is surrounded closer supra. Adams, See also Swann n. S., 445-446; v. 385 U. Court did not err that the finding plain- Thus the District that Feldman Plan showing tiffs had met their burden as did not come as nearly practicable population equality.

<1 itself, the discussion does not establish that By foregoing the Feldman Plan is unconstitutional. Rather, appellees’ that the Feldman Plan was not the success proving product of a effort to achieve means population good-faith equality the burden shifted to the State to only prove to achieve necessary deviations its were White v. Weiser demon state legitimate objective. some are to defer to state willing legislative poli strates that we norms, are consistent with constitutional cies, they so long if small differences in the of con they even require Upham 795-797; cf. U. gressional districts. Seamon, Finch, Connor 431 U. (1982); 456 U. S. S. (1977). Any number of consistently applied variance, including, some legislative might justify policies districts instance, respecting municipal making compact, districts, and avoid cores of boundaries, preserving prior incumbent As long contests between ing Representatives. Light Gomillion see nondiscriminatory, as the criteria are foot, (1960), 364 U. these are all legitimate objectives S. minor devi could showing justify population on proper g., Virginia Liberties Civil Union West See, e. ations. Oradell, Norwood, Mont- Similarly, Edge, River Seventh District. com- barely contiguous districts and almost present ville are with their Further im- pletely by suggested the new above. surrounded districts computer accomplished the aid of a provement could doubtless be with 5, supra. also n. detailed census data. See not, validity incorporating these course, prejudge plan of a We do represent good-faith changes, plan effort nor do we indicate that a cannot improvements. point minor We them a court can conceive of whenever could have been achieved only out illustrate that further reductions framework of the Feldman Plan. within basic (SD 1972) Supp. (ap 336F.

Rockefeller, 398-400 W. Va. proving plan justified by with 0.78% maximum deviation as compactness provision Constitution); Reynolds in State cf. (1964); v. Sims, 377 U. S. Burns v. Richardson, 73, 89, U. S. and n. 16 The State must, however, specificity particular objective show with some that a re quired specific plan, simply deviations in its rather than relying general showing jus required on assertions. The tify population depending deviations is flexible, on the size of importance the deviations, the of the State’s interests, consistency which as a whole reflects those availability might interests, and the of alternatives that sub stantially yet approximate popula vindicate those interests equality closely. By necessity, tion more whether devi justified requires case-by-case ations are attention to these factors. possibility justify that a State could small variations

the census-based of its districts on legitimate, consistently applied policy the basis of some was recognized Kirkpatrick case, itself. ad- Missouri theory, vanced the echoed in dissent, see Justice White post, at that district-to-district in the 771-772, differences eligible projected population jus- number voters, shifts, tified the deviations that case. U. rejected arguments 534-535. We its not because those fac- impermissible apportionment tors were in the considerations process, resulting but devi- rather because the size of haphazard “[a]t ations and because Missouri best. . . made adjustments population,” to a scheme based made on total attempt” “no districts, for the to account same factors all findings generally thoroughly and to and failed document its apply “throughout systematic, an them the in a not ad State hoc, at 535.11 Id., manner.”

11 Kirkpatrick very princi on which relied made clear that the cases equality entirely ple population preclude small deviations caused did Adams, See Swann v. 385 U. S. by policies. adherence to consistent state 742

The District Court found that did not properly appellants in the deviations this case. At justify population argument Court, in before the District Court and on this appeal appel- one for the Feldman emphasized only justification lants the preserving voting strength Plan’s deviations — submitted affidavits from minority They of racial groups.12 (1967); Reynolds Sims, 533, 440, 444 377 U. v. S. 579 District consistently Kirkpatrick recognized the applying Courts standard have See, White, justified. g., small could e. Doulin v. 528 that deviations (ED 1982) 1323, (rejecting projected population F. Ark. Supp. 1330 shifts plan largest justification for with 1.87% maximum deviation because projected Virginia largest growth); district also had West Civil Liber (SD 1972). Rockefeller, 395, Supp. ties Union 336 F. 398-400 W. Va. v. Furthermore, using Kirkpatrick proposed courts standard evaluate often, apportionments v. remedies unconstitutional have as White Weiser, rejected with the lowest deviation in favor of slightly plans higher poli deviations that reflected consistent state (NJ See, Cahill, 1972); g., Supp. e. David 342 F. Skolnick v. cies. v. 463 Illinois, F. Supp., State Electoral Board 336 at 842-846. A number of applying apportionments Kirkpatrick District Courts test state legislatures, disapproved practice this Court in Mahan v. How before (1973), ell, justification 410 315 also of small devia U. S. understood very possibility. g., Kelly Supp. 340 F. Bumpers, tions was real E. (ED aff’d, (1973); 1972), summarily Ark. 413 571 U. S. 901 Ferrell v. (WD Hall, Okla.), summarily Supp. ex rel. 339 F. Oklahoma aff’d, Tammany (1972); Jury, St. Police 406 U. S. 939 Sewell v. Parish 338 (ED 1971). Barnes, Supp. 252, F. 255 La. The court Graves 343 (WD 1972) Supp. applying F. 704 Tex. reversed this Court for - later (1973) all, Kirkpatrick Regester, at White v. U. S. - characterized by Kirkpatrick “The remains inquiry required as follows: critical issue variances, small, any justified the same: Has the and all however on State consistent, 713; policy.” Supp., F. see the basis of a rational State id., at 713-716. argument Court, appellants oral stated that the drafters of At this objectives as Feldman Plan were concerned with a number of other preserve well, namely preserve existing and “to “to cores of districts” Arg. municipal Tr. of 14. also Answer boundaries.” Oral (Record 17). ¶ Doc. No. on Behalf of Alan J. Karcher Counterclaim legislature Similarly, suggests was Speaker Karcher’s affidavit entirely Bergen concerned that the Ninth District should lie within justifications presented the Dis County. App. None these was 84. way, however, any general trict Court or this Court but the most Mayors Kenneth Gibson of Newark and Thomas of Cooke Orange, discussing importance having large East of a ma- jority App. of black voters Newark’s Tenth District, 86- an 92, as well as affidavit from Woodson, Jr., S. Howard Mayor comparing candidate of Trenton, for the Feldman Plan’s treatment of black voters in the Trenton and Camden with of areas the Reock Plan, id., at 93-96. See also (affidavit Karcher). id., at 82-83 of A. The District Court found, however:

“[Appellants] attempted not have to demonstrate, nor they any relationship can demonstrate, causal between goal minority preserving voting strength in the population Tenth District and the variances in the other goal preserving . districts. . . We find that the minor- ity voting strength in Tenth District related any way population to the in the deviations Fourth and Supp., Sixth Districts.” 535 F. at 982. largest

Under the Plan, Feldman districts are the Districts, Fourth and Ninth and the Third smallest are the supra, and Sixth. at 728. See None these districts bor- only on Tenth, ders one—the Fourth —is even men- appellants’ preserving minority tioned in discussions of vot- ing strength. large appellants suggest Nowhere do that the population necessary preserve of the Fourth District was minority strength; voting fact, deviation between Fourth District and other districts the effect of dilut- has ing including district, the votes of all residents mem- compared minorities, bers racial with other districts minority completely with fewer voters. The record is silent relationship voting preserving minority on the between presented question by appellants the relevant to this excludes them: Court policy minority voting strength jus- legislative preserving “Whether the equality tifies small deviations from census Furthermore, reapportionment plan.” Brief Appellants i. several plans legislature significantly before the lower deviations kept municipalities County Ninth all-Bergen intact and had an District. App. 66-74. strength populations of the Third and and the small Sixth findings easily Therefore, Districts. the District Court's pass “clearly the erroneous” test.

Y two-part properly applied of The District Court the test Jersey’s apportionment Kirkpatrick v. Preisler to New Representatives. States House districts for United plan correctly population deviations It held that functionally equal law, as a matter of and it found not were good-faith plan effort was not a achieve that equality using It available census data. also cor- best rectly rejected attempt justify appellants’ judgment supported the evidence. The deviations as not is Court, therefore, of the District

Affirmed. Jersey Congressional [Map this Districts follows New page.] Stevens, concurring.

Justice appellees ground affirmance, the con- As an alternative argument configuration of bizarre tended at oral Jersey’s congressional districts is sufficient to demon- New adopted “good faith.” This strate that the argument, was is a claim that the district it, understand they product are the because boundaries are unconstitutional my political gerrymandering. is vote decisive Since explain argument appropriate how this case, this it seems my analysis question divides the Court. influences gerrymandering pointed previously out, As I have Equal proscribed by species that is of “vote dilution” one analysis judicial adequate Protection Because an Clause.1 (CA7) City Chicago, Cousins v. Council F. 2d 848-853 Mobile denied, (1972); (Stevens, J., dissenting), 409 U. S. 893 cert. (1980) (Stevens, Bolden, judgment); J., concurring U. S. (1982) (Stevens, J., dissenting). Rogers Lodge, 458 U. S. special problems, gerrymandering I com- claim raises shall legal length gerrymandering the at some on basis for

ment judging claim, the standards for such a and their rele- claim, present the case. vance to

I interpreted Relying Constitution, in I, §2, on Art. (1964), subsequent Wesberry Sanders, 376 U. S. successfully appellees challenged congressional cases, Jersey districting plan adopted by Legislature. the New opinion in For the reasons stated Justice Brennan’s join, requires the doctrine of stare decisis Court, which holding demonstrated, however, result. It can be that the Wesberry, holding today, as well as our has firmer roots § provided by than those Art. 2. I, Constitution

The constitutional mandate in Art. con- I, §2, contained Representatives “appor- cerns the number of that shall be among says nothing tioned the several States.”2 The section composition about districts within a places Indeed, text State.3 of that section no restriction power any group on the whatsoever State to define the persons may particular within State who vote for candi- registered If a dates. State should divide its voters into separate by alphabetical classes defined order of their by by age, period initials, their their residence by State, or even their affiliation, such a classifica- § tion would be barred the text of Art. if I, 2, even the widely classes contained different numbers voters. I, §2, provides, part: Article “Representatives and direct Taxes apportioned among shall be the sev- eral may States which Union, be included according within this to their respective Numbers, which shall by adding be determined to the whole Persons, Number free including those bound to Service for a Term of Years, excluding taxed, Indians not three fifths of all other Persons.” Const., I, §2, U. S. Art. (emphasis cl. 3 supplied). 3During years the first 50 history, of our Nation’s widespread it was a practice to elect Members of the Representatives House of group as a on a Wesberry Sanders, statewide basis. 376 U. S. in his opinion out dissenting Harlan pointed As Justice Wesberry, of numerical the principle to the Civil War prior the by contradicted actually was of representation equality of that the “whole Number I, § which provided text of Art. should counted, certain Indians be that free Persons” should of all other Persons” only “three-fifths excluded, be and that Constitution, the to the total.4 analyzing be added should that, originally fact as regrettable we cannot ignore of On slavery. it tolerated the institution framed, expressly changes the basic hand, ignore other neither can we Civil They planted War Amendments. by caused firmly today’s holding. roots that support slavery citizenship The abolition of guarantees Fourteenth, Thirteenth, contained rights and voting §2’s effectively Amendments Art. repealed I, and Fifteenth than given greater that some votes be requirement weight true, however, I, §2, It remains that Art. does not others. itself contain any equality representation. guarantee The source of must be found But guarantee elsewhere. Justice partial as Clark noted his concurrence perceptively

4“Representatives apportioned among were on to be States the basis population plus population. of free three-fifths of the slave no slave Since voted, apportion inclusion of three-fifths their number in the basis gave representation voting ment the favored their States far excess of If, population. then, representation, slaves were intended to without exactly says prohibited: ‘weighted’ Article I did what the it Court now it Alternatively, the vote of voters in might the slave States. it have been thought Representatives speak by elected free men of a would State only also for representation the slaves. But since the slaves added to the State, Representatives of their own from the slave States could have been thought speak only States, indicating for the slaves of their own both that the possible Representative Convention believed it for a elected group speak nonvoting group one Representatives another and that in large degree thought were still speaking of as for the whole Id., a State.” at 27-28. Reading person, requirement I, 2,§ a “one one vote” is histori- into Art. cally textually as Kelly, well unsound. See Clio and the Court: An Affair, Illicit Love 1965 S. Ct. Rev. 135-136.

in had earlier in his as Justice Black written Wesberry —and Colegrove Green, (1946) in v. 328 569 dissent U. S. —that in the Protection Clause Equal guarantee firmly grounded Even Harlan’s pow- of the Fourteenth Amendment.5 Justice Wesberry in no in that analysis. erful dissent could find flaw In its of state laws dis- redefining congressional review Wesberry Sanders, to v. has not tricts Court subsequent to on the Protection Clause. rely found it necessary Equal the basis for has, however, provided apply- That Clause the “one one vote” standard to other electoral ing person, g., Carr, e. Baker v. 369 See, (1962); districts. U. S. 186 Reynolds Avery Sims, 377 U. S. v. Midland (1964); v. 533 County, §2, 390 474 Even if Art. I, U. S. were the “one one vote” rule would person, wholly disregarded, to action state officials con- apply defining unquestionably it districts does to state action gressional just defining districts.6 state legislative

5 permit pick qualified That Clause “does not to out certain cit States deny right groups izens or of citizens and them the to vote at all.... No deny equal protection prohibit a one would clause would also law expressly give that would certain citizens a half-vote and others a full vote. probable Apportionment coming The effect of the 1901 Act in the State citizens, among appellants, election will be that certain them the will in only choosing repre some instances have votes one-ninth in as effective Congress discriminatory sentatives to as the votes of other citizens. Such legislation exactly protection me equal seems to the kind that the clause Green, prohibit.” Colegrove (Black, intended to was v. 328 U. at J., Wesberry Sanders, dissenting), quoted part supra, (Clark, in in J., concurring part dissenting in part). rule, person, The “one one vote” the Equal like Protection Clause in firmly grounded, provides which it is protection against more than one form of discrimination. In the developed, cases which the rule was first district significantly weight boundaries accorded less individual votes populous the most it districts. But was also clear that those boundaries political strength political

maximized the of rural voters and diluted the power Hacker, Congressional Districting: urban voters. See A. The Equal Representation generally Issue of see Standards for (1963); Congressional (Apportionment), Districts Hearings before Subcommittee Judiciary No. of the House Committee on the H. on R. H. R. gov- every requires Equal State Protection Clause governing adopts rules its impartially. a State ern When defining machinery those boundaries, electoral or election community. of the entire serve interests rules must they supra, Reynolds If serve no at 565-566. Sims, segment purpose racial, one other than favor —whether may occupy religious, economic, ethnic, —that strength particular point position time, or to dis- at a community, they advantage politically segment of the weak guarantee equal protection. the violate constitutional (1960), Lightfoot, the 339, In v. 364 U. S. 340 Gomillion change city in the boundaries Tuske- Court invalidated twenty-eight- gee, square to an uncouth Alabama, “from a figure” excluding virtually city’s all the black voters. sided right opinion vio- The Court’s identified the that had been group right: lated as a legislature singles readily out a isolated

“When thus segment minority special discriminatory of a racial In treatment, it violates Fifteenth Amendment. no involving unequal weight voting case in distribution that has come before the Court did the decision sanction a dif- whereby approval given on ferentiation racial lines was solely unequivocal from withdrawal of the vote colored Id., citizens.” at 346.

Although explicitly the Court- its decision rested on analysis Fifteenth Amendment, Justice Whittaker’s concurring opinion Wesberry Justice Clark’s —like —is equally coherent, see 364U. at 349. Moreover, Court subsequently though has treated Gomillion as it had been equal protection grounds. on decided v. See Whitcomb (1971). 403 149 Chavis, 124, U. S. (1959). 8266, 8473, Cong., Sess., pri- H. R. and H. R. 86th 1st 65-90

mary consequence of protection the rule has been its of the individual voter, provided but it has identifying curtailing also one mechanism for against cognizable groups discrimination of voters.

749 complete geographical involved of a Gomillion exclusion racially group. arising identified But case after case Equal suggested under the Protection Clause Court has voting strength cognizable 'political that “dilution” of the groups may. Thus, as well as racial be unconstitutional. question Dorsey, in Fortson reserved v. 379 S. 439 433, U. (1965), apportionment might “oper related to an scheme that voting strength ate to minimize or cancel out the of racial political voting population.” or elements of See also Gaffney Cummings, (1973); 735, 751, v. U. S. White Regester, (1973); 412 U. S. Whitcomb supra, Chavis, 143-144; Richardson, Burns v. 384 U. S. separate opinion In 73, 88-89 his in Williams v. (1968), Douglas pointed 393 U. Justice Rhodes, S. out Equal protects “voting rights that the Protection Clause political groups units, ... as well as economic commu racial and other nities, entities.” And Abate v. Mundt, (1971), any U. the Court S. noted absence particular tending “built-in bias to favor interests geographic dissenting opinion today, areas.” his Justice Jersey’s agree seems to that New would violate White Equal “invidiously Protection Clause if it discriminated against political group.” Post, racial or 783. only Equal

There is one Protection Clause. Since the groups equal Clause does not make some more citizens (1982) than Zobel others, see 457 Williams, U. S. 71 concurring), against protection J., its dilu- vote (Brennan, groups. pro- long tion cannot be confined to racial As as it gerrymandering against groups, proscription scribes its such provide cognizable comparable protection must groups for other previously I

of voters as well. As have written: line-drawing process, religious, “In the racial, ethnic, political gerrymanders species and economic are all gerrymanders. standpoint groups

“From are of voters that important line-drawing process, affected it is also 750 gaining group’s or recognize interest in it is the power

maintaining mere political stake. The that is at ethnic, a common of citizens share a number fact that religious background the need does not create racial, or only gerrymandering. protection against It is when strong enough mani- to be interests are their common political For the that the need arises. action fested group strength ethnic, political a function of its of a is not composition; religious it is function of rather racial, persons specifically who will the number of numbers— vote way.” 446 U. Bolden, Mobile S. the same (1980)(concurring judgment). Chicago, City 466 F. 2d 851- Cousins v. Council (CA7) (Stevens, dissenting), denied, 409 U. S. J., cert. (1972).7 II preoccu- judicial I am convinced that Like Justice White, perfect population equality

pation goal an is inade- constitutionality judging apportion- quate of an method gerrymander plan. I that an obvious ment wholly would hold simply it immune from attack because comes closer to every competing plan. perfect population equality than On any proposal hand, the other do not find virtue Wesberry subsequent forth in relax the standard set ignore population disparities cases, and to after some arbi- trarily has been As one defined threshold crossed.8 com-

7Similarly, gerrymander the motivation for the turns on the strength voting patterns, group, of members of the derived from cohesive 2d, of their common F. rather than on the source interests. at 852. appear The former would to be consistent with what the Court has writ case, ante, 734-735, 6; ten in this n. the latter would be consistent dissent, post, what with Justice White has written at 780-783. Either approaches gerrymandering of these would the to leave door unrestricted open. Engstrom, Supreme Gerry Equipopulous wide See The Court and mandering: Quest Remaining A in the for Pair and Effective Obstacle 285-286, Baker, Representation, 296; Quantita- 1976 Ariz. L. J. State experience,

mentator has written: as well as tells us “Logic, . . . that there can be no total sanctuaries the political will shift thicket, else unfairness from one form to simply Rather, another.”9 we should supplement standard additional criteria that are no less “judi- equality ” In cially evaluating challenges manageable. equal protection to as in such attacks on other districting just resolving plans, action, forms of would consider whether the discriminatory has a adverse on an plan significant political identifiable impact whether has indicia of plan objective irregular- group, then, whether the is able to ity, produce convincing State neutral, evidence that the nevertheless serves legitimate interests of the as a whole. community

Until two decades constrained its fear of ago, entering thicket, standardless the Court abstained simply from any constitutionality attempt judge legislative even apportionment plans, popu- when districts varied lation 914,053 112,116. Green, from Colegrove S.,U. at 557. In Baker v. Carr, (1962), 369 U. S. 186 Reynolds Sims, 377 U. S. 533 (1964), Court abandoned that extreme form of restraint judicial and enunciated “one one person, vote” That standard is principle. “judi- cially because census data are concrete and rea- manageable” reliable and because sonably judges can and divide. multiply Even as a basis voters in their individual protecting ca- the “one one vote” has its pacity, person, shortcom- approach are ings. Although population disparities easily quantified, the standard significance no measure of the provides any numerical It difference. the element of easy recognize Descriptive tive and Gerrymandering, Guidelines to Minimize 219 Annals (1973) (“If N. Y. guidelines Acad. Sci. specific more to minimize gerrymandering forthcoming, great are not principle— then a democratic man, one one degenerated simplistic vote —will have into a arithmetical scale”). discriminatory facade for cartography on an extensive 9Dixon, Court, Man, Vote,” People, and “One in Reappor One (N. 1971). 1970s, tionment p. Polsby ed. Congress- allowing one voters to elect 112,116 unfairness signifi- by 914,053. But how elected another is man while 527,472 counts census is difference between cant the mortality the tran- rate, rate, the birth Given 523,798? acknowledged society, and the of modern sient character may such all know that differences in the we census, errors and the date of the date of the census vanish between impossible population equality is election. Absolute next achieve. equality important, is not a sufficient mere numerical

More Although directly pro- representation. guarantee equal it indirectly groups only protects it at best. individuals, tects may Reynolds supra, A at 561. voter chal- Sims, ground gives lenge apportionment it scheme on an pur- weight voters; than of other for that vote less his pose plaintiff combined with not matter whether is it does might group separated who share his affili- from others plainly It is unrealistic to assume that a smaller ation. districting disparity always produce will a fairer numerical plan. correctly Harlan observed in Indeed, as Justice Wells a standard “of abso- 394 U. S. Rockefeller, (1969), compatible ‘gerrymandering’ equality perfectly with lute computer may grind A sort. out district lines worst totally popular which can frustrate the will on an overwhelm- ing wrote, number of critical issues.” Justice Since Harlan computer developments technology have made the task gerrymanderer post, even easier. See at (WHITE, dissenting).10 J., 10Computers possible generate large make it now a of alterna number plans,

tive consistent equal guidelines other and various criteria, relatively period time, analyze short and to contrast, characteristics of each one detail. “[i]n considerable reapportionment, 1970's round of barely generate some states were able to single reapportionment plan in the time allotted to the task.” National Legislatures, Conference of State Reapportionment: Technology Law and (June 1980); Engstrom, supra see also n. at 281-282. imperfections The in the numerical standard do not, of provides render it course, useless. It one neutral criterion evaluating districting plan. disparities may Numerical provide shifting justification sufficient basis for the burden of equi- to the Moreover, State. if all other factors were in poise, proper plan it would be to conclude that the that most nearly goal complete equality attains the of would be the plan. major shortcoming fairest The of the numerical stand- is ard its failure to take of account other indeed, relevant — important relating group more of the fairness —criteria participation political process. may in the To extent, that it counterproductive. Gaffney Cummings, indeed be See S.,U. at 748-749.11 cognizance To a limited extent the Court has taken of dis- criminatory groups path treatment of The voters. political Court has sometimes used to enter this thicket is finding majority marked the label “intent.” A that deliberately sought minority group to make it difficultfor a representatives may provide holding elect a sufficientbasis for objectively plan that an is neutral electoral unconstitutional. (1982). Rogers Lodge, See For U. S. 616-617 rea- length, already sons forth that have set this standard is inadequate. dissenting); id., J., at 642-650 (Stevens, concurring J., Mobile v. 446 U. at 83 in Bolden, (Stevens, judgment). legislature’s districting I would not condemn a plan discriminatory impact simply in the absence because proponents part, by partisanship its were motivated, group Legislators politicians; are, all, animus. after it is un- proscribe attempt realistic to all considerations essentially political process redistricting. long In the adjudication premised case-by- run, constitutional on a subjective appraisal case intent local decisionmakers Edwards, Gerrymander Man, Vote,” 11 See and “One One Prometheus, Proteus, Pandora, (1971); Elliott, N. Y. L. U. Rev. 879 Consequences Reapportionment, Procrustes Unbound: The Political Engstrom, supra (1970); L. U. Chi. Rev. n. 8. adminis- satisfy impartial

cannot possibly requirement Equal tration of the law that is embodied Protection Amendment. hand, Clause of the Fourteenth On other if a defined has a adverse significant impact upon an that it dramati- showing departs additional political group, from criteria should suffice to shift the task of cally neutral to the state defendants. justification this is can reasons, plaintiffs

For a number of a burden matter, As a cases. threshold relatively plain- meet few of tiffs that are members an identifiable they must show strength They whose has been diluted. political voting group class, to a salient they belong politically must first that prove supra, one whose distribution is 749-750, see at geographical it could have been taken into ascertainable sufficiently Second, in boundaries.12 must drawing they account district or in the State the relevant district districts prove their influence has been whole, proportionate voting as a Third, scheme.13 challenged plain- affected adversely affiliation, race, generally groups will be based on Identifiable status, origin, religion, or but other char group, ethnic national economic particular may politically significant context. acteristics become Gerrymander Explorations in The Clinton, Further the Political Thicket: (1973) Constitution, 1, (cognizable 59 Iowa L. Rev. 38-39 interest and the Comment, legislative policy); Political group with coherent and identifiable Statutory Compactness an Antidote for Gerrymandering: A Standard as (1974)(clearly 398, Impotence, 41 L. Rev. 407-408 identifi Judicial U. Chi. group). and stable able difficulty making showing from the existence of alter this stems circumstances, Depending vote strategies of vote dilution. on the native group mem may if a concentration of dilution be demonstrated districts, group if of have fragmented among or members bers has been percent single greatly in a in excess of the been overconcentrated district Bolden, Mobile v. age elect a candidate of their choice. See needed to Hacker, judgment); (Stevens, J., concurring in and n. 13 U. 46-50; Note, Reapportionment, 25 Compensatory n. cf. Racial supra (1972) shortcomings of (pointing to the several L. Rev. Stan. including opportunity swing votes and political strength, of to cast tests representative group). of their own opportunity to elect Act, developed Voting have litigation Rights under the federal courts familiarity identifying measuring dilution of problems of some with the tiffs must make a prima facie showing that raises a rebuttable of presumption discrimination.

One standard method by which of members a disadvan taged political group may establish a dilution their voting rights by reliance on the “one person, one vote” principle, which on depends a statewide statistical But analysis. prima facie evidence of gerrymandering can be surely presented other ways. One obvious type of evidence is the shape district configurations themselves. One need not use Justice Stewart’s classic definition obscenity know it when —“I see it”14—as an ultimate standard for judging constitu of a tionality gerrymander recognize that dramatically irregular may have sufficient shapes probative force to call for an explanation.15

Substantial divergences from a mathematical standard of compactness may symptoms illegitimate gerrymander- As Dr. ing. Reock, Ernest Jr., Rutgers University has written: “Without some requirement compactness, boundaries a district may twist wind their across way in fantastic fashion in map order absorb scattered group voting strength. concepts developed racial of the Some for statu- tory purposes might applied adjudicating be in constitutional claims types political groups. of showing may other of harm threshold more for political party difficult adherents of a than for members of a racial group, however, possible because there are of a number base-line meas- party’s strength, including registration past ures for voter and vote- getting performance generally in one or more election contests. See Backstrom, Robins, Eller, Gerrymandering: Exploratory Issues in An & Minnesota, of Gerrymandering Applied Measure Partisan 62 Minn. L. (1978). 1121, Rev. 14 (1964). Ohio, 184, U. 197 Jacobellis 378 S. 15 defining gerrymander quite properly against Professor Dixon warns ing Dixon, shapes. Representation: in of Democratic terms odd See R. (1968). time, At the Reapportionment Law and Politics 459-460 same however, recognizes compactness contiguity, rule of and “if used he that a districts, merely explanation odd-shaped have much to force an for can Id., Tribe, 760 merit.” at 460. L. American Constitutional Law See (1978) (oddity shape, coupled of district’s with racial distribution State). justification population, the burden of should shift 756 extent, To some support.”16 geographi- of

pockets partisan serves it facilitates independent values; cal compactness electoral and constituent campaigning, organization, A number state statutes and Constitu- representation.17 districts to be and These compact contiguous. tions require because have not utility standards have been of limited they with and Yet applied rigor precision.18 been defined and set forth a number Professor Reock and other scholars have that can be measuring compactness computed of methods as a the same virtually degree precision of a true, course, count.19 It is significance par- 16 Ap Reock, Measuring Compactness Requirement Legislative as a (1961). 70, Backstrom, Robins, portionment, 5 Midwest J. Pol. Sci. 71 Cf. 1126, 1137 Eller, supra (compactness n. standard cannot eliminate & may of discretion available to those gerrymandering but reduce the band boundaries). possible group’s district It is of course to dilute drawing relatively compact. Engstrom, if all are voting strength even districts 8,n. supra 280. Evaluating Taylor, Shape A New Measure for Electoral District Patterns, Compactness is not to be 67 Am. Pol. Sci. Rev. Reynolds Sims, physical As 377 U. S. confused with area. we stated (1964): transporta developments improvements “Modern mid-1960’s, hollow, rather most tion and communications make population-based validly be representation from can claims that deviations allowing solely Arguments for such geographical on considerations. based representation sparsely in order insure effective settled deviations legislative becoming large that the prevent areas and to districts from so impaired availability representatives of citizens to their are of access Nevertheless, although today, part, unconvincing.” low for the most districts, density may differ require geographically extensive presented by with distorted questions ent are the creation districts shapes irregular, boundaries. indented explicitly require that dis state statute and 21 State Constitutions One explicitly pro compact; statutes and Constitutions tricts be two state territory. Congressional contiguous of See vide that be formed districts Service, Statutory and Provisions Concern Research State Constitutional (June 1981). Redistricting Legislative But ing Congressional and State (ineffective enforcement); Comment, supra Clinton, 12, at 2 supra see n. 12, n. at 412-413. 19 scholarly suggests of different mathematical literature a number rela- compactness, focusing each on different variables. One measures of

757 compactness may ticular measure evaluate, be difficultto but figures may as the demonstrate, this case the same be said population disparities. although of In addition, some devi- compactness may inescapable ations from because of the geographical configuration population density or uneven of a particular degrees compactness the State,20 relative of of dif- tively simple relationship method to measure the between the area of the possible district and the of circumscribing area the smallest circle. See supra Reock, 16, particularly n. 71. This calculation is sensitive the degree elongation given shape. of simple of a Another method is to deter- figure’s perimeter mine the ratio of a to the circumference of the smallest possible circle, circumscribing a measurement that is well suited to meas- uring degree the of indentation. Schwartzberg, Reapportionment, Gerrymanders, and “Compactness,” the Notion of L. 50 Minn. Rev. 443- (1966). Other compactness aggregate measures of are on the based of geometrical the from the or population-weighted distances district’s cen- points, Kaiser, of gravity Objective ter to each of its see An for Method Legislative Districts, Establishing (1966); 10 Midwest J. Pol. Sci. 200-223 Hess, A Nonpartisan Districting: Development Weaver & Procedure for of Computer 288, (1963); degree L. Techniques, 73 Yale J. 296-300 Taylor, supra district, indentation of the boundaries of a nonconvex see 17; length boundaries, Cause, aggregate n. of district see Common System Representation” (1977); of “Fair and Toward a Effective Adams, Continuing A Apportionment Statute: Model Process: The State Representation,” 825, Quest Legis. for “Fair and Effective 14 Harv. J. 875-876, (1977); Edwards, 894; Walker, supra and n. 184 n. One Ideal, Hastings Q. In Man-One Vote: Pursuit Of an Elusive Const. L. (1976); to minimum and the ratio of maximum the diameters district, Morrill, Redistricting Geographic Theory R. Political measurement, case, compact the more each the smaller the 4b(3)c Acts, § (setting district districts. See also 1980 Iowa ch. compactness geometrical determining forth relative alternative tests plans: districting alternative the absolute value of the difference between district, length dispersion of the and width “ratio population dispersion district to the about center district”). geographic about the center of the boundaries, oddly shaped If a have adher State’s subdivisions ing may geographical compactness. See to these boundaries detract from (1980) §§2-2-105, explanations that (legislative Rev. 2-2-203 Colo. Stat. compactness shape county bound variations from were caused “the ary lines, lines, boundaries, population den- census enumeration natural ferent can As maps always district with the compared. standard, numerical it seems fair to conclude that drastic de- from a partures signal may are compactness something be amiss.

Extensive deviation from established boundaries political for is another basis a facie possible prima showing gerry Reynolds v. Sims: “Indiscrimi As we wrote mandering. nate without districting, any regard political subdivision or or boundary lines, may natural historical be little more than an invitation to open partisan gerrymandering.” S., at 578-579.21 boundaries tend to remain U. Subdivision stable over time. Residents of units such as town cities, often a ships, develop and counties inter community an est, when the subdivision role particularly plays important In services. provision governmental addition, legis lative districts that do not cross subdivision boundaries are administratively convenient and less to confuse the likely from Although significance voters.22 of deviations sub districts”); sity, adjacent Adams, compactness and the need to retain 19, 875-876, supra n. n. 184. addition, compactness may sociopolitical

In from com- geographic differ pactness. Baker, 8, supra geographer n. As one has noted: 205. uneven, strung many regions, perhaps along “In out roads may cheaper railroads. Travel be easier and some directions than others, elongated major transport such that an district astride a corridor might compact in fact be the most in the of minimum travel time for a sense so, representative to travel around the district. If then a modified crite- time, rion, the minimum travel would be a the ratio of the maximum to preferred Morrill, supra n. at 22.- measure.” 526, 534, (1969), Preisler, n. 4 Kirkpatrick 894 U. S. Court correctly prevent boundaries could not noted that adherence subdivision attempt to gerrymandering. But there it was concerned with the State’s adhering existing justify population disparities by policy subdivision partisan gerrymander My discussion here is directed toward boundaries. ing relatively To the extent that equipopulous in scheme with districts. Kirkpatrick reject respecting dicta in the notion that subdivision boundaries gerrymandering, respectfully disagree. See n. will not inhibit infra. 22Morrill, supra n. at 25. division boundaries will vary with the number of legislative seats and the number, size, and of the shape State’s sub divisions, the number can be counted23 and alternative plans can be compared.

A standard, procedural although less obviously precise, also be If may the for enlightening. process and formulating adopting a excluded plan divergent re- viewpoints, openly flected the use of partisan criteria, and no provided explana- tion of the reasons for one over selecting plan another, it would seem to conclude that an appropriate adversely af- plaintiff fected is entitled the group to have majority explain its action.24 On the other if hand, neutral decisionmakers de- on the if veloped criteria, basis neutral there plan was an adequate and opportunity presentation consider- ation of if view, and differing points guidelines used a were a selecting plan explained, strong presumption should attach validity to whatever such a plan process produced. a scheme fact

Although voting worsens the a position and its particular group,25 though configuration or geographic 23See, g., 315, 319, Howell, (1973);Backstrom, e. Mahan v. 410 U. S. Robins, Eller, 13, 71; Morrill, 19, supra supra & n. n. n. 25. average The smaller the of a subdivision relative to the district population, the among more dubious it is to divide it two or more districts. particularly suspect particular political It is also divide a subdivision among territory more than two districts which also contain other subdivisions. (1964) 24 See, Rockefeller, g., Wright (Goldberg, e. 376 U. S. (the

J., Edwards, dissenting); supra congres n. at 881 1961 New York redistricting plan by party sional up majority leg was drawn members of a participation any islative of the committee staff without member held; plan opposition party; public hearings no was released to the were straight public day adoption; approved by party-line it a before its was legislature; single extraordinary session of the vote afternoon at an signed day). the Governor the bill the same may grounds 25 The on the that this element has not been State defend challenge adequately example, plaintiffs’ if the is based on a shown. For districts, may particular district or be able to show State genesis sufficiently irregular to violate one or more constitutionally just it will nevertheless be discussed, criteria if the that the as a whole valid State can demonstrate objectives. acceptable, neutral The same kinds embodies accepts legitimate justification in the con- that the Court population disparities text of would also be available when- shape, compactness, boundaries, ever the criteria of up warning flags. decisionmaking procedures have sent invalidity, prima In order to facie case overcome may “legitimate incident to the State adduce considerations Reynolds policy,” Sims, of a rational state v. 377 effectuation may S.,U. also particular objective specificity that

“show with some specific plan, requires in than the deviations its rather general showing simply relying . . . on assertions. depending the the flexible, deviations, is on size of consistency importance interests, of the the the State’s plan interests, reflects those with which a whole availability might of alternatives that substan- and the yet popula- tially approximate vindicate those interests closely.” equality Ante, more at 741.26 tion prima respond plaintiff’s facie If a State is unable to to a by showing plan supported by adequate neu- case that its properly I could conclude that criteria, tral believe court entirely challenged totally scheme is either irrational or in whole. Even if the group’s voting strength is not diluted the State as a reduced, previous may plan group’s voting strength has in fact been Backstrom, Robins, gerrymandered generally in its favor. have been supra Eller, (discussing possible standards of “fair & n. at 1134-1137 representation”). 26 justifica determining In the State has carried its burden whether interests tion, give weight importance of the State’s greater I would to than to the size consistency which those interests are served and the perspective implied in the I of the deviations. Thus do not share Preisler, Kirkpatrick v. justifications in purported Court’s discussion S., 394 U. at 533-536.

761 the political motivated a desire curtail strength This not mean affected does that federal political group. review every courts should invalidate even apportionment have been affected to some extent may by partisan But am convinced that Judi- legislative maneuvering.27 is not a constitutional ciary powerless provide remedy cases.28 egregious

Ill In it this case is not necessary go beyond the reasoning Wesberry Sanders, 376 U. S. in the Court’s v. opinions Kirkpatrick Preisler, (1964), S. (1969), U. large potentially political groups, Given the number of affected even a neutral, justifiable plan may change position groups well of some addition, the worse. inevitably some “vote dilution” will from result Backstrom, Robins, Eller, supra patterns; residential see & n. at 1127. Although may plan, State of course adduce this factor defense its prima enough the criteria for a should demanding they facie case every are in the See Mobile apportionment plan. not satisfied case of Bolden, (“the J., at concurring judgment) (Stevens, 446 U. every impact standard cannot condemn adverse on more one or groups spawning litigation without more dilution than judiciary can id., manage”); (quoting opinion at n. 15 93, Baker of Justice Frankfurter Carr, (1962)). v. 186, 369 U. S. 267 28 Lightfoot, (1960) See Gomillion v. 364 U. 341 (noting S. that alle gations “abundantly would 140 ordinary establish that Act was not an geographic redistricting gerry measure even within familiar abuses of mandering”). in Gomillion had Tuskegee If the map virtually all excluded Republicans city limits, rather than blacks from the the Constitution would also have been violated. gives comparably Professor Tribe egregious a hypothetical: numerical example, jurisdiction

“For if consisting a Republicans and 460 randomly districts, Democrats were Republicans subdivided into 10 would probably However, elected six or more districts. if malevolent Dem- ocrats could precision, they might draw district lines with be able to isolate Republicans one district and win all the other district elections a margin votes, of one or capturing two legislature thus 90% of the state commanding only while supra popular Tribe, 46% of the vote.” n. 756, n. 2. Hacker, supra n. at 47-50. (1973), to reach correct 412 U. Weiser, S. White I have mentioned criteria that None of the additional result. propriety any hold- on the Court’s doubt would cast Although ing I need not decide whether in this case. regarding shape compactness, shortcomings plan’s sub- decisionmaking would estab- and neutral boundaries, division certainly strengthen prima these factors case, a facie lish Equal Jersey plan my violates the the New conclusion that Clause. Protection followingp. glance map, district 744, shows ante,

A at the deserving descriptive adjec- configurations the kind of well traditionally have and “bizarre”30—that tives—“uncouth”29 acknowledged gerrymanders. I have used to describe been compactness applied measures of to the the mathematical not likely Jersey map, would I think it New but disregarding geographical In addition, fare well. while disregards redistricting wantonly compactness, scheme example, county in the of a com- For words boundaries. cartographic fancy, Legislature flight “In mentator: packed many Jersey Republicans North into new district body long from call ‘the Swan.’ Its neck and twisted stretch upper rural reaches of the Dela- the New York suburbs to the segments Fifth, district, River.” That contains ware The same commentator described least seven counties. parts comprised of five District, counties, Seventh tracing curving partisan path through “a industrial Eliza- largely Marl- beth, liberal, academic Princeton Jewish Lightfoot, supra, Gomillion at 339. 30Indeed, very map this was so described in a recent article entitled New Jersey Map Imaginative Gerrymander, appearing Congressional in the Quarterly: Jersey’s congressional map gerryman “New new is a four-star in bizarrely shaped der that boasts some of the most districts to be found (1982). Congressional Quarterly glance quick the nation.” A districting maps for the other 49 States lends credence Congressional Directory this conclusion. 1983-1984 Official County. resulting monstrosity in The boro Monmouth was by Congressional Quar- called ‘the Fishhook’ detractors.” (1982).31 terly 1193-1195 process map prompts inquiry a an into the that led

Such plan sponsored adoption. by leadership was the its Party, the which Democratic controlled both houses of the legislature state as well as the office, Governor’s was signed day inauguration Republi into law the the before a legislators formally explained can Governor. The never guidelines formulating selecting used in their inor it plans. rejected plans over other available Several nearly equal popula contained districts that were more compact, more tion, and more consistent with subdivision including by recognized expert, boundaries, one a submitted impartiality Dr. Reock, Jr., Ernest whose and academic cre challenged. dentials were not The District Court found that rejected the Reock Plan “was it did because not reflect the leadership’s partisan Daggett concerns.” v. Kimmelman, (NJ 1982). Supp. 535 F. 978, 982 This conclusion, which naturally persuasive justifications from the arises absence of rejection by for the Plan, Reock buttressed letter Speaker written to Dr. Reock the Democratic of the New Jersey Assembly. frankly explained General This letter importance taking advantage to the Democrats of of their opportunity redistricting to control after the 1980 census. Speaker justified partisanship by The ing his own overt describ considerations that had motivated the Re publican majority adoption plans in the of district New

31The same commentator described the Thirteenth District this man ner: “In an effort to a ‘dumping ground’ create for Republican votes trou bling Howard, Hughes to Democrats Legislature established a 13th District that map, stretches all over the Philadelphia from the suburbs County in Camden to the County.” New York suburbs in Monmouth Congressional Quarterly, argument, at 1198. At oral we observed yet likeness between boundaries of another district —the Fourth —and shape running Arg. back. Tr. of Oral 21. In the present.32 in other States in the past Jersey —and decisionmaking process that the indicates the record sum, far from was challenged plan to adoption leading the number of Demo to increase It was designed neutral. New of Republicans, number crats, and to decrease in future Congress years.33 to would send Jersey’s voters justifica any legitimate does not show the record Finally, although in the New Jersey plan, for irregularities tions theory on tried a different case was concededly Court. District other dis- study I have not made comparative

Because had has not the opportu- and because State tricting plans, redistricting Jersey also be viewed from “Congressional in New must Republican party only is perspective. The the more national broad-based Congress. With a shift short of absolute control 27 votes traditionally Congressional from the Democratic consequently and seats Republican dominated sun-belt states industrial states to the more urban redistricting process by Republicans opportunity as an viewed entirely.” margin, perhaps or even overcome it close that 27 vote Supp., F. at 991. legislators. Democratic

Copies of the letter were sent to all Judge disagreed holding with the of the Dis Although Circuit Gibbons case, paragraphs dissenting opinion of his concluding trict Court in this identifying difficulty he this as a unambiguously imply that would have no disadvantage an in which the district lines were drawn in order case political group. identifiable He wrote: me, apportionment map produced by P. L. c.l leaves as a

“The Jersey, citizen of New disturbed. It creates several districts which are contiguous only anything compact, at least one district which is for but yachtsmen. maintained, municipal have been has While boundaries there having community been effort little to create districts interests. districts, stations, example, some different television and radio differ- systems newspapers, transportation ent different northern serve the harshly partisan Speaker and southern localities. Moreover the tone Christopher Reock, disedifying, say Jackman’s letter to Ernest Jr. is C. plain, well, partisanship produced bulges It is least. artificial appendages place Congressmen of two districts so as to the residences running against they and Courier in districts incum- Smith where would be Id., bents.” at 984. nity justifications specifically to offer directed toward the I I discussed, additional concerns have cannot conclude with Jersey certainty plan absolute that New was an unconsti- partisan gerrymander. agreement I tutional But am in full holding plan that, with the Court’s because the embodies de- justified equality viations from that have been by any objective, neutral state it cannot if Further, stand. population equality only provides political gerry- check on mandering, virtually impossible it would be to fashion a fair remedy shape and effective a case like if this. For legislative entirely districts is unconstrained, the dominant majority respond judgment could no doubt to an unfavorable by providing grotesque-appearing map an even more acceptable equality greater polit- reflects numerical with even judges inequality. prevent ical If federal can that conse- by taking quence shape things a hard look at the to come remedy hearing, they believe can also scrutinize the original map with sufficient care to determine whether dis- any have tortions rational basis neutral criteria. Other- promise Reynolds wise, the v. Baker Carr and v. Sims— judicially manageable “[f]ull standards can assure by participation may effective all citizens,” 377 U. at 565— never be fulfilled.

Justice White, with whom The Chief Justice, Justice Rehnquist Powell, and Justice join, dissenting. congressional reapportionment This case concerns Jersey. districting plan by New enacted the New Jer- sey Legislature signed by into law the Governor on Janu- ary 19, 1982, 1, Pub. L. ch. reduced the number of con- gressional required by districts in the from State 15 to figures. the 1980census The 14 districts cre- by legislature average ated have an deviation of 0.1384% largest and a maximum deviation between the and smallest districts of 0.6984%. In words, other this case concerns a in districts encom- 3,674 individuals maximum difference The New Jersey million people. than a half more passing District because a divided Court invalidated was “ despite were not ‘unavoidable variances these Daggett absolute equality.’” effort achieve good-faith Kirk- (NJ Kimmelman, 1982), 535 F. Supp. quoting patrick Preisler, Today, 394 U. S. thereby striking the District Court’s decision affirms Court in Court’s a experience legislative the first time the for an and maximum average congressional districting plan variance of under 1%. in- from the Court’s unreasonable dissent respectfully in of on an unattainable perfection equalizing sistence is not today The Court’s decision districts. congressional Kirkpatrick supra, and White Preisler, v. v. by compelled infra, Weiser, if the see Part and (1973), I, 412 U. S. 783 that our demand the result reached convinced cases Court precedents. time has arrived to reconsider these today, event, In an affirmance the decision below is inconsist- any Kirkpatrick own “modifications” of majority’s ent with the and White which minimum, further consider- at a require, IV, ation of this case the District Court. See Part infra. of fair and effective for all achieving representation “[T]he citizens is aim of concededly basic legislative apportion- Reynolds Sims, ment.” v. S. 565-566 U.

One must to believe that the draco- credulity Court’s suspend nian to a 0.6984% maximum deviation response trifling pro- motes “fair and effective for the representation” people Jersey. New The as is nearly practi- “as requirement man’s in cable one vote election is to be worth congressional Wesberry Sanders, as much another’s,” as 376 U. S.

(1964), must be in light understood malapportion- Wesberry ment in the States at the time was decided. Wesberry

plaintiffs were voters district congressional 823,680) (population encompassing Atlanta was three (272,154) larger Georgia’s than times smallest district average more size of than double the an district. Because reapportioned years, the State had not for 30 the Atlanta Georgia’spopulation only possessing District one-fifth had Congressmen. Georgia atypical; one-tenth of the was country congressional throughout districts had not been redrawn for decades deviations over 50% were the rule.1 These substantial differences district size dimin- representativeness congressional ished, sense, in a real profoundly elections. The Court’s invalidation of these un- equal precise districts should not be read a demand equality mathematical between the Indeed, districts. sensibly may possible [for “it Court observed that not be States] to draw districts with mathematical precision.” Reynolds supra, Id., Sims, at 18. at 577, requirement decided Term, the same the Court disavowed a legislative of mathematical exactness for districts even explicit more terms: practical impossibility arrange

“We realize that it is a legislative districts so that each one has an identical residents, number of or or citizens, voters. Mathemati- precision hardly cal or exactness a workable constitu- requirement.” tional responded Wesberry by eliminating gross

The States to disparities between districts. Nevertheless, redistricting plans by smaller far variations were struck years Kirkpatrick supra, the Court five Preisler, later v. (1969). companion, and its Wells v. 394 542 Rockefeller, U. S. redistricting The statutes before the Court contained total percentage respectively. deviations of 5.97% and 13.1%, 1 By 35 out of 42 among States had variances their of districts 100,000. (1964) Wesberry Sanders, (Harlan, over v. 376 U. S. J. dissenting). recognized significance Court has the fact that “enormous variations” in early legislative district size were at issue in the Gaffney Cummings, apportionment 735, 744, cases. 412 U. S. and n. 9 768 nearly practicable” read Wesberry’s standard was as “as

But precise good-faith numerical require effort to achieve “a objections of equality.” 530-531. Over the 394 at U. (Fortas, concurring); id., at J., id., Justices, four see dissenting); (Harlan, joined by id., Stewart, J., J., rejected argument dissenting), Kirkpatrick J., (White, percentage vari- a fixed numerical that there is satisfy enough de minimis and to to be considered ance small Kirkpatrick’s, nearly practicable” rule standard. the “as supra, applied to in- Weiser, in White v. the Court was redistricting had maximum scheme which validate Texas’ variance 4.13%. require Kirkpatrick, Kirkpatrick Wesberry Just as did today. ineluctably Al to the decision does not lead Court’s nonarbitrary though “no stated that it could see Court way” point, pick deviation a de minimis the maximum eight large Kirkpatrick, more than times as small, while was Kirkpatrick posed here. the deviation Moreover, as that officiallyaccepted range argued was not to fall within enough, imprecision Interestingly statistical the census. redistricting plan approved Kirkpatrick after Missouri virtually the same deviation contained a deviation of 0.629%— Secretary in this case. Preisler v. declared unconstitutional (WD Mo.), Supp. 341 F. sum Missouri, 1158, 1162 State marily Preisler, aff’d nom. sub U. S. Danforth (1972).2 today Accordingly, do not view Court’s decision by Kirkpatrick Apparently nei as foreordained and Weiser. staying ther did Justice the District who, Brennan Court’s wrote: order, appeal appear present important

“The question would thus adop- Kirkpatrick requires v. Preisler

whether plan precise tion that achieves the most math- 2 upheld plans District Courts have with similar deviations. selected (court (ED 1982) g., White, e. Doulin See, Supp. 535 F. 451 Ark. implementation plan despite ordered with 0.78% deviation alternative 0.13%). plan with deviation of

769 Kirkpatrick exactitude, ematical or whether left some Jersey Legislature recognize latitude for the New to by considerations taken into account it as a basis for choosing among plans, arguably several each with ‘sta- tistically insignificant’ variances from the constitutional (1982). precision.” ideal absolute 455U. S. 1303,1305 question There can be little but that the variances Jersey plan “statistically insignificant.” Although New are the Government strives to make the decennial census as ac- humanly possible, curate as the Census Bureau has never perfect intimated that the results are a count of the American population. The Bureau itself estimates inexactitude in taking figure 1970 census at 2.3%,3 which is con- siderably larger than the 0.6984%maximum variance in the Jersey plan, New and which dwarfs the 0.2470% difference between the maximum deviations of the selected and the leading plan, suggested by alternative Professor Reock. undercounting Because the amount of from differs district to point district, there is no for a court of law to act under an unproved assumption tiny that such differences between re- districting plans population. actual reflect differences As expert Dr. Trussel, James an these and whose matters, testimony accept, purports ante, 735-736, Court explained: Jersey

“The distribution the undercount New is ob- viously also unknown, and see no reason believe that 3 Census, Guide, of the Population U. S. Bureau Users’ 1980 Census of (Mar. 1982). Housing Academy has esti The National Sciences mated that the national undercount in the census was 2.5%. Panel on Plans, Counting Appraisal An of Cen People Decennial Census in 1980: sus Plans error in the 1980 One estimate is that the undercount nationwide, likely App. census is be more people than million (Dr. Passel, Trussel), may Siegel, high million. J. J. & be as as 5 Census, Robinson, Coverage Population J. in the 1980 National Demographic Analysis Age, Sex, Preliminary by and Race: Estimates (Nov. 1981) (Record 31). Doc. No. over spread it would be all uniformly municipalities. reasons, For these one cannot make dis- congressional if of size one relies on census counts. truly equal tricts one Nor is it to rank meaningful redistricting plan are to another when differences district size superior enu- In districts whose opinion, small. my professional from than merated differ one another less populations by should,be to in size. equal one considered be To percent for this is an beyond point numerical push equality 103-104.4 exercise illusion.” App. Court, professing purposes argument after to for “[a]ssum[e] correct,” ante, 735-736, pro [Dr.

that each of statements is Trussel’s] paragraph denigrate guesswork following ceeds in the to his calculation as e., i. imprecision, undercounting margin because the of statistical precisely. quantify uncertainty persons, cannot be known The failure ques exactly pretending does not excuse that it does not exist. When the range tion whether error or 2.5% and the deviation at is is 1% 2% or larger 0.6984%, question practi than hand is no than more academic Moreover, margin of error required, cal. if fixed benchmark were officiallyrecognized by the Census estimated at 2.3%—could Bureau —last easily selected. precise also much of the fact that amount varia-

The Court makes certainty. undercounting among tion in districts cannot be known with however, point, variances The relevant is that these district-to-district impossible opting it whether make to determine with statistical confidence ameliorating ag- for with the smallest maximum deviation is addition, gravating equality population among actual the districts. upon per depends the count of individuals district the Census Bureau’s geographic by “Data from group selection of boundaries which to data. by compiled congressional equating census for districts the 1980 have been component summing all data geographic census areas to each district and geographic for census area areas coded the district. Where the smallest maps the area split boundary, was district the census majority were reviewed determine which district fell, and the entire area was coded to that district.” U. S. Bureau Cen- (1983) (preliminary sus, Congressional Congress A-l Districts of the 98th draft). effects, Thus, it is obvious completely undercounting aside from for con- figures numerical the census equality that even absolute between size. gressional equal districts of districts does reflect Even if the 0.6984% deviation here is not encompassed within the of the scope imprecision statistical it census, is miniscule when compared with other variations the among districts inherent census translating numbers into citizens’ votes. the First, census “is more of an event than a proc- Gaffney Cummings, ess.” (1973). 412 U. S. “It measures at population only instant in time. single Dis- trict are populations constantly often changing, at different Ibid. direction, rates either or down.” up As Court admits, “the well-known restlessness the American people means that counts for particular localities are out- completed.” Ante, dated long they before are 732.5 Sec- ond, far differences districts larger among are introduced because a substantial total too percentage population is

Finally, concept the Court dismisses the entire of statistical error with sophistic say if certainty “[e]ven comment that one cannot with larger merely higher one district is than another it because has a census count, say certainty one larger can that the district with a census likely larger count is more to be than the other district than it is to be Ante, or the degree certainty, smaller same size.” The at 738. however, speculative. is The relevant is consideration not whether Dis- Six, larger, and, is larger trict Four than District but how much much how larger plan. less under the selected vis-á-vis an alternative More- over, undercounting variable and differences between census units and dis- may having higher trict lines in other result districts maximum deviations. general point small, The that when the it no is numbers become so makes gradations sense to concentrate on ever finer when one cannot even be cer- doing tain whether so increases or decreases actual variances. Jersey, example, population growth during New en 1970’s 26%, larged by up some districts while other lost districts up population. Census, Congres to 8.7% of their 1970 U. S. Bureau of Gaffney Congress sional Districts of the also 98th Cummings, 412 U. n. 11. point.

Justice Stevens makes same rate, rate, mortality character of mod- “Given birth transient census, society, acknowledged ern and the in the we all know that errors may the date of such vanish between the date of the census and differences impossible to achieve.” population equality the next election. Absolute Ante, (concurring opinion). at 752 by alienage.6 young register disqualified Third, or is proportion figures cannot account for the of all those census eligible register.7 fail to individuals who otherwise eligible per in the number voters district for differences the minimal variations attributable these reasons overwhelm districting plan to the itself.8 plans

Accepting districting census, that the and the which perfect represents upon no it, back- are based cannot assuring equal repre- sliding in our fair and commitment agree Congress. views sentation in the election of with the Judge Court, dissented in the District Gibbons, who against “prohibition Kirkpatrick not be read as a should population have de minimis variances which toleration of representation.” statistically no relevant effect on relative plus-minus Supp., Daggett 535 F. at 984. A Kimmelman, surely category. this deviation of 0.6984% falls within unjustified today’s simply produced standard If decision an enough. import, practical it bad Unfor- with little would be tunately, I that “there are no fear that the Court’s insistence practicably population which could be variations, de minimis Art. avoided, I, but which nonetheless meet the standard of § justification,” further liti- ante, at invites 2, without plan every congressional redistricting gation virtually 6 years and over dif Jersey, example, In 18 old New significantly districts. In District 10 among fers 429,000. 282,000 individuals, U. while District had S. had but such (1979). Census, Metropolitan Area Data Book 549 Bureau State Gaffney Cummings, supra, n. 18. also Nation, voting age population 7 Throughout approximately 71% of the Metropolitan Census, Area registers to vote. Bureau of State and U. S. (1982). Data Book 567 factors, registered many As of all these as well as the failure a result inev ballots, weight of a vote in one district voters to cast citizen’s example, the total number of itably different from that in others. For significantly Jersey congressional races differed votes cast the 1982 New 186,879 in Dis districts, 92,852 10 to ranging from in District between Congressional Quarterly 391 trict 9. 41 completed the Nation. At least 12 States which have re- districting adopted plans on the basis of census have higher presented with a deviation than that and 4 oth- here, quite Jersey’s.9 ers have deviations similar to New Of course, under the Court’s rationale, even Rhode Island’s two districts have a deviation of 0.02% or about —whose subject people to constitutional attack. —would legislatures pressed cases, all such state will be hard justify preference plan. good-faith their for the A selected population equality enough effort to achieve is not if the population variances are not “unavoidable.” The court must consider whether the differences could have been altogether.” Ante, further “reduced or eliminated at 730. computers, generally theWith assistance of there will be a plan with an even more minimal deviation from the math- Then, ematical “the ideal. State must bear the burden of proving significant that each variance between districts was necessary legitimate goal.” Ante, to achieve some 731. literally any As this illustrates, case variance dis- between “significant.”10 tricts will be considered The State’s burden easily justifying “the will not be met: State bears the burden of 9 (2.45%); larger (2.96%); States with deviations are Indiana Alabama (2.40%); (2.00%); (1.81%); Georgia Virginia Tennessee North Carolina (1.64%); (1.39%); (1.30%); (1.76%); Kentucky Washington York New (0.78%). (1.09%); (0.87%); New Mexico States Massachusetts Arkansas (0.68%); (0.60%); with maximum are Okla similar deviations Ohio Nevada (0.49%). (0.58%); Virginia Governments & homa West Council of State Legislatures, Reapportionment of Informa National Conference State 1982). (Nov. 12, Update tion only vari language suggests 10 The that not must the maximum Court’s every from plan supported, ance in a but that also deviation absolute be justified. Ante, staggering na equality be at 740. must so Consider any two imposed: of between ture the burden Each difference justified, apparently plans if none of the in a must be even districts State have reduced the difference. legislature before the or commission would n. infra. Ante, 7B9. When particularity.” differences with its the result burden, generally fails to sustain will State an alternative The choice will plan. that a court must select g., e. very election, see, until the eve of an often be disputed curiam), Upham (per Seamon, (1982) U. S. in a and voters state confiision. candidates leaving or commission can only way legislature bipartisan will be to dismiss all other hope litigation legitimate to avoid for the automatically plan with opt districting concerns Yet seriously deviation.11 no one can contend the smallest insistence mathematical exact- upon that such an inflexible “fair and representa- ness will serve effective promote of Kirk- today’s likely tion.” The more result extension patrick Justice fulfilling is to move closer to Fortas’ prophecy “a have the boundaries legislature might ignore district line down sense, running common of the corridor an house or even apartment the middle of a house two single-family the residents between dividing 538. 394 U. Such sterile and mecha- districts.” one man, nistic “one only brings principle application vote” into disrepute.

II the Court had reasons to force might strong One expect offers this task the States. Yet the Court Sisyphean upon deviation, a by choosing legislature Even with the smallest avoiding challenge. In this commission cannot be assured constitutional find was avoidable be case the does not 0.6984% deviation Court Jersey plans Legislature there were other before the New cause appellants’ smaller maximum variations. Nor does Court counter by record, plans position, supported evidence in the that these alternative Instead, hand at disqualifying other faults. the Court tries its own had Jersey redistricting by moving around 13 New New and concludes that Jersey subdivisions, the maximum deviation could be reduced 0.449%. Ante, 739-740, is it message legislatures clear: is at n. 10. The for state enough plan any plans introduced superior that the chosen actual alternatives, any must also be better than conceivable alterna as' judge tive a federal can devise. positive pre-

no virtues that will from its follow decision. No path Reynolds tense is made that this case in follows Wesberry insuring representa- and in the “fair and effective expended tion” of citizens. No effort is I, to show that Art. § requirement Congressmen “by peo- 2’s that be elected ple,” Wesberry (1964), Sanders, v. U. S. demands the Any invalidation of deviations this level. such requirement, if exist, absolute it did would be irreconcilable recognition justifications popu- with the Court’s certain for express lation variances. at 740. no ante, Given con- holding, showing stitutional basis for the Court’s and no that objectives by representation compromised of fair are disparities, up- these minimal the normal course would be to legislature fulfilling hold the actions of the its constitution- prescribe ally delegated responsibility to the manner of hold- § ing Representatives. elections for Art. 4. I, Senators Doing keeping oft-expressed so would be with the Court’s recognition apportionment primarily leg- that is a matter for judgment. Upham supra, Seamon, 41; islative v. at White Reynolds 795; Sims, S., U. v. 377 U. Weiser, legislature “[A] at 586. state is the institution that is identify then far the situated to reconcile traditional best constitutionally policies frame- state within mandated Finch, work .. ..” Connor U. S. purely support of its

Instead defensive the.Court adopt any fixed numerical decision. The Court refiises inter- the federal courts would not standard, below which equality asserting principle population “[t]he vene, socially unjust proved or districts has not economically experience.” Ante, at 733. Of harmful unjust; equality principle is not course, application principle rub. of this is the the unreasonable applied Leaving principle has been aside that the never including many, vengeance today, there are witnessed self-congratulatory myself, with the Court’s who take issue assumption Kirkpatrick First, a a success. has been Kirkpatrick experience “the decade with has shown that compatible equality perfectly ‘gerry- is with rule of absolute mandering’ sort.” Rockefeller, of the worst Wells (Harlan, dissenting). S., U. at 551 With ever more so- J., legislators plans phisticated computers, can draw countless having population equality, for absolute political but each its own Although neither a rule of absolute ramifications. equality equality prevent nor of substantial can alone de- one legisla- partisan gerrymandering, the former offers liberate ready geographical justification disregarding and tors a political I in

boundaries. remain convinced what said Kirkpatrick “[Those] . . and decisions . dissent Wells: greater downgrade potential on far threat to a restraint representation, gerrymander. Legislatures equality of minimizing representation on of selected intent ignore political groups and or racial are invited to boundaries long they population compact equal- as districts so adhere to ity among using they districts standards which we know and quite know are sometimes incorrect.” 349 at 555. U. pre- There is now evidence that Harlan was correct to Justice past, “[e]ven are dict that more than district lines likely political advantage to maximize the be drawn public party temporarily Id., dominant affairs.” at 552.12 deviations, political gerrymandering 12 Unlike does not lend Nevertheless, proof. reviewing re itself to arithmetic after the recent following districting throughout country, commentator offered the one assessment: ‘one-man, increasing nobly principle coming

“The aimed one-vote’ is into gerrymandering. weapon legislators partisan use as a for state bent on between, Jersey points Republicans From California to New highly remaps by demonstrating justifying partisan Democrats alike are respect Supreme congres- for the Court mandate equal possible. sional as Mean- districts within states must be made as while, preservation redistricting other interests at stake in —such community con- grouping boundaries and the of constituencies similar one-man, being emphasis cerns —are aside on one-vote brushed .... many only permits encourages states it gerrymandering, it it. for addition to a of the providing patina respectability equipopulous gerrymander, Kirkpatrick’s assured regime judiciary legislative extensive intrusion of the into business. impossible approach population equality crossing city, county to without township legislature recognizes move must and lines. Once made, only step drawing it is a short further to the of a line that dances jaggedly through every region interests, of state. Local informed that whole-county longer legally permissible it is no to draw a states, likely map object they in most are far less to than in the were past.... reject tiny The court’s decision to deviation in favor of an may encourage hairsplitting even smaller one further game numbers partisan country.” given gerrymanders that has rise to all over the Con- Inc., gressional Quarterly, Redistricting State Politics and Engstrom, Supreme Gerrymander- Equipopulous See also Court and ing: Remaining Quest Representa- A Fair Obstacle and Effective (“Not tion, only 1976 Ariz. L. J. State has the Court failed to de- velop practice gerrymandering, checks of pursuing effective on the but in goal point satiety actually of to a it equality of has facilitated Baker, Man, Vote, practice”); Fairness,” One One and “Political (1974) (hereafter Baker) Emory (“Priority L. typically J. was given expense any recognition to miniscule variations of at the Charges partisan gerrymandering of subdivisions. of were more widespread past major than decades for two the extent reasons: redistricting activity among fifty states, emphasis all and the lack on lines”). compactness boundary former norms of and adherence to local eyes commentators, experience

In the of some of New York in the aftermath Wells v. is instructive. Rockefeller “Subsequent congressional districting in possible New York became a prototype for the ‘equal-population gerrymander.’ Whereas the former pattern by district Supreme nullified Court had been the result of bipartisan compromise house, major party controlling by with each one Republicans legislative 1970 the governor- held both houses as well as the (and ship. The majority assistant counsel to the senate leader chief co- redistricting) Supreme ordinator of candidly ‘The is remarked: Court just making gerrymandering only easier than it used to Not was New be.’ City subjected major upstate York cartographic surgery, to but cities were fragmented, portions joined also being with and suburban rural areas attempt an Baker, dilute concentrations Democrats.” at 712-713. Yet, plan, under the new no district per- deviated more than than 490 average, configuration sons from the and the of district boundaries re- generally vealed compact contiguous Baker, Gerrymander- contours. dealing funda [re]apportionment as it must with “[T]he task, representation,' Burns the nature about mental ‘choices leg primarily political is S., 384 U. Richardson, Cummings, Gaffney 749. process.” 412 U. islative legislative reappor Gaffney respect said What we apropos here: tionment [is not] goal representation

“[T]he fair and effective reapportionment by making the standards furthered satisfy reapportionment re- the task is to that so difficult curringly performed legislative from hands and removed politi- by make must the federal courts which themselves plan accept necessary formulate a or cal decisions plaintiffs may by reapportionment who have made those wholly goals in the from those embodied official different plan.” Ibid. experience Kirkpatrick demon-

More than a decade’s with only precise equality numerical that insistence on strates political refight their lost in the arena to invites those who Consequently, “[m]ost are in federal court. estimates battles percent percent 25 and 35 of current house dis- that between Bar Asso- trict lines were drawn the Courts.” American (1981). Redistricting Congressional As I ciation, have already by extending Kirkpatrick noted, to deviations below redistricting plan every in level, even the 1% the State single Representative more than is vulnerable rendered by anyone complaint to after-the-fact attack with a and a calculator. ultimately refuge seeks in stare decisis. do Court slight respect g., due, see, that e. doctrine White Privileged

ing: Sanctuary Target?, Reapportionment Next in in Judicial 1971). (N. Wells, 1970s, p. Polsby Ironically, plain- ed. David successfully challenged tiff pattern, who the former district returned to February federal court 1970 to ask if the old could be restored. Dixon, Man, Next?,” Happens “One One Vote —What 60 Nat. Civic Rev. (1973), it Weiser, U. S. 783 but is not at least ironic to protect Kirkpatrick stare find decisis invoked as the Court very proceeds holdings overrule itself other that deci- Kirkpatrick, squarely rejected argu- sion? In Court slight proper ment in district variances size were fragmenting political to avoid order subdivisions: legally acceptable argument “[W]e do not find justified necessarily they are if variances result from a attempt fragmenting political State’s to avoid subdi- by drawing congressional along visions district lines existing county, municipal, or other subdivision boundaries.” 394 U. at 533-534.13 pages rejected equally uncategori- later, Several Court may justified cal terms the idea that variances in order to compact. Id., make districts more preference at 535-536. “A State’s pleasingly shaped districts,” Court con- hardly justify population Id., cluded, “can variances.” Kirkpatrick 536. In Justice Fortas’ “re- words, the Court ject^], every type justification seriatim, of that has been— possibly, every one that Id., could be—advanced.” at 537. today contrary

Yet the Court—with no mention the of “Any holdings Kirkpatrick opines: in number of consist- — ently applied legislative policies might justify some variance, including making compact, respecting instance, for districts municipal preserving prior boundaries, the districts, cores avoiding Representatives.” and contests between incumbent Howell, (1973) 13 Seealso Mahan (Brennan, J., 410 U. S. (“What concurring part dissenting part) in and our decisions have made clear is pertinent legislative that certain state interests that are re to apportionment possible districting. can have no relevance to Thus, preserve the integrity political political need to subdivisions may, justify instances, popula subdivisions in some small variations legislators tion of districts from which state are elected. But interest hardly justification malapportioned congressional can asserted in be dis Preisler, Kirkpatrick supra”). tricts. overruling of the Court’s I, course, at 740. welcome Ante, holdings Kirkpatrick. There should these ill-considered legislatures may question account but that state be no preserve geographic tra- political in order to boundaries compact contiguous subdivisions and achieve ditional recognizes that courts should districts. Justice Stevens importance “give greater weight inter- to the State’s consistency which those interests are with ests and Ante, of the deviations.” than to size served majority appears ready apply n. of the Court Thus, 26. a strong “with measure of deference to this new standard legitimate Post, concerns of the n. State.” dissenting). (Powell, J., legislatures room order that have to accommodate these range popula-

legitimate of minimis factors, noncensus de legislative permitted reappor- like deviation, tion required. The insistence that cases, tionment Court’s justified speci- every small, no matter how be deviation, ficity legislatures considering “legiti- discourages from these making plans, justification mate” factors in their lest the wanting, judicially invalidated, found and a drawn put place. requirement Moreover, substitute its precise equality mathematical invite continues to those who bury political opposition employ equipopulous would their gerrymanders. range preclude A de minimis would not gerrymanders such but least would at force the car- tographer justify his work on its own terms.

h-Hr-H dealing legislative apportionment Our cases with state approach. recognized have taken a more sensible We have ordinarily that certain small in themselves, deviations do not, prima Gaffney constitute a facie constitutional violation. v. Cummings, (1973); Regester, 412 U. S. 735 v. White upheld plans U. S. Moreover, we have with rea- necessary political sonable variances were to account for (1973), pre- subdivisions, Howell, Mahan v. 410 U. S. strength minority voting groups, serve the and to insure political Gaffney Cummings, supra. fairness, What we Gaffney Cummings legislative apportionment in held for is fully applicable congressional redistricting:

“ achieving ‘[T]he representation of fair and effective worthy goal, surely all is’... and citizens vital but its any way depend attainment does not commonsense upon eliminating insignificant population variations representation involved in this case. Fair and effective may destroyed by gross among variations apparent representation districts, but it that such solely depend equality among does on mathematical populations overemphasis district .... An unrealistic figures, on raw a mere nose count may submerge districts, these other and considerations ready ignoring day- itself furnish a tool for factors that to-day operation important acceptable repre- are to an apportionment arrangement.” sentation and 412 U.S., at 748-749.

Bringing together legislative congressional our state imply overlooking cases does not relevant differences be- normally larger leg- tween the two. States draw a number of accordingly greater margin require islative districts, which geographical “[Con- to account for boundaries. gressional freighted districts are not so intertwined and strictly legislative local interests as are state districts.” White v. Weiser, Furthermore, U. at 793. because generally larger are districts much than state legislative percentage point repre- districts, each of variation commensurately greater people. sents a number But degree. They suggest are these differences that the level challenges districting at which courts should entertain plans, absent unusual circumstances, should be lower 782 Al nonexistent.14 cases, altogether but not

congressional in of light figure, I not wedded to precise am though deviations, a 5% cutoff appears of current range absent judicial challenges, I not entertain would reasonable. the maximum deviation circumstances, where extraordinary if deviations, rationally greater 5%. Somewhat is less than also be interest, may permissib an state related to important dis of maintaining compact, contiguous Certainly, le.15 subdivisions, and efforts of tricts, respécting Gaffney Cummings, supra, fairness, e. v. g., assure political constitute such interests. a model reflection Jersey’s New up

I would not hold here Nevertheless, the deviation involved interests. of such minimis, and, regardless is de of other infirmities the what 14 I, developed, congressional are in Art the law has our cases rooted As rely legislative upon Equal Pro § of the while our cases Constitution however, aware, of the Fourteenth Amendment. am not tection Clause justifies, requires, provisions which let alone anything respective in the of emerged in that has between the two lines deci difference treatment cross-cited, early frequently were and the formulation sions. Our cases Reynolds appears in v. nearly equal practicable” “as as is Sims, Wesberry Sanders, S., S., 589, as in v. 376 at 377 U. well as U. Kirkpatrick paths re differing 7-8. The the cases have taken since must rejection legitimacy considering nonpopu- that of the sult from decision’s Howell, See Mahan congressional redistricting. in lation factors (Brennan, J., concurring part dissenting part). in in U. and holding Kirkpatrick, any today’s long-awaited overruling in With remaining justification disappears ap a marked difference in such our legislative reapportionment. proach to 15Experience legislative following apportionment in the field our allow de minimis range convincing proof ance of a variance is that we need not goal equal population fear that the districts will receive less than de minimis prediction tolerating popu its due. Justice Brennan’s “jeopardize very gains” lation variances would substantial made (1973) Regester, White equalizing legislative districts, 412 U. S. part dissenting and, (concurring part), proved, has not in been deed, prediction analysis legislative redistricting is refuted an undertaken after the 1980 census. See Council of State Governments & Legislatures, Reapportionment National Conference of State Informa (Nov. 1982). Update tion plan may or there no have, otherwise, constitutional §2 viola- It I, tion Art. sole issue before us. would, —the appellees a if course, be different matter could demonstrate Jersey’s plan invidiously against that New discriminated political group. Regester, supra; racial See White Gaffney Cummings, supra, 751-754; at Whitcomb v. (1971); Lightfoot, Chavis, 403 U. Gomillion v. S. U. S. 339

IV dispo- Even if the Court’s view of the law correct, were its minimum, sition of case is not. At a the Court should of the vacate decision District Court and remand for fur- previously indicated, ther consideration. As nally recognizes today fi- Court respecting that considerations such as avoiding and subdivisions contests between incum- Representatives might justify bent small vari- “any Indeed, ances. the Court indicates that number of con- sistently applied legislative policies” might Ante, do so. suggest There 740. is evidence the record to that the Jersey Legislature New was concerned with such consider- “many problems ations.16 The Court itself notes: Jersey Legislature drawing the New encountered districts equal population from the . . . stemmed decision not to any municipalities divide between two dis- n. Ante, tricts.” 5. But even if there were no evi- given record, dence in the the State should be a chance to de- Surely, rely fend its this on basis. on Court cannot appellants only justification the fact that have advanced one plan’s population preserving voting for the deviations— strength minority groups. Relying Kirkpatrick of racial on supra, appellants and v. Weiser, White no doubt concluded justifications that other were and foreclosed that the intro- proof duction of such would be futile. 16See, g., e. (Record 39) (concern Deposition, Feldman at 91-94 Doc. No. (Record incumbents);

with fairness to Deposition, Jackman at 91-92 Doc. 40) (concern subdivisions). No. with preserving political Powell, dissenting. Justice dissenting opinion, White’s excellent Justice join that “the Constitu- doubt expressed my previously

reaffirm two centuries nearly charter after living tion —a vital key of its flexibility provisions of the wise because —could exactitude legisla- of mathematical rule read to require Weiser, 412 U. S. v. White tive reapportionment.” (1973) I write separately express (concurring opinion). relation and its on thoughts gerrymandering some additional were thought factors that presumably to apportionment (1969). Preisler, 526 394 U. S. under Kirkpatrick relevant r-H

The invalidates New Kirkpatrick, Court, following today because various alternative Jersey’s redistricting plan solely Reock, the one Professor had plans, proposed by principally smaller de- “appreciably what the Court views Ante, the largest viations between and smallest districts.” all of the the maximum vari- plans, at 728. Under under 1%. these differences as neither ances were view nor As JUSTICE constitutionally significant. “appreciable” ante, at 769-772 (dissenting opinion), demonstrates, White insistence on mathematical is self- precise equality Court’s the inherent inaccuracies census data deluding, given and the other difficulties in measuring voting population See Kirk- years. a district that will exist for 10 period (Fortas, at 538 patrick, supra, J., pre- concurring) (pursuit cise “is a search for a will-o’-the-wisp”). Moreover, equality for it has become clear that leaves no room proper Kirkpatrick factors, such legislative consideration of other as preserva- boundaries, tion of political and that are geographic plainly Gaffney decisions,1 relevant to rational see reapportionment “[a]ny consistently legislative The Court applied holds that number of policies instance, might justify variance, including, making some dis compact, boundaries, respecting municipal preserving tricts the cores of Cummings, Mahan v. Howell, 412 U. (1973); S. U. S. As Justice White correctly observes, ante, at 775-776, decade of experience has con- Kirkpatrick firmed the fears of the dissenters that an uncom- on promising emphasis numerical equality would serve to encourage legitimate even most outrageously parti- san see 394 gerrymandering, S.,U. at 551-552 (Harlan, J., id., at 555 dissenting); (WHITE, J., dissenting). plain fact is the computer this age, type political and discriminatory can be gerrymandering accomplished entirely with districts of consistently equal population.2 districts, prior avoiding Represent- contests between incumbent Ante, Although atives.” at 740. it is remarkable that the Court thus Kirkpatrick silently important simultaneously discards features of while ante, stare decisis to defend the remainder of that invoking decision, see (White, J., dissenting), change I welcome this in the law. It is to hoped strong new applied this standard will be with a measure ante, legitimate deference to the concerns of the State. See n. 26 (Stevens, J., concurring) (recognizing “give greater that courts should weight importance consistency to the of the State’s interests and the with deviations”). which those interests are than to served the size of the 2An the congressional redistricting illustration is recent in Illinois. Legislature After the Illinois reapportionment plan, had failed to enact a a among plans three-judge varying District Court chose four from 0.02851% Following Kirkpatrick, to 0.14797%in maximum majority the deviation. plan deviation, of the court chose the with the smallest one that was a plan” designed voting strength “Democratic to maximize Democratic Congressional In re Illinois Districts expense Republicans. of (ND Cases, Reapportionment 1981), summarily No. 81-C-3915 Ill. aff’d (1982). Ryan Otto, sub nom. v. 454 1130 A commentator noted: U. S. victory part sophisticated computer “The Democratic to a was due exactly program possible having that made the creation of districts almost equal population. only people populous The most district has 171 more populous accuracy court, impress than the least one. That seemed expressed which no concern the new divided cities and district lines up Surprise Map Unpleasant carved counties all over the Illinois state.” GOP, Congressional Quarterly for (Colo. Lamm, See also Carstens v. 68, 73-74, Supp. 543 F. and n. 1982) major redistricting plans, five (three-judge District Court reviewed I per- therefore continue to believe that Constitution of recognition mits variations from “theoretical 'exactitude’ Kirkpatrick rule as well as impracticality applying Weiser, state interests.” White deference legitimate supra, when a J., concurring). Certainly at 798 (Powell, with an popula- a districting plan average State has adopted 0.1384%, and a maximum deviation of tion deviation the Constitution’s mandate it has with 0.6984%, complied districts. equally among apportioned r-H of the New map Jersey congressional

The extraordinary ante, following p. me comment districts, prompts see on the deliberate question gerrymandering separate —“the of district boundaries arbitrary populations distortion Kirkpatrick, partisan personal purposes,” supra, in full am (Fortas, J., concurring). agreement at 538 observation more than a decade ago with Justice White’s far threat greater potential “a presents that gerrymandering achieve than State’s failure to to equality representation” plan larg- legislature’s a difference between including Republican with e., persons, i. a maximum deviation est and smallest districts of seven difference, 0.0015%, plan 15-person Governor’s with a and the Democratic Brier, e., Supp. i. O’Sullivan 0.0031%); 540 F. a maximum deviation (Kan. 1982) (three-judge to choose be- District Court asked Republi- plan a 0.11% maximum deviation and a tween Democratic deviation). can with a 0.09% maximum side effect of Kirk- illustrate an additional unfortunate These cases also patrick: tendency legislators increasing and Governors— of state any redistricting plan is to after-the- have that “vulnerable who learned ante, calculator,” 778 by complaint a and a at anyone fact attack drawing simply J., dissenting) spurn compromise favor of (White, —to appears consistent with the up partisan plan the most that merely review the equality longer do district courts criterion. No federal Rather, many constitutionality redistricting plan. cases of a State’s redistricting plan in first they placed position choosing a are instance.

787 “precise admittedly figures.” adherence to inexact census (1969) (dissenting v. Rockefeller, Wells opinion). 394 U. S. injuries I also believe that the that from result may gerrymandering rise to constitutional dimensions. As observes, if a State’s electoral rules “serve Justice Stevens purpose segment no other than to favor one racial, —whether religious, may occupy ethnic, or economic, a —that position strength particular point at a in time, or to disad- vantage politically segment community, they weak guarantee equal protection.” violate the constitutional (concurring opinion). gerry- Ante, at Moreover, most mandering produces any regard politi- districts “without boundary Reyn- cal subdivision or lines,” natural historical (1964), Sims, olds v. 377 U. S. is result that profoundly apportionment goal destructive of “fair and representation,” legislator id., effective at A 565. cannot represent properly his constituents can from a voters —nor fragmented intelligently district exercise the ballot a—when voting nothing district more than an artificial unit divorced from, and indeed often in with, conflict the various com- map munities established in the State.3 The attached to opinion the Court’s this better illustrates far than words can describe. prepared

I therefore am to entertain chal- constitutional lenges partisan gerrymandering level that reaches the discrimination described ante, Justice Stevens. (concurring opinion). suggest shape I do not of a Lamm, supra, Carstens three-judge District Court noted city identity”: preserving voting an entire one “voter district facilitated city county they in, “Most know live likely voters what but fewer are they split to know what district live in if the districts coun congressional district, ties and cities. If a voter knows he is more his presumably likely representative to know who his is. would This lead voting.” likely Supp., more informed F. n. 78. It also is to lead Representative to a who of his district and is more knows needs responsive to them.

districting map invariably irregu- dispositive. itself is Some larity degree shape irregularity inevitable, political depending primarily geographic on and bound- its State, aries within as well as distribution of population. political parti- Moreover, considerations, even system. A ones, court, san are inherent in a democratic “attemp[t] impossible task extir- therefore, should not essentially proc- pating politics are the from what Gaffney, sovereign 754. esses of the Finally, States.” U. at legislative suggest reapportionment I that a do plan plan might is invalid whenever an alternative be viewed partisan apportionment or more in accord with various as less legislature necessarily must have discre- criteria. state competing tion to accommodate considerations. that the believe, however, do constitutional mandate of Reynolds, representation,” supra,

“fair at and effective proscribes apportionment plans purpose that have the and substantially disenfranchising groups identifiable effect Generally, presumptive existence of such uncon- voters. districting stitutional discrimination will indicated appear their face lit- the boundaries of which on to bear relationship any legitimate purpose. As tle or no state “dramatically shapes may irregular states, Justice Stevens explanation,” probative to call for an have sufficient force (concurring opinion); departures from ante, at “drastic something may compactness signal be amiss,” are a po- “[e]xtensive ante, from established deviation 758; prima possible basis for a facie litical boundaries is another showing gerrymandering,” circumstances, a ibid. In such legitimate required provide nondis- should be State districting criminatory explanation lines it has drawn. (the apportionment “presented Reynolds, supra, crazy lacking quilts, completely in rational- little more than alone”). ity, and could be found invalid on that basis rationally that the New believe case, In this one cannot Jersey Legislature other than the most considered factors

789 partisan political goals population equality. hardly It suggested, example, could be for that the contorted Districts any attempt 3, 5, and 7 reflect to follownatural, or historical, local boundaries.4 Nor do these district lines reflect any likely quality repre- consideration of the effect on the of they sentation when the boundaries are so artificial that are likely Congressmen Judge to confound the themselves. As eloquently stated Gibbons his dissent below: apportionment map produced by “The L. 1 P. c. Jersey, me, leaves aas citizen of New It disturbed. cre- anything compact, ates several districts which are but contiguous only and at least one district which is for yachtsmen. municipal While boundaries have been maintained, there has been little effort to create districts having community districts, interests. some for example, different stations, television and radio differ- newspapers, transportation systems ent and different serve the northern and southern localities. Moreover harshly partisan Speaker Christopher tone of Jack- disedifying, say Reock, man’s letter to Ernest C. Jr. is plain, partisanship pro- It well, least. is bulges appendages duced artificial of two so districts place Congressmen toas the residences of Smith and they running against Courter districts would be where Daggett Supp. incumbents.” 535 F. Kimmelman, (NJ 1982). 978, 984 summary by Judge Gibbons,

This statement resident Jersey, powerful persuasive support a con- New for (the may noted, adopted by Jersey example, plan It for that the New Plan) fragments. plan Feldman 21 counties divided State’s into 55 Hardwick, by by proposed Reock, Assemblyman Professor cre introduced county fragments, existing congressional ated 45 and the districts divided fragments. App. (Appendix the counties into 42 A to Affidavit Alito, Legislative A. Executive Samuel Director Office Services Jersey Legislature). the New Legislature’s Jersey redistricting

elusion that New gerrymander. is an Cf. ante, unconstitutional n. concurring). precise J., Because this issue was (Stevens, it Court, however, not addressed the District need not be population equality, reached here. As to the issue dis forth sent the reasons set above and Justice White’s dissenting opinion.

Case Details

Case Name: Karcher v. Daggett

Court Name: Supreme Court of the United States

Date Published: Jun 22, 1983

Citation: 462 U.S. 725

Docket Number: 81-2057

Court Abbreviation: SCOTUS