delivered the opinion of the Court.
In Wesberry v. Sanders, 376 U. S. 1 (1964), we held that “[w]hile it may not be possible [for the States] to draw congressional districts with mathematical precision,” id., at 18, Art. I, § 2, of the Constitution requires that “as nearly as is practicable one man’s vote in a con*528gressional election is to be worth as much as another’s.” Id., at 7-8. We are required in these cases to elucidate the “as nearly as practicable” standard.
The Missouri congressional redistricting statute challenged in these cases resulted from that State’s second attempt at congressional redistricting since Wesberry was decided. In 1965, a three-judge District Court for the Western District of Missouri declared that the Missouri congressional districting Act then in effect was unconstitutional under Wesberry but withheld any judicial relief “until the Legislature of the State of Missouri has once more had an opportunity to deal with the problem . . . .” Preisler v. Secretary of State of Missouri, 238 F. Supp. 187, 191. Thereafter, the General Assembly of Missouri enacted a redistricting statute, but this statute too was declared unconstitutional. The District Court, however, retained jurisdiction to review any further plan that might be enacted. Preisler v. Secretary of State of Missouri, 257 F. Supp. 953 (1966), aff’d, sub nom. Kirkpatrick v. Preisler, 385 U. S. 450 (1967). In 1967, the General Assembly enacted the statute under attack here, Mo. Rev. Stat., c. 128 (Cum. Supp. 1967), and the Attorney General of Missouri moved in the District Court for a declaration sustaining the Act and an order dismissing the case.
Based on the best population data available to the legislature in 1967, the 1960 United States census figures, absolute population equality among Missouri’s 10 congressional districts would mean a population of 431,981 in each district. The districts created by the 1967 Act, however, varied from this ideal within a range of 12,260 below it to 13,542 above it. The difference between the least and most populous districts was thus 25,802. In percentage terms, the most populous district was 3.13% *529above the mathematical ideal, and the least populous was 2.84% below.1
The District Court found that the General Assembly had not in fact relied on the census figures but instead had based its plan on less accurate data. In addition, the District Court found that the General Assembly had rejected a redistricting plan submitted to it which provided for districts with smaller population variances among them. Finally, the District Court found that the simple device of switching some counties from one district to another would have produced a plan with markedly reduced variances among districts. Based on these findings, the District Court, one judge dissenting, held that the 1967 Act did not meet the constitutional standard of equal representation for equal numbers of people “as nearly as practicable,” and that the State had failed to make any acceptable justification for the variances. 279 F. Supp. 952 (1967). We noted *530probable jurisdiction but stayed the District Court’s judgment pending appeal and expressly authorized the State “to conduct 1968 congressional elections under and pursuant to [the] 1967 . . . Act . . . 390 U. S. 939 (1968). We affirm.
Missouri’s primary argument is that the population variances among the districts created by the 1967 Act are so small that they should be considered de minimis and for that reason to satisfy the “as nearly as practicable” limitation and not to require independent justification. Alternatively, Missouri argues that justification for the variances was established in the evidence: it is contended that the General Assembly provided for variances out of legitimate regard for such factors as the representation of distinct interest groups, the integrity of county fines, the compactness of districts, the population trends within the State, the high proportion of military personnel, college students, and other nonvoters in some districts, and the political realities of “legislative interplay.”
I.
We reject Missouri’s argument that there is a fixed numerical or percentage population variance small enough to be considered de minimis and to satisfy without question the “as nearly as practicable” standard. The whole thrust of the “as nearly as practicable” approach is inconsistent with adoption of fixed numerical standards which excuse population variances without regard to the circumstances of each particular case. The extent to which equality may practicably be achieved may differ from State to State and from district to district. Since “equal representation for equal numbers of people [is] the fundamental goal for the House of Representatives,” Wesberry v. Sanders, supra, at 18, the “as nearly as practicable” standard requires that the State make a good-faith effort to achieve precise mathematical *531equality. See Reynolds v. Sims, 377 U. S. 533, 577 (1964). Unless population variances among congressional districts are shown to have resulted despite such effort, the State must justify each variance, no matter how small.
There are other reasons for rejecting the de minimis approach. We can see no nonarbitrary way to pick a cutoff point at which population variances suddenly become de minimis. Moreover, to consider a certain range of variances de minimis would encourage legislators to strive for that range rather than for equality as nearly as practicable. The District Court found, for example, that at least one leading Missouri legislator deemed it proper to attempt to achieve a 2% level of variance rather than to seek population equality.
Equal representation for equal numbers of people is a principle designed to prevent debasement of voting power and diminution of access to elected representatives. Toleration of even small deviations detracts from these purposes. Therefore, the command of Art. I, § 2, that States create congressional districts which provide equal representation for equal numbers of people permits only the limited population variances which are unavoidable despite a good-faith effort to achieve absolute equality, or for which justification is shown.
Clearly, the population variances among the Missouri congressional districts were not unavoidable. Indeed, it is not seriously contended that the Missouri Legislature came as close to equality as it might have come. The District Court found that, to the contrary, in the two reapportionment efforts of the Missouri Legislature since Wesberry “the leadership of both political parties in the Senate and the House were given nothing better to work with than a makeshift bill produced by what has been candidly recognized to be no more than ... an expedient political compromise.” 279 F. Supp., at 966. Legisla*532tive proponents of the 1967 Act frankly conceded at the District Court hearing that resort to the simple device of transferring entire political subdivisions of known population between contiguous districts would have produced districts much closer to numerical equality. The District Court found, moreover, that the Missouri Legislature relied on inaccurate data in constructing the districts, and that it rejected without consideration a plan which would have markedly reduced population variances among the districts. Finally, it is simply inconceivable that population disparities of the magnitude found in the Missouri plan were unavoidable.2 The New York apportionment plan of regions divided into districts of almost absolute population equality described in Wells v. Rockefeller, post, at 545-546, provides striking evidence that a state legislature which tries can achieve almost complete numerical equality among all the State’s districts. In sum, “it seems quite obvious that the State could have come much closer to providing districts of equal population than it did.” Swann v. Adams, 385 U. S. 440, 445 (1967).
We therefore turn to the question whether the record establishes any legally acceptable justification for the population variances. It was the burden of the State “to present . . . acceptable reasons for the variations among the populations of the various . . . districts . . . .” Swann v. Adams, supra, at 443-444.
*533II.
We agree with the District Court that Missouri has not satisfactorily justified the population variances among the districts.
Missouri contends that variances were necessary to avoid fragmenting areas with distinct economic and social interests and thereby diluting the effective representation of those interests in Congress. But to accept population variances, large or small, in order to create districts with specific interest orientations is antithetical to the basic premise of the constitutional command to provide equal representation for equal numbers of people. “[N] either history alone, nor economic or other sorts of group interests, are permissible factors in attempting to justify disparities from population-based representation. Citizens, not history or economic interests, cast votes.” Reynolds v. Sims, supra, at 579-580. See also Davis v. Mann, 377 U. S. 678, 692 (1964).
We also reject Missouri’s argument that “[t]he reasonableness of the population differences in the congressional districts under review must ... be viewed in the context of legislative interplay. The legislative leaders all testified that the act in question was in their opinion a reasonable legislative compromise. ... It must be remembered . . . that practical political problems are inherent in the enactment of congressional reapportionment legislation.”3 We agree with the District Court that “the rule is one of ‘practicability’ rather than political ‘practicality.’ ” 279 F. Supp., at 989. Problems created by partisan politics cannot justify an apportionment which does not otherwise pass constitutional muster.
Similarly, we do not find legally acceptable the argument that variances are justified if they necessarily result from a State’s attempt to avoid fragmenting political *534subdivisions by drawing congressional district lines along existing county, municipal, or other political subdivision boundaries. The State’s interest in constructing congressional districts in this manner, it is suggested, is to minimize the opportunities for partisan gerrymandering. But an argument that deviations from equality are justified in order to inhibit legislators from engaging in partisan gerrymandering4 is no more than a variant of the argument, already rejected, that considerations of practical politics can justify population disparities.
Missouri further contends that certain population variances resulted from the legislature’s taking account of the fact that the percentage of eligible voters among the total population differed significantly from district to district — some districts contained disproportionately large numbers of military personnel stationed at bases maintained by the Armed Forces and students in attendance at universities or colleges. There may be a question whether distribution of congressional seats except according to total population can ever be permissible under Art. I, § 2. But assuming without deciding that apportionment may be based on eligible voter population rather than total population, the Missouri plan is still unacceptable. Missouri made no attempt to ascertain the *535number of eligible voters in each district and to apportion accordingly. At best it made haphazard adjustments to a scheme based on total population: overpopulation in the Eighth District was explained away by the presence in that district of a military base and a university; no attempt was made to account for the presence of universities in other districts or the disproportionate numbers of newly arrived and short-term residents in the City of St. Louis. Even as to the Eighth District, there is no indication that the excess population allocated to that district corresponds to the alleged extraordinary additional numbers of noneligible voters there.
Missouri also argues that population disparities between some of its congressional districts result from the legislature’s attempt to take into account projected population shifts. We recognize that a congressional district-ing plan will usually be in effect for at least 10 years and five congressional elections. Situations may arise where substantial population shifts over such a period can be anticipated. Where these shifts can be predicted with a high degree of accuracy, States that are redistricting may properly consider them. By this we mean to open no avenue for subterfuge. Findings as to population trends must be thoroughly documented and applied throughout the State in a systematic, not an ad hoc, manner. Missouri’s attempted justification of the substantial underpopulation in the Fourth and Sixth Districts falls far short of this standard. The District Court found “no evidence . . . that the . . . General Assembly adopted any policy of population projection in devising Districts 4 and 6, or any other district, in enacting the 1967 Act.” 279 F. Supp., at 983.
Finally, Missouri claims that some of the deviations from equality were a consequence of the legislature’s attempt to ensure that each congressional district would be geographically compact. However, in Reynolds v. Sims, *536supra, at 580, we said, “Modern developments and improvements in transportation and communications make rather hollow, in the mid-1960’s, most claims that deviations from population-based representation can validly be based solely on geographical considerations. Arguments for allowing such deviations in order to insure effective representation for sparsely settled areas and to prevent legislative districts from becoming so large that the availability of access of citizens to their representatives is impaired are today, for the most part, unconvincing.” In any event, Missouri’s claim of compactness is based solely upon the unaesthetic appearance of the map of congressional boundaries that would result from an attempt to effect some of the changes in district lines which, according to the lower court, would achieve greater equality. A State’s preference for pleasingly shaped districts can hardly justify population variances.
Affirmed.
[For dissenting opinion of Mr. Justice Harlan, see post, p. 549.] [For dissenting opinion of Mr. Justice White, see post, p. 553.]The redistricting effected by the 1967 Act, based on a population of 4,319,813 according to the 1960 census, is as follows:
% Variation
District No. Population. From, Ideal.
One 439,746 + 1.80
Two 436,448 + 1.03
Three 436,099 + 0.95
Four 419,721 - 2.84
Five 431,178 - 0.19
Six 422,238 - 2.26
Seven 436,769 + 1.11
Eight 445,523 + 3.13
Nine 428,223 - 0.87
Ten 423,868 - 1.88
Ideal population per district. 431,981
Average variation from ideal. 1.6%
Ratio of largest to smallest district. 1.06 to 1
Number of districts within 1.88% of ideal. 7
Population difference between largest and smallest districts.. 25,802
Contrary to appellants’ assertion, we have not sustained the constitutionality of any congressional districting plan with population variances of the magnitude found in the Missouri plan. In Connor v. Johnson, 386 U. S. 483 (1967), the only issue presented to this Court was whether the districting plan involved racial gerrymandering. Alton v. Tawes, 384 U. S. 315 (1966), and Kirk v. Gong, 389 U. S. 574 (1968), involved situations where the lower courts themselves had reapportioned the districts on an emergency basis, and our affirmances were based on agreement with the use of the plans in that circumstance, and not on any view that the plans in question achieved equality as nearly as practicable.
Brief for Appellants 37-38.
It is dubious in any event that the temptation to gerrymander would be much inhibited, since the legislature would still be free to choose which of several subdivisions, all with their own political complexion, to include in a particular congressional district. Besides, opportunities for gerrymandering are greatest when there is freedom to construct unequally populated districts. “[T]he artistry of the political cartographer is put to its highest test when he must work with constituencies of equal population. At such times, his skills can be compared to those of a surgeon, for both work under fixed and arduous rules. However, if the mapmaker is free to allocate varying populations to different districts, then the butcher's cleaver replaces the scalpel; and the results reflect sharply the difference in the method of operation.” A. Hacker, Congressional Districting 59 (1964 rev. ed.).