DEPARTMENT OF COMMERCE v. UNITED STATES HOUSE OF REPRESENTATIVES

Justice O'Connor delivered the opinion of the Court, except as to Part III—B.

The Census Bureau (Bureau) has announced a plan to use two forms of statistical sampling in the 2000 Decennial Census to address a chronic and apparently growing problem of "under counting" certain identifiable groups of individuals. Two sets of plaintiffs filed separate suits challenging the legality and constitutionality of the Bureau's plan. Convened as three-judge courts, the District Court for the Eastern District of Virginia and the District Court for the District of Columbia each held that the Bureau's plan for the 2000 census violates the Census Act, 13 U. S. C. § 1 et seq. , and both courts permanently enjoined the Bureau's planned use of statistical sampling to determine the population for purposes of congressional apportionment. 19 F.Supp.2d 543 (ED Va. 1998); 11 F.Supp.2d 76 (DC 1998). We noted probable jurisdiction in both cases, 524 U.S. 978 (1998); 525 U.S. 924 (1998), and consolidated the cases for oral argument, 525 U.S. 924 (1998). We now affirm the judgment of the District Court for the Eastern District of Virginia, and we dismiss the appeal from the District Court for the District of Columbia.

I

A

Article I, § 2, cl. 3, of the United States Constitution states that "Representatives . . . shall be apportioned among the several States . . . according to their respective Numbers." It further requires that "[t]he actual Enumeration shall be made within three Years after the first Meeting of the Congress of the United States, and within every subsequent Term of ten Years, in such Manner as they shall by Law direct." Ibid. Finally, § 2 of the Fourteenth Amendment provides that "Representatives shall be apportioned among the several States according to their respective numbers, counting the whole number of persons in each State, excluding Indians not taxed."

Pursuant to this constitutional authority to direct the manner in which the "actual Enumeration" of the population shall be made, Congress enacted the Census Act (hereinafter Census Act or Act), 13 U. S. C. § 1 et seq. , delegating to the Secretary of Commerce (Secretary) authority to conduct the decennial census. § 4. The Act provides that the Secretary "shall, in the year 1980 and every 10 years thereafter, take a decennial census of population as of the first day of April of such year." § 141(a). It further requires that "[t]he tabulation of total population by States . . . as required for the apportionment of Representatives in Congress among the several States shall be completed within 9 months after the census date and reported by the Secretary to the President of the United States." § 141(b). Using this information, the President must then "transmit to the Congress a statement showing the whole number of persons in each State . . . and the number of Representatives to which each State would be entitled." 2 U. S. C. § 2a(a). Within 15 days thereafter, the Clerk of the House of Representatives must "send to the executive of each State a certificate of the number of Representatives to which such State is entitled." 2 U. S. C. § 2a(b) (1994 ed., Supp. III).

The instant dispute centers on the problem of "undercount" in the decennial census. For the last few decades, the Bureau has sent census forms to every household, which it asked residents to complete and return. The Bureau followed up on the mailing by sending enumerators to personally visit all households that did not respond by mail. Despite this comprehensive effort to reach every household, the Bureau has always failed to reach—and has thus failed to count—a portion of the population. This shortfall has been labeled the census "undercount."

The Bureau has been measuring the census undercount rate since 1940, and undercount has been the subject of public debate at least since the early 1970's. See M. Anderson, The American Census: A Social History 221-222 (1988). It has been measured in one of two ways. Under one method, known as "demographic analysis," the Bureau develops an independent estimate of the population using birth, death, immigration, and emigration records. U. S. Dept. of Commerce, Bureau of the Census, Report to Congress: The Plan for Census 2000, p. 2, and n. 1 (Aug. 1997) (hereinafter Census 2000 Report). A second method, first used in 1990, involves a large sample survey, called the "Post-Enumeration Survey," that is conducted in conjunction with the decennial census. The Bureau compares the information gathered during the survey with the information obtained in the census and uses the comparison to estimate the number of unenumerated people in the census. See National Research Council, Modernizing the U. S. Census 30-31 (B. Edmonston & C. Schultze eds. 1995).

Some identifiable groups—including certain minorities, children, and renters—have historically had substantially higher undercount rates than the population as a whole. See Census 2000 Report 3-4. Accordingly, in previous censuses, the Bureau sought to increase the number of persons from whom it obtained information. In 1990, for instance, the Bureau attempted to reach out to traditionally undercounted groups by promoting awareness of the census and its importance, providing access to Spanish language forms, and offering a toll free number for those who had questions about the forms. Id., at 4. Indeed, the 1990 census was "better designed and executed than any previous census." Id., at 2. Nonetheless, it was less accurate than its predecessor for the first time since the Bureau began measuring the undercount rate in 1940. Ibid.

In a further effort to address growing concerns about undercount in the census, Congress passed the Decennial Census Improvement Act of 1991, which instructed the Secretary to contract with the National Academy of Sciences (Academy) to study the "means by which the Government could achieve the most accurate population count possible." § 2(a)(1), 105 Stat. 635, note following 13 U. S. C. § 141. Among the issues the Academy was directed to consider was "the appropriateness of using sampling methods, in combination with basic data-collection techniques or otherwise, in the acquisition or refinement of population data." Ibid. Two of the three panels established by the Academy pursuant to this Act concluded that "[d]ifferential undercount cannot be reduced to acceptable levels at acceptable costs without the use of integrated coverage measurement," a statistical sampling procedure that adjusts census results to account for undercount in the initial enumeration, Census 2000 Report 7-8, and all three panels recommended including integrated coverage measurement in the 2000 census, id., at 29. See National Research Council, Preparing for the 2000 Census: Interim Report II (A. White & K. Rust eds. 1997) (report of Panel to Evaluate Alternative Census Methodologies); Modernizing the U. S. Census, supra (report of Panel on Census Requirements in the Year 2000 and Beyond); U. S. Dept. of Commerce, Bureau of the Census, Census 2000 Operational Plan (1997).

In light of these studies and other research, the Bureau formulated a plan for the 2000 census that uses statistical sampling to supplement data obtained through traditional census methods. The Bureau plan provides for two types of sampling that are the subject of the instant challenge.1 First, appellees challenge the proposed use of sampling in the Nonresponse Followup program (NRFU). Under this program, the Bureau would continue to send census forms to all households, as well as make forms available in post offices and in other public places. The Bureau expects that 67 percent of households will return the forms. See Census 2000 Report 26. The Bureau then plans to divide the population into census tracts of approximately 4,000 people that have "homogenous population characteristics, economic status, and living conditions." Id., at 27. The Bureau would then visit a randomly selected sample of nonresponding housing units, which would be "statistically representative of all housing units in [a] nonresponding tract." Id., at 28. The rate of nonresponse followup in a tract would vary with the mail response rate to ensure that the Bureau obtains census data from at least 90 percent of the housing units in each census tract. Ibid. For instance, if a census tract had 1,000 housing units and 800 units responded by mail, the Bureau would survey 100 out of the 200 nonresponding units to obtain information about 90 percent of the housing units. However, if only 400 of the 1,000 housing units responded by mail, the Bureau would visit 500 of the 600 nonresponding units to achieve the same result. Id., at 29. The information gathered from the nonresponding housing units surveyed by the Bureau would then be used to estimate the size and characteristics of the nonresponding housing units that the Bureau did not visit. Thus, continuing with the first example, the Bureau would use information about the 100 nonresponding units it visits to estimate the characteristics of the remaining 100 nonresponding units on which the Bureau has no information. See ibid.

The second challenged sampling procedure—which would be implemented after the first is completed—is known as Integrated Coverage Measurement (ICM). ICM employs the statistical technique called Dual System Estimation (DSE) to adjust the census results to account for under count in the initial enumeration. The plan requires the Bureau to begin by classifying each of the country's 7 million blocks into "strata," which are defined by the characteristics of each block, including state, racial, and ethnic composition, and the proportion of homeowners to renters, as revealed in the 1990 census. Id., at 30. The Bureau then plans to select blocks at random from each stratum, for a total of 25,000 blocks, or an estimated 750,000 housing units. Ibid. Enumerators would then conduct interviews at each of those 750,000 units, and if discrepancies were detected between the pre-ICM response and ICM response, a followup interview would be conducted to determine the "true" situation in the home. Ibid. The information gathered during this stage would be used to assign each person to a poststratum—a group of people who have similar chances of being counted in the initial data collection—which would be defined by state geographic subdivision (e. g., rural or urban), owner or renter, age, sex, race, and Hispanic origin. Id., at 31.

In the final stage of the census, the Bureau plans to use DSE to obtain the final count and characteristics of the population. The census plan calls for the Bureau to compare the dual systems of information—that is, the data gathered on the sample blocks during the ICM and the data gathered on those same blocks through the initial phase of the census— to produce an estimation factor for each poststratum. The estimation factors would account for the differences between the ICM numbers and the initial enumeration and would be applied to the initial enumeration to estimate the total population and housing units in each poststratum. Id., at 31-32. The totals for the poststrata would then be summed to determine state and national population totals. Id., at 32.

The Bureau's announcement of its plan to use statistical sampling in the 2000 census led to a flurry of legislative activity. Congress amended the Census Act to provide that, "[n]otwithstanding any other provision of law, no sampling or any other statistical procedure, including any statistical adjustment, may be used in any determination of population for purposes of the apportionment of Representatives in Congress among the several States," H. R. Conf. Rep. No. 105-119, p. 67 (1997), but President Clinton vetoed the bill, see Message to the House of Representatives Returning Without Approval Emergency Supplemental Appropriations Legislation, 33 Weekly Comp. of Pres. Doc. 846, 847 (1997). Congress then passed, and the President signed, a bill providing for the creation of a "comprehensive and detailed plan outlining [the Bureau's] proposed methodologies for conducting the 2000 Decennial Census and available methods to conduct an actual enumeration of the population," including an explanation of any statistical methodologies that may be used. 1997 Emergency Supplemental Appropriations Act for Recovery From Natural Disasters, and for Overseas Peacekeeping Efforts, Including Those in Bosnia, Tit. VIII, 111 Stat. 217. Pursuant to this directive, the Commerce Department issued the Census 2000 Report. After receiving the Report, Congress passed the 1998 Departments of Commerce, Justice, and State, the Judiciary, and Related Agencies Appropriations Act, § 209, 111 Stat. 2482, which provides that the Census 2000 Report and the Bureau's Census 2000 Operational Plan "shall be deemed to constitute final agency action regarding the use of statistical methods in the 2000 decennial census." The Act also permits any person aggrieved by the plan to use statistical sampling in the decennial census to bring a legal action and requires that any action brought under the Act be heard by a three-judge district court. Ibid. It further provides for review by appeal directly to this Court. Ibid.

B

The publication of the Bureau's plan for the 2000 census occasioned two separate legal challenges. The first suit, styled Clinton v. Glavin, was filed on February 12, 1998, in the District Court for the Eastern District of Virginia by four counties (Cobb County, Georgia; Bucks County, Pennsylvania; Delaware County, Pennsylvania; and DuPage County, Illinois) and residents of 13 States (Arizona, California, Connecticut, Florida, Georgia, Illinois, Indiana, Montana, Nevada, Ohio, Pennsylvania, Virginia, and Wisconsin), who claimed that the Bureau's planned use of statistical sampling to apportion Representatives among the States violates the Census Act and the Census Clause of the Constitution. They sought a declaration that the Bureau's plan is unlawful and/or unconstitutional and an injunction barring use of the NRFU and ICM sampling procedures in the 2000 census.

The District Court held that the case was ripe for review, that the plaintiffs satisfied the requirements for Article III standing, and that the Census Act prohibited use of the challenged sampling procedures to apportion Representatives. 19 F. Supp. 2d, at 547, 548-550, 553. The District Court concluded that, because the statute was clear on its face, the court did not need to reach the constitutional questions presented. Id., at 553. It thus denied defendants' motion to dismiss, granted plaintiffs' motion for summary judgment, and permanently enjoined the use of the challenged sampling procedures to determine the population for purposes of congressional apportionment. Id., at 545, 553. We noted probable jurisdiction on October 9, 1998. 525 U.S. 924.

The second challenge was filed by the United States House of Representatives on February 20, 1998, in the District Court for the District of Columbia. The House sought a declaration that the Bureau's proposed use of sampling to determine the population for purposes of apportioning Members of the House of Representatives among the several States violates the Census Act and the Constitution. The House also sought a permanent injunction barring use of the challenged sampling procedures in the apportionment aspect of the 2000 census.

The District Court held that the House had Article III standing, the suit was ripe for review, equitable concerns did not warrant dismissal, the suit did not violate separation of powers principles, and the Census Act does not permit the use of the challenged sampling procedures in counting the population for apportionment. 11 F.Supp. 2d, at 93, 95, 97, 104. Because it held that the Census Act does not allow for the challenged sampling procedures, it declined to reach the House's constitutional challenge under the Census Clause. Id., at 104. The District Court denied the defendants' motion to dismiss, granted the plaintiffs' motion for summary judgment, and issued an injunction preventing defendants from using the challenged sampling methods in the apportionment aspect of the 2000 census. Id., at 79, 104. The defendants appealed to this Court and we noted probable jurisdiction on September 10, 1998, 524 U.S. 978, and consolidated this case with Clinton v. Glavin, No. 98-564, for oral argument, 525 U.S. 924 (1998).

II

We turn our attention first to the issues presented by Clinton v. Glavin, No. 98-564, and we begin our analysis with the threshold issue of justiciability. Congress has eliminated any prudential concerns in this case by providing that "[a]ny person aggrieved by the use of any statistical method in violation of the Constitution or any provision of law (other than this Act), in connection with the 2000 census or any later decennial census, to determine the population for purposes of the apportionment or redistricting of Members in Congress, may in a civil action obtain declaratory, injunctive, and any other appropriate relief against the use of such method." § 209(b), 111 Stat. 2481. In addition, the District Court below correctly found that the case is ripe for review, and that determination is not challenged here. 19 F. Supp. 2d, at 547; see Abbott Laboratories v. Gardner, 387 U.S. 136, 149 (1967). Thus, the only open justiciability question in this case is whether appellees satisfy the requirements of Article III standing.

We have repeatedly noted that in order to establish Article III standing, "[a] plaintiff must allege personal injury fairly traceable to the defendant's allegedly unlawful conduct and likely to be redressed by the requested relief." Allen v. Wright, 468 U.S. 737, 751 (1984). See also Lujan v. Defenders of Wildlife, 504 U.S. 555, 560-561 (1992); Valley Forge Christian College v. Americans United for Separation of Church and State, Inc., 454 U.S. 464, 472 (1982). To prevail on a Federal Rule of Civil Procedure 56 motion for summary judgment—as opposed to a motion to dismiss—however, mere allegations of injury are insufficient. Rather, a plaintiff must establish that there exists no genuine issue of material fact as to justiciability or the merits. See Lujan v. National Wildlife Federation, 497 U.S. 871, 884 (1990). See also id., at 902 (Blackmun, J., dissenting). Here, the District Court, considering a Rule 56 motion, held that the plaintiffsappellees, residents from 13 States, had established Article III standing to bring suit challenging the proposed method for conducting the 2000 census because they had made "[g]eneral factual allegations of injury resulting from Defendant's conduct." 19 F. Supp., at 548-550. The court did not, however, consider whether there was a genuine issue of material fact as to standing.

Nonetheless, because the record before us amply supports the conclusion that several of the appellees have met their burden of proof regarding their standing to bring this suit, we affirm the District Court's holding. See Director, Office of Workers' Compensation Programs v. Perini North River Associates, 459 U.S. 297, 303-305 (1983) (holding that presence of one party with standing assures that controversy before Court is justiciable); Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252, 264, and n. 9 (1977) (same). In support of their motion for summary judgment, appellees submitted the affidavit of Dr. Ronald F. Weber, a professor of government at the University of Wisconsin, which demonstrates that Indiana resident Gary A. Hofmeister has standing to challenge the proposed census 2000 plan.2 Affidavit of Dr. Ronald F. Weber, App. in No. 98-564, pp. 56-79 (hereinafter Weber Affidavit). Utilizing data published by the Bureau, Dr. Weber projected year 2000 populations and net undercount rates for all States under the 1990 method of enumeration and under the Department's proposed plan for the 2000 census. See id., at 62-63. He then determined on the basis of these projections how many Representatives would be apportioned to each State under each method and concluded that "it is a virtual certainty that Indiana will lose a seat . . . under the Department's Plan." Id., at 65.

Appellants have failed to set forth any specific facts showing that there is a genuine issue of standing for trial. See Fed. Rule Civ. Proc. 56(e). Appellants have submitted two affidavits that detail various deficiencies in the statistical analysis performed by Dr. Weber. See Declaration of Signe I. Wetrogan, Assistant Division Chief for Population Estimates and Projections, United States Bureau of the Census, App. in No. 98-564, pp. 92-99 (hereinafter Wetrogan Declaration); Declaration of John H. Thompson, Associate Director for the Decennial Census, United States Bureau of the Census, id., at 100-110 (hereinafter Thompson Declaration). Appellants' experts do not, however, demonstrate that any alleged flaw in Dr. Weber's analysis calls into question his ultimate conclusion that Indiana is virtually certain to lose a seat. One expert, for example, claims that Dr. Weber's statement that Indiana is virtually certain to lose a seat is "of dubious credibility," but she fails to provide any specific factual support for this assertion. Wetrogan Declaration 97. She claims that Dr. Weber used outdated population numbers, but she does not demonstrate the impact that using more recent population data would have on Dr. Weber's ultimate conclusion about Indiana. Id., at 97-98. Neither of the appellants' experts reestimates the populations of the States using more "accurate" or "up-to-date" data to show that this data would produce different results. Indeed, the Associate Director for the Decennial Census specifically admits in his declaration that Dr. Weber used precisely the same data that the Bureau uses "to help it estimate expected error rates for Census 2000." Thompson Declaration 106. Appellants have therefore failed to raise a genuine issue of material fact regarding Indiana's loss of a Representative.

Appellee Hofmeister's expected loss of a Representative to the United States Congress undoubtedly satisfies the injury-in-fact requirement of Article III standing. In the context of apportionment, we have held that voters have standing to challenge an apportionment statute because "[t]hey are asserting `a plain, direct and adequate interest in maintaining the effectiveness of their votes.' " Baker v. Carr, 369 U.S. 186, 208 (1962) (quoting Coleman v. Miller, 307 U.S. 433, 438 (1939)). The same distinct interest is at issue here: With one fewer Representative, Indiana residents' votes will be diluted. Moreover, the threat of vote dilution through the use of sampling is "concrete" and "actual or imminent, not `conjectural' or `hypothetical.' " Whitmore v. Arkansas, 495 U.S. 149, 155 (1990). It is clear that if the Bureau is going to alter its plan to use sampling in the 2000 census, it must begin doing so by March 1999. See Oversight of the 2000 Census: Putting the Dress Rehearsals in Perspective, Hearing before the Subcommittee on the Census of the House Committee on Government Reform and Oversight, 105th Cong., 2d Sess., 84 (1998) (statement of James F. Holmes, Acting Director of the Bureau) ("I must caution that by this time next year [i. e., March 1999] the train for census 2000 has to be on one track. If the uncertainty continues, if our staff continues to have to do two jobs,. . . [the census] will truly be imperiled"). See also § 209, 111 Stat. 2480 (providing that the Bureau's plan to use statistical sampling in the 2000 census constitutes "final agency action"). And it is certainly not necessary for this Court to wait until the census has been conducted to consider the issues presented here, because such a pause would result in extreme—possibly irremediable—hardship. In addition, as Dr. Weber's affidavit demonstrates, Hofmeister meets the second and third requirements of Article III standing. There is undoubtedly a "traceable" connection between the use of sampling in the decennial census and Indiana's expected loss of a Representative, and there is a substantial likelihood that the requested relief—a permanent injunction against the proposed uses of sampling in the census—will redress the alleged injury.

Appellees have also established standing on the basis of the expected effects of the use of sampling in the 2000 census on intrastate redistricting. Dr. Weber indicated in his affidavit that "[i]t is substantially likely that voters in Maricopa County, Arizona, Bergen County, New Jersey, Cumberland County, Pennsylvania, LaSalle County, Illinois, Orange County, California, St. Johns County, Florida, Gallatin County, Montana, Forsyth County, Georgia, and Loudoun County, Virginia, will suffer vote dilution in state and local elections as a result of the [Bureau's] Plan." Weber Affidavit 77-78. Several of the appellees reside in these counties,3 and several of the States in which these counties are located require use of federal decennial census population numbers for their state legislative redistricting. The New Jersey Constitution, for instance, requires that state senators be apportioned among Senate districts "as nearly as may be according to the number of their inhabitants as reported in the last preceding decennial census of the United States." Art. IV, § 1, ¶1. Similarly, the Pennsylvania Constitution requires that "[i]n each year following the year of the Federal decennial census, a Legislative Reapportionment Commission shall be constituted for the purpose of reapportioning the Commonwealth." Art. 2, § 17(a). Several of the other States cited by Dr. Weber have comparable laws.4 Moreover, States use the population numbers generated by the federal decennial census for federal congressional redistricting. See Karcher v. Daggett, 462 U.S. 725, 738 (1983) ("[B]ecause the census count represents the `best population data available,' . . . it is the only basis for good-faith attempts to achieve population equality" (citation omitted)). Thus, the appellees who live in the aforementioned counties have a strong claim that they will be injured by the Bureau's plan because their votes will be diluted vis-à-vis residents of counties with larger "undercount" rates. Neither of appellants' experts specifically contested Dr. Weber's conclusion that the nine counties were substantially likely to lose population if statistical sampling were used in the 2000 census. See Wetrogan Declaration 92-99; Thompson Declaration 100-110. The experts' general assertions regarding Dr. Weber's methodology and data are again insufficient to create a genuine issue of material fact. For the reasons discussed above, see supra, at 332-333 and this page, this expected intrastate vote dilution satisfies the injury-in-fact, causation, and redressibility requirements. Accordingly, appellees have again carried their burden under Rule 56 and have established standing to pursue this case.

III

We accordingly arrive at the dispute over the meaning of the relevant provisions of the Census Act. The District Court below examined the plain text and legislative history of the Act and concluded that the proposed use of statistical sampling to determine population for purposes of apportioning congressional seats among the States violates the Act. We agree.

A

An understanding of the historical background of the decennial census and the Act that governs it is essential to a proper interpretation of the Act's present text. From the very first census, the census of 1790, Congress has prohibited the use of statistical sampling in calculating the population for purposes of apportionment. The First Congress enacted legislation requiring census enumerators to swear an oath to make "a just and perfect enumeration" of every person within the division to which they were assigned. Act of Mar. 1, 1790, § 1, 1 Stat. 101. Each enumerator was required to compile a schedule of information for his district, listing by family name the number of persons in each family that fell into each of five specified categories. See id., at 101-102. Congress modified this provision in 1810, adding an express statement that "the said enumeration shall be made by an actual inquiry at every dwelling-house, or of the head of every family within each district, and not otherwise," and expanding the number of specifications in the schedule of information. Act of Mar. 26, 1810, § 1, 2 Stat. 565-566. The requirement that census enumerators visit each home in person appeared in statutes governing the next 14 censuses.5

The current Census Act was enacted into positive law in 1954. It contained substantially the same language as did its predecessor statutes, requiring enumerators to "visit personally each dwelling house in his subdivision" in order to obtain "every item of information and all particulars required for any census or survey" conducted in connection with the census. Act of Aug. 31, 1954, § 25(c), 68 Stat. 1012, 1015. Indeed, the first departure from the requirement that the enumerators collect all census information through personal visits to every household in the Nation came in 1957 at the behest of the Secretary. The Secretary asked Congress to amend the Act to permit the Bureau to use statistical sampling in gathering some of the census information. See Amendment of Title 13, United States Code, Relating to Census: Hearing on H. R. 7911 before the House Committee on the Post Office and Civil Service, 85th Cong., 1st Sess., 4-8 (1957) (hereinafter 1957 Hearing). In response, Congress enacted § 195, which provided that, "[e]xcept for the determination of population for apportionment purposes, the Secretary may, where he deems it appropriate, authorize the use of the statistical method known as `sampling' in carrying out the provisions of this title." 13 U. S. C. § 195 (1970 ed.). This provision allowed the Secretary to authorize the use of sampling procedures in gathering supplemental, nonapportionment census information regarding population, unemployment, housing, and other matters collected in conjunction with the decennial census—much of which is now collected through what is known as the "long form"—but it did not authorize the use of sampling procedures in connection with apportionment of Representatives. See also 1957 Hearing 7-8 ("Experience has shown that some of the information which is desired in connection with a census could be secured efficiently through a sample survey which is conducted concurrently with the complete enumeration of other items").

In 1964, Congress repealed former § 25(c) of the Census Act, see Act of Aug. 31, 1964, 78 Stat. 737, which had required that each enumerator obtain "every item of information" by personal visit to each household, 68 Stat. 1015. The repeal of this section permitted the Bureau to replace the personal visit of the enumerator with a form delivered and returned via the Postal Service. Pursuant to this new authority, census officials conducted approximately 60 percent of the census through a new "mailout-mailback" system for the first time in 1970. See M. Anderson, The American Census: A Social History 210-211 (1988). The Bureau then conducted followup visits to homes that failed to return census forms. Thus, although the legislation permitted the Bureau to conduct a portion of the census through the mail, there was no suggestion from any quarter that this change altered the prohibition in § 195 on the use of statistical sampling in determining the population for apportionment purposes.

In 1976, the provisions of the Census Act at issue in this case took their present form. Congress revised § 141 of the Census Act, which is now entitled "Population and other census information." It amended subsection (a) to authorize the Secretary to "take a decennial census of population as of the first day of April of such year, which date shall be known as the `decennial census date', in such form and content as he may determine, including the use of sampling procedures and special surveys." 13 U. S. C. § 141(a). Congress also added several subsections to § 141, among them a provision specifying that the term "census of population," as used in § 141, "means a census of population, housing, and matters relating to population and housing." § 141(g). Together, these revisions provided a broad statement that in collecting a range of demographic information during the decennial census, the Bureau would be permitted to use sampling procedures and special surveys.

This broad grant of authority given in § 141(a) is informed, however, by the narrower and more specific § 195, which is revealingly entitled, "Use of Sampling." See Green v. Bock Laundry Machine Co., 490 U.S. 504, 524 (1989). The § 141 authorization to use sampling techniques in the decennial census is not necessarily an authorization to use these techniques in collecting all of the information that is gathered during the decennial census. We look to the remainder of the law to determine what portions of the decennial census the authorization covers. When we do, we discover that, as discussed above, § 195 directly prohibits the use of sampling in the determination of population for purposes of apportionment.6

When Congress amended § 195 in 1976, it did not in doing so alter the longstanding prohibition on the use of sampling in matters relating to apportionment. Congress modified the section by changing "apportionment purposes" to "purposes of apportionment of Representatives in Congress among the several States" and changing the phrase "may, where he deems it appropriate" to "shall, if he considers it feasible." 90 Stat. 2464. The amended section thus reads: "Except for the determination of population for purposes of apportionment of Representatives in Congress among the several States, the Secretary shall, if he considers it feasible, authorize the use of the statistical method known as `sampling' in carrying out the provisions of this title." 13 U. S. C. § 195. As amended, the section now requires the Secretary to use statistical sampling in assembling the myriad demographic data that are collected in connection with the decennial census. But the section maintains its prohibition on the use of statistical sampling in calculating population for purposes of apportionment.

Absent any historical context, the language in the amended § 195 might reasonably be read as either permissive or prohibitive with regard to the use of sampling for apportionment purposes. Indeed, appellees and appellants each cite numerous examples of the "except/shall" sentence structure that support their respective interpretations of the statute. See, e. g., Brief for Appellee Glavin et al. in No. 98-564, p. 36, n. 36 (citing § 2 of the Fourteenth Amendment, which provides that "when the right to vote . . . is denied to any of the male inhabitants of such State . . . except for participation in rebellion, or other crime, the basis of representation therein shall be reduced in the proportion which the number of such male citizens shall bear to the whole number of male citizens twenty-one years of age in such State" (emphasis added)); Brief for Federal Appellant et al. in No. 98-404, p. 29, n. 15 (citing 2 U. S. C. §§ 179n(a)(1) and 384(a) and 5 U. S. C. § 555(e), which contain the "except/shall" formulation in contexts where appellants claim "the exception cannot reasonably be construed as prohibiting the excepted activity"). But these dueling examples only serve to illustrate that the interpretation of the "except/shall" structure depends primarily on the broader context in which that structure appears. Here, the context is provided by over 200 years during which federal statutes have prohibited the use of statistical sampling where apportionment is concerned. In light of this background, there is only one plausible reading of the amended § 195: It prohibits the use of sampling in calculating the population for purposes of apportionment.

In fact, the Bureau itself concluded in 1980 that the Census Act, as amended, "clearly" continued the "historical precedent of using the `actual Enumeration' for purposes of apportionment, while eschewing estimates based on sampling or other statistical procedures, no matter how sophisticated." See 45 Fed. Reg. 69366, 69372 (1980). That same year, the Solicitor General argued before this Court that "13 U. S. C. 195 prohibits the use of statistical `sampling methods' in determining the state-by-state population totals." Application for Stay in Klutznick v. Young, O. T. 1979, No. A-533, p. 14, n. 7. See also Young v. Klutznick, 652 F.2d 617, 621 (CA6 1981) (noting that the Census Director and other officials explained at trial that "since 1790 the census enumeration has never been adjusted to reflect an estimated undercount and that in their opinion Congress by statute had prohibited such an adjustment in the figures used for purposes of Congressional apportionment"), cert. denied sub nom. Young v. Baldrige, 455 U.S. 939 (1982); Philadelphia v. Klutznik, 503 F.Supp. 663, 678 (ED Pa. 1980) (noting that the Bureau argued that "Congress has clearly rejected the use of an adjustment figure in the Census Act"); Carey v. Klutznik, 508 F.Supp. 404 (SDNY 1980) ("Defendants [including the Secretary of Commerce and the Director of the Bureau of the Census] [contend that the] Census Act preclude[s] utilization of statistical adjustment for the purpose of apportioning representatives"), rev'd, 653 F.2d 732 (CA2 1981), cert. denied, 455 U.S. 999 (1982). The administration did not adopt the contrary position until 1994, when it first concluded that using statistical sampling to adjust census figures would be consistent with the Census Act. Memorandum for the Solicitor General from Assistant Attorney General Dellinger 1 (Oct. 7, 1994). In light of this history, appellants make no claim to deference under Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc. , 467 U.S. 837 (1984), on behalf of the Secretary's interpretation of the Census Act. Reply Brief for Federal Appellant et al. in No. 98-404, p. 11, n. 10.

In holding that the 1976 amendments did not change the prohibition on the use of sampling in determining the population for apportionment purposes, we do not mean to suggest, as Justice Stevens claims in dissent, that the 1976 amendments had no purpose. See post, at 360-362. Rather, the amendments served a very important purpose: They changed a provision that permitted the use of sampling for purposes other than apportionment into one that required that sampling be used for such purposes if "feasible." They also added to the existing delegation of authority to the Secretary to carry out the decennial census a statement indicating that despite the move to mandatory use of sampling in collecting nonapportionment information, the Secretary retained substantial authority to determine the manner in which the decennial census is conducted.

Justice Stevens argument reveals a rather limited conception of the extent and purpose of the decennial census. The decennial census is "the only census that is used for apportionment purposes," post, at 359, but the decennial census is not only used for apportionment purposes. Although originally established for the sole purpose of apportioning Representatives, the decennial census has grown considerably over the past 200 years. It now serves as "a linchpin of the federal statistical system by collecting data on the characteristics of individuals, households, and housing units throughout the country." National Research Council, Counting People in the Information Age 1 (D. Steffey & N. Bradburn eds. 1994). Thus, to say that the 1976 amendments required the use of sampling in collecting nonapportionment information but had no effect on the way in which the Secretary could determine the population for the purposes of apportionment is to say that they had a purpose—just not the purpose that Justice Stevens imagines.

Justice Breyer's interpretation of § 195 is equally unpersuasive. Justice Breyer agrees with the Court that the Census Act prohibits the use of sampling as a substitute for traditional enumeration methods. But he believes that this prohibition does not apply to the use of sampling as a "supplement" to traditional enumeration methods. This distinction is not borne out by the language of the statute. The Census Act provides that sampling cannot be used "for the determination of population for purposes of apportionment of Representatives in Congress among the several States." 13 U. S. C. § 195. Whether used as a "supplement" or as a "substitute," sampling is still used in "determining"—that is, in "the act of deciding definitely and firmly." Webster's Ninth New Collegiate Dictionary 346 (1983). Under the proposed plan, the population is not "determined," not decided definitely and firmly, until the NRFU and ICM are complete. That the distinction drawn by Justice Breyer is untenable is perhaps best demonstrated by his own inability to apply it consistently. He acknowledges that the NRFU uses statistical sampling "to determine the last 10% of the population in each census tract," post, at 355 (emphasis added), yet he nonetheless finds that it is a supplement to the headcount and thus permitted by the Act.

B

The conclusion that the Census Act prohibits the use of sampling for apportionment purposes finds support in the debate and discussions surrounding the 1976 revisions to the Census Act. At no point during the debates over these amendments did a single Member of Congress suggest that the amendments would so fundamentally change the manner in which the Bureau could calculate the population for purposes of apportionment. See 122 Cong. Rec. 35171-35175 (1976); id., at 9792-9803, 32251-32253, 33128-33132, 33305— 33307, 33815; Mid-Decade Census Legislation: Hearing on S. 3688 and H. R. 11337 before the Subcommittee on Census and Statistics of the House Committee on the Post Office and Civil Service, 94th Cong., 2d Sess. (1976). See also H. R. Rep. No. 94-944 (1976); H. R. Conf. Rep. No. 94-1719 (1976); S. Rep. No. 94-1256 (1976). This is true despite the fact that such a change would profoundly affect Congress by likely shifting the number of seats apportioned to some States and altering district lines in many others. Indeed, it tests the limits of reason to suggest that despite such silence, Members of Congress voting for those amendments intended to enact what would arguably be the single most significant change in the method of conducting the decennial census since its inception. That the 1976 changes to §§ 141 and 195 were not the focus of partisan debate, see post, at 360-361 (Stevens, J., dissenting), is almost certainly due to the fact that the Members of Congress voting on the bill read the text of the statute, as do we, to prohibit the use of sampling in determining the population for apportionment purposes. Moreover, it is hard to imagine that, having explicitly prohibited the use of sampling for apportionment purposes in 1957, Congress would have decided to reverse course on such an important issue by enacting only a subtle change in phraseology.

IV

For the reasons stated, we conclude that the Census Act prohibits the proposed uses of statistical sampling in calculating the population for purposes of apportionment. Because we so conclude, we find it unnecessary to reach the constitutional question presented. See Spector Motor Service, Inc. v. McLaughlin, 323 U.S. 101, 105 (1944) ("If there is one doctrine more deeply rooted than any other in the process of constitutional adjudication, it is that we ought not to pass on questions of constitutionality . . . unless such adjudication is unavoidable"); Ashwander v. TVA, 297 U.S. 288,

Scalia, J., concurring in part 347 (1936) (Brandeis, J., concurring) ("[I]f a case can be decided on either of two grounds, one involving a constitutional question, the other a question of statutory construction or general law, the Court will decide only the latter"). Accordingly, we affirm the judgment of the District Court for the Eastern District of Virginia in Clinton v. Glavin, No. 98— 564. As this decision also resolves the substantive issues presented by Department of Commerce v. United States House of Representatives, No. 98-404, that case no longer presents a substantial federal question. The appeal in that case is therefore dismissed. Cf. Sanks v. Georgia, 401 U.S. 144, 145 (1971).

It is so ordered.

Justice Scalia, with whom Justice Thomas joins, and with whom The Chief Justice and Justice Kennedy join as to Part II, concurring in part.

I

I join the opinion of the Court, excluding, of course, the plurality's resort in Part III—B to what was said by individual legislators and committees of legislators—or more precisely (and worse yet), what was not said by individual legislators and committees of legislators. I write separately to respond at somewhat greater length to Justice Stevens analysis of 13 U. S. C. § 141(a), to add several additional points of textual analysis, and to invoke the doctrine of constitutional doubt, which is a major factor in my decision.

II

Section 141(a) requires the Secretary of Commerce to conduct a "decennial census of population . . . in such form and content as he may determine, including the use of sampling procedures and special surveys." Justice Stevens reasons that a reading of § 195 that would prohibit sampling for apportionment purposes contradicts this provision. It seems to me there is no conflict at all. The phrase "decennial census of population" in § 141(a) refers to far more than the "tabulation of total population by States . . . as required for the apportionment of Representatives in Congress among the several States . . . ." § 141(b). See U. S. Const., Art. I, § 2. It also includes "a census of population, housing, and matters relating to population and housing." § 141(g). The authorization of sampling techniques in the "decennial census of population" is not necessarily an authorization of such techniques in all aspects of the decennial census—any more than it is necessarily an authorization of all sampling techniques (for example, those that would violate the Fourth Amendment). One looks to the remainder of the law to determine what techniques, and what aspects of the decennial census, the authorization covers.

If, for example, it were utterly clear and universally agreed that the Constitution prohibits sampling in those aspects of the census related to apportionment, it would be strange to contend that, by authorizing the Secretary of Commerce to use sampling in his census work, § 141(a) "contradicts" the Constitution. The use of sampling it authorizes is lawful use of sampling, and if this does not include the apportionment aspect then the authorization obviously does not extend that far. I think the situation the same with regard to the legal impediment imposed by § 195. Justice Stevens would be correct that the Court is not interpreting § 195 "consistently with § 141(a)," post, at 358, if the latter provision specifically authorized sampling in "all aspects of the decennial census." But since it does not, the Court's interpretation is entirely harmonious.

Justice Stevens interpretation of this statute creates a palpable absurdity within § 195 itself. The "shall" of that provision is subject to not one exception, but two. The first, which is at issue here, is introduced by "Except." The second is contained within the phrase "if he considers it feasible." The Secretary is under no command to authorize

Scalia, J., concurring in part sampling if he does not consider it feasible. Is it even thinkable that he may (though he need not ) authorize sampling if he does not consider it feasible? The clear implication of "shall," as applied to this exception, is that where the exception applies he shall not. It would be strange to draw the different implication of "may" when the word is applied to the other exception.

And finally, Justice Stevens interpretation creates a statute in which Congress swallows a camel and strains out a gnat. Section 181 of the statute requires the Secretary to compile annual and biennial "interim current data"—a useful but hardly indispensable function. The Secretary is authorized to use sampling in the performance of this function only if he determines that it will produce "current, comprehensive, and reliable data." § 181(a). The statute Justice Stevens creates is one in which Congress carefully circumscribes the Secretary's discretion to use sampling in compiling "interim current data," but leaves it entirely up to the Secretary whether he will use sampling for the purpose most important (and closest to the Congress's heart): the apportionment of Representatives.

Even if one is not entirely persuaded by the foregoing arguments, and the more substantial analysis contained in the opinion of the Court, I think it must be acknowledged that the statutory intent to permit use of sampling for apportionment purposes is at least not clear. In these circumstances, it is our practice to construe the text in such fashion as to avoid serious constitutional doubt. See, e. g., Edward J. DeBartolo Corp. v. Florida Gulf Coast Building & Constr. Trades Council, 485 U.S. 568, 575 (1988). It is in my view unquestionably doubtful whether the constitutional requirement of an "actual Enumeration," Art. I, § 2, cl. 3, is satisfied by statistical sampling.

Dictionaries roughly contemporaneous with the ratification of the Constitution demonstrate that an "enumeration" requires an actual counting, and not just an estimation of number. Noah Webster's 1828 American Dictionary of the English Language defines "enumerate" as "[t]o count or tell, number by number; to reckon or mention a number of things, each separately"; and defines "enumeration" as "[t]he act of counting or telling a number, by naming each particular," and "[a]n account of a number of things, in which mention is made of every particular article." Samuel Johnson's 1773 Dictionary of the English Language 658 (4th ed.) defines "enumerate" as "[t]o reckon up singly; to count over distinctly; to number"; and "enumeration" as "[t]he act of numbering or counting over; number told out." Thomas Sheridan's 1796 Complete Dictionary of the English Language (6th ed.) defines "enumerate" as "[t]o reckon up singly; to count over distinctly"; and "enumeration" as "[t]he act of numbering or counting over." The notion of counting "singly," "separately," "number by number," "distinctly," which runs through these definitions is incompatible (or at least arguably incompatible, which is all that needs to be established) with gross statistical estimates.

One must also be impressed by the facts recited in the opinion of the Court, ante, at 335: that the Census Acts of 1790 and 1800 required a listing of persons by family name, and the Census Acts of 1810 through 1950 required census enumerators to visit each home in person. This demonstrates a longstanding tradition of Congress's forbidding the use of estimation techniques in conducting the apportionment census. Could it be that all these Congresses were unaware that (in the words of Justice Stevens dissent) estimation techniques "will make the census more accurate than an admittedly futile attempt to count every individual by personal inspection, interview, or written interrogatory"? Post, at 364. There were difficult-to-reach inhabitants in the early 1800's, just as there are today—indeed, perhaps a greater proportion of them, since the society was overwhelmingly composed of farmers, and largely of frontiersmen. And though there were no professional statisticians,

Scalia, J., concurring in part it must have been known that various methods of estimating unreachable people would be more accurate than assuming that all unreachable people did not exist. (Thomas Jefferson's 1782 estimate of the population of Virginia based upon limited data and specific demographic assumptions is thought to have been accurate by a margin of one-to-two percent. H. Alterman, Counting People: The Census in History 168— 170 (1969).) Yet such methods of estimation have not been used for over two centuries. The stronger the case the dissents make for the irrationality of that course, the more likely it seems that the early Congresses, and every Congress before the present one, thought that estimations were not permissible. See, e. g., Printz v. United States, 521 U.S. 898, 905 (1997) (historical evidence that "earlier Congresses avoided use of [the] highly attractive power [to compel state executive officers to administer federal programs]" gave us "reason to believe that the power was thought not to exist").

Justice Stevens reasons from the purpose of the Census Clause: "The census is intended to serve the constitutional goal of equal representation. That goal is best served by the use of a `Manner' that is most likely to be complete and accurate." Post, at 364 (internal quotation marks and citation omitted). That is true enough, and would prove the point if either (1) every estimate is more accurate than a headcount, or (2) Congress could be relied upon to permit only those estimates that are more accurate than headcounts. It is metaphysically certain that the first proposition is false, and morally certain that the second is. To give Congress the power, under the guise of regulating the "Manner" by which the census is taken, to select among various estimation techniques having credible (or even incredible) "expert" support is to give the party controlling Congress the power to distort representation in its own favor. In other words, genuine enumeration may not be the most accurate way of determining population, but it may be the most accurate way of determining population with minimal possibility of partisan manipulation. The prospect of this Court's reviewing estimation techniques in the future, to determine which of them so obviously creates a distortion that it cannot be allowed, is not a happy one. (I foresee the new specialty of "Census Law.") Indeed, it is doubtful whether— separation-of-powers considerations aside—the Court would even have available the raw material to conduct such review effectively. As pointed out by the appellants in the present cases, we will never be able to assess the relative accuracy of the sampling system used for the 2000 census by comparing it to the results of a headcount, for there will have been no headcount.

For reasons of text and tradition, fully compatible with a constitutional purpose that is entirely sensible, a strong case can be made that an apportionment census conducted with the use of "sampling techniques" is not the "actual Enumeration" that the Constitution requires. (Appellant Commerce Department itself once argued that case in the courts. See, e. g., Young v. Klutznick, 497 F.Supp. 1318, 1332 (ED Mich. 1980), rev'd 652 F.2d 617 (CA6 1981).) And since that is so, the statute before us, which certainly need not be interpreted to permit such a census, ought not be interpreted to do so.

Justice Breyer, concurring in part and dissenting in part.

I join Part II of the majority opinion concerning standing, and I join Parts II and III of Justice Stevens dissent. I also agree with Justice Stevens conclusion in Part I that the plan for the 2000 census presented by the Secretary of Commerce is not barred by the Census Act. In my view, however, the reason that 13 U. S. C. § 195 does not bar the statistical sampling at issue here is that § 195 focuses upon sampling used as a substitute for traditional enumeration methods, while the proposal at the heart of the Secretary's plan for the 2000 census (namely, Integrated Coverage Measurement, or ICM) is not so intended. Rather, ICM uses statistical sampling to supplement traditional enumeration methods in order to achieve the very accuracy that the census seeks and the Census Act itself demands. See, e. g., Decennial Census Improvement Act of 1991, § 2(a)(1), 105 Stat. 635, note following 13 U. S. C. § 141 (directing the Secretary to contract with the National Academy of Sciences to study "means by which the Government could achieve the most accurate population count possible").

The language of § 195 permits a distinction between sampling used as a substitute and sampling used as a supplement. The literal wording of its "except" clause focuses upon the use of sampling "for the determination of population for purposes of apportionment of Representatives in Congress among the several States." 13 U. S. C. § 195 (emphasis added). One can read those words as the majority does—applying to apportionment-connected sampling irrespective of use or kind. But one can also read them as applicable only to the use of sampling in place of the traditional "determination of population for purposes of apportionment." The "except" clause does not necessarily apply to every conceivable use of statistical sampling any more than, say, a statutory rule forbidding "vehicles" in the park applies to everything that could possibly be characterized as a "vehicle." See generally H. Hart, The Concept of Law 124-136 (2d ed. 1994) (discussing the "open texture of law"). Context normally informs the meaning of a general statutory phrase and often limits its scope.

The history and context of § 195 favor an interpretation that so limits the scope of that section. Cf. Brief for Appellants in No. 98-404, p. 36, n. 19; Brief for Appellees Gephardt et al. in No. 98-404, pp. 9-10, 22-23, 33-38; Young v. Klutznick, 497 F.Supp. 1318, 1335 (ED Mich. 1980) ("All that § 195 does is prohibit the use of figures derived solely by statistical techniques. It does not prohibit the use of statistics in addition to the more traditional measuring tools to arrive at a more accurate population count"), rev'd on other grounds, 652 F.2d 617 (CA6 1981); Carey v. Klutznick, 508 F.Supp. 404, 415 (SDNY 1980) (Census Act permits sampling in the context of apportionment as long as it is used only in addition to more traditional methods of enumeration). In the 1940's the Census Bureau began using statistical sampling in the collection of a variety of demographic information. U. S. Dept. of Commerce, Bureau of the Census, 200 Years of Census Taking: Population and Housing Questions, 1790-1990, p. 5 (Nov. 1989). Thus, during the 1940's and 1950's, each American family was asked to complete a short form containing a few information-gathering questions. In addition, the Bureau also used a long form that contained additional questions about individuals and families, but it asked only 1 family in 20 to complete this form. Ibid.; R. Jenkins, Procedural History of the 1940 Census of Housing and Population 13-15 (1985). The Census Bureau used those long-form answers, from 5% of the population, as a basis for extrapolating statistics and trends, about, say, unemployment or housing conditions, for the Nation as a whole.

In 1957 Congress focused upon this kind of sampling—a long form completed by only 1 American household in 20— as a model of what § 195 would authorize the Secretary to do—"[e]xcept for the determination of population for purposes of apportionment." 13 U. S. C. § 195. When explaining the need for the proposed § 195, the Secretary of Commerce spoke of a "sample enumeration or a sample census [that] might be substituted for a full census." Amendment of Title 13, United States Code, Relating to Census, Hearing on H. R. 7911 before the House Committee on Post Office and Civil Service, 85th Cong., 1st Sess., 7 (1957) (Statement of Purpose and Need) (emphasis added). He added that "[e]xperience has shown that some of the information which is desired in connection with a census could be secured efficiently through a sample survey . . . [and] that in some instances a portion of the universe to be included might be efficiently covered on a sample rather than a complete enumeration basis . . . ." Ibid. The House Report spoke in the same terms: "The purpose of section 195 in authorizing the use of sampling procedures is to permit the utilization of something less than a complete enumeration, as implied by the word `census,' when efficient and accurate coverage may be effected through a sample survey." H. R. Rep. No. 1043, 85th Cong., 1st Sess., 10 (1957) (emphasis added); accord, S. Rep. No. 94-1256, p. 1 (1976) (1976 amendments added new language "to direct the Secretary . . . to use sampling and special surveys in lieu of total enumeration in the collection of statistical data whenever feasible" (emphasis added)). The discussion thus linked the authorization—and hence the exception—to sampling as a substitute for a headcount.

Census Bureau practice also helps to support this limited interpretation of the section's scope. Both before and after § 195 was enacted in 1957, the census has used sampling techniques in one capacity or another in connection with its determination of population, most often as a quality check on the headcount itself. See, e. g., Declaration of Margo J. Anderson ¶12, App. in No. 98-404, p. 348 (first postenumeration survey was performed following the 1950 census to check for inaccuracies).

The Census Bureau has also used a form of statistical estimation to adjust or correct its actual headcount. Since at least 1940, the Census Bureau has used an estimation process called "imputation" to fill in gaps in its headcount. U. S. Dept. of Commerce, Bureau of the Census, Report to Congress: The Plan for Census 2000, p. 23 (Aug. 1997) (hereinafter Census 2000 Report). When an enumerator believes a residence is occupied but is unable to obtain any information about how many people live there, the Census Bureau "imputes" that information based upon the demographics of nearby households. Imputation was responsible, for example, for adding 761,000 people to the Nation's total population in 1980 and 53,590 people in 1990. Ibid. In 1970, when the Census Bureau discovered at the last minute that it had mistakenly assumed that a significant number of housing units were vacant, it adjusted the headcount to add 1,068,882 people, or 0.5% of the total population. Ibid.

Integrated Coverage Measurement would not substitute for, but rather would supplement, a traditional headcount, and it would do so to achieve the basic purpose of the statutes that authorize the headcount—namely, accuracy. The Census Bureau has learned over time that certain portions of the population—for example, children, racial and ethnic minorities, and those who rent rather than own their homes—are systematically undercounted in a traditional headcount. Id., at 2-4; see also Wisconsin v. City of New York, 517 U.S. 1, 6-8 (1996). The ICM program is the Census Bureau's effort to correct for this problem. As I understand it, this proposal would use statistical sampling to check headcount results, State by State, by intensively investigating sample blocks in each State, comparing the results from that investigation with the results of the headcount, and using that information to estimate to what extent different groups of persons were undercounted during the headcount. The undercount rates—which will be calculated separately for every State in the Union—will then be used to adjust the headcount totals in an effort to correct for those inaccuracies.

I recognize that the use of statistical sampling to correct or reduce headcount inaccuracies is a complicated matter. An overall national improvement in accuracy does not necessarily tell the whole story. Apportionment demands comparable accuracy State by State. A count that reflected evenly distributed error (say, if the population in every State were undercounted by 20%) would produce the same congressional apportionment as a perfectly accurate count; a count that is less comparatively accurate could make matters worse. Although earlier attempts at ICM-like adjustments apparently failed to take some of these difficulties into account, the Secretary believes the present proposal does so. Census 2000 Report 30 (strata crossed state lines in 1990, but in 2000, strata will be defined on a state-by-state basis); cf. id., at 29 (explaining that the ICM methodology, which was used in the past two censuses to evaluate census quality, has "undergone substantial review and improvement" and "is generally accepted as the most reliable method to improve census results"). And, as I understand it, ICM will help to uncover and to correct undercounting not only among minority but also among majority populations. Any special emphasis the Census Bureau might place on including racial and ethnic minority neighborhoods among its samples would be justified as an effort to ensure proper counts among groups that history shows have been undercounted. Although some amici express concerns about the possibility of error in the execution of the statistical program, the Census Bureau itself, aware of potential difficulties, has created an expert panel of statisticians and social scientists, which will guide the Census Bureau's execution of its plan for the 2000 census, particularly with respect to its use of sampling. See id., at 49-51. And, of course, unadjusted headcounts are also subject to error or bias—the very fact that creates the need for a statistical supplement. See, e. g., id., at 3-4 (describing the problem of differential undercount under the traditional headcount method); id., at 37 (without ICM, the 2000 census will be less accurate than the 1990 census).

Finally, as Justice Stevens points out, Congress has changed the statute considerably since it enacted § 195 in 1957. Each change tends to favor the use of statistical sampling. In 1964, for example, Congress repealed former § 25(c) of the Census Act, see Act of Aug. 31, 1964, 78 Stat. 737, which had required that each enumerator obtain "every item of information" through a personal visit to each household, 68 Stat. 1015, thereby permitting census taking by mail. In 1976, Congress amended § 141(a) ("Population and other census information") to authorize the Secretary to "take a decennial census of population . . . in such form and content as he may determine including the use of sampling procedures and special surveys." At the same time, Congress strengthened § 195's position on sampling, providing that the Secretary "shall" use sampling for purposes other than "for the determination of population for purposes of apportionment." 13 U. S. C. § 195. Given the legal need to interpret subsections of a single statute as creating a single coherent whole, these changes strengthen the case for an interpretation that restricts the scope of § 195 to the kind and use of sampling that called it into being, placing beyond its outer limits a conceptually different (i. e., supplementary) use needed to achieve that statute's basic goal—greater census accuracy.

The Secretary's further proposal, the Nonresponse Followup program, uses statistical sampling not simply to verify a headcount, but to determine the last 10% of population in each census tract. I concede that this kind of statistical "followup" is conceptually similar to the kind of sampling that was before Congress in 1957, in the sense that it involves determining a portion of the total population based upon a sample. But one can consider it supplementary for a different reason—because it simply does not have a great enough impact upon the headcount to be considered a "substitute" falling within § 195's "except" clause.

I note that the Census Bureau has never relied exclusively upon headcounts to determine population. As discussed above, for example, the Census Bureau has supplemented its headcounts with imputation to some degree for at least the last 50 years. Section 195 of the Census Act, at least in my view, could not have been intended as a prohibition so absolute as to stop the Census Bureau from imputing the existence of a living family behind the closed doors of an apparently occupied house, should that family refuse to answer the bell. Similarly, I am not convinced that the Act prevents the use of sampling to ascertain the existence of a certain number of the families that fail to mail back their census forms.

The question, then, is what "number" of housing units will be assigned a population through sampling. Whether the Nonresponse Followup program is sufficiently like imputation in terms of its degree of impact so as to be a supplement to the headcount—or rather whether it is more like the way in which the Bureau uses sampling in connection with the "long form," as a substitute for a headcount—is here a matter of degree, not kind. Is the use of that method in the Nonresponse Followup, limited to the last 10%, sufficiently small, as a portion of the total population, and sufficiently justified, through the need to avoid disproportionately prohibitive costs, that it remains, effectively, a "supplement" to the traditional headcount?

For each census tract (made up of roughly 1,700 housing units), the Nonresponse Followup program will assign population figures to no more than 170 housing units. Census Bureau enumerators will personally visit enough of the housing units in each census tract to ensure that 90% of all housing units have been counted either by mail or in person. The Census Bureau will then use the information gathered from the housing units that the census enumerators actually visited in that tract to arrive at a number for the remaining 10%. See generally Census 2000 Report 26-29. The primary advantage of this program is financial; it is considerably cheaper than a personal search by enumerators to take account of the last few of the households that do not respond by mail. See, e. g., National Research Council, Panel to Evaluate Alternative Census Methods, Counting People in the Information Age 100 (D. Steffey & N. Bradburn eds. 1994). But the Secretary also believes that this program addresses other concerns—concerns related to the immense difficulties involved in personally visiting every home that does not respond by mail—and that, overall, the Nonresponse Followup plan "will increase the accuracy of the census as a whole." Reply Brief for Appellants in No. 98-564, p. 4; see also Census 2000 Report 27; id., at 7 (quoting the National Academy of Sciences Panel on Requirements as concluding that "[i]tis fruitless to continue trying to count every last person with traditional Census methods of physical enumeration").

In answering the question whether this use of sampling remains a "supplement" because of its limited impact on the total headcount, I would give considerable weight to the views of the Secretary, to whom the Act entrusts broad discretionary authority. See 13 U. S. C. § 141(a). The Secretary's decision to draw the line at the last 10%, rather than at the last 5% or 1%, of each census tract's population may well approach the limit of his discretionary authority. But I cannot say that it exceeds that limit. Consequently, I would not set aside the Census Bureau's Nonresponse Followup proposal on this basis.

For these reasons, I respectfully dissent.

Justice Stevens, with whom Justice Souter and Justice Ginsburg join as to Parts I and II, and with whom Justice Breyer joins as to Parts II and III, dissenting.

The Census Act, 13 U. S. C. § 1 et seq., unambiguously authorizes the Secretary of Commerce to use sampling procedures when taking the decennial census. That this authorization is constitutional is equally clear. Moreover, because I am satisfied that at least one of the plaintiffs in each of these cases has standing, I would reverse both District Court judgments.

I

The Census Act, as amended in 1976, contains two provisions that relate to sampling. The first is an unlimited authorization; the second is a limited mandate.

The unlimited authorization is contained in § 141(a). As its text plainly states, that section gives the Secretary of Commerce unqualified authority to use sampling procedures when taking the decennial census, the census used to apportion the House of Representatives. It reads as follows:

The limited mandate is contained in § 195. That section commands the Secretary to use sampling, subject to two limitations: he need not do so when determining the population for apportionment purposes, and he need not do so unless he considers it feasible. The command reads as follows:

Although § 195 does not command the Secretary to use sampling in the determination of population for apportionment purposes, neither does it prohibit such sampling. Not a word in § 195 qualifies the unlimited grant of authority in § 141(a). Even if its text were ambiguous, § 195 should be construed consistently with § 141(a). Moreover, since § 141(a) refers specifically to the decennial census, whereas § 195 refers to the use of sampling in both the mid-decade and the decennial censuses, the former more specific provision would prevail over the latter if there were any conflict between the two. See Edmond v. United States, 520 U.S. 651, 657 (1997). In my judgment, however, the text of both provisions is perfectly clear: They authorize sampling in both the decennial and the mid-decade censuses, but they only command its use when the determination is not for apportionment purposes.

A comparison of the text of these provisions with their predecessors in the 1957 Census Act further demonstrates that in 1976 Congress specifically intended to authorize the use of sampling for the purpose of apportioning the House of Representatives. Prior to 1976, the Census Act contained neither an unlimited authorization to use sampling nor a limited mandate to do so. Instead, the 1957 Act merely provided that the Secretary "may" use sampling for any purpose except apportionment. 13 U. S. C. § 195 (1958 ed.). In other words, it contained a limited authorization that was coextensive with the present limited mandate. The 1976 amendments made two changes, each of which is significant. First, Congress added § 141(a), which unambiguously told the Secretary to take the decennial census "in such form and content as he may determine, including the use of sampling procedures and special surveys." Second, Congress changed § 195 by replacing the word "may" with the word "shall." Both amendments unambiguously endorsed the use of sampling. The amendment to § 141 gave the Secretary authority that he did not previously possess, and the amendment to § 195 changed a limited authorization into a limited command.

The primary purpose of the 1976 enactment was to provide for a mid-decade census to be used for various purposes other than apportionment. Section 141(a), however, is concerned only with the decennial census. The comment in the Senate Report on the new language in § 141(a) states that this provision was intended "to encourage the use of sampling and surveys in the taking of the decennial census." S. Rep. No. 94-1256, p. 4 (1976). Given that there is only one decennial census, and that it is the only census that is used for apportionment purposes, the import of this comment in the Senate Report could not be more clear. See ibid. ("It is for the purpose of apportioning Representatives that the United States Constitution establishes a decennial census of population").

Nevertheless, in an unusual tour de force, the Court concludes that the amendments made no change in the scope of the Secretary's authority: Both before and after 1976, he could use sampling for any census-related purpose, other than apportionment. The plurality finds an omission in the legislative history of the 1976 enactment more probative of congressional intent than either the plain text of the statute itself or the pertinent comment in the Senate Report. For the plurality, it is incredible that such an important change in the law would not be discussed in the floor debates. See ante, at 342-343.1 It appears, however, that even though other provisions of the legislation were controversial,2 no one objected to this change. That the use of sampling has since become a partisan issue sheds no light on the views of the legislators who enacted the authorization to use sampling in 1976.3 Indeed, the bill was reported out of the House Committee by a unanimous vote, both the House and Senate versions easily passed, and the Conference was unanimous in recommending the revised legislation.4 Surely we must presume that the legislators who voted for the bill were familiar with its text as well as the several references to sampling in the Committee Reports.5 Given the general agreement on the proposition that "sampling and surveys" should be encouraged because they can both save money and increase the reliability of the population count, it is not at all surprising that no one objected to what was perceived as an obviously desirable change in the law.6

What is surprising is that the Court's interpretation of the 1976 amendment to § 141 drains it of any meaning.7 If the Court is correct, prior to 1976 the Secretary could have used sampling for any census-related purpose except apportionment, and after 1976 he retained precisely the same authority. Why, one must wonder, did Congress make this textual change in 1976?8 The substantial revision of § 141 cannot fairly be dismissed as "only a subtle change in phraseology." Ante, at 343. Indeed, it "tests the limits of reason to suggest" that this change had no purpose at all. Ibid.

II

Appellees have argued that the reference in Article I of the Constitution to the apportionment of Representatives and to direct taxes on the basis of an "actual Enumeration" precludes the use of sampling procedures to supplement data obtained through more traditional census methods. U. S. Const., Art. I, § 2, cl. 3. There is no merit to their argument.

In 1787, when the Constitution was being drafted, the Framers negotiated the number of Representatives allocated to each State because it was not feasible to conduct a census.9 See Department of Commerce v. Montana, 503 U.S. 442, 448, and n. 15 (1992). They provided, however, that an "actual Enumeration shall be made within three Years after the first Meeting of the Congress of the United States, and within every subsequent Term of ten Years, in such Manner as they shall by Law direct." U. S. Const., Art. I, § 2, cl. 3. The paramount constitutional principle codified in this Clause was the rule of periodic reapportionment by means of a decennial census. The words "actual Enumeration" require post-1787 apportionments to be based on actual population counts, rather than mere speculation or bare estimate, but they do not purport to limit the authority of Congress to direct the "Manner" in which such counts should be made.

The July 1787 debate over future reapportionment of seats in the House of Representatives did not include any dispute about proposed methods of determining the population. Rather, the key questions were whether the rule of reapportionment would be constitutionally fixed and whether subsequent allocations of seats would be based on population or property. See 1 Records of the Federal Convention of 1787, pp. 57-71, 542, 559-562, 566-570, 578-579, 579-580, 586, 594 (M. Farrand ed. 1911); see also Declaration of Jack N. Rakove, App. 387 ("What was at issue . . . were fundamental principles of representation itself . . . not the secondary matter of exactly how census data was to be compiled"); J. Rakove, Original Meanings: Politics and Ideas in the Making of the Constitution 70-74 (1996). The Committee of Style, charged with delivering a polished final version of the Constitution, added the term "actual Enumeration" to the draft reported to the Convention on September 12, 1787—five days before adjournment. 2 Records, supra, at 590-591. This stylistic change did not limit Congress' authority to determine the "Manner" of conducting the census.

The census is intended to serve "the constitutional goal of equal representation." Franklin v. Massachusetts, 505 U.S. 788, 804 (1992). That goal is best served by the use of a "Manner" that is most likely to be complete and accurate. As we repeatedly emphasized in our recent decision in Wisconsin v. City of New York, 517 U.S. 1, 3 (1996), our construction of that authorization must respect "the wide discretion bestowed by the Constitution upon Congress." Methodological improvements have been employed to ease the administrative burden of the census and increase the accuracy of the data collected. The "mailout-mailback" procedure now considered a traditional method of enumeration was itself an innovation of the 1970 census.10 Requiring a face-to-face headcount would yield absurd results: For example, enumerators unable to gain entry to a large and clearly occupied apartment complex would be required to note zero occupants. For this reason, the 1970 census introduced the Postal Vacancy Check—a form of sampling not challenged here—which uses sample households to impute population figures that have been designated vacant but appear to be occupied.11 Since it is perfectly clear that the use of sampling will make the census more accurate than an admittedly futile attempt to count every individual by personal inspection, interview, or written interrogatory, the proposed method is a legitimate means of making the "actual Enumeration" that the Constitution commands.

III

I agree with the Court's discussion of the standing of the plaintiffs in No. 98-564. I am also convinced that the House of Representatives has standing to challenge the validity of the process that will determine the size of each State's congressional delegation. See Powell v. McCormack, 395 U.S. 486, 548 (1969) ("Unquestionably, Congress has an interest in preserving its institutional integrity"). As the District Court in No. 98-404 correctly held, the House has a concrete and particularized "institutional interest in preventing its unlawful composition" that satisfies the injury in fact requirement of Article III. 11 F.Supp.2d 76, 86 (DC 1998). Accordingly, I respectfully dissent in both cases. I would reverse both judgments on the merits.

Justice Ginsburg, with whom Justice Souter joins, dissenting.

I agree with the Court that Indiana resident Hofmeister, an appellee in No. 98-564, has standing to challenge the Census 2000 plan on the ground that Indiana would lose a Representative in Congress under the Census Bureau's proposed sampling plan. I also agree with the Court's conclusion that the appeal in No. 98-404 should be dismissed. I would not decide whether other appellees in No. 98-564 have established standing on the basis of the expected effects of the sampling plan on intrastate redistricting. Respecting the merits, I join Parts I and II of Justice Stevens' dissent.