Carter v. United States Department of Commerce

186 F. Supp. 2d 1147, 2001 U.S. Dist. LEXIS 23664

• Court: District Court, D. Oregon
• Decided: November 20, 2001
• Docket: 3:01-cv-00868
• Status: Published

Opinion

OPINION AND ORDER

REDDEN, District Judge.

This is an action under the Freedom of Information Act, 5 U.S.C. § 552 (FOIA) by two Oregon legislators against the United States Department of Commerce (the Department). Plaintiffs challenge the denial of their request for statistically adjusted data from the 2000 census (adjusted data). The adjusted data were withheld on the basis of Exemption 5 of the FOIA. Plaintiffs have filed a motion for summary judgment. The court heard oral argument on the motion October 18, 2001. The Department requested an opportunity to file supplemental briefing, and the motion was granted. The court heard additional oral argument on November 8, 2001. Plaintiffs’ motion for summary judgment is granted.

Factual Background

The census in general and the problem of undercounting

The Census Clause of the United States Constitution requires Congress to conduct a census every ten years to apportion congressional representation among the states. U.S. Const, art. I, § 2, cl. 3. Through the Census Act, 13 U.S.C. § 1 et seq., Congress delegated this responsibility to the Secretary of Commerce, authorizing him to conduct the census “in such form and content as he may determine, including the use of sampling procedures and special surveys.” 13 U.S.C. § 141(a). The Secretary is assisted by the Bureau of the Census (the Bureau).

It has long been recognized that census data are inaccurate. See Wisconsin v. City of New York, 517 U.S. 1, 6, 116 S.Ct. 1091, 134 L.Ed.2d 167 (1996) and cases cited therein. “Persons who should have been counted are not counted at all or are counted at the wrong location; persons who should not have been counted (whether because they died before or were born after the decennial census date, because they were not residents of the country, or because they did not exist) are counted; and persons who should have been counted only once are counted twice.” Id. It is thought that these inaccuracies have resulted in a net “undercount” of the actual population in every decennial census. Id.

The undercount is “not thought to be spread consistently across the population: some segments of the population are ‘un- dercounted’ to a greater degree than are others, resulting in a phenomenon termed the ‘differential undercount.’” Id. at 7, 116 S.Ct. 1091. The undercounted groups include certain minorities, children, and renters. Department of Commerce v. United States House of Representatives, 525 U.S. 316, 322-23, 119 S.Ct. 765, 142 L.Ed.2d 797 (1999), citing U.S. Dept. of Commerce, Bureau of the Census, Report to Congress: The Plan for Census 2000, at 3-4 (Aug.1997). According to the Secretary of Commerce, the official 1990 census counted about 98% of the people living in the United States. Assembly of the State of California v. United States Department of Commerce, 797 F.Supp. 1554, 1556 (E.D.Cal.1992)(Assembly I), aff'd, 968 F.2d 916 (9th Cir.1992)(Assembly II) citing Decision of the Secretary of Commerce on Whether a Statistical Adjustment of the 1990 Census of Population and Housing Should be Made for Coverage Deficiencies Resulting in Overcount or Undercount of the Population, 56 Fed.Reg. 33582 (1991). However, “census participation and coverage was lower than average among certain segments of the population.” Id. Based on the Department of Commerce’s estimates, blacks appeared to have been undercount-ed in the 1990 census by 4.8%, Hispanies by 5.2%, Asian Pacific Islanders by 3.1%, and American Indians by 5.9%. Id.

One effect of undercounting is that un-dercounted communities fail to receive their proportionate share of government and private grants based on census data. According to the United States General Accounting Office (GAO), the federal government allocated $185 billion of grants to states and localities based on census data. GAO, Formula Grants: Effects of Adjusted Popttlation Counts on Federal Funding to States, GAO/HEHS-99-69 at 1-2 (Feb. 1999), attached as Exhibit B to Declaration of Ashley N. Bailey (Bailey Declaration). However, the GAO also found that as a direct consequence of the 1990 undercount, states and localities with the largest un-dercount failed to receive approximately $449 million from 15 federal programs. Id. at 4. So, for example, according to the GAO correction of the undercount in Oregon would have resulted in approximately 24% more funds from those federal programs. Id.

Since the 1980 census, the Bureau has studied the efficacy of using statistical sampling to adjust the results of the “actual headcount” or initial enumeration (also referred to as “unadjusted data”), which is based primarily on information obtained through mail-in surveys and personal visits. (Although the initial enumeration itself also involves some statistical adjustment or “imputation,” derived from demographic characteristics of neighboring households, when census takers are unable to obtain information about a household. Approximately 1.3% of the initial enumeration for 2000 was based on imputations. See Hogan, YL.rA.ccuracy and Coverage Evaluation: Data and Analysis to Inform the ESCAP Report at 21-22, attached as Exhibit M to Bailey Declaration.)

In preparation for the 1990 census, the Bureau studied the feasibility of using large-scale statistical adjustment in a capture-recapture methodology to compensate for the undercount and the differential un-dercount. City of New York, 517 U.S. at 8, 116 S.Ct. 1091. The Supreme Court has explained this methodology:

the initial enumeration of the entire population (the “capture”) would be followed by the post-enumeration survey (PESXthe “recapture”) of certain representative geographical areas. The Bureau would then compare the results of the PES to the results of the initial enumeration for those areas targeted by the PES, in order to determine a rate of error in those areas for the initial enu meration (i.e., the rate at which the initial enumeration undercounted people in those areas). That rate of error would be extrapolated to the entire population, and thus would be used to statistically adjust the results of the initial enumeration.

517 U.S. at 9, 116 S.Ct. 1091. This process was completed for each of the almost five million inhabited blocks in the United States. Assembly I, 797 F.Supp. at 1556. Department officials then undertook a large-scale intensive evaluation of the accuracy of the adjusted estimates. Id. at 1557. Experts differed on whether the adjusted census figures were more reliable than the initial enumeration figures. Id. On July 15, 1991, the Secretary announced his decision that the initial enumeration should not be changed by statistical adjustment. Id. In reaching that decision, the Secretary determined that the adjusted estimates did not accurately distribute the United States population across and within the states. Id.

The 2000 census

For the initial enumeration in the 2000 census the Bureau divided the country into seven million blocks of citizens or households, and attempted to obtain information about every household through mail-in surveys and personal visits. If a household did not return a survey, enumerators contacted either a household member or another person, such as a neighbor, who could provide information about the household’s residents; failing this, the Department used imputation. This was the means by which the Bureau obtained its unadjusted data.

To obtain adjusted data, the Bureau undertook a capture-recapture sampling methodology called Accuracy and Coverage Evaluation (ACE). It was ultimately decided that a sample constituting less than 1/4 of 1% of the population (314,000 households) would supply the necessary data.

Individuals in both the initial enumeration and the ACE sample were categorized by demographers and mathematical statisticians into 448 groupings called post-strata, based on demographic and geographic variables such as age, race, gender, and tenure (homeowner or renter). The Bureau compared the results of the ACE sample to the results from the corresponding portions of the initial enumeration to create “coverage correction factors.” The coverage correction factors were designed to determine the amount by which the results of the initial enumeration for each post-stratum would undercount or overcount. By applying the coverage correction factors to the initial enumeration total for each post-stratum, on a block by block basis, the Bureau arrived at estimates for each demographic group or post-stratum in the United States. The totals for each group were totaled to arrive at adjusted population totals for every block in the United States. The adjusted block level data were then tabulated to generate adjusted data at the national, state and local levels. On March 15, 2001, the Bureau’s Acting Director, William Barron, confirmed that the Bureau had produced 2000 adjusted population totals for the entire country and the states.

Only unadjusted data can be used for apportioning seats in the United States House of Representatives. 13 U.S.C. § 195; Department of Commerce, 525 U.S. at 340, 119 S.Ct. 765 (holding that 13 U.S.C. § 195 precluded the use of sampling to apportion Representatives, but did not bar the use of sampling for other purposes). However, for purposes of intrastate redistricting and the allocation of state and federal funds, including federal grants to states and localities, the Secretary has discretion to release either the initial enumeration totals (i.e., the unadjusted data), or the numbers that have been adjusted on the basis of sampling procedures (the adjusted data).

In March 2000, Acting Director Barron, joined by a Bureau committee that had been formed to evaluate the accuracy of the adjusted numbers called Executive Steering Committee for Accuracy and Coverage Evaluation Policy (ESCAP), recommended release of the unadjusted data as the official state redistricting data. The basis for the recommendation to use the unadjusted data was unresolved discrepancies between the ACE totals and those produced by the Demographic Analysis, as well as other technical concerns. The Demographic Analysis uses records of births, deaths, documented immigration, and Medicare enrollment, and estimates of emigration and undocumented immigration, as an independent benchmark to validate the accuracy of both unadjusted and adjusted data. The Demographic Analysis indicated that the initial enumeration for Census 2000 contained an overcount of the national population by 1.8 million people, while the ACE found a net undercount of 3.3 million people.

Nevertheless, ESCAP reported that 1) further research was likely to establish that adjustment based on the ACE would result in improved accuracy; 2) the ACE was an efficient and effective operation that produced high quality data; 3) under what ESCAP considered “reasonable assumptions,” state, congressional district, and county level analyses (except for counties with fewer than 100,000 people) showed a marked improvement for adjustment; and 4) given more time to analyze the results, ESCAP could “well have reached a different recommendation,” and urged release of the adjusted numbers. The ESCAP report also stated that “the majority of the evidence indicates both the continued existence of a differential under-count of the population and the superior accuracy of the adjusted numbers.” Report of the Executive Steering Committee for Accuracy and Coverage Evaluation Policy: Recommendation Concerning the Methodology to be Used in Producing the Tabulations of Population Reported to States and Localities Pursuant to IS U.S.C. § 14.1(c) at i, iii, 9 and Recommendation (March 1, 2001) (ESCAP Report).

The Bureau maintains a website at “www.census.gov” at which it has released the ESCAP recommendations and other information about the 2000 census and the ACE. The full text of the ESCAP Report, with attachments, has also been released to the public. The information released includes detailed and technical information about the adjusted data, including specifications for the ACE methodology, quality indicators for both the initial enumeration and ACE, factors affecting the accuracy of the initial enumeration and ACE, equations used to measure error in the initial enumeration and the ACE, the criteria upon which the decision to use unadjusted data were based, undercount percentages for each race category, males and females, renters and homeowners, and various age groups, the percentage undercounted as measured by the Demographic Analysis, synthetic estimation (carrying down within the post-strata the estimated coverage correction factors for each of the post-strata), “other issues” regarding census adjustment which did not concern the ESCAP, and minutes of the ESCAP meetings.

On March 6, 2001, Secretary Donald Evans announced his decision to release the unadjusted data as the official 2000 state redistricting data. Although the Secretary reviewed ESCAP’s recommendations about the accuracy of the adjusted data, he did not review the adjusted block-level data before making his decision.

On October 17, 2001, the Bureau announced that unadjusted data would also be used to apportion funds for federal programs. The adjusted data have not been released to the public.

Plaintiffs make their FOIA request

On April 20, 2001, the plaintiffs made a request under the FOIA for the adjusted data in the form of computer data files. The plaintiffs’ specific request was “the 2000 census data required under 13 U.S.C. § 141(c), adjusted using the ACE ... for each of the jurisdictions covered by the entire 2000 Decennial Census.” On May 7, 2001, the Bureau denied the request, asserting that the information was protected under the “deliberative process” privilege of Exemption 5 of the FOIA.

Plaintiffs appealed to the Department’s Assistant General Counsel for Administration on May 9, 2001. On June 8, 2001, the Office of General Counsel upheld the Bureau’s decision, again relying on the “deliberative process” exemption.

Standards

The district court’s review of the agency’s decision is de novo. 5 U.S.C. § 552(a)(4)(B).

Summary judgment is appropriate “if the pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any material fact and that the moving party is entitled to a judgment as a matter of law.” Fed.R.Civ.P. 56(c). At summary judgment, the court views the evidence in the light most favorable to the nonmoving party: if direct evidence produced by the moving party conflicts with direct evidence produced by the nonmoving party, the court assumes the truth of the evidence set forth by the nonmoving party with respect to that fact. T.W. Electrical Service, Inc. v. Pacific Electrical Contractors Assoc., 809 F.2d 626, 630 (9th Cir.1987).

Discussion

The FOIA and Exemption 5

The FOIA “mandates a policy of broad disclosure of government documents.” Church of Scientology v. Dep’t of the Army, 611 F.2d 738, 741 (9th Cir.1979); Maricopa Audubon Society v. United States Forest Service, 108 F.3d 1082, 1085 (9th Cir.1997). When a FOIA request is made, an agency may withhold a document, or portions of a document, only if the material falls within one of the nine statutory exemptions found in § 552(b). Maricopa Audubon Society, 108 F.3d at 1085. These exemptions are to be narrowly construed “in light of FOIA’s dominant objective of disclosure, not secrecy.” Dept. of the Air Force v. Rose, 425 U.S. 352, 361, 96 S.Ct. 1592, 48 L.Ed.2d 11 (1976); see also Weatherhead v. United States, 157 F.3d 735, 738 (9th Cir.1998) and S.Rep. No. 813, 89th Cong., 1st Sess. 9 (1965)(exemption 5 to be applied “as narrowly as consistent with efficient Government operation.”) The FOIA imposes on agencies the burden of proving that withheld materials are exempt from disclosure. 5 U.S.C. § 552(a)(4)(B); Maricopa Audubon Society, 108 F.3d at 1085.

Exemption 5 permits nondisclosure of “inter-agency or intra-agency memorandums or letters which would not be available by law to a party other than an agency in litigation with that agency.” 5 U.S.C. § 552(b)(5). This provision shields “those documents, and only those documents, normally privileged in the civil discovery context.” NLRB v. Sears, Roebuck 6 Co., 421 U.S. 132, 149, 95 S.Ct. 1504, 44 L.Ed.2d 29 (1975). Exemption 5 thus covers the attorney-client privilege, the attorney work-product privilege, and the executive “deliberative process” privilege on which the Department relies here. Maricopa Audubon Society v. United States Forest Service, 108 F.3d 1089, 1092 (9th Cir.1997).

The purpose of the deliberative process privilege is to allow agencies “freely to explore possibilities, engage in internal debates, or play devil’s advocate without fear of public scrutiny.” Assembly II, 968 F.2d at 920. The deliberative process exemption reflects Congress’s “concern that if agencies were forced to operate in a fishbowl, candid exchange of ideas within an agency would cease and the quality of decisions would suffer.” Id. (internal quotations and citations omitted).

To qualify for exemption 5 under the “deliberative process” privilege, a document must be both predecisional and deliberative. Assembly II, 968 F.2d at 920. These twin requirements “recognize that the underlying purpose of this privilege is to protect the consultative functions of government by maintaining the confidentiality of advisory opinions, recommendations, and deliberations comprising part of a process by which governmental decisions and policies are formulated”—ie., the “give and take that occurs among agency members in the formulation of policy.” National Wildlife Federation v. United States Forest Service, 861 F.2d 1114, 1117 (9th Cir.1988)

Are the adjusted data predecisional?

The Department asserts that an agency record is “predecisional” if it was created antecedent to the adoption of the agency’s policy, citing National Wildlife, 861 F.2d at 1117. It argues that the adjusted data satisfy that standard because the data were generated before the Secretary’s March 6, 2001 decision to use the initial enumeration as the official figures for Census 2000.

I disagree with the Department that the predecisional requirement is merely chronological. A predecisional document is one “prepared in order to assist an agency decisionmaker in arriving at his decision.” Assembly II, 968 F.2d at 920, quoting Renegotiation Board v. Grumman Aircraft Engineering Corp., 421 U.S. 168, 184, 95 S.Ct. 1491, 44 L.Ed.2d 57; Maricopa Audubon Society, 108 F.3d at 1094. Such documents include “recommendations, draft documents, proposals, suggestions, and other subjective documents which reflect the personal opinions of the writer rather than the policy of the agency.” Assembly II, 968 F.2d at 920 (internal quotations and citation omitted). “Material which predates a decision chronologically, but did not contribute to that decision, is not predecisional in any meaningful sense.” Id. at 921.

The Department contends that the adjusted data did, as a factual matter, “contribute” to the deliberative process because the adjusted data were created for the precise purpose of allowing the decision makers to be able to adopt either the unadjusted or the adjusted data. I do not find this argument persuasive. The fact that the adjusted data were the subject of the Secretary’s decision—ie., one of the alternatives in an either/or choice—does not satisfy the Department’s burden of proving that the adjusted data was part of the process which led to the final decision on the issue of whether to use unadjusted or adjusted data.

Further, in this case, as in Assembly I, the Department has acknowledged that the Secretary did not review the adjusted data before making his decision. See, e.g., Defendant’s Response to Plaintiffs’ Motion for Summary Judgment at 12: “The Secretary relied upon, among other things, the ESCAP recommendation, the Acting Director’s concurrence, and the recommendation of ... non-Government statisticians and demographers. Although the Secretary reviewed these recommendations regarding the accuracy of ‘adjusted’ data, the Secretary did not review the adjusted block-level data.” Material that is not examined by the decision maker cannot be said to have contributed to the decision making process.

The Department also argues that the adjusted data are predecisional because the ESCAP Report makes it clear that the decision to use unadjusted data was “driven by unexplained inconsistencies in the [adjusted] data themselves.” Defendant’s Supplemental Response at 3 (emphasis in original). However, as plaintiffs point out, the determination of whether the adjusted data are predecisional cannot turn on “what reason the Department chose to articulate for not adopting the adjusted data.” Plaintiffs’ Supplemental Reply Memorandum at 3. Moreover, the Department has stated elsewhere in its motion papers that the Secretary’s decision was driven, not by inconsistencies contained within the adjusted data, but rather by inconsistencies between the adjusted data and the Demographic Analysis, and by technical concerns about the ACE methodology used to generate the adjusted data. See Defendant’s Response at 10, 11. Here, as in Assembly I, the decision whether to use unadjusted or adjusted data turned on the reliability of the statistical factors applied to the original count to arrive at the adjusted data — not the adjusted numbers themselves. See 797 F.Supp. at 1563.

The Department also argues that the adjusted data are predecisional because the underlying statistical and demographic methodology — the ACE — is actively under consideration for future uses other than the official census. The Department cites language from the Supreme Court’s Sears Roebuck decision to the effect that “memo-randa containing recommendations which do not ripen into agency decisions” may also be considered predecisional documents. 421 U.S. at 151 n. 18, 95 S.Ct. 1504. The force of this argument is vitiated by the Secretary’s October 2001 decision not to use the adjusted data for allocation of federal funding. Moreover, the same argument was rejected in Assembly II, where the court said, “Any memorandum always will be ‘predecisional’ if referenced to a decision that possibly may be made at some undisclosed time in the future .... Characterizing these documents as ‘predecisional’ simply because they play into an ongoing audit process would be a serious warping of the meaning of the word.” 968 F.2d at 921 (internal quotations and citation omitted).

The Department argues that the Assembly II case is distinguishable because there the Court of Appeals discounted a generalized assertion that the adjusted data may be used in the future, while in this case the Bureau did, in fact, “continue to carefully review the adjusted estimates after the initial adjustment decision.” Defendant’s Supplemental Response at 5. I do not find this factual difference, if it is one, to be legally significant now that the Secretary has made the decision not to use the adjusted data for other purposes.

Are the adjusted data deliberative?

While the “predecisional” label focuses on the role of the entire document in the decision making process, the deliberative process criterion may be applied to distinguishing between privileged and nonprivi-leged material within a single predecisional document. Assembly I, 797 F.Supp. at 1565. A predecisional document is deliberative if disclosure of the document would “expose an agency’s decisionmaking process in such a way as to discourage candid discussion within the agency and thereby undermine the agency’s ability to perform its functions.” Id. (internal quotations and citation omitted).

In Assembly II, the court noted that the earliest cases to examine the deliberative process privilege had contrasted “factual” and “deliberative” materials, the distinction being that agencies had “no legitimate interest in keeping the public ignorant of the facts the agencies worked from, while they did have a legitimate interest in shielding their preliminary opinions and explorations.” 968 F.2d at 921. However, this distinction was found wanting in some cases, and the key inquiry became “whether revealing the information exposes the deliberative process.” Id. (internal citation omitted). “The factual/dehberative distinction survives, but only as a useful rule of thumb when such a separation is feasible.” Id.

The adjusted data at issue in this case are clearly not suggestions, recommendations, or advice; they are not subjective or personal, and they do not reflect “agency give-and-take” by which a decision was made. The district court in Assembly I found that the requested tapes were factual.

The numerical data contained on these tapes, like the official census data, provides a list of numbers of people, broken down by race, associated with census tracts. The court finds the material is purely factual and in no way divulges the reasoning process through which the data was derived or in any way explains any recommendation or decision not to adjust the census.

797 F.Supp. at 1566.

The Department argues that ideas which are expressed in the form of numbers are not necessarily facts, relying on Quarles v. Department of the Navy, 893 F.2d 390, 392-93 (D.C.Cir.1990). In Quarles, the Navy initiated a search for “homeports” for 130 ships that it proposed to add to the United States fleet. One battleship group was slated for the Gulf of Mexico. The Navy initially planned to deploy the Gulf Coast group from a single harbor, and convened a special study team, composed of representatives of various command and technical units, to evaluate seven finalist sites on operational, logistic, environmental and other criteria. The team’s final report included the material at issue in the case— cost estimates for each site, including the costs of land, ship berthing, dredging, buildings and facilities, and utilities.

After completion of the study, the Secretary decided not to designate a single homeport but rather to assign the vessels among different harbors. Pursuant to a FOIA request, the Navy produced “truly factual information” such as the composition of the battleship task force, physical dimensions of the ships, personnel complement, and photographs of the proposed ports, but excised the rest — analysis, conclusions and cost estimates.

The court concluded that the military cost estimates, although numerical, were protected from disclosure because they derived from a “complex set of judgments,”, and partook of “just that elasticity that has persuaded courts to provide shelter for opinions generally.” 893 F.2d at 392-93.

The Court of Appeals discussed the Quarles opinion in Assembly II, and agreed that numbers could sometimes derive from a complex set of judgments and demonstrate the elasticity of opinions. However, the Court of Appeals agreed with the district court that the adjusted data fell closer to fact “along the continuum of deliberation and fact,” and would not “reveal the agency’s protectable thought processes.” Assembly II, 968 F.2d at 922. The Court of Appeals also noted, “As is typical for factual matter, release of the adjusted block-level data would not enable the public to reconstruct any of the protected deliberative process,” because, first, it was not possible to recreate the adjusted data solely from the raw data and the adjusted formulas and, second, the adjusted data would not reveal anything about “the most sensitive decision the Secretary had to make, namely which set of figures (adjusted or unadjusted) should be adopted as the official United States census.” Id. The court concluded, “That decision involved the Secretary’s judgment as to which figures best estimated the actual population of the United States. The bare numbers reveal nothing about the process informing that judgment.” Id.

The Department argues that the adjusted data are “the product of’ the “subjective and enormously complex statistical operations used to generate the estimates.” Defendant’s Memorandum at 23. According to the declaration of John Thompson, Principal Associate Director for Programs at the Bureau, the adjusted data are not “raw numbers,” but rather a “statistical construct that embodies a host of calculations, assumptions, and hypotheses analyzed by professional employees,” such as “mathematicians, statisticians, demographers and experts,” and that as a result, the adjusted data are necessarily both “imprecise” and “controversial.” Id. at 23-24.

This same argument was made to the district court in Assembly I, where the Department had asserted that the 1990 adjusted numbers were “not raw data but a statistical construct that embodies a host of calculations, assumptions and hypotheses.” 797 F.Supp. at 1566. In that case, as here, the Department contended that the adjusted data were “completely different from the unadjusted figures because they are created rather than discovered.” Id.

The district court rejected that argument in Assembly I, and I reject it here for the same reasons. As the court pointed out in Assembly I, the adjusted data are “no more numerical recommendations than the official census data which defendant contrasts as an example of purely factual and diselosable data. Those numbers too are elastic and far from fixed.” Id. The factual record in this case, like the one in Assembly I, reveals that 1.3% of the initial enumeration is based on imputations and information obtained from third parties such as relatives and neighbors, while the Demographic Analysis is also the result of projections. As the court said in Assembly I,

In the court’s view, the two processes are not so fundamentally different as to render the product of one “classic, objective, non-deliberative material that must be disclosed under FOIA” and the product of the other deliberative material that must not be disclosed [citation omitted]. Both methods share common shortcomings, inadequacies and traits, and yield similar numerical data prepared for distribution to the states in precisely the same format.

797 F.Supp. at 1567. Further, as the court noted in Assembly I, even if the adjusted data contained numerical estimates fundamentally different from the initial enumeration, release of the adjusted data would not “reveal the issues” the Department considers important or “provide telling clues” as to the Department’s proposed course of action. Id.,quoting National Wildlife Federation, 861 F.2d at 1121.

It is noteworthy here, as it was in the two Assembly cases, that the adjusted data were prepared in anticipation of, and intended for, public release if the Secretary had chosen to adjust the initial enumeration. Moreover, the Department has already disclosed a significant amount of the calculations, assumptions, hypotheses, equations, analysis, and discussion relevant to the adjusted data. The Department argues that this case differs from the Assembly cases because the ESCAP report confirms that in this case the Department has not already released all of the signifi cant adjusted 2000 population estimates, while in the Assembly cases, the Department had released not only national adjusted data, but also adjusted data for state, city and county levels and adjusted block-level estimates for half of the blocks. The Department argues that this difference is “critical” because national adjusted estimates, even if accurate, cannot determine the reliability of adjusted estimates at these “smaller and more significant units of aggregation.” Thus, the Department argues, the “compelled release of these estimates would reveal additional and politically sensitive information about an adjustment option that the Secretary has specifically rejected, and about a post-censal opinion that ESCAP has specifically recommended against.”

Like the district court in Assembly /, I conclude that “disclosing the adjusted census figures would not reveal anything more about the deliberative process than has already been disclosed” by the Department. See 797 F.Supp. at 1567. While release of the adjusted data may subject the Department to some political controversy over the decision that was made, or generate some additional departmental embarrassment over numerical errors which have already been publicly disclosed, these are the very consequences which the FOIA intends: robust political debate, agency accountability, and public disclosure rather than secrecy.

For similar reasons, I reject the Department’s argument that disclosure of the adjusted data would make “candid assessments far more difficult” and that, accordingly, the “chilling effects” from the compelled release of the adjusted data would hamper the Department’s “efforts to assess competing population estimates and the underlying statistical methodologies, or other new and potentially innovative processes.” Defendant’s Response at 28. As plaintiffs point out, the compelled disclosure of the 1990 adjusted data in 1992 does not seem to have had a detrimental effect on the Department’s efforts to construct, implement and assess the viability of ACE.

The Department takes the position that Assembly II was wrongly decided or distinguishable on the facts. I find Assembly II both controlling and compelling, and that it is not distinguishable.

Conclusion

In view of the narrow scope of exemption 5 and the strong policy of public disclosure embodied in the FOIA, I conclude that the Department has not sustained its burden of proving that the data requested by the plaintiffs are protected from disclosure. The adjusted data was prepared in anticipation of possible public dissemination, did not contribute to the deliberations which culminated in the Department’s decision to use unadjusted data, and contain factual information which reveals nothing about the subjective thought processes involved in deciding whether to release unadjusted or adjusted data. The data sought are numbers. It may be that a deliberative process led to the methodology which generated the numbers, but the numbers are the result of the deliberative process. They are not the process.

Plaintiffs’ motion for summary judgment (doc. # 5) is GRANTED

IT IS SO ORDERED.