Detailed Information on the
Census Bureau: Decennial Census Assessment

Continue to examine all key cost factors to identify potential areas for savings.

Ensure there is adequate oversight of contractors developing critical 2010 IT systems.

Measure: Meet constitutional and legislative mandates by implementing a re-engineered 2010 Census that is cost-effective, provides more timely data, improves coverage accuracy and reduces operational risk.

Explanation:This encompasses all three of the strategic outcome measures for the effort to reengineer the 2010 Decennial Census Program. The outcome depends on successful completion of all three program components--ACS, MTEP, and the 2010 Census.

Measure: Percent of Weigted response rate for annual implementation of the ACS.

Explanation:Response rates influence the quality of annual data, and thus can be used to assess the ability of ACS to produce annual data that replaces decennial-long form data.

Measure: Number of counties with improved GPS location (the information shown is the annual number of counties updated for that fiscal year).

Explanation:All 3,232 counties will be complete by FY 2008. The FY 2009 target is to enable geographic partners to update/correct TIGER features via the Internet.

Measure: Percent of Census Test Objectives achieved

Explanation:Annual performance measures and targets will evolve over the decade as the Census Bureau implements a multi-year research, testing, and development program for a short-form 2010 Census.

Measure: ACS Cost per Household.

Explanation:Unit cost for data collection per household. Data in targets and actual results section reflects unit costs for 4 modes of data collection, in the following order: Mail, Phone, Personal Visit, and Group Quarters (Group Quarters from FY 2008)

Is the program purpose clear?

Explanation: The goal of the decennial census is to provide comprehensive and useful demographic information about all people living in the United States, including population information used to apportion the seats in the Congress and define the districts that each member will represent. The decennial census is constitutionally mandated and the sole legal source of data for Congressional apportionment. In addition, Federal law requires that decennial census data be used for Congressional redistricting, and other Federal laws require decennial census data be used for program fund allocations by many agencies.

Evidence: Constitution of the United States (Article 1, Section 2), Public Law 94-171, Title 13, U.S. Code, Various Federal laws that mandate use of decennial census data for fund allocations. These needs also help determine the questions to be asked by the decennial census.

Does the program address a specific and existing problem, interest, or need?

Explanation: The decennial census provides the official count of the U.S. population used for apportionment and redistricting of Congressional seats. Data from the census are used, by law, in the administration of federal programs such as the Commission of Civil Rights, Environmental Protection Agency (EPA), and the Departments of Justice, Education, Health and Human Services, Housing and Urban Development, Labor, and Veterans Affairs. The Decennial Census also provides comprehensive and unique demographic information about all people in the United States, Puerto Rico, and the Island Areas.

Evidence: Constitution of the United States (Article 1, Section 2), Public Law 94-171, Title 13, U.S. Code, Various Federal laws that mandate use of decennial census data for fund allocations.

Is the program designed so that it is not redundant or duplicative of any other Federal, state, local or private effort?

Explanation: The decennial program collects and disseminates data mandated by the U.S. Constitution and U.S. Code for use in apportionment and redistricting. There is no other source for these data. In addition, based on a variety of laws, approximately $283 billion in Federal dollars are distributed to states and localities each year based on Decennial Census.

Evidence: Constitution of the United States (Article 1, Section 2), Public Law 94-171, Title 13, U.S. Code, Various Federal laws that mandate use of decennial census data for fund allocations

Is the program design free of major flaws that would limit the program's effectiveness or efficiency?

Explanation: In general, Census 2000 was a success, with operations completed on time, while achieving the lowest net coverage error of recent censuses. However, in order to improve the design for 2010 (by reducing risks, controlling costs, producing more timely data, and improving coverage), the Census Bureau is in the midst of a multi-year strategic effort to reengineer the 2010 Census program. That effort includes three components: the American Community Survey (ACS) to replace the decennial long-form, the MAF/TIGER Enhancements Program (MTEP, to align Census geographic data with GPS), and the short-form only 2010 Census testing and planning. The ACS is fully operational and is now releasing data annually. This demonstrates improved program effectiveness to the data users who previously received decennial long-form data once a decade. MTEP is expected to be completed on schedule and within budget. Testing in 2003, 2004, 2005, and 2006 have revealed no fatal flaws in the Census Bureau's 2010 plan. The Dress Rehearsal in 2008 will allow for fine tuning and one last opportunity to find and fix serious flaws before going live in 2010. This reduction in risk for the Census was a major strategic goal for 2010, and the basis for the multi-year research, development, and testing program.

Evidence: Census Bureau Strategic Plan, Exhibit 300s for MTEP, ACS, and the 2010 Census (available at www.commerce.gov/exhibit300/)

Is the program design effectively targeted so that resources will address the program's purpose directly and will reach intended beneficiaries?

Explanation: Overall, the decennial census is not a targeted program, in that the goal is to enumerate the entire U.S. population, both for apportionment and also to help Federal agencies target funding to intended beneficiaries. However, to achieve that goal, resources and activities are targeted to meet specific operational difficulties in counting historically harder to enumerate populations and areas. Among the 2010 Census's long-term goals is to improve the accuracy of Census coverage. The Census Bureau devoted significant efforts to reduce both overcounts and undercounts of different population groups. Coverage improvement research has studied coverage probes, residence rules, and clarification of instructions on the questionnaires.

Evidence: Constitution of the United States (Article 1, Section 2), Public Law 94-171, Title 13, U.S. Code

Does the program have a limited number of specific long-term performance measures that focus on outcomes and meaningfully reflect the purpose of the program?

Explanation: The reengineered 2010 Decennial Census Program is a multi-year effort with four long-term strategic performance outcomes. These are documented in the Census Bureau's Strategic Plan, under Strategic Objective 1: "Meet constitutional and legislative mandates by implementing a reengineered 2010 Census that is cost-effective, provides more timely data, improves coverage accuracy, and reduces operational risks." The specific long-term measures are 1) Improve the relevance and timeliness of census long-form data; 2) Reduce operational risk; 3) Improve the accuracy of census coverage; and 4) Contain costs.

Evidence: FY 2007 Annual Performance Plan, FY 2008 Annual Performance Plan, Census Bureau Strategic Plan, FY 2008 Congressional Budget Submission Exhibits (http://www.census.gov/main/www/FY_2008_Congressional_Submission.pdf)

Does the program have ambitious targets and timeframes for its long-term measures?

Explanation: In general, the targets for the 2010 goals are ambitious: 1) Improve the relevance and timeliness of census long-form data compared to Census 2000 by implementing the ACS to produce long-form type data each year; 2) Reduce operational risk compared to Census 2000 by completing a multi-year program of research, development, and testing, ending with a dress rehearsal of 2010 Census methods and systems in FY 2008 and by completing the MAF/TIGER Enhancements program for all 3,232 counties by FY 2008; 3) Improve the accuracy of census coverage compared to Census 2000 by reducing the measured number of geographic coding errors by at least 50%, reducing the measured number of duplicates by at least 50%, and reducing the measured overall net coverage error at the national level to less than one-half of one percent; 4) Contain costs by conducting all three components of the reengineered census for an amount (estimated at $11.5 billion in February 2007) that is less than the cost of repeating the methodology used in the 2000 Census, in large part through the use of hand held computers. While the use of handheld computers and other improvements will allow the Census Bureau to realize some cost avoidance, costs for decennial censuses have risen significantly over time. The Census Bureau should continue to identify additional efficiencies where possible.

Evidence: FY 2007 Annual Performance Plan, FY 2008 Annual Performance Plan, Census Bureau Strategic Plan, FY 2008 Congressional Budget Submission Exhibits (http://www.census.gov/main/www/FY_2008_Congressional_Submission.pdf), Estimated Life Cycle Costs for the Reengineered 2010 Census of Population and Housing, Milestone Schedule for the Reengineered 2010 Census, American Community Survey Operational Plan

Does the program have a limited number of specific annual performance measures that can demonstrate progress toward achieving the program's long-term goals?

Explanation: Ultimately, the "outcome" performance of the decennial census program will be determined when the census is conducted in 2010. To ensure the program is on-track for implementation in 2010, the Annual Performance Plan and yearly Congressional Budget request include specific performance milestones for each fiscal year which are linked to the long-term performance goals. These annual output measures include: Percent of Weighted response rate for annual implementation of the ACS; Number of counties with improved GPS location; and Percent of Census Objectives achieved. Tracking the response rate of the ACS helps monitor the quality of the survey as a replacement for the decennial long-form, while completion of GPS-aligned counties and census test objectives shows progress towards implementation of a fully reengineered census in 2010. Targets for these measures are set and monitored each year for each of the component programs within the reengineering effort--the ACS, the MAF/TIGER Enhancement Program, and the short-form only Census. Census also has an efficiency measure for the ACS and MAF/TIGER Enhancements Project.

Evidence: FY 2008 Annual Performance Plan, Census Bureau Strategic Plan, FY 2008 Congressional Budget Submission Exhibits (http://www.census.gov/main/www/FY_2008_Congressional_Submission.pdf)

Does the program have baselines and ambitious targets for its annual measures?

Explanation: Census has yearly and quarterly performance milestones for accomplishment of the three components of the 2010 reengineering effort. Annual targets for FY 2008 include a 92 Percent of Weighted response rate for annual implementation of the ACS; 367 counties with improved GPS location; and 100 percent of Census objectives achieved. These three components are linked to the strategic goal of implementing a 2010 Census that is more cost-effective, provides more timely data, improves coverage, and reduces operational risk.

Evidence: FY 2008 Annual Performance Plan, Census Bureau Strategic Plan, Milestone Schedule for the Reengineered 2010 Census

Do all partners (including grantees, sub-grantees, contractors, cost-sharing partners, and other government partners) commit to and work toward the annual and/or long-term goals of the program?

Explanation: Key program partners are the Congress, OMB and other Federal agencies, state, tribal, and local governments, Census Advisory Committees, data users, and other stakeholders concerned about the apportionment, redistricting, and Federal fund allocation purposes of the decennial census. The Census Bureau routinely consults program partners in developing the goals for the 2010 Census. Partners have shown support for the goals and the research, testing, and development efforts planned and completed to date. The Census Bureau also incorporates decennial program goals into contracts with its IT partners working on segments of the reengineering effort.

Evidence: Census Bureau Advisory Committees have expressed support for the goals of the 2010 Census, and continue to assess and advise Census on efforts to meet those goals. The various Federal agencies that will use the ACS and 2010 Census results also have endorsed the goals through inter-agency efforts led by OMB. Goals and progress requirements are documented in contracts now in place, such as with the Harris Corporation and Lockheed Martin.

Are independent evaluations of sufficient scope and quality conducted on a regular basis or as needed to support program improvements and evaluate effectiveness and relevance to the problem, interest, or need?

Explanation: Several external evaluations were conducted on Census 2000 by the DOC Office of the Inspector General, the Government Accountability Office (GAO), and the National Academy of Sciences (NAS), and these types of evaluations continue for the 2010 reengineering effort. Census also conducts regular internal evaluations of the decennial program. As these internal and external evaluations are completed, Census incorporates findings and recommendations into the reengineering plans for the 2010 Census. For example, the ACS program is built on a decade of research and evaluation, major evaluations also are part of the MAF/TIGER Enhancements Program, and the research, testing, and development efforts for the short-form only 2010 Census is a multi-year effort designed to test, evaluate, and modify all aspects of that program.

Evidence: Commerce Inspector General evaluations (http://www.oig.doc.gov/oig/reports/census_bureau/index.html), Reports and Recommendations by GAO, NAS, and the Census Bureau Advisory Committees

Are Budget requests explicitly tied to accomplishment of the annual and long-term performance goals, and are the resource needs presented in a complete and transparent manner in the program's budget?

Explanation: Budget requests for these programs are tied to the accomplishment of performance goals and are presented in an integrated manner within the budget. Annual and long-term performance targets are integrated with budget resources. Specific performance measures, targets, and unit costs are embedded in the budget. In addition, performance goals contained in the Census Bureau's Strategic and Annual Performance Plans are linked to budget performance goals.

Evidence: FY 2007 Annual Performance Plan, FY 2008 Annual Performance Plan, Census Bureau Strategic Plan, FY 2008 Congressional Budget Submission Exhibits (http://www.census.gov/main/www/FY_2008_Congressional_Submission.pdf)

Has the program taken meaningful steps to correct its strategic planning deficiencies?

Explanation: The Census Bureau's strategic planning efforts include periodic reviews by senior staff and soliciting comments from staff and data users to identify and correct any strategic planning deficiencies that may be found. After Census 2000, the Census Bureau identified a number of areas for improvement. The re-engineered 2010 Decennial Census Program consists of three highly integrated components designed to take advantage of opportunities for innovations made possible through the expanded use of technology, major changes in the business process for data collection, and the use of focused coverage improvement procedures. The plan for the 2010 Census is focused on making significant improvements to process and cost efficiencies, data quality, and respondent reporting. These include earlier planning, an integrated testing and evaluation plan, and the use of the ACS to collect long-form type data, thus simplifying the decennial into a short-form only census.

Evidence: FY 2008 Annual Performance Plan, Census Bureau Strategic Plan, FY 2008 Congressional Budget Submission (http://www.census.gov/main/www/FY_2008_Congressional_Submission.pdf)

Does the agency regularly collect timely and credible performance information, including information from key program partners, and use it to manage the program and improve performance?

Explanation: Census regularly collects performance information for use in managing and improving performance. Throughout the period leading up to a census, progress completing quarterly and yearly milestones and information from stakeholders and contractors are collected and used to inform program management. Following completion of a decennial census, information on coverage accuracy, household response rates, and the costs of collecting, processing, and disseminating data are collected for use in evaluation and planning the next census. During ACS, MTEP, and 2010 Census operations (including tests and dress rehearsal) the bureau collects and analyzes monthly (and for some critical operations weekly and even daily) cost and progress information to ensure programs are on schedule or to take prompt action on issues/problems.

Evidence: FY 2008 Annual Performance Plan, Census Bureau Strategic Plan, FY 2008 Congressional Budget Exhibits, FY 2005 Performance and Accountability Report, FY 2006 Performance and Accountability Report, Milestone Schedule for the Reengineered 2010 Census, Monthly Earned Value Management Reports from Contractors

Are Federal managers and program partners (including grantees, sub-grantees, contractors, cost-sharing partners, and other government partners) held accountable for cost, schedule and performance results?

Explanation: The program has identified the managers who are responsible for achieving key program results, and annual performance plans for these managers include goals and measures linked to the strategic goals of the program, which are linked to the Census Bureau and DOC strategic goals. Individual performance plans contain specific performance standards and targets in areas for which the manager is responsible. Results towards meeting performance goals are measured annually through the Annual Performance Plan process, and program managers are held accountable to these plans. Contracts are designed, managed, and monitored using such tools as earned value management to support specific projects relevant to performance goals. Contractors are held accountable for progress as part of their award fee. The two major IT contracts for 2010 (for field data collection and automated data processing) also have a percentage of the contractor's award fee tied to how well they work together on critical interface activities.

Evidence: SES Performance Agreement, 2010 Census Decision Memorandum 4: Targets for 2010 Census Strategic Goal Performance plans and evaluations that are part of each program managers annual review. Goals and progress requirements are documented in contracts now in place, such as with the Harris Corporation and Lockheed Martin.

Are funds (Federal and partners') obligated in a timely manner, spent for the intended purpose and accurately reported?

Explanation: All funds are obligated and spent in a timely manner. The decennial program has obligation plans for its frameworks. At the end of each fiscal year, the 2010 Census has had, on average, an unobligated balance of less than 2 percent of its appropriated funds. The bureau's major IT contracts (valued at $1.455 billion) require the use of earned value management to ensure timely and effective completion, relative to budget and schedule.

Evidence: FY 2005 Performance and Accountability Report, FY 2006 Performance and Accountability Report, Monthly Earned Value Management Reports from Contractors

Does the program have procedures (e.g. competitive sourcing/cost comparisons, IT improvements, appropriate incentives) to measure and achieve efficiencies and cost effectiveness in program execution?

Explanation: The Census Bureau has cost efficiency measures for the ACS and unit cost information for the counties realigned in MAF/TIGER Improvement initiative. Census has competitively sourced a wide variety of contracts related to decennial activities to improve efficiencies. These contractors are required to use earned value management in order to ensure that they make efficient and effective use of program funds. This decade, the Decennial Census program also is making use of Capability Maturity Model Integration (CMMI) principles and best practices in providing support for the planning and coordination of the broad range of activities that are required to successfully conduct the 2010 Census. Overall, the reengineering effort includes cost savings initiatives such as a short-form only questionnaire, automated field data collection, and targeted second mailing to reduce the amount of paper and staff in local census offices, and to reduce the amount of time spent by enumerators in non-response follow up work. In 2001, the bureau estimated the total savings from this reengineering to be $455 million. However, even with these savings, the total 2010 Census lifecycle cost will be roughly double that of Census 2000. This significant cost increase is the result of a number of factors, including inflation and increased workload, public resistance to answering surveys, and increasing diversity. Census should continue to explore additional efficiencies for implementation in 2010.

Evidence: FY 2008 Annual Performance Plan, FY 2008 Congressional Submission Exhibits, Exhibit 300s for MTEP, the American Community Survey (ACS), and for 2010 Census planning and testing (available at www.commerce.gov/exhibihit300/)

Does the program collaborate and coordinate effectively with related programs?

Explanation: The Bureau is involved in cross-cutting activities within the Department of Commerce and with other Federal Agencies in order to achieve the performance goals of the 2010 Census. Decennial programs work closely with these partners to ensure that Census data, and the strategic goals and efforts for the 2010 reengineering program, are meeting the needs of as broad a constituency as possible. Some of these partnerships include MAF/TIGER Partnership Program with State, Local and Tribal governments; Collaboration with Department of Justice on the Voting Rights Act; and the OMB's Federal Agency Council, which discusses ACS data requirements from multiple federal agencies.

Evidence: DOC Strategic Plan, Census Bureau Strategic Plan, OMB-led efforts concerning Federal agency data needs from the Census MAF/TIGER Partnership Program with State, Local and Tribal governments Collaboration with Department of Justice on enforcing Voting Rights Act. Internal Steering Committees and Working Groups that ensure coordinated goals and efforts for all components of the reengineering program.

Does the program use strong financial management practices?

Explanation: Census Bureau financial information is accurate and timely and performance and financial information are integrated to support day-to-day operations. Spending reports are reviewed monthly to ensure accuracy. The Census Bureau's financial management practices have resulted in a clean opinion on its financial audit since 1999. Contractor cost information provided in cost and performance reports for major 2010 Census contracts is compared to cost information in the Commerce Business System (accounting system of record). This provides an independent source of validation of Earned Value Management Data supplied by the contractors.

Evidence: Department of Commerce's FY 2005 and 2006 Consolidated Financial Statements, FY 2005 and 2006 Performance and Accountability Report

Has the program taken meaningful steps to address its management deficiencies?

Explanation: The program has developed and maintained more rigorous (based on industry best practices) management structures and processes for the 2010 Census to monitor progress, communicate effectively and coordinate planning to ensure performance goals are accomplished. New practices for the 2010 Census include full requirements definition and change control management, as the Census Bureau moves towards mature Capability Maturity Model Integration (CMMI) and earned value management levels. A report from the Office of the Inspector General for the Department of Commerce listed 10 lessons learned from Census 2000. The Census Bureau utilized this list to strengthen its reengineering plan for the 2010 Census. A GAO report on Planning a more cost effective 2010 Census included recommendations, most of which were agreed to by the Census Bureau and were also utilized in strengthening the reengineering plan.

Evidence: FY 2008 Annual Performance Plan, FY 2008 Congressional Budget Exhibits, 2010 Decennial Census Risk Management Plan, Exhibit 300s for MTEP, ACS and the 2010 Census, Reports and recommendations by DOC IG, GAO, NAS, and the Census Bureau Advisory Committees. GAO Report 03-40, Lessons learned for planning a more cost effective 2010 Census (http://www.gao.gov/new.items/d0340.pdf), OIG Report #14431, Improving our Measure in America: What Census 2000 can teach us in planning for 2010 (http://www.oig.doc.gov/oig/reports/2002/ESA-Census-OIG-14431-03-2002.pdf)

Has the program demonstrated adequate progress in achieving its long-term performance goals?

Explanation: The program is generally on track in meeting the long-term performance goals to reduce operational risk, improve accuracy, provide more timely data, and contain costs. With the full implementation of the ACS, the Decennial Program has begun to achieve some of its long term performance goals. The ACS now produces data products annually for areas and population groups of 65,000 or greater. Populations greater than 20,000 will be available annually, beginning in FY2008 and areas of all sizes and census tracts will be available annually, beginning in FY2010. The MTEP is expected to be completed within budget and on schedule in FY2008. For the 2010 Census, successful tests were completed in 2003, 2004, 2005, and 2006, and the dress rehearsal is on schedule for 2008. While the 2010 Census improvements should provide some cost avoidance, the total lifecycle cost has increased slightly, from $11.3 billion (the original 2001 estimate) to $11.5 billion (a 2% increase).

Evidence: FY 2007 Annual Performance Plan, FY 2008 Annual Performance Plan, FY 2008 Congressional Budget Submission Exhibits, 2010 National Census Test Memoranda Series, 2003 National Census Test, No. 3, FY 2005 Performance and Accountability Report, FY 2006 Performance and Accountability Report, 2010 Census Test Memoranda Series, 2004 Census Test, No. 45

Does the program (including program partners) achieve its annual performance goals?

Explanation: During FY 2006 all annual performance targets for the 2010 Decennial Census Program were met. These included the 97.6% response rate for the annual implementation of ACS, bringing an additional 700 counties into alignment with global positioning system (GPS), and completing 100% of Census test objectives.

Evidence: FY 2008 Annual Performance Plan, Department of Commerce Strategic Plan, Census Bureau Strategic Plan, FY 2006 Performance and Accountability Report

Does the program demonstrate improved efficiencies or cost effectiveness in achieving program goals each year?

Explanation: The Decennial Census is a cyclical program where it is most useful to evaluate efficiency and cost effectiveness over the life cycle of the program. Since 1970, the costs of each decennial census have risen significantly. Currently, the total lifecycle cost of the 2010 Census ($11.5 billion as of February 2007) is nearly double the cost of Census 2000. This increase is partly due to inflation and increased workload, the need for highly accurate data at low levels of geography, increased public resistance to answering surveys, and an increasingly diverse population. Recognizing the need to contain costs, Census has implemented a number of improvements that should slow the increase. The use of GPS-enabled handheld computers will allow the Census Bureau to more accurately locate housing units, reduce time and mileage costs, reduce follow-up workload, and reduce the amount of paper and the people and space necessary to handle the paper. A short-form only census will also allow the bureau to conduct a targeted second mailing to non-responding households, thereby reducing the costly non-response follow-up workload. While the 2010 Census will be significantly more expensive than Census 2000, these improvements are expected to slow the rate of cost increase. The average cycle-over-cycle rate of increase in unit cost for the last 3 censuses (1980, 1990, & 2000) was 70.7 percent. The rate of increase in unit cost forecasted for the 2010 Census is 33.1 percent.

Evidence: Estimated Life Cycle Costs for the Reengineered 2010 Census of Population and Housing, 2010 Decision Memo 10: Decision to Use Mobile Computing Equipment (MCE) for Field Data Collection During the 2010 Census http://www.census.gov/acs/www/

Does the performance of this program compare favorably to other programs, including government, private, etc., with similar purpose and goals?

Explanation: There are no similar programs that provide population data for apportionment and redistricting or program fund allocations by many agencies. Further, international counterparts do not face the same consistutional and political challenges, particularly with respect to coverage accuracy.

Evidence: Census Bureau Strategic Plan, Constitution of the United States (Article 1, Section 2), Public Law 94-171, Title 13, U.S. Code, Various Federal laws that mandate use of Decennial Census data for fund allocations.

Do independent evaluations of sufficient scope and quality indicate that the program is effective and achieving results?

Explanation: Evaluations by the Department of Commerce Inspector General, the Census Monitoring Board, GAO, the National Academy of Sciences, and the Census Advisory Committees, as well as internal evaluations, indicate that Census 2000 was well executed in many respects, reduced the undercount as compared to the 1990 Census, and data were released on schedule. Similar external and internal reviews have continued for the effort to reengineer the 2010 Census program, and in general, these evaluations indicate that the program is progressing well and is on track for implementation in 2010. However, both GAO and DOC IG evaluations have found some risks still remain, and there are opportunities to improve plans and operations for implementation in 2010.

Evidence: Reports by DOC IG, GAO, NAS, and Census Advisory Committees. Examples of areas identified by the Commerce IG and GAO that could be improved or need continued focus include enumeration of non-traditional housing units (http://www.oig.doc.gov/oig/reports/2006/Census-IPE-18046-09-06.pdf) and contract management (http://www.gao.gov/new.items/d06277.pdf).