This is the accessible text file for GAO report number GAO-03-178 entitled 'Formula Grants: 2000 Census Redistributes Federal Funding Among States' which was released on March 27, 2003. This text file was formatted by the U.S. General Accounting Office (GAO) to be accessible to users with visual impairments, as part of a longer term project to improve GAO products’ accessibility. Every attempt has been made to maintain the structural and data integrity of the original printed product. Accessibility features, such as text descriptions of tables, consecutively numbered footnotes placed at the end of the file, and the text of agency comment letters, are provided but may not exactly duplicate the presentation or format of the printed version. The portable document format (PDF) file is an exact electronic replica of the printed version. We welcome your feedback. Please E-mail your comments regarding the contents or accessibility features of this document to Webmaster@gao.gov. Report to Congressional Requesters: United States General Accounting Office: GAO: February 2003: FORMULA GRANTS: 2000 Census Redistributes Federal Funding Among States: GAO-03-178: GAO Highlights: Highlights of GAO-03-178, a report to Congressional Requesters: February 2003: FORMULA GRANTS: 2000 Census Redistributes Federal Funding Among States: Why GAO Did This Study: In fiscal year 2000, about $283 billion in federal grant money was distributed to state and local governments by formula, about half of it through four formula grant programs—Medicaid, Foster Care Title IV-E, Adoption Assistance, and the Social Services Block Grant (SSBG). States receive money based in part on factors such as annual population estimates derived from the previous decennial census, which is conducted by the Department of Commerce, Bureau of the Census. GAO was asked to measure the effect that using the 2000 census data has on redistributing funding for federal formula grant programs. To do this, GAO analyzed the change in the U.S. and state populations between 1999 and 2000 that was the result of correcting prior population estimates and estimated for the four programs the extent of any redistribution of federal funding among states. What GAO Found: The 2000 census count of 281.4 million people exceeded the 1999 population estimate by 8.7 million people, or 3.2 percent. Three-quarters of this 1-year population increase, 6.8 million people, was the result of correcting errors in population estimates over the preceding decade; the remaining portion of the increase, 1.9 million people, was the result of population growth from 1999 to 2000. Every state’s population had been underestimated during the 1990s, but the extent varied, from the smallest correction in West Virginia—0.3 percent—to the largest in the District of Columbia—10.2 percent. Twenty-eight states had a correction below the national average of 2.5 percent, and 23 states had a correction above the national average. Correcting population estimates for the 2000 census redistributes among states about $380 million in federal grant funding for Medicaid, Foster Care, Adoption Assistance, and SSBG. Funding for the 28 states that had below- average corrections to their populations decreases by an estimated $380.3 million; funding for the 23 states that had above-average corrections increases by an estimated $388.8 million. Most of the change in funding is concentrated in states with larger populations. However, changes in funding are smaller in several large states because the matching rates for Medicaid, Foster Care, and Adoption Assistance are limited by statute—matching rates cannot fall below 50 percent. Some higher-income states would receive matching rates below 50 percent if not for this limitation. Most of the shift in funding occurs in fiscal year 2003 when federal matching rates for the Medicaid, Foster Care, and Adoption Assistance programs are based on population estimates derived from the 2000 census.A small portion of the shift occurred in fiscal year 2002 because that is when the SSBG began using the 2000 census counts. The Department of Commerce provided technical comments on a draft of this report. www.gao.gov/cgi-bin/getrpt?GAO-03-178. To view the full report, including the scope and methodology, click on the link above. For more information, contact Kathryn G. Allen at (202) 512-7114. Contents: Letter: Results in Brief: Background: Most of Population Difference Between 1999 and 2000 Resulted from Correction of Errors That Occurred During 1990s: 2000 Census Correction of Population Estimates Redistributes an Estimated $380Million Among States for Four Formula Grant Programs: Agency Comments: Appendix I: Data for Population Estimates, Decennial Census Population Counts, and the Error of Closure: Appendix II: Analysis of Estimated Funding Changes for Four Formula Grant Programs: Medicaid, Foster Care, and Adoption Assistance Social Services Block Grant: Tables: Table 1: Definition of Population Terminology Used in This Report: Table 2: Federal Formula Grant Program Funding for Fiscal Year 2000: Table 3: Population Data Used in Four Selected Formula Grant Programs, by Fiscal Year: Table 4: Estimated Changes in Federal Funding as a Result of the Correction in Population, by Grant Program: Table 5: Comparison of the 1999 Postcensal Population Estimates and the 2000 Census Counts: Table 6: Comparison of the 2000 Postcensal Population Estimates and the 2000 Census Counts to Determine the Error of Closure and the Percentage Correction in Population: Table 7: Actual and Estimated FMAPs for the Medicaid, Adoption Assistance, and Foster Care Programs for Fiscal Year 2003, by State: Table 8: Medicaid Program Exenditures and Estimated Federal Payments, by State: Table 9: Foster Care Program Expenditures and Estimated Federal Payments, by State: Table 10: Adoption Assistance Program Expenditures and Estimated Federal Payments by State: Table 11: SSBG State Allocations, Actual and Estimated, for Fiscal Year 2002: Figure: Figure 1: Percentage Difference in Population Due to the Correction of the Error in Population Estimates, by State on April 1, 2000: Abbreviations: FMAP: Federal Medical Assistance Percentage: SSBG: Social Services Block Grant: This is a work of the U.S. Government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. It may contain copyrighted graphics, images or other materials. Permission from the copyright holder may be necessary should you wish to reproduce copyrighted materials separately from GAO’s product. Letter: February 24, 2003: The Honorable Tom Davis Chairman, Committee on Government Reform House of Representatives: The Honorable Adam H. Putnam Chairman, Subcommittee on Technology, Information Policy, Intergovernmental Relations, and the Census Committee on Government Reform House of Representatives: The Honorable Dave Weldon House of Representatives: In fiscal year 2000, the federal government obligated about $332 billion in grants to state and local governments to help fund an array of programs ranging from Medicaid to Highway Planning and Construction. Over 85 percent, or about $283 billion, of this grant money was distributed to state[Footnote 1] and local governments using formulas that are based on data such as state population and personal income. For example, the $196 billion federal-state Medicaid program finances health care to low-income families with children and aged, blind, and disabled individuals through a statutory formula based on state per capita income--the ratio of total personal income to state population. To calculate grant amounts, formula grant programs generally rely on annual population estimates for each state developed by the Bureau of the Census. State populations are estimated by adding to the prior year’s population estimate the number of births and immigrants and subtracting the number of deaths and emigrants. These estimates are subject to error, mainly because migration between states and between the United States and other countries is difficult to measure. By the end of each decade, when the decennial census is taken, a significant gap may have arisen between the population estimate and the census population count for the same day of the year, such as April 1, 2000. When population data based on a new census enter into federal formula grant calculations, states gain or lose funding depending on how the gaps between their population estimates and their census counts compare with the U.S. average gap. The larger the gap between a state’s population estimate and its census count, the larger the shift in funding is. For formula grant programs that distribute a set amount of federal funding, the gains in states with increased funding are offset by the losses in states with decreased funding. For open-ended formula grant programs, such as Medicaid, states with increased funding do not necessarily offset states with decreased funding. To measure the effect of using the 2000 census on the distribution of formula grant funding among the states, you asked us to examine (1) the change in the U.S. and each state’s population between 1999 and 2000 due to correcting prior population estimates and (2) the extent of any redistribution of federal funding among the states for four selected formula grant programs as a result of the 2000 census. To address these objectives, we used information on annual state population estimates that were derived from the 1990 census and state estimates that were derived from both the 1990 and 2000 censuses, as reported by the Census Bureau. To estimate the error in population estimates, we compared the April 1, 2000, population estimates based on the 1990 census with the April 1, 2000, census counts. To determine the effect of correcting the errors in population estimates on the distribution of formula grant funding to the states, we analyzed 4 federal formula grant programs of the 172 such programs identified in the Catalog of Federal Domestic Assistance--Social Services Block Grant (SSBG), Medicaid, Foster Care Title IV-E, and Adoption Assistance.[Footnote 2] We chose these 4 programs because their formulas use population estimates to distribute federal assistance, and they represented almost half of all formula grant funding (46 percent) in fiscal year 2000. The SSBG distributes a set appropriation exclusively on the basis of population data. The 3 entitlement programs, Medicaid, Foster Care, and Adoption Assistance, use per capita income--the ratio of personal income to state population--in identical formulas to determine federal matching rates. We obtained information on the formulas for these programs from the Department of Health and Human Services, and we used funding data for each program for the fiscal year in which the program first used population data derived from the 2000 census to calculate grant awards. To calculate the change in formula funding resulting from correcting population estimates, we compared what funding would be if formula grant amounts were calculated using two different population estimates for the same year, one based on the 1990 census and the other on the 2000 census. We conducted our work from July 2001 through January 2003 in accordance with generally accepted government auditing standards. Results in Brief: The 2000 census count of 281.4 million people exceeded the 1999 population estimate by 8.7 million people, or 3.2 percent. Three- quarters of this 1-year population increase, 6.8 million people, was the result of correcting errors in population estimates over the preceding decade; the remaining portion of the increase, 1.9 million people, was the result of population growth from 1999 to 2000. The error corrected by the 2000 census was substantially larger than the error reported for the 1990 census--2.5 percent compared with 0.6 percent. The Census Bureau attributed the increase in the 2000 “error of closure” to underestimates in the measurement of net international migration and the increased accuracy of the 2000 census--it counted people who were probably missed in the 1990 census. Every state’s population had been underestimated during the 1990s, but the extent varied widely: the largest correction was in the District of Columbia- -10.2 percent--and the smallest, West Virginia--0.3 percent. Twenty- eight states had a correction below the national average of 2.5 percent, and 23 states had a correction above the national average. Of the four Census regions (Northeast, South, Midwest, and West), only the Midwest showed a pattern: all 12 midwestern states were close to or below the U.S. average correction to the population. Overall, the Midwest’s correction was the smallest of the four regions--1.5 percent. Correcting population estimates based on the 2000 census redistributes about $380 million in federal grant funding among states for the four programs we examined. We estimate that funding for the 28 states that had below-average corrections to their populations decreases by $380.3 million in the first year the new population numbers are factored into the formula grants; funding for the 23 states that had above-average corrections in their population increases by an estimated $388.8 million. Most of the change in funding is concentrated in states with larger populations. However, several large states have only minor changes in funding because the funding formula used by Medicaid, Foster Care, and Adoption Assistance limits the effect of the population correction for high-income states by applying a minimum 50 percent federal matching rate. Some higher-income states would receive matching rates below 50 percent, but because of the minimum they are guaranteed a rate no lower than 50 percent. Most of the shift in funding occurs in fiscal year 2003 when federal matching rates for the Medicaid, Foster Care, and Adoption Assistance programs are based on population estimates derived from the 2000 census. A minor portion of the shift occurred for fiscal year 2002 because the census counts were used in the SSBG that year. The Department of Commerce provided technical comments on a draft of this report, which we incorporated as appropriate. Background: The Census Bureau counts the U.S. population once every decade through its decennial census. For the years in between, the Bureau estimates states’ populations from annual data on changes in births, deaths, and net migration (including net movements of military personnel). These annual population estimates are called postcensal population estimates because they are based on the prior census (see table 1 for definitions of different population counts used in this report). This process of making annual postcensal population estimates continues until the next census. Once the new census is taken, the Bureau compares the population estimates to the census population counts for the same date. The difference between the population estimate and the census count is called the error of closure. Subsequently, annual population estimates are revised for the prior decade using the counts from the new census. For example, after the 2000 census, the annual population estimates from the 1990s were revised to be consistent with both the 1990 and 2000 censuses. These revised population estimates are called the intercensal population estimates because they rely on the preceding and the succeeding censuses.[Footnote 3] Table 1: Definition of Population Terminology Used in This Report: Term: Census population count; Description: A population count is made at the beginning of each decade as of April 1. It is based on a count of the entire population. The latest census counted the population as of April 1, 2000.. Term: Postcensal population estimate; Description: Population estimates are made annually throughout a decade, usually as of July 1 of each year. Such estimates are based on the prior census, and include annual population changes due to births, deaths, and domestic and international migration. The postcensal population estimates for July 1, 2001, were based on the April 1, 2000, census and the population change between April 1, 2000, and June 30, 2001.. Term: Error of closure; Description: The error of closure is the difference between the postcensal population estimate and census population count for the same date. For example, the error of closure for April 1, 2000, is the difference between the postcensal population estimate and the census population count for April 1, 2000.. Term: Intercensal population estimate; Description: Once a new census is completed, the annual population estimates of the prior decade (the postcensal population estimates) are adjusted to reflect the new census counts. The resulting population estimates, known as intercensal population estimates, are calculated using a mathematical formula that distributes the error of closure across the postcensal population estimates for the prior decade. Intercensal population estimates thus have been adjusted according to counts at both the beginning and the end of the decade. The intercensal population estimates for 1990 through 1999 were issued in April 2002.. Source: Department of Commerce, Bureau of the Census. [End of table] Of the four programs we analyzed, Medicaid is the largest, comprising 43 percent of all federal formula-based programs and 94 percent of the total funding for the four programs analyzed for this report (see table 2). Table 2: Federal Formula Grant Program Funding for Fiscal Year 2000: Program: Medicaid; Fiscal year 2000 federal obligations[A] (millions): $121,809; Percentage of total federal obligations: 43.0. Program: Foster Care Title IV-E; Fiscal year 2000 federal obligations[A] (millions): 4,536; Percentage of total federal obligations: 1.6. Program: Adoption Assistance; Fiscal year 2000 federal obligations[A] (millions): 1,008; Percentage of total federal obligations: 0.4. Program: SSBG; Fiscal year 2000 federal obligations[A] (millions): 1,775; Percentage of total federal obligations: 0.6. Program: Remaining 168 formula programs; Fiscal year 2000 federal obligations[A] (millions): 154,221; Percentage of total federal obligations: 54.4. Program: Total obligations[B]; Fiscal year 2000 federal obligations[A] (millions): 283,348; Percentage of total federal obligations: 100.0. Source: U.S. General Services Administration, Catalog of Federal Domestic Assistance (Washington, D.C.: December 2001 edition) (CD-ROM version). Note: Federal obligations do not add to total because of rounding. [A] The obligated amounts shown here will differ slightly from the amounts allocated by formula. The obligations of the allocations may occur in years other than when the allocations occurred. [B] Total obligations include 23 programs that are both formula and project grants. [End of table] The SSBG formula allocates an amount of funding, set by annual appropriation, directly to the states. A state’s allocation is proportional to its share of the total U.S. population. State allocations for fiscal year 2002 used the April 2000 census, and allocations for prior years used postcensal population estimates that were based on the 1990 census. In contrast with the SSBG’s fixed appropriation, the Medicaid, Foster Care, and Adoption Assistance programs are open-ended entitlement programs--the states determine the level of program expenditures, and the federal government reimburses a share of their expenditures according to matching rates, called the Federal Medical Assistance Percentages (FMAP), set by statutory formula. All three programs use the same formula, which is based on a 3-year average of state per capita income--the ratio of aggregate personal income to state population. As a state’s per capita income increases, its matching rate decreases, and vice versa. In addition, unless a state experiences changes in aggregate personal income, its federal payment generally declines if the state’s population growth is less than the national average. Matching rates range from a minimum of 50 percent to a maximum of 83 percent of a state’s Medicaid expenditures. The minimum 50 percent rate affects only the high per capita income states. For fiscal year 2002, for example, a high-income state such as Connecticut would receive a 15 percent federal matching rate if the 50 percent minimum was not in place. For fiscal year 2002, the federal matching rates for Medicaid, Foster Care, and Adoption Assistance were based on a 3-year average of per capita income from 1997 through 1999. Rates for fiscal year 2003 are based on a 3-year average from 1998 through 2000. Although the formulas use overlapping years, the state population numbers used to compute per capita income differ depending on which fiscal year the grant is for. For these three programs, the fiscal year 2002 formula calculations used postcensal population estimates derived from the 1990 census for 1997 through 1999 to calculate per capita income. Fiscal year 2003 formula calculations used population estimates for 1998 through 2000 derived from the 2000 census.[Footnote 4] Thus, the 2000 census affects matching rates for these programs beginning in fiscal year 2003 (see table 3). Table 3: Population Data Used in Four Selected Formula Grant Programs, by Fiscal Year: SSBG: Fiscal year allocation or payment: 2001[A]; Data used: July 1998 postcensal state population estimates[B]. Fiscal year allocation or payment: 2002[C]; Data used: April 2000 decennial census by state. Medicaid, Foster Care, and Adoption Assistance: Fiscal year allocation or payment: 2002[A]; Data used: July 1997, 1998, and 1999 postcensal state population estimates[B]. Fiscal year allocation or payment: 2003[C]; Data used: July 1998, 1999, and 2000 state population estimates[D]. Sources: Department of Health and Human Services, Administration for Children and Families; and Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation. [A] The last year the population estimates based on the 1990 census were used in the formula. [B] These postcensal population estimates are based on the 1990 census. [C] The first year that the counts based on the 2000 census were used in the formula. [D] These population estimates were published by the Department of Commerce’s Bureau of Economic Analysis and were based on the 2000 census. [End of table] Most of Population Difference Between 1999 and 2000 Resulted from Correction of Errors That Occurred During 1990s: The difference between the 2000 census count and the 1999 postcensal population estimate was 3.2 percent, which is large compared with the 1 percent average annual growth rate estimated over the preceding decade. Most of the difference was due to the correction of the error that had occurred during the 1990s. According to the Census Bureau, the size of the error was the result of an underestimate in the measurement of net international migration during the 1990s and the improved coverage of the 2000 census compared with the 1990 census. Consequently, the postcensal population estimate for 2000 was smaller than the 2000 census count. Every state’s population growth was underestimated and needed correction, but the correction amounts varied widely. Among the four Census regions, only the Midwest[Footnote 5] showed a consistent pattern: all 12 states were close to or below the national average correction. California, Florida, and New York accounted for a high percentage of the correction in population estimates in their respective regions. Correcting Errors in Population Estimates Accounted for Three-Quarters of the Difference Between 1999 to 2000: The 2000 census count of 281.4 million people as reported by the Census Bureau exceeded the 1999 postcensal population estimate by 8.7 million people, or 3.2 percent. Slightly more than three-quarters of this difference (2.5 percent) was the result of correcting errors in the population estimates that occurred over the decade, called the error of closure (see app. I for detailed data for all states). The error of closure was 6.8 million people, substantially larger than the 1.5 million error of closure associated with the 1990 census. The error of closure for the 2000 census was four times the corresponding percentage error for the 1990 census (2.5 percent compared with 0.6 percent). The large error of closure in 2000 was due to underestimating the annual growth in population during the 1990s and to the improved coverage of the 2000 census over the 1990 census. The postcensal population estimates for the decade grew an average 1.0 percent annually. However, the 2000 census showed that the average annual growth rate in population was 0.2 percent higher than the estimated rate, or 1.2 percent. The Census Bureau revised its annual population estimates upward when it released its intercensal population estimates in the spring of 2002. The Census Bureau cited two reasons for the size of the error in its postcensal estimated population growth through the 1990s. First, the net international migration was underestimated during the decade, especially for the Hispanic population. The Hispanic population was underestimated by approximately 10 percent, four times higher than the national average population underestimate, 2.5 percent.[Footnote 6] Second, the 2000 census was more accurate than the 1990 census. The population undercount from the 2000 census was much smaller compared with the 1990 census (1.18 percent, compared with 1.62 percent, making the 2000 census more accurate[Footnote 7]); the 2000 census counted people who were probably missed in the 1990 census. Size of Population Correction Differed Widely Across States: The error of closure shows a wide variation across states. For example, West Virginia and Michigan had the smallest percentage corrections, 0.27 and 0.34 percent, respectively. The District of Columbia and Nevada had the largest percentage corrections in their population estimates, 10.2 percent and 7.5 percent, respectively. Twenty-eight states had a lower-than-average percentage difference, and 23 states had a greater-than-average percentage difference (see fig. 1 for the correction percentages for all states). Among the four Census regions, the Midwest had the smallest correction in population, 1.5 percent; all 12 Midwest states had corrections close to or below the national average.[Footnote 8] In the other three regions, a single state accounted for a large share of the population change for the region. For example, in the South, Florida’s correction in population of 4.7 percent constituted about 25 percent of the correction for the entire region. Similarly, New York’s correction was 44 percent of the northeastern states’ correction, and California’s correction was 26 percent of the correction for the western states. Figure 1: Percentage Difference in Population Due to the Correction of the Error in Population Estimates, by State, on April 1, 2000: [See PDF for image] [End of figure] 2000 Census Correction of Population Estimates Redistributes an Estimated $380 Million Among States for Four Formula Grant Programs: The correction to the population estimates generally redistributes federal funding for the four programs we analyzed from the states with the smallest corrections to those having the largest. Federal funding for the 28 states that had below-average corrections decreases by an estimated $380.3 million. In contrast, federal funding in the 23 states with above-average corrections to their population estimates increases by an estimated $388.8 million. Most of the change in funding is concentrated in states with larger populations. Michigan and Ohio, for example, account for 57 percent of the total decrease in funding for states with below-average population corrections. A number of high- income states, including California and New York, are largely unaffected by the correction in their populations because their matching rates for the Medicaid, Foster Care, and Adoption Assistance programs cannot decrease below the minimum 50 percent matching rate. Without this minimum, more funding would be shifted among the states. While the redistribution of funding in the four programs began to occur in fiscal year 2002, almost all of it occurs in fiscal year 2003, when the 2000 census data are used to determine federal matching rates in the three open-ended entitlement programs. Population Correction Causes Significant Funding Changes for Many States: The correction in state populations resulting from the 2000 census causes significant changes in the funding levels among the states for the four programs we examined. We estimate that the funding for the 28 states that had below-average corrections in their populations decreases by a total of $380.3 million. Conversely, funding for the 23 states that had above-average corrections in their populations increases by an estimated $388.8 million (see table 4). Table 4: Estimated Changes in Federal Funding as a Result of the Correction in Population, by Grant Program: Dollars in thousands: [See PDF for image] Sources: GAO calculations based on data obtained from the Department of Health and Human Services and the Department of Commerce, Bureau of the Census. Note: Totals may not add because of rounding. [End of table] These results are dominated by a few highly populated states whose corrections were among the largest--meaning they are estimated to receive the most additional money or to lose the most. For example, Michigan, the eighth most populous state,[Footnote 9] has an estimated $119 million decline in funding because of its 0.34 percent correction in population. Michigan’s federal funding decrease accounts for about one-third of the decreases for the 28 states with a below-average correction in population. Moreover, when Michigan’s decrease is combined with that of Ohio, the seventh most populous state, the two states account for 57 percent of the estimated total decline in funding from the corrections of the population estimates. Conversely, Florida, the fourth most populous state, has the largest estimated increase in funding (about $126 million) because of the 4.7 percent correction in its population estimate. This is almost double the national average correction and accounts for about one-third of the estimated increase for the 23 states with an above-average correction in population. Funding changes did not occur in some states and were muted in others because the states’ federal matching rates were fixed by the minimum 50 percent rate for the three open-ended entitlement programs. For example, on the basis of its fiscal year 2000 spending levels, California would receive an estimated $305 million less in matching aid in the three entitlement programs if its matching rate were allowed to fall below the minimum. Because of the 50 percent minimum federal matching rate, however, California only receives an estimated $2.8 million decrease--all of it linked to the SSBG. For the three entitlement programs, the correction in population had no effect in 11 states that were affected by the 50 percent minimum, and for 2 states the correction in population had a diminished effect because of the floor.[Footnote 10] The funding changes due to the population corrections showed little regional pattern except in the Midwest, where all 12 states had a correction in population estimates close to or below the national average that resulted in an estimated $289.5 million loss in funding owing to the correction in their populations. Medicaid Accounts for Most of the Change in Program Funding: Most of the change in funding resulting from the corrections in population estimates is the result of changes in Medicaid funding. The federal share of total Medicaid payments was approximately $111 billion in fiscal year 2000 and constituted 96 percent of the share of funding to the states for the four programs and approximately 96 percent of the total estimated change in funding as well.[Footnote 11] The SSBG distributed $1.69 billion for fiscal year 2002, representing 1.5 percent of the funding we analyzed. It accounted for a slightly higher percentage, 2.2 percent, of the estimated funding changes. Finally, the Foster Care and Adoption Assistance programs represented 1.6 and 0.6 percent of the funding, respectively. They account for 1.4 and 0.7 percent, respectively, of the estimated funding changes for 2003. The earliest effect of the 2000 census on any of the four programs we analyzed occurred when it was used to calculate fiscal year 2002 SSBG grants. For the Medicaid, Foster Care, and Adoption Assistance programs, the 2000 census is first used for fiscal year 2003 payments. Agency Comments: We provided the Department of Commerce a draft of this report for comment. The department provided technical comments, which we have incorporated where appropriate. As arranged with your offices, unless you release its contents earlier, we plan no further distribution of this report until 30 days after its issuance date. At that time, we will send copies of this report to interested congressional committees; the Secretary of Commerce; the Secretary of Health and Human Services; and the Director, Bureau of the Census. We will also make copies available to others on request. In addition, the report will be available at no charge on GAO’s Web site at http://www.gao.gov. If you or your staffs have questions about this report, please call me at (202) 512-7114 or Jerry Fastrup at (202) 512-7211. Major contributors to this report are Gregory Dybalski, Elizabeth T. Morrison, and Michael Rose. Kathryn G. Allen Director, Health Care--Medicaid and Private Health Insurance Issues: Signed by Kathryn G. Allen: [End of section] Appendix I Data for Population Estimates, Decennial Census Population Counts, and the Error of Closure: This appendix compares the postcensal population estimates for July 1, 1999, with the census count for April 1, 2000 (table 5), and compares the April 1, 2000, postcensal population estimates (based on the 1990 census) with the census counts (table 6). States are listed in tables 5 and 6 by the magnitude of the percentage correction in population. Table 5: Comparison of the 1999 Postcensal Population Estimates and the 2000 Census Counts: [See PDF for image] Source: Department of Commerce, Bureau of the Census, Population Estimates Division (Washington, D.C.), http://www.census.gov (downloaded Oct. 23, 2001). Notes: The states are listed in order of increasing percentage of population correction. (See table 6.): Totals may not add because of rounding. [End of table] The census is a population count made at the beginning of each decade as of April 1. It is based on a count of the entire population. Postcensal population estimates are made annually throughout a decade, usually as of July 1 of each year. Such estimates are based on the prior census and include annual population changes due to births, deaths, and domestic and international migration. Table 6: Comparison of the 2000 Postcensal Population Estimates and the 2000 Census Counts to Determine the Error of Closure and the Percentage Correction in Population: [See PDF for image] Sources: The postcensal population estimates for April 1, 2000, are from unpublished data provided by Department of Commerce, Bureau of the Census, Population Estimates Division. The April 1, 2000, census counts are from the Bureau of the Census, http://www.census.gov (downloaded Oct. 23, 2001). Notes: The states are listed in order of increasing percentage of population correction. Totals may not add because of rounding. [End of table] The census is a population count that is made at the beginning of each decade as of April 1. It is based on a count of the entire population. Postcensal population estimates are made annually throughout a decade, usually as of July 1 of each year. Such estimates are based on the prior census and include annual population changes due to births, deaths, and domestic and international migration. The error of closure is the difference between the postcensal population estimate and the census population count for the same date. The percentage correction in population is calculated by dividing the error of closure by the July 1, 1999, postcensal population estimate. [End of section] Appendix II Analysis of Estimated Funding Changes for Four Formula Grant Programs: This appendix contains the supporting data for our calculations of the estimated change in funding due to correcting the population estimates. Specifically, for each state, we provide the funding amounts for the four programs and the estimated funding changes due to the correction in population estimates. States are listed in tables 7 through 11 by the magnitude of the percentage correction in population. Medicaid, Foster Care, and Adoption Assistance: The Medicaid, Foster Care, and Adoption Assistance programs are open- ended entitlement programs for which states determine the level of program expenditures. The federal government reimburses states for a share of eligible state spending based on state per capita income. To calculate the effect of the population correction on the Federal Medical Assistance Percentages (FMAP)--also called federal matching rates--we compared actual matching rates for fiscal year 2003,[Footnote 12] based on the 2000 census, with the estimated matching rates based on the 1990 census (shown in table 7). Subtracting the estimated rates from the actual fiscal year 2003 rates shows the effect on the matching rates of correcting population estimates. In general, the states that had a below-average correction in population have a decrease in federal matching rates, while the states that had an above-average correction in population have an increase in matching rates. For 13 high-income states, the correction in population had no effect or had a diminished effect because of the minimum 50 percent matching rate. (Under the matching rate formula, no state can receive less than a 50 percent matching rate.) In our analysis, 11 states receive the 50 percent matching rate for fiscal year 2003; hence, under the estimated rates, the correction in population shows no change in these states’ matching rates. Two additional states, Washington and Nevada, are partially affected. Washington’s actual fiscal year 2003 matching rate is at the 50 percent minimum, while its estimated matching rate is slightly above the 50 percent minimum. Conversely, Nevada’s actual fiscal year 2003 matching rate is above the minimum, and its estimated matching rate is at the 50 percent minimum. The 70 percent matching rate for the District of Columbia is established by a special statutory provision. Accordingly, the District of Columbia’s matching rate remains unchanged, and the correction in population has no effect on funding. Table 7: Actual and Estimated FMAPs for the Medicaid, Adoption Assistance, and Foster Care Programs for Fiscal Year 2003, by State: [See PDF for image] Source: 66 Fed. Reg. 59792 (2001) and GAO calculations of Department of Commerce, Bureau of the Census and Bureau of Economic Analysis data. Notes: The states are listed in order of increasing percentage of population correction (see table 6). [End of table] The census is a population count made at the beginning of each decade as of April 1; it is based on a count of the entire population. Postcensal population estimates are made annually throughout a decade, usually as of July 1 of each year. Such estimates are based on the prior census and include annual population changes due to births, deaths, and domestic and international migration. Analysis of Funding Changes for Medicaid for Fiscal Year 2003: To measure the effect of the correction in the population estimates on federal payments, we estimated what federal payments would be using matching rates calculated on the basis of postcensal population estimates derived from the 1990 census. Specifically, multiplying the two sets of state matching rates in table 7 by program expenditures (fiscal year 2000 Medicaid expenditures) yields the estimated payments. The 2000 program expenditures were the latest year for which the data were available. (See table 8.): Overall, the states that had a below-average correction in population show a decrease in payments, while the states that had an above-average correction in population show an increase in payments. As discussed in the previous section, 11 states show no effect, and 2 states show a partial effect because of the minimum 50 percent federal matching rate. The District of Columbia is also unaffected because of its special statutorily set matching rate. Table 8: Medicaid Program Expenditures and Estimated Federal Payments, by State: [See PDF for image] Source: Fiscal year 2000 program expenditures obtained from the Department of Health and Human Services, Centers for Medicare & Medicaid Services. GAO computed the estimated payments. Notes: States are listed in order of increasing percentage of population correction. Totals may not add because of rounding. [A] Excludes administrative expenditures. [End of table] Analysis of Funding Changes for Foster Care and Adoption Assistance for Fiscal Year 2003: The effects on the funding for Foster Care and Adoption Assistance are similar to the effects on the Medicaid programs because these programs use the same matching rates. Table 9 shows the Foster Care program expenditures for fiscal year 2000, the estimated federal payments, and changes in funding for Foster Care based on these estimated payments. Table 10 shows the Adoption Assistance program expenditures for fiscal year 2000, the estimated federal payments, and the changes in funding for the program based on the estimated payments. Table 9: Foster Care Program Expenditures and Estimated Federal Payments, by State: [See PDF for image] Source: Fiscal year 2000 program expenditures obtained from the Department of Health and Human Services, Administration for Children and Families. GAO computed the estimated payments. Note: States are listed in order of increasing percentage of population correction. Totals may not add because of rounding. [A] Excludes administrative expenditures. [End of table] Table 10: Adoption Assistance Program Expenditures and Estimated Federal Payments, by State: [See PDF for image] Source: Fiscal year 2000 program expenditures obtained from the Department of Health and Human Services, Administration for Children and Families. GAO computed the estimated payments. Notes: The states are listed in order of increasing percentage of population correction. Totals may not add because of rounding. [A] Excludes administrative expenditures. [End of table] Social Services Block Grant: The fiscal year 2002 formula allocations for the SSBG are based on the April 1, 2000, decennial census population counts. To calculate the effect of the correction in population estimates, we compared fiscal year 2002 allocations that were calculated using the April 1, 2000, decennial census (actual allocations) with allocations using the 1990 postcensal population estimates for April 1, 2000 (estimated allocations). The differences in these allocations represent the effect of the population correction reflected in the 2000 census. The change in funding is directly proportional to the percentage correction in population because the SSBG allocations are calculated exclusively on the basis of population data (see table 11). Table 11: SSBG State Allocations, Actual and Estimated, for Fiscal Year 2002: [See PDF for image] Source: Department of Health and Human Services, Administration for Children and Families (Washington, D.C.), http://www.acf.hhs.gov (downloaded July 19, 2002). GAO computed the allocations for fiscal year 2002 based on the April 1, 2000, postcensal population estimates. Notes: The states are listed in order of increasing percentage of population correction. Totals may not add because of rounding. [End of table] The census is a population count that is made at the beginning of each decade as of April 1. It is based on a count of the entire population. Postcensal population estimates are made annually throughout a decade, usually as of July 1 of each year. Such estimates are based on the prior census and include annual population changes due to births, deaths, and domestic and international migration. FOOTNOTES [1] For this report, we use “state” to refer to the 50 states and the District of Columbia. [2] U.S. General Services Administration, Catalog of Federal Domestic Assistance (Washington, D.C., December 2001 edition) (CD-ROM version). [3] For more information about Census population estimates see Department of Commerce, Bureau of the Census, Population Estimates: Concepts and Geography (Washington, D.C.: Department of Commerce, Dec. 26, 2001), http://eire.census.gov/popest/archives/place/concepts.php (downloaded Jan. 31, 2003). [4] These population estimates were developed as interim estimates by the Department of Commerce’s Bureau of Economic Analysis. [5] The 12 Midwest states are Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, and Wisconsin. [6] J. Gregory Robinson, Accuracy and Coverage Evaluation: Demographic Analysis Results (Washington, D.C.: Department of Commerce, Bureau of the Census, March 2001), 9-11, http://landview.census.gov/dmd/www/ ReportRec.htm (downloaded Aug. 29, 2002). [7] The percentages are the net undercounts for the 1990 and 2000 censuses for household population from the 1990 Post-Enumeration Survey and 2000 Accuracy and Coverage Evaluation. Howard Hogan, Accuracy and Coverage Evaluation: Data and Analysis to Inform the ESCAP Report (Washington, D.C.: Department of Commerce, Bureau of the Census, March 2001), 12-14, http://landview.census.gov/dmd/www/ReportRec.htm (downloaded Jan. 15, 2003). [8] Nebraska and South Dakota were 0.03 and 0.04 percentage points above the national average, respectively. [9] State population rankings are based on the 2000 census. [10] The 11 states are California, Colorado, Connecticut, Delaware, Illinois, Maryland, Massachusetts, Minnesota, New Hampshire, New Jersey, and New York. The two states partially affected are Nevada and Washington. In addition, the District of Columbia receives a special federal matching rate of 70 percent and consequently is unaffected by the correction in population. [11] See appendix II for additional detail, by state, on the changes in federal matching rates and estimated shifts in funding under each of the four programs. [12] The matching rates for fiscal year 2003 are for the first year in which population estimates based on the 2000 census are used. GAO’s Mission: The General Accounting Office, the investigative arm of Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO’s commitment to good government is reflected in its core values of accountability, integrity, and reliability. Obtaining Copies of GAO Reports and Testimony: The fastest and easiest way to obtain copies of GAO documents at no cost is through the Internet. GAO’s Web site ( www.gao.gov ) contains abstracts and full-text files of current reports and testimony and an expanding archive of older products. The Web site features a search engine to help you locate documents using key words and phrases. You can print these documents in their entirety, including charts and other graphics. Each day, GAO issues a list of newly released reports, testimony, and correspondence. GAO posts this list, known as “Today’s Reports,” on its Web site daily. The list contains links to the full-text document files. To have GAO e-mail this list to you every afternoon, go to www.gao.gov and select “Subscribe to daily E-mail alert for newly released products” under the GAO Reports heading. Order by Mail or Phone: The first copy of each printed report is free. Additional copies are $2 each. A check or money order should be made out to the Superintendent of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more copies mailed to a single address are discounted 25 percent. Orders should be sent to: U.S. General Accounting Office 441 G Street NW, Room LM Washington, D.C. 20548: To order by Phone: Voice: (202) 512-6000: TDD: (202) 512-2537: Fax: (202) 512-6061: To Report Fraud, Waste, and Abuse in Federal Programs: Contact: Web site: www.gao.gov/fraudnet/fraudnet.htm E-mail: fraudnet@gao.gov Automated answering system: (800) 424-5454 or (202) 512-7470: Public Affairs: Jeff Nelligan, managing director, NelliganJ@gao.gov (202) 512-4800 U.S. General Accounting Office, 441 G Street NW, Room 7149 Washington, D.C. 20548: