This Article Abstract Submit a response to this article Services E-mail this article to a friend Alert me to new issues of the journal

H E A L T H  T R A C K I N G
T R E N D S 

26 January 2005  
  

Understanding The Recent Growth 
In Medicaid Spending, 2000–2003

Enrollment growth fueled Medicaid spending—
growth that likely kept the number of uninsured people
from rising even more than it did.

By John Holahan and Arunabh Ghosh

ABSTRACT:

Growth in Medicaid spending averaged 10.2 percent per year between 2000 and 2003, resulting in a one-third increase in program spending. Spending growth was lower from 2002 to 2003 because of slower growth in enrollment and in spending per enrollee, particularly for acute care services, and declines in disproportionate-share hospital (DSH) payments and upper payment limit (UPL) programs. For the entire 2000–2003 period, Medicaid spending increases were largely driven by enrollment growth, much of which was attributable to the economic downturn. Increases in spending per enrollee over the period were faster than inflation but slower than increases in private insurance spending.

Medicaid spending (federal and state) grew by about one-third, from $205.7 billion to $275.5 billion, between fiscal years 2000 and 2003 (Exhibit 1). This explosive growth has caused fiscal problems for state governments faced with depressed revenues. At the federal level, Medicaid spending is potentially a major target in the 2005 budget debate. It is expected that the Bush administration will propose policies to sharply curtail the growth in Medicaid spending.¹ The nation’s governors, in anticipation of federal Medicaid cutbacks, have written to congressional leaders arguing against dramatic action.² Efforts to constrain program growth require understanding of the reasons why this growth occurred.

This paper explores changes that took place during the 2000–2003 period. Over this period spending grew by 10.2 percent annually, with growth in spending on medical services of 11.3 percent (Exhibit 1). Between 2000 and 2002 Medicaid spending increased at annual rates of 11.8 percent, resulting in an increase in spending of $51.5 billion. Medicaid spending growth slowed from 2002 to 2003 (Exhibit 1), increasing by 7.1 percent or $18.3 billion. The increase for medical services (excluding disproportionate-share hospital [DSH] payments, administrative expenses, and adjustments) alone was 8.2 percent from 2002 to 2003, following annual increases of 12.9 percent in the previous two years. The reduction in spending growth between 2002 and 2003 was attributable to somewhat lower rates of growth in enrollment, in spending per enrollee for most services (primarily nursing home care and prescription drug costs), and in DSH payments.

Background

The slow growth from 2002 to 2003 returns Medicaid spending growth close to levels of the late 1990s. From this perspective, the spending increase of about 12 percent per year between 2000 and 2002 seems to be an exception. Between 1995 and 1998 Medicaid spending grew 3.6 percent per year, and between 1998 and 2000, 7.8 percent per year. The 3.6 percent growth between 1995 and 1998 was also an aberration, but on the low side. Because of welfare reform and the strong economy, there was an actual decline in Medicaid enrollment. At the same time, health care inflation was low throughout the entire health care sector.

In the late 1990s (1998–2000) Medicaid spending growth increased to 7.8 percent per year. Medicaid enrollment increased as the initial effects of welfare reform subsided. Moreover, several states adopted the State Children’s Health Insurance Program (SCHIP), and several used the Section 1931(b) provisions of welfare reform legislation or Section 1115 waivers to expand coverage. States also expanded home and community-based waiver programs and developed prescription drug programs for the elderly. Meanwhile, health care inflation began to rise, particularly for prescription drugs, which grew at double-digit rates. Hospital costs for both inpatient and outpatient care increased more rapidly than in the mid-1990s. The use of upper payment limit (UPL) programs accelerated, contributing to rising Medicaid spending.³

Medicaid spending growth accelerated between 2000 and 2002 for several reasons. Medicaid enrollment increased both because of the previous expansions and because the economic downturn made more people eligible for the program. Health care costs also rose more rapidly, led by prescription drug and hospital costs. Finally, Medicaid managed care was no longer providing states with the same savings it had provided in the mid-1990s.⁴

This paper examines the 2002–03 experience in more detail, placing it in the context of the entire 2000–2003 period of recession and slow recovery. The National Bureau of Economic Research (NBER) officially records the recession as lasting between March and November 2001, but state revenues were depressed and affected states’ decision making during the entire period.⁵

Data Sources And Methods

Data sources. We relied on three different data sources. First, we used enrollment data from the Kaiser Commission on Medicaid and the Uninsured collected by Health Management Associates (HMA) directly from all fifty states and the District of Columbia.⁶ Enrollment data are collected in whatever way states maintain their data systems; thus, there is some inconsistency among states. HMA could not collect detailed data from most states for the four major enrollment groups: the elderly; disabled; children; and nonelderly, nondisabled adults. However, for forty-four states the data are reasonably consistent, first on enrollment of the aged and disabled and second for families (which includes children and nondisabled adults). The Kaiser Commission also has data on total enrollment for all states. For this report we used the total enrollment data from the remaining states and allocated enrollment to the aged and disabled and families in the same proportions as reported by the forty-four states.

Second, data on spending by service comes from Medicaid Financial Management Reports (Form 64) from the Centers for Medicare and Medicaid Services (CMS) for 1995– 2003.⁷ These data are provided by state for each of the major services covered by Medicaid. The CMS Form 64 does not have the data by type of enrollee. The CMS data also have some inconsistencies across time, and we have made some edits to the data based on information provided by states themselves.

Third, we used data from the Medicaid Statistical Information System (MSIS) for 2000 on Medicaid spending by eligibility group. The MSIS has detailed person-level data by service and eligibility group, but the data were available for 2000 only.⁸ For some of the analyses in this paper, we projected these expenditures to increase by the growth rates that we observed in the CMS-64 data for 2000–2003. The MSIS data report lower spending than the CMS-64 data because the MSIS does not include payments to Medicare and DSH payments. In addition, some expenditures cannot be attributed to specific enrollees or, in other cases, to specific services. We also made some edits to the MSIS data to address inconsistencies.

Study Findings

Growth in enrollment. The Kaiser Commission’s Medicaid enrollment data provides enrollment in June of each year (Exhibit 2). Enrollment among the aged and disabled grew about 2.9 percent per year during the entire period. Enrollment for families increased 11.6 percent between 2000 and 2002 and another 7.1 percent between 2002 and 2003. These increases are consistent with the growth in Medicaid enrollment reported by the Current Population Survey (CPS).⁹

The notable growth in enrollment among nondisabled adults and children is largely related to the economic slowdown that occurred between 2000 and 2003. Because of the economic decline, many people experienced job losses and income declines. The number of people below 200 percent of the federal poverty level greatly increased; thus, more people became eligible for Medicaid under existing eligibility standards. The growth during the period may also reflect some of the state coverage expansions that occurred in the late 1990s. During the past few years, several states cut back coverage and curtailed their efforts to increase participation rates. But these cutbacks were overwhelmed by the large enrollment increases resulting from three years of slow economic growth. In addition to job and income losses, there were also reductions in rates of employer-sponsored coverage. These occurred in part from reductions in employment in large firms and in industries with high rates of employer-based coverage, but also from rapid increases in insurance premiums.¹⁰

It is also important to note that while the growth in enrollment among the aged and disabled was much slower than for children and nondisabled adults, it was much faster than rates of growth in the U.S. population. This is important because of the high cost of serving these populations. The 2.9 percent annual growth in enrollment of aged and disabled populations is not well understood. The increased enrollment of the aged may be attributable to higher participation rates in Medicaid, possibly because the rapid rise in the cost of prescription drugs made Medicaid participation more attractive. The enrollment in Medicaid of people using home and community-based services, which, as we show below, has still continued to grow, could also be a factor.

Some share of the growth could be attributable to the effects of life-saving medical technology that lengthens lives but leaves many people with disabilities. The contribution of drugs to the increased longevity of people with HIV/AIDS may be a factor as well. Another may be the increased recognition of chronic problems as disabilities. Finally, the baby boom that will eventually affect the size of the elderly population is now entering the 55–64 age range, ages at which the likelihood of disabilities increases. Some combination of these factors is causing the aged and disabled populations to increase faster than the rate of the overall population growth; this is likely to continue during the rest of the decade.¹¹

Growth in spending. Growth in Medicaid spending, either between 2002 and 2003 or over the entire period, was fairly well spread across all services, with a few exceptions (Exhibit 3).¹² Between 2002 and 2003, spending for acute care increased 11.2 percent; for long-term care, it increased only 3.8 percent. During the entire period, acute care services grew 13.4 percent and long-term care, 8.4 percent. The remarkably slower growth in long-term care from 2002 to 2003 is attributable to an absolute-dollar decline in nursing home spending. The faster growth in acute care spending relative to long-term care was no doubt caused by the faster growth in enrollment among children and nondisabled adults, groups far more likely to use acute care than long-term care.

The period also saw very rapid growth in prescription drugs. Medicaid drug spending rose from $16.6 billion in 2000 to $26.6 billion in 2003—an average annual rate of growth of 17.1 percent. Drug spending slowed from 2002 to 2003, however, relative to the previous two years. Spending on services provided by prepaid/managed care plans increased from $26.5 billion in 2000 to $41.5 billion in 2003—an average annual increase of 16.1 percent. This reflects increased enrollment in managed care plans as overall program enrollment increased, but it may also reflect a continuing movement from fee-for-service (FFS) to managed care within Medicaid. If the latter, growth rates of all other acute care services would be lower as a result.

Spending on nursing home care increased 4.0 percent annually over the 2000–2003 period but fell $2.9 billion (6.0 percent) between 2002 and 2003. Nursing home spending grew 9.5 percent annually between 2000 and 2002, probably because the increased use of UPL programs. These programs function by transferring funds from local governments or providers to the state government. The states in return make higher rate payments to providers—in this case, nursing homes. The local share is usually returned to the local entity making the transfer to the state. Thus, the state generates federal matching payments with little or no state or local contribution. The drop in spending by 2003 may reflect some movement of care out of nursing homes, but it more likely reflects decline in the use of UPL programs. The Benefit Improvements and Protection Act (BIPA) of 2002 limits states’ ability to use these mechanisms, and, as new regulations are phased in, the effects have shown up as reduced spending on nursing home care.

Home/personal care services continue to increase rapidly, growing 15.9 percent over the entire period with no slowdown in 2003. Spending on home/personal care increased from $22.3 billion in 2000 to $34.7 billion in 2003. Such spending is still about $10 billion less than spending on nursing home care, but the absolute-dollar differences between institutional and noninstitutional care are clearly diminishing. For example, the annual rate of growth in spending during 2000–2003 for home/personal care was 15.9 percent, compared with 4.0 percent for nursing facilities and 5.0 percent per year for intermediate care facilities for the mentally retarded (ICF-MRs).

Finally, it is noteworthy that DSH payments fell 2.9 percent over the period. The Balanced Budget Act (BBA) set out state-specific federal DSH allotments between 1998 and 2002. BIPA froze DSH allotments in 2001 and 2002 at 2000 BBA levels but then returned them to levels specified in the BBA for 2003, causing sizable reductions in several states. DSH spending also fell in 2003 because of limits on DSH payments to mental institutions. Federal law restricting DSH payments was eased by the Medicaid Prescription Drug, Improvement, and Modernization Act (MMA) of 2003; in fact, DSH payments were actually allowed to increase in 2004.

Medicaid payments to Medicare grew from $4.7 billion to $6.3 billion during the period.¹³ This includes payments for Medicare premiums for all Medicaid dual-eligibles, as well as some cost sharing that cannot be allocated to specific services. Spending increased 9.5 percent between 2000 and 2002 and 10.4 percent between 2002 and 2003. The large increases may reflect the rising costs of prescription drugs. Participation in Medicaid may have become more attractive because of the Medicaid drug benefit.

Spending growth in acute versus long-term care. Exhibit 4 shows how the growth in each period was distributed across acute and long-term care services. Between 2000 and 2002, nearly two-thirds of health spending growth was for acute care services and payments to Medicare for premiums and copayments. From 2002 to 2003, more than 80 percent of Medicaid spending growth was attributable to acute care services and payments to Medicare. Growth in long-term care services accounted for less than 20 percent. During the entire period, a little more than two- thirds of the growth in Medicaid spending was attributable to acute care services.

Spending growth by eligibility group. We next attempt to estimate spending growth by eligibility group (that is, families versus the aged and disabled). The distribution of spending by service for families and the aged and disabled in FY 2000 is taken from the FY 2000 MSIS data. In that year, children and families accounted for more than 40 percent of spending on inpatient hospital, physician, lab and x-ray, and outpatient hospital services and more than 60 percent of the spending on prepaid/managed care, but only a very small share of all long-term care spending. Aged and disabled enrollees accounted for more than 70 percent of all spending, including 85 percent of spending on prescription drugs. The aged and disabled account for more than half of the spending on inpatient and outpatient hospital services and nearly all spending for long-term care. Thus, enrollment growth among children and families is particularly likely to affect acute care services, while enrollment growth among the aged and disabled is likely to affect virtually all services.

Changes in spending per enrollee. To analyze the contributions of enrollment growth and changes in spending per enrollee to overall Medicaid spending growth, we calculated a measure of spending per enrollee that controls for the effect of the changing composition of the Medicaid caseload. Simply dividing changes in spending by changes in enrollees would give a biased (downward) estimate of the growth in spending per enrollee. Estimates would be biased downward because the caseload had shifted toward a less costly group of people because of the faster enrollment growth among children and nondisabled adults.

We used MSIS data for 2000 on Medicaid spending by service and project this spending to increase by the same growth rates that we observed in the CMS-64 data for 2000–2003. To develop an unbiased measure of spending growth per enrollee, we divided the percentage change in spending for each service by a measure of enrollment growth that reflects the eligibility groups that use the service. For example, for a service largely used by the aged and disabled, the relevant measure of enrollment growth should be largely based on enrollment increases in these groups. Thus, to calculate a measure of enrollment growth specific to prescription drugs, we used MSIS data on the share of spending attributable to the aged and disabled (0.85) and families (0.15). For hospitals, enrollment growth among the aged and disabled was given a weight of 0.555 and families, 0.445. These service-specific weights for each group were then multiplied by the enrollment growth observed for each group to obtain a service-specific enrollment growth rate. Enrollment growth for each service was then divided into the growth in spending for the service to calculate the increase in spending per enrollee.

The results are shown in Exhibit 5; they reflect changes in spending per enrollee that are not affected by changes in the covered population. Overall growth in spending per enrollee between 2002 and 2003 was remarkably low: 4.2 percent per year, about two-thirds the annual rate of growth in spending between 2000 and 2003 (6.1 percent). The slower rate of growth in Medicaid spending per enrollee was largely attributable to the slower growth in long-term care spending per enrollee: 1.0 percent from 2002 to 2003 compared with 7.2 percent from 2000 to 2002. Acute care spending per enrollee increased 6.4 percent from 2002 to 2003, only slightly below its rate in 2000–2002 (7.3 percent). The rate of growth in Medicaid spending per enrollee for acute care compares favorably with estimates of the growth in spending per privately insured person (7.4 percent) and in private insurance premiums (13.9 percent) from 2002 to 2003.¹⁴

Exhibit 5 also shows growth in spending per enrollee for specific services.¹⁵ Spending on some acute care services (per enrollee) increased faster from 2002 to 2003 than from 2000 to 2002, and some rose at a slower rate. Inpatient hospital spending increased 6.1 percent from 2002 to 2003 compared with 3.7 percent from 2000 to 2002. Spending per enrollee in prepaid or managed care services increased 9.9 percent from 2002 to 2003 compared with 6.9 percent from 2000 to 2002. In contrast, spending on physician services declined from 4.5 percent from 2000 to 2002 to 1.4 percent from 2002 to 2003, and outpatient/clinic services, from 6.2 percent to –0.8 percent. Some of the slow growth in physician and outpatient services may be attributable to a shift from FFS to prepaid managed care, but it also very likely reflects the widespread reductions in provider reimbursement rates and benefits that were part of cost containment strategies enacted because of budget pressures during the economic downturn.¹⁶ Prescription drug spending per enrollee also grew at a slower rate from 2002 to 2003: 10.0 percent, compared with 13.9 percent from 2000 to 2002. This may also reflect state efforts to curtail prescription drug use and prices.

There was a particularly sharp decline in spending per enrollee for long-term care services from 2002 to 2003. Spending per enrollee on nursing home care increased 6.2 percent between 2000 and 2002 but fell 8.4 percent between 2002 and 2003. As described above, this may reflect some reduction in nursing home use but most likely reflects a decline in use of UPL programs. Spending per enrollee for care provided by ICF-MRs continued a very low rate of growth in spending per person. Per enrollee spending on home/personal care services actually increased from 11.2 percent in 2000–2002 to 13.9 percent in 2002–2003.

Growth in enrollment versus spending per enrollee. We next decomposed the growth of spending into growth in enrollment and growth in spending per enrollee (Exhibit 6). Spending totals and annual rates of growth differ somewhat from the spending growth in Exhibits 1 and 3 because we began with MSIS data on spending by eligibility group in 2000 and applied growth rates for each service from the CMS-64 data.

During 2000–2002, enrollment in Medicaid increased 9.1 percent per year; spending per enrollee, 3.8 percent; and overall spending, 13.2 percent. The slower growth in spending per enrollee shown in Exhibit 6 compared with Exhibit 5 is related to the fact that Exhibit 6 reflects the changing composition of the Medicaid caseload, which shifted toward more low-cost families. Spending on families increased much faster because of faster enrollment growth. Overall spending on families increased 19.0 percent, while spending on the aged and disabled increased 10.9 percent.

This pattern roughly continued during 2002–2003. Overall, Medicaid enrollment increased 5.9 percent, while spending per enrollee increased only 2.1 percent. Total spending increased 8.2 percent. Medicaid spending on the aged and disabled increased only 5.5 percent, while spending on families increased 14.2 percent. The slower growth in spending on the aged and disabled was attributable to both slower enrollment growth and slower growth in spending per enrollee, the latter largely because of the decline in nursing home spending. The 14.2 percent increase in spending on families was largely driven by enrollment growth of 7.1 percent. The growth in spending per enrollee of 6.6 percent per year for families was slower than the rate of growth in spending per person for private coverage.¹⁷ The growth in overall spending per enrollee (2.1 percent) was low because of a combination of a slower growth of spending per enrollee (4.2 percent independent of caseload changes) and a shift in Medicaid enrollment toward a lower-cost mix of enrollees.

Over the entire 2000–2003 period, Medicaid spending grew 11.4 percent per year. Medicaid enrollment grew 8.0 percent per year, and spending per enrollee, when changes in caseload are accounted for, increased only 3.1 percent annually. Spending for the aged and disabled grew 9.0 percent, and for families, 17.3 percent. The faster growth in spending for families, once again, was attributable to the much faster growth in enrollment. Over the entire period, spending per enrollee for the aged and disabled grew 5.9 percent, and for families, 6.5 percent. Both of these figures are relatively modest by national standards.¹⁸

Spending growth for families versus the aged/disabled. More than half of Medicaid spending growth over the 2000–2003 period was attributable to the aged and disabled (56 percent), compared with 44 percent for families (Exhibit 7). Between 2000 and 2002, almost 60 percent of the growth was attributable to the aged and disabled, compared with about 41 percent for families. This pattern was somewhat reversed between 2002 and 2003 largely because of the decline in nursing home spending attributable to the reduction in UPL programs. Thus, despite the much faster growth in enrollment, increases in Medicaid spending on families accounted for less than half of the growth in program spending over the 2000–2003 period. The aged/disabled, while growing more slowly in numbers, are so much more costly per person that they account for more than half of the spending growth.

Conclusions And Policy Implications

Medicaid spending increased by about one-third between 2000 and 2003, largely driven by enrollment growth. The program saw rapid increases in enrollment of children and non-disabled adults. Many more Americans were covered by Medicaid at the end of the period than at the beginning. There were few expansions of and even some reductions in eligibility standards during this period. Thus, the growth in enrollment was largely due to the economic downturn, which led to declines in incomes and made more people eligible under existing eligibility standards.

The enrollment growth among children and adults in Medicaid (as well as SCHIP) undoubtedly kept the uninsurance rate from increasing more than it otherwise would have.¹⁹ The large increase in Medicaid and SCHIP enrollment for children offset the decline in enrollment in employer-sponsored insurance; as a result, the number of uninsured children declined slightly. Medicaid enrollment among adults also rose, but the increase was much less. As a result, it only partially offset the decline in employer-sponsored coverage, and there was a large increase in the number of uninsured adults.

In general, health spending per Medicaid enrollee grew at a rate below that experienced in the private sector. States made policy choices that affected benefits and provider reimbursement rates, and in the end they controlled the rate of spending growth. The large growth in Medicaid spending over the period—10.2 percent per year—nonetheless represents fairly rapid growth in spending in an already high-cost program. Medicaid played its role as a safety net, providing coverage to those facing economic declines and loss of employer-sponsored coverage, but the result was a sharp increase in program costs.

Medicaid spending growth placed heavy burdens on state budgets and has contributed to the federal budget deficit. Early indications are that the 109th Congress will seriously consider reining in Medicaid spending growth. But it is important to recognize, as this paper has shown, that tight caps on Medicaid spending growth would not have allowed the enrollment increases we have seen. Without these enrollment increases in Medicaid, the number of uninsured Americans would have grown much more than it did, and there would have been strong pressure on local hospitals and clinics to increase the amount of free care provided. Cities, counties, and states would have had to finance this care with no federal matching payments.

Research for this paper was conducted with financial support provided by the Kaiser Commission on Medicaid and the Uninsured.

NOTES

1. S. Lueck, “Leavitt Is Veteran of Medicaid Overhaul,” Wall Street Journal, 17 December 2004.
2. Letter from the National Governors Association to Senate and House majority and minority leaders, 22 December 2004.
3. B. Bruen and J. Holahan, Medicaid Spending Growth Remained Modest in 1998 but Likely Headed Upward, Pub no. 2230 (Washington: Kaiser Commission on Medicaid and the Uninsured, February 2001); and B. Bruen and J. Holahan, Acceleration of Medicaid Spending Reflects Mounting Pressures, Pub no. 4056 (Washington: Kaiser Commission, May 2002).
4. B. Bruen and J. Holahan, Medicaid Spending: What Factors Contributed to the Growth between 2000 and 2002? Pub. no. 4139 (Washington: Kaiser Commission, September 2003).
5. National Bureau of Economic Research, “Report of the Business Cycle Dating Committee” (Cambridge, Mass.: NBER, 17 July 2003); and D. Boyd, The Current State Fiscal Crisis and Its Aftermath (Washington: Kaiser Commission, September 2003).
6. E. Ellis, V. Smith, and D. Rousseau, Medicaid Enrollment in Fifty States—June 2003 Data Update, Pub. no. 7237 (Washington: Kaiser Commission, October 2004).
7. The CMS-64 is updated periodically. We used the April 2004 version, and results may differ as more current data become available.
8. We now have access to 2001 MSIS person-level data for most states but have not completed the necessary data edits.
9. Overall, the CPS shows enrollment growth in Medicaid and SCHIP of 6.9 percent between 2000 and 2002, and 8.4 percent between 2002 and 2003. Enrollment of children increased 7.7 percent between 2000 and 2002 and grew even faster (10.4 percent) between 2002 and 2003. Enrollment of adults grew 5.6 percent between 2000 and 2002 and 5.2 percent between 2002 and 2003. These data are reasonably consistent with data from Health Management Associates (HMA), which show somewhat faster growth among families than the CPS, but the latter includes the disabled, which were growing at a somewhat slower rate. It is not possible in the CPS to reliably exclude the disabled, so they are included with both children and adults. If it were possible to do so, the CPS data would be closer to those of HMA.
10. J. Holahan and A. Ghosh, The Economic Downturn and Changes in Health Insurance Coverage 2000–2003 (Washington: Kaiser Commission, September 2004).
11. Congressional Budget Office March 2004 Baseline.
12. For the purposes of this paper, the “entire period” of analysis is defined as 2000–2003.
13. Payments to Medicare in 2003 use an estimate for Michigan based on growth in all other states over the 2002–2003 period. Michigan’s reported Medicare payments in 2003 are unusually high compared with those of earlier years (most of the spike being in Medicare copayments) and appear to be an error. The difference in the reported spending and our estimate is transferred to adjustments (in Exhibit 5).
14. B.C. Strunk and P.B. Ginsburg, “Tracking Health Care Costs: Trends Turn Downward in 2003,” Health Affairs, 9 June 2004,
content.healthaffairs.org/cgi/content/abstract/hlthaff.w4.354 (3 January 2005); S. Heffler et al., “Health Spending Projections through 2013,” Health Affairs, 11 February 2004, content.healthaffairs.org/cgi/content/abstract/hlthaff.w4.79 (3 January 2005); and J. Gabel et al., “Health Benefits in 2004: Four Years of Double-Digit Premium Increases Take Their Toll on Coverage,” Health Affairs 23, no. 5 (2004): 200–209.
15. The measure of enrollment growth includes all enrollees, not just enrollees who use the service.
16. V. Smith et al., The Continuing Medicaid Budget Challenge: State Medicaid Spending Growth and Cost Containment in Fiscal Years 2004 and 2005: Results from a Fifty-State Survey, Pub no. 7190 (Washington: Kaiser Commission, October 2004).
17. Strunk and Ginsburg, “Tracking Health Care Costs”; Heffler et al., “Health Spending Projections through 2013”; and Gabel et al., “Health Benefits in 2004.”
18. The growth in average annual health care spending per capita between 2000 and 2003 was 9 percent. Strunk and Ginsburg, “Tracking Health Care Costs.”
19. Holahan and Ghosh, The Economic Downturn.

John Holahan (jholahan{at}ui.urban.org) directs the Urban Institute's Health Policy Center in Washington, D.C. Arunabh Ghosh is a research assistant at the center.

DOI: 10.1377/hlthaff.w5.52
©2005 Project HOPE–The People-to-People Health Foundation, Inc.