urban institute nonprofit social and economic policy research

Market Competition and Uncompensated Care Pools

Read complete document: PDF


PrintPrint this page
Share:
Share on Facebook Share on Twitter Share on LinkedIn Share on Digg Share on Reddit
| Email this pageE-mail
Document date: March 01, 2000
Released online: March 01, 2000

The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders.

Note: For those wishing to print this report it is recommended that they use the PDF Version.


About the Series

Assessing the New Federalism is a multi-year Urban Institute project designed to analyze the devolution of responsibility from the federal government to the states for health care, income security, employment and training programs, and social services. Researchers monitor program changes and fiscal developments. In collaboration with Child Trends, Inc., the project studies changes in family well-being. The project aims to provide timely, nonpartisan information to inform public debate and to help state and local decisionmakers carry out their new responsibilities more effectively.

Key components of the project include a household survey, studies of policies in 13 states, and a database with information on all states and the District of Columbia, available at the Urban Institute's Web site. This paper is one in a series of occasional papers analyzing the information from these and other sources.


Contents

Introduction
The Uninsured in the Changing Medical Marketplace
The Evolution of Three States' Pools
Post-Deregulation Targeting of Support
The Challenges of Maintaining Pool Funding under Hospital Price Competition
Recent Developments
Assessing Pool Finance
Conclusion: Pools in Policy Context
Notes
References
About the Authors

The number of uninsured Americans has been rising for many years (Blumberg and Liska 1996).  Even in the midst of unprecedented growth in the economy and in employment and despite numerous state efforts to expand access to coverage, 44 million people, or 16 percent of the population, lacked coverage for all of 1998.  Over two-thirds of the uninsured had incomes less than twice the federal poverty level (Holahan and Kim 2000).  The main public safety net for the most urgent form of medical care is the willingness and ability of hospitals to provide free or reduced-fee care to the uninsured or underinsured, especially the poor (Baxter and Mechanic 1997).  Indeed, hospitals are the only medical provider legally required to see patients in extreme need (Bovbjerg and Kopit 1986; Jain and Hoyt 1992).  Most hospitals' charitable capabilities have been eroded by increased price competition, and the few hospitals in each area that provide large amounts of safety net care are especially burdened in the competition for paying patients (Fishman 1997; Norton and Lipson 1998).

This paper reports on three states that have long used pools to pay for uncompensated or charitable care, which spread costs beyond the hospital providing the care.  Massachusetts, New Jersey, and New York began pooling under comprehensive hospital regulation in the 1980s, but the pools remain important in the competitive era of the 1990s.  Information comes from literature review and personal interviews (Holahan, Bovbjerg, et al. 1997; Holahan, Evans, et al. 1997; Bovbjerg et al. 1998).  The key themes are:

  • These pools originally had the expansive goal of making all hospitals economically indifferent to patients' insurance status, relying on rate-setting regulation to control costs.
  • With deregulation, these states moved to rely on competition to control costs, but they also chose to continue mandatory pool funding because they recognized that competition would undercut hospitals' ability to self-fund safety net services.
  • The states found new sources of funding, in the process setting important legal precedent on states' legal authority over health care payers under federal ERISA legislation.
  • The states shifted to more limited goals for pools, focusing support more on charity than on bad debt and giving disproportionate assistance to the neediest safety net hospitals.
  • States face several tensions in the continuing redesign of their pools: maintaining broad political support for redistribution toward a small number of safety net hospitals, deciding how much to fund last-resort hospital care as against general access through prepaid insurance, and creating incentives for efficient substitution of ambulatory for hospital services in the absence of health plan enrollment.

We conclude with a policy discussion of these pools' efforts to raise and redistribute funds equitably and efficiently.  Pools are an intermediate policy option between broad, insurance-based strategy and narrow reliance on public hospitals.  They seem of most utility to states with a tradition of relying on private as well as public hospitals for safety net care.


The Uninsured in the Changing Medical Marketplace

The recent massive shift of paying patients into managed care has increased price competition and incentives for efficiency (Jensen et al. 1997; Holahan, Wiener, and Wallin 1998).  Managed care holds down prices and utilization of hospital care, to the benefit of paying customers (Levit et al. 1998) but often to the detriment of hospitals' ability to serve the uninsured.  It has long been recognized that increased hospital competition tends to drive out the cross-subsidies that cover the uninsured or underinsured (Bovbjerg and Curtis 1987; Pauly 1988).  Legal obligations create a minimum floor of access for the poor: nonprofits must provide a reasonable level of charity care for the poor, and all hospitals accepting Medicare or Medicaid must at least examine and stabilize all patients seeking care in emergency or active labor (Marsteller, Bovbjerg, and Nichols 1998; Jain and Hoyt 1992).  However, hospitals that provide more uncompensated care than their competitors' because of their location, traditional access patterns, or mission - incur higher costs and face added difficulty meeting their competitors' prices.  Only those with market power can readily earn extra surplus to cover their extra uncompensated care.

If there were complete insurance coverage, there would be much less need for a backstop of hospital charitable capability.  Then, everyone would enjoy broad access to all kinds of medical care, and medical services markets could function without the need to solve the problem of the uninsured and underinsured at the same time.  Competition could weed out inefficiency, but not at the expense of the poor.  However, in practice coverage is far from complete.  Some states are attempting to expand coverage in a number of ways, including Medicaid and related expansions (Holahan, Wiener, and Wallin 1998), new children's health coverage (Bruen and Ullman 1998), insurance market reforms (Bovbjerg 1992; Blumberg and Nichols 1997), and purchasing coalitions (IHPS 1999).  Until recently, declines in private coverage outweighed increases in public coverage (Holahan, Winterbottom, and Rajan 1995).  Since welfare reform, increases in employer-sponsored coverage have been more than offset by reductions in Medicaid and in private nongroup coverage (Holahan and Kim 1999).  The net result has been steady increases in the uninsured population.  In 1997-98, when 16.2 percent of Americans were uninsured, only five states had fewer than 10 percent uninsured residents, and six were above 20 percent (Campbell 1999).  Moreover, many of the insured lack comprehensive coverage, and even the well-insured can exceed their policy limits (Sulvetta and Swartz 1986; Swartz 1989), a fact recently highlighted by the advocacy of Christopher Reeve (Reeve 1998).  Accordingly, hospital access for the uninsured is a continuing problem.

Many localities traditionally served the uninsured by funding public hospitals (Friedman 1987; Altman et al. 1989).  Public hospitals provide access but give the uninsured no choice of hospital.  In addition, many areas lack public hospitals, and geographically unequal distributions of poverty heavily burden those jurisdictions that do fund charity care.  Finally, public hospitals themselves have difficulty funding themselves, given Medicare and Medicaid cuts and increases in managed care.

An alternative is to pool the burdens of uncompensated or charity care statewide, redistributing funds to hospitals that provide the most care.  Massachusetts, New Jersey, and New York have run such pools for many years.


The Evolution of Three States' Pools

These states' policies on uncompensated hospital care have evolved over time, as have the characteristic issues addressed by policymakers (table 1).  This section covers state action through the rate-setting era.  The next sections consider the complex changes made after hospital deregulation.

Early Policy: Hospital-Specific Markups

Traditionally, hospitals financed uncompensated care out of net revenues earned from paying patients, as well as other revenue, including public funds in the case of public hospitals and philanthropic contributions in the case of nonprofits.  In the early 1980s, the longstanding steady growth in insurance coverage ended, and the rate of uninsurance began to rise in response to recession, rapid health care cost inflation, and structural changes in labor markets.  These trends simultaneously increased the demand for uncompensated care and reduced hospitals' ability to pass their costs through to payers.

Hospital concerns about rising uncompensated care first attracted policy attention nationally in the early 1980s (Sloan, Blumstein, and Perrin 1986).  Then, too, hospital finances were under pressure in much of the country.  Uncompensated care was a special problem in New Jersey, which has no public general hospitals to serve the uninsured and which in the late 1970s had enacted an unusual requirement that hospitals treat all comers, regardless of ability to pay (Bovbjerg et al. 1998).

Hospitals in Massachusetts, New Jersey, and New York, and in other states with mandatory rate setting, by law could add a markup for uncompensated care to rates that payers were obligated to pay.  Starting in 1983, obligations applied even to Medicare, the largest hospital payer, under federal waivers for the three states to regulate all payers.  Thus, all insurers contributed to financing each hospital's uncompensated care.  Even so, care for the uninsured was highly concentrated.  In New York in 1981, 40 percent of care for the uninsured came from only 7 percent of hospitals, compared with 16 percent of hospitals for the insured (Spencer 1998).  Hospital-specific markups naturally varied according to the balance of uninsured to insured patients at each hospital (Gaskin 1997).  In New Jersey, for example, the hospital markups ranged from 1 percent to 25 percent of rates.  The approach was problematic for hospitals with high levels of uncompensated care.  In Massachusetts, a private payer won a lawsuit over the very high markup at one public hospital.

The Invention of Pools under Rate Setting

To avoid such disparities across hospitals, all three states pooled hospitals' uncompensated care and shared the funding burden across hospitals.  To replace the old hospital-specific add-on to rates, a uniform surcharge was applied.  The resulting funds were then pooled and redistributed to hospitals according to their amounts of uncompensated care.  Rate setting allowed the states to make hospitals with little uncompensated care charge paying patients rates higher than the hospitals otherwise would have.  The pools then accumulated the surcharges from all hospitals and distributed funds to hospitals to cover nonpaying patients.  Hospitals with low loads of uncompensated care were net contributors.  Those with high levels were net recipients.  The goals of pooling were to improve the financial condition of hospitals with high uninsured care loads, more equitably fund uncompensated care, and improve access for the uninsured by removing disincentives for hospitals, particularly private hospitals, to treat uninsured patients.  Surcharges to payers were thus attractive relative to having to budget state tax funds for hospital subsidies.

Massachusetts created its pool in 1985, and New Jersey followed suit two years later. Both set uniform statewide add-ons to hospitals' rates.1 At least initially, pool payments were usually intended to cover full (rate-setting-approved) costs at hospitals, thus greatly increasing the attractiveness of serving patients at risk of generating uncompensated care.  The uniform surcharge rate was computed periodically based on total uncompensated care compared with covered patient revenues, and payouts followed hospitals' shares of uncompensated care in a prior period.

New York's all-payer system, meanwhile, had incorporated an uncompensated care pool from its beginnings in 1983 (Thorpe 1987).  There, however, pooling occurred on a regional rather than statewide basis, mainly to keep subsidies across New York City institutions separate from those elsewhere.  New York did not cover all uncompensated care costs.  By design, public hospitals received much lower uncompensated care payment rates than private institutions, as they had other public funding for this purpose.

By and large, the pools worked as intended during the rate-setting period.  Research on the mid- to late 1980s impact of the pools has found that they did indeed increase provision of uncompensated care (Rosco 1990; Thorpe and Spencer 1991; Gaskin 1997; Spencer 1998).  Pools were not without problems, however, and policy shifted in response.

Problems under Rate Setting

The primary issue early on was the total amount of pool spending.  Although the pools redistributed the burden of uncompensated care across hospitals, their existence did nothing to control the average level of uncompensated care costs.  In fact, the availability of nearly open-ended pool reimbursement reduced hospitals' incentive to seek full payment from payers.  How much such moral hazard affected pool finance is unclear.  In New Jersey, one empirical study found that hospitals were better paid for uncompensated care than for insured care (Kronick 1990); another, however, suggested that state review of hospitals' collections efforts counterbalanced any moral hazard (Gaskin 1997).

In any event, states saw the aggregate amount of uncompensated care rise over time.  The add-on in New Jersey rose especially rapidly, to 19.1 percent in 1991, about a billion dollars in all.  The New York surcharge began at only 2 percent on average in 1983 ($150 million in aggregate) and rose to 5.48 percent by 1995 ($630 million) near the end of rate setting.  Creating explicit add-ons for uncompensated care, moreover, made those costs very visible and accordingly a target of increasing public scrutiny, especially as the redistributions grew in magnitude.  Particularly in Massachusetts and New Jersey, calls for reform increased.

Problems also arose in hospitals' ability to pass through costs to payers.  Soon after the start of pools, Medicare, the largest hospital payer, ceased to be included in these formerly all-payer state programs of rate setting.  Medicare's shift to its own prospective rates was advantageous for many hospitals - paying them as much or more than state regulators had allowed - but it eliminated a major contributor to pool funding.  In New Jersey, the add-on for remaining payers rose sharply.  New York imposed a new assessment on hospitals' inpatient Medicare revenues in order to make up the pool shortfall caused by Medicare's withdrawal from the main pool (Thorpe and Spencer 1991).  Funds went largely to hospitals deemed financially distressed; a new statewide pool made the allocations.  This assessment was later changed to a 1 percent assessment on all hospital inpatient revenue.

HMOs in Massachusetts posed a similar issue.  They had always been exempted from mandatory rate setting as a means of encouraging their growth (McDonough 1992; McDonough, Hager, and Rossman 1997).  Accordingly, they made no earmarked pool contribution, paying instead as little as they could negotiate.  This helped HMOs grow rapidly in the late 1980s, and the financial condition of the state's previously dominant Blue Cross plan began to deteriorate.

In all three states, Blues plans had been central to rate setting and the main contributor of pool surcharges.  Even beyond the surcharges, every Blues plan had difficulty adjusting to increased price competition.  In these states, the Blues plans also traditionally served as the insurer of last resort, facing adverse selection in enrollment as well.  The plans' growing problems also raised questions about the long-term stability of pool funding.

Political resistance to redistribution across hospitals also posed problems for pools.  By its nature, the pool surcharges for uncompensated care made costs very visible and, accordingly, a target of increasing public scrutiny, especially as the redistributions grew.  Particularly in Massachusetts and New Jersey, calls for reform increased.  Under increased fiscal pressure themselves, especially from drops in utilization, relatively better off hospitals were less willing to see revenues from their patients shifted to safety net institutions.  The split was clearest in New Jersey.  In the early 1990s, during the battle over systems reform, urban, often safety net institutions in New Jersey broke away from the state hospital association, largely over lobbying for pool funding (Bovbjerg et al. 1998).

A final development affecting pools was declining political support for rate setting.  Rate setting, when first adopted, seemed a very effective tool of cost control relative to the uncontrolled fee-for-service payment typical of indemnity insurance (Eby and Cohodes 1985; Schramm, Renn, and Biles 1986; National Hospital Rate-Setting Study 1988; Hadley and Swartz 1989).  It seemed much less effective relative to Medicare's administrative price setting under prospective payment and to the managed-care-driven price competition that had become a viable alternative in many states by the late 1980s (Robinson and Luft 1988).  Moreover, political sentiment was generally turning against many forms of government regulation.

Policy Begins to Shift under Rate Setting

A search for broader funding mechanisms began even under rate setting, where the add-ons were the dominant funding modality.  New York's assessment on Medicare revenues outside of rate setting was an early example.  A much larger financing consideration involved Medicaid disproportionate share hospital (DSH) payments.  Federal DSH participation is available for state Medicaid funds for hospitals but not other types of providers or managed care plans.  All three states in the early 1990s (beginning with Massachusetts) redesignated hospital payments from their pools as Medicaid DSH payments and received federal Medicaid matching payments on their pools.  New York changed its main pool's surcharges from regional to statewide operations in 1995 in large part to meet federal DSH requirements that provider taxes be uniform and broad-based.  In all three cases, federal matching payments revert to the state treasury, not to hospitals, which partly explains some of the states' continuing interest in and support for pools, separate from rate setting.  In 1995, all three were among the top nine states in amount of federal DSH spending (Liska et al. 1997).2

Massachusetts was the first state to deal with the issue of overall pool spending.  It did so in the context of insurance expansion.  The state passed a universal insurance coverage statute in 1988, planning to phase in a mandate on employers.  Big business supported the law but only on condition that their pool exposure be capped, and lawmakers recognized that the insurance expansion would reduce the need for a pool.  The law immediately capped private-sector obligations to the uncompensated care pool, but coverage mandates were to be phased in over time.  The first cap was set at $325 million a year and was scheduled to decline by 2 percent per year thereafter.  Pool funds began a long period of decline, especially relative to increasing need, as the coverage mandate was first postponed and then repealed (Special Commission 1997).

By the late 1980s, all three states were increasingly emphasizing insurance.  They were not merely implementing mandatory increases in Medicaid coverage for children but also developing their own initiatives.  At a minimum, this diverted political attention away from pool finance.  Sometimes funding was diverted as well.  The universal coverage bill in Massachusetts was the strongest example, but in the other states pool funds were increasingly used to fund insurance subsidies.  New York started its children's health insurance program (CHIP), which covered primary care and ambulatory services for young children.  In New Jersey, a policy recommendation from a 1991 gubernatorial task force also suggested insurance approaches in lieu of the pool.

Another shift in policy under rate setting was targeting greater assistance to certain hospitals rather than treating all equally under the pool.  New York went furthest here, creating additional pooling mechanisms apart from its main pool.  As just noted, its assessment on Medicare revenues was initially ear-marked for financially distressed hospitals.  After those pooled funds were partly diverted to help fund insurance, the state in 1991 helped financially distressed hospitals by adding another payer add-on, whose level rose from 0.235 percent to 0.325 percent ($37 million in 1995).  In order to aid public hospitals, which received less from the main pool than other hospitals with similar uncompensated care loads, New York added two other, non-pool subsidy programs.  One supplemental adjustment was made to Medicaid rates, mostly for public hospitals, beginning in 1989, and another adjustment followed in 1991.  The adjustments for public hospitals totaled $160 million and $261 million, respectively, in 1995.  In both cases, the locality puts up funding for the payment surcharges and, because the adjustment qualifies as a Medicaid DSH payment, the federal government matches it.3 Recall that public hospitals received proportionately less from the pool than others on the grounds that they received other local or state support.  Of the three states, New York has the largest non-pool subsidies relative to its main pool's size.

The End of Hospital Rate Setting

In the 1990s, all three states deregulated hospital rates.  In each case, the move to competitive pricing came after major electoral changes.  In Massachusetts, a strongly promarket Republican governor took office in 1991.  Deregulation was enacted on December 31st of that year, effective immediately (McDonough 1992; Holahan, Bovbjerg, et al. 1997).  In New Jersey, a 1991 "tax revolt" gave Republicans control of the legislature starting in 1992, with a veto-proof majority (Bovbjerg et al. 1998).  Then, early in 1992, a federal trial-level court struck down the New Jersey pool's surcharge for self-insured employers, ruling that it was preempted by the federal Employee Retirement and Income Security Act of 1974 (ERISA).  That adverse decision precipitated deregulation, which continued even after the decision was overturned on appeal.4   Unwilling to leave out self-insured payers and surcharge only those with commercial insurance, New Jersey dismantled all-payer rates and revised its pool as part of a comprehensive set of reforms, also affecting insurance regulation and the status of Blue Cross-Blue Shield (Bovbjerg et al. 1998).  Change came last to New York.  Adverse court decisions also occurred there, but a state appeal ended in a landmark victory in 1995.5   Reform came in 1996 legislation, in the second year of a new, conservative Republican administration (Holahan, Evans, et al. 1997).


Post-Deregulation Targeting of Support

When these three states deregulated hospital rates in the 1990s, all three kept their pools.  That they chose to do so is remarkable.  They recognized that the problems addressed by pools would not end with the end of state rate setting.  As New York's task force report recommending deregulation stated, "New York should continue to finance public policy goals unlikely to be addressed through market forces" (Pataki Task Force 1995).  After deregulation, distribution of pool funding continued to pose problems for the reasons just discussed.  In addition, the issue arose of how to raise pool funds under a system of competitive hospital pricing, which was handled differently in each state.  (See "Targeting of Support to Safety Net Hospitals" below.)

Concerned about spending levels, the states increasingly targeted which costs and which hospitals may receive reimbursement through the pools, continuing a trend started under rate setting.  Two states, Massachusetts and New Jersey, shifted their focus from uncompensated care to charity care, cutting back pool support for bad debts (Massachusetts continued to cover emergency bad debt).  In all three states, separate non-pool subsidies were also created in order to aid certain facilities.  Massachusetts, which had previously capped its pool, maintained the cap in its 1991 reform law and increased it somewhat in an additional 1997 reform.  New Jersey's 1992 reforms downsized pool support over time; the 1997 revisions increased it, but not to prior levels.  New York's single 1996 reform kept support at very similar levels.

Targeting of Support to Safety Net Hospitals

The initial 1991 reform in Massachusetts changed the distribution of pool funds in order to protect hospitals with high proportions of charity care, which were most affected by the previously set cap on the pool.  Although hospitals were able to set their own rates starting in 1991, these safety net hospitals would have needed very high charges in order to make up for the increasing shortfall in the pool.  It was assumed that other hospitals also treated a large number of uninsured but had greater ability to shift costs to payers.  Under the new formula, hospitals are allocated a share of the shortfall in the pool (the difference between total uncompensated care costs and pool funds) and a hospital's pool payment is reduced by its share of the shortfall.  The shortfall is distributed among all hospitals based on total patient care costs (effectively, hospital size).  Because two hospitals of equal size receive the same shortfall allocation, the hospital with greater uncompensated care costs has a higher proportion of those costs reimbursed.  This method was termed the "greater proportional requirement" method.

Universal coverage was never implemented in Massachusetts, although Massachusetts did create a number of new public programs covering, for example, disabled persons and unemployed adults (McDonough et al. 1997).  Over time, however, the uninsured rate in Massachusetts rose and the cap on the free care pool remained in place (Special Commission 1997).  Consequently, the gap between total uncompensated care and funds available through the pool widened and reached $170 million by 1997.  Because of the proportional requirement and the growing shortfall in the pool, the percentage of hospitals receiving pool funds on net declined from 31 percent in 1991 to 18 percent in 1996.  After the formula change, two hospitals, Boston City Hospital (BCH) and The Cambridge Hospital (TCH), received the bulk of the payments (51 percent of pool payments in 1997).

In addition to the free care pool, Massachusetts created two separate subsidy programs to aid certain publicly oriented hospitals (table 2).  Over $50 million was targeted to public hospitals or those serving mainly public beneficiaries.

Half of the funding came from state general revenue, the other half from federal matching payments through the Medicaid disproportionate share program.  The other program ($32 million in 1996) assisted the two city hospitals, TCH and BCH, by using intergovernmental transfers from the local governments in order to receive another $16 million in federal Medicaid matching payments, which were forwarded back to the hospitals.  A state-owned university hospital receives $10 million through a similar Medicaid DSH mechanism.  In total, non-pool subsidies are about one-sixth the size of the uncompensated care pool in Massachusetts, which is lower than in New Jersey or New York.

The 1997 reform did not affect these additional programs.  It did reduce the skewing of pool funding toward TCH and BCH by replacing some of their pool funds with new Medicaid demonstration funds (see "Recent Developments" on page 22), which also added further initiatives to increase insurance coverage.

New Jersey, like Massachusetts, used its 1992 deregulatory reform to create a more targeted pool for charity cases only (table 3a).  The rate-setting pool totaled $912 million in 1991.  After reform, the main pool was cut to $500 million in 1993 (plus $100 million for pool-like "other uncompensated care"); it subsequently was funded at $300 million in 1997.6

With reduced pool funds available, the distribution formula was adjusted to focus pool support on higher-needs hospitals.  Hospitals had to be in the top 80 percent in terms of charity care as a percentage of costs in order to qualify for funds.  Once eligible for funds, they were reimbursed for charity care costs at Medicaid payment rates up to the maximum available in the pool.  To provide a disincentive to inappropriate emergency care (i.e., care that could have been provided in another setting), these encounters were reimbursed at primary care rates.

Shortly after the 1993 changes in pool funding and distribution, New Jersey instituted additional special subsidies for hospitals with very high levels of uncompensated care.  One funding mechanism covers institutions with particularly high caseloads of patients with AIDS, TB, mental illness, substance abuse, or high-risk pregnancies.  The other covers those serving the mentally ill and developmentally disabled.  These two Hospital Relief Subsidy Funds totaled $110 million in 1993, rose to $143 million in 1994, and stayed about level through 1997.7  They became relatively more important as the main pool was reduced.

Pool funding declined over the next several years, and the gap between total uncompensated care and the charity pool's funding rose.  In addition, Medicaid payments were set in 1993 at $152 million less in aggregate than Medicaid rates under the all-payer system.  In 1995, with the main pool then at $400 million, New Jersey changed its distribution again in order to maintain funding for the highest-need hospitals.

The new distribution took into account measures of payer mix (or shiftability) and profitability.  Charity care funding was mainly dependent on a hospital's percentage of nongovernmental payers relative to other hospitals.  In 1996, this formula was modified in order to eliminate bad debt from the percentage calculation, but the profitability and shiftability criteria remained in effect.

New Jersey's 1997 reform did not change targeting (table 3b).

New York's charity care policy took a somewhat different direction.  New York did not formally cap its largest pool either during rate setting or after deregulation, although surcharges were set to meet estimated need and often fell short.  The state's targeting of funds to hospitals had begun early on, through the addition of supplementary pools and subsidies.  The state also has consistently continued to fund both charity and bad debt (table 4).

Targeting of the main, bad debt and charity care (BDCC) pool occurred with deregulation.  Distribution of funds based on hospitals' need for subsidy was changed under the new "Indigent Care Pool." The need percentage is defined for private hospitals as the ratio of inpatient bad debt and charity care plus outpatient deficits to inpatient and outpatient revenue.  Funds are distributed to institutions on a sliding scale according to need.  The higher the need percentage, the higher the proportion of costs reimbursed through the pool, just as income tax brackets apply increasing marginal rates at different brackets of income - except of course that the scale is used to distribute funds rather than to raise them.  Prior to reform, as of 1995, hospitals received 35 percent payment for their need in the 0 to 1 percent bracket, 50 percent for the next 1 percent level of need, and so forth.  Public hospitals' need was computed differently; they received proportionately less from the pool under a different formula in light of their access to more direct forms of public subsidy.  After reform, the payment rates relative to need were shifted to favor hospitals with higher uncompensated care (table 5).

Compared with the prior formula, the new approach requires hospitals to have higher uncompensated care loads to be eligible for any payment, but the pool pays a higher portion of uncompensated care costs for eligible hospitals and does best for very needy hospitals.  The new formula is phased in over three years for most hospitals.  For hospitals designated as financially distressed, the new formula is only partially phased in.  In 1999, for example, 50 percent of pool payments to these hospitals is based on the old formula and 50 percent is based on the new formula.

As already noted, New York under rate setting targeted support to particular hospitals through additional programs outside of the original BDCC pool, notably to hospitals deemed financially distressed and to public hospitals, which received low levels of BDCC support.  Other hospitals with high indigent loads became eligible for adjustments in mid-1991, but most of the funds ($265 million) have gone to public hospitals (table 4).  New York has the largest non-pool subsidies of the three states, relative to its BDCC pool size.

When New York dismantled its all-payer rate-setting system (New York Health Care Reform Act of 1996), it also fundamentally reformed its pools.  These reforms are effective from 1997 to 1999.  Approximately $1.2 billion is used for two pools, Indigent Care Pool and Health Care Initiatives Pool (HANYS 1996).  Total uncompensated care support to hospitals is maintained.  Hospitals will receive approximately $738 million from the pool, community health centers $48 million.  Approximately one-third of the funds go to CHIP ($109 million in 1997), other insurance subsidies ($24 million in 1997), and public health initiatives ($136 million) and assistance for providers to make the transition to a new competitive system, for example, through worker retraining (up to $100 million).  In addition, coverage through CHIP was expanded to include coverage for inpatient hospital care, offsetting some hospital uncompensated care.

Under the new Indigent Care Pool, public hospitals will receive the same amount as they did in 1996 from the BDCC pool and will continue receiving supplemental Medicaid adjustments, funded with local funds and federal Medicaid matching payments separate from the $738 million Indigent Care Pool.  The subsidy for financially distressed hospitals is being phased down over multiple years; need was redefined, replacing outpatient deficits with outpatient bad debt and charity care; and the need formula has become more skewed, favoring high-indigent-load hospitals.  An additional $36 million continues to be set aside for high-need adjustments to other hospitals.

Targeting to Charity Rather Than Bad Debt

Two of the states also moved to target funding to charitable services.  When Massachusetts dismantled its all-payer rate setting at the end of 1991, it cut back support for hospital bad debt.  Reimbursing hospitals for bad debt, it was felt, provided a disincentive for hospitals to pursue collections from individuals and payers.  Massachusetts therefore changed its pool, excluding bad debt reimbursement unless it was generated by emergency services provided to uninsured patients.8

New Jersey cut out bad debt altogether.  Bad debt was perceived as a major reason for the huge growth in the pool toward the end of rate setting.  Even before deregulation, the state began to require extensive documentation and standardized procedures for debt collection.  With deregulation in 1992, bad debt was excluded altogether from the pool, although hospitals' difficulties obtaining documentation of eligibility for charity care led to 1995 legislation to allow coverage of some bad debt for the prior year.

Demographic information about pool patients is required in both Massachusetts and New Jersey, because eligibility for charity is related to patient income.  In both states, patients with incomes above 200 percent of the federal poverty level (FPL) are subject to cost-sharing.  For example, in New Jersey, patients with family income between 275 and 300 percent of the FPL must pay 80 percent of a hospital's charges, with a cap at 30 percent of the family's income.  Persons above 300 percent of the FPL do not qualify for subsidy.  In Massachusetts, partial free care is available to families with incomes between 200 and 400 percent of the FPL but none is provided to families above 400 percent of the FPL except in extreme hardship cases.  Thus, the pool operates in some sense as a means-tested substitute for insurance coverage, paying for services as provided.

Targeting Nonhospital Care and Coverage

All three states in the early 1990s began using pool funds for providers and services beyond hospitals.  In Massachusetts and New York, community health centers became eligible for pool funds.9  New Jersey covers community health center charity through a different program created under the Health Care Cost Reduction Act of 1991.  In New Jersey and New York, add-ons were created to fund primary care and similar health care services.  Although the amounts are less than 10 percent of the pool, this reflects an increasing emphasis on improved access to coordinated primary and preventive care, in contrast to potentially avoidable emergency room and inpatient care provided by hospitals.  Hospitals were not viewed as having - and most did not have - adequate outpatient capacity to address the primary and preventive care needs of the uninsured.10

Overall, pool costs are not high per uninsured person.  In 1997, support per uninsured person was highest in Massachusetts ($475) and lowest in New York ($277).  Adding in similar non-pool support for charity in hospitals reduces the ratio between high and low (table 6).


The Challenges of Maintaining Pool Funding under Hospital Price Competition

Reform required refunding the pools.  Rate setting had facilitated funding because it allowed mandatory collection from all covered payers at state-set levels.  Under competition, each hospital negotiates on its own, and many payers can resist contributing toward uncompensated care.  So the largest challenge for these states after deregulation has been to identify a reliable source of revenue that can increase with rising levels of charity care, appears equitable, and is satisfactory to local residents and providers.  Traditionally, imposing state requirements directly on payers was problematic under judicial interpretations of ERISA.  Each state has devised a somewhat different approach: Massachusetts and New Jersey have each reformed pool finance twice, mainly for funding reasons; New York has had one reform.

Federal ERISA Limitations on State Authority

A major issue for states for many years has been broad court interpretations of the federal ERISA statute, like the one in New Jersey that prompted deregulation.11  ERISA preempts state regulation of employee benefit plans other than through insurance regulation12 and thus clearly bars mandates on employers to provide health coverage as a way of addressing uncompensated care.  The state of Hawaii constitutes the only exception, for its mandate on employers predates ERISA.

For many years, it also appeared that nearly any state action that affected such health plans was federally preempted, including even judicial imposition of personal injury awards (Employee Benefit Research Institute 1995; Mariner 1996).  Moreover, while under ERISA a state can regulate licensed insurers, it cannot directly regulate self-insured plans, which constituted half or more of the market in these three states as of the late 1990s (Holahan, Bovbjerg, et al. 1997; Holahan, Evans, et al. 1997; Bovbjerg et al. 1998).  States can raise funds through premium taxes on insurance, for example, but such taxes may not be imposed on self-insured plans.  Political reality also imposes constraints.  According to one New York interviewee, each state also hesitates to enact unusually large premium taxes because it fears other states will retaliate with similar taxes against its carriers operating in those states.  Nonetheless, many states want broad payer responsibility to help fund charity care or other state health initiatives.

Recent court rulings give states more latitude in funding.  The U.S. Supreme Court issued its landmark Travelers decision in April 1995, unanimously and somewhat surprisingly upholding New York's rate-setting surcharges on commercial payers and HMOs.13   The court saw the rates as an allowable exercise of general health care regulation pursuant to states' inherent power to promote health and welfare, which Congress did not mean to preempt.  The surcharges did not unacceptably affect the basic structure of the affected plans or their uniform administration; they had only "indirect economic effects" on employee benefit plans that were not substantial enough to trigger preemption, even though they totaled 24 percent.14

In June 1997, the Supreme Court upheld another New York pool-funding mechanism, a very small tax on provider gross receipts (0.6 percent), even though the health care facility challenging the tax was owned by an ERISA benefit plan.15   After these two decisions, ERISA no longer seems an insurmountable barrier to broadly funding uncompensated care through indirect levies on health payers.  Of course, states still have the same power they have always had to impose a direct tax of more general application, such as a payroll or income tax levy.

Pool Funding in Massachusetts

When deregulating in 1992, Massachusetts shifted pool contributions from a surcharge on rates, seen as essentially an imposition on private payers, to an assessment on hospitals' private-payer revenue, which was clearly a hospital obligation.  The rationale was that ERISA would have preempted any assessment on payers.  The pool funding cap stayed at $315 million.

In 1997, pool financing was reformed again, as political opposition to redistribution had grown, demand for a rise in the cap had increased, and state options had expanded in the wake of the Supreme Court's loosening of ERISA restrictions.  Prior to the second reform, hospitals argued that it had become more and more difficult to add their assessment into their charges, since insurers were negotiating lower rates without regard to hospital obligations.  The shortfall grew larger, and hospitals faced with competitive forces began to feel a greater pinch from uncompensated care.  Consequently, they became less willing to support the pool.  As the two largest safety net hospitals, BCH and TCH, claimed ever larger shares of the limited pool, support for redistribution ebbed.

Political dispute over reform culminated in a 1996-97 special blue-ribbon commission.  The resulting 1997 legislation lowered hospital assessments to $215 million in aggregate.16   To make up the shortfall, private payers are to contribute $100 million directly to the state (table 2).  Beginning in 1998, private payers - including individuals and third-party administrators paying on behalf of self-insured plans - are being billed by hospitals but must remit approximately 2 to 3 percent of hospital payments directly to the state.17  If they do not, they face a higher levy in the form of a sales tax on hospitals that the legislation authorizes hospitals to collect from payers, plus a collection fee paid to the hospital.

In addition, pool funding for the two big safety net hospitals was cut by some $70 million, increasing the amount available to other hospitals that provide charity care.  The two hospitals instead got that amount in separate additional funding from a managed care initiative under a new Medicaid demonstration.  Other hospitals will have to contribute less into the pool than previously and, because BCH and TCH are taken out of the pool, will receive more in return.  Over time, the amount that hospitals will receive from the pool will decline from $330 million in 1990 to $231 million in 2002, as free care is expected to decline because of coverage expansions in the demonstration.

Pool Funding in New Jersey

New Jersey took a different approach.  After the adverse ERISA decision on its hospital surcharge, reformers turned to other funding sources for its slimmed-down pool (table 3a).  For the first three years after deregulation, the pool was funded from surpluses in the state unemployment insurance (UI) trust fund, on a provisional basis.  The use of the UI trust fund was set to expire at the end of 1995, but after heated debates and a short lapse, its use was extended two more years, through 1997.  Business and labor particularly opposed tapping the UI trust fund, but partly because UI premiums had been declining, the transfer continued.  Governor Whitman proposed cigarette taxes as a funding source in 1995 but lacked bipartisan support.

Ultimately, in mid-December of 1997, a second round of pool reform solidified future funding.  Tobacco taxes were doubled as the main funding source for the next five years of the pool, 1998 through 2002, with state general revenues expected gradually to supplant UI funds over that period (American Health Line 1997; Jackson 1997).  Thereafter, pool funding is expected to be a regular budget item, subject to the normal appropriations process each year.  New Jersey's pool-like Hospital Relief Funds - DSH-matched, additional subsidy programs created for targeted hospitals after deregulation - have relied on somewhat different mixes of funding sources (tables 3a and 3b).

Among the three states, debates over financing of charity care were arguably the most acrimonious in New Jersey, where the pool had become very large by the end of rate setting, mainly to cover bad debt.  In New Jersey, debates were so heated that pool funding has lapsed twice, once in 1992 under rate setting and again in early 1996 after deregulation.

Pool Funding in New York

New York's 1996 reform sought to tap both hospitals and payers to fund uncompensated care and other initiatives,18 somewhat similarly to the second Massachusetts pool reform that followed in 1997.  Although the financing changed, the total amount of funds to hospitals did not (HANYS 1996).  Starting in 1997, hospitals are to pay a 1 percent assessment on inpatient revenue.  In addition, all payers, except Medicare, are requested to register and pay a patient services assessment on payments for hospital inpatient or outpatient care, clinics, and laboratories (but not physicians' services), as part of the cost of purchasing such institutional services.  Paid directly to the state pool, the amount of the patient services assessment is 5.98 percent for Medicaid and 8.18 percent for private payers and workers compensation.  Private payers negotiate payment rates directly with providers and may elect to pay the service assessment on an encounter basis without paying the state directly.  In that case, however, hospitals are entitled to charge an additional 24-percentage-point surcharge to such payers (on top of the 8.18 percent assessment), a figure similar to the combined effect of the two surcharges upheld in Travelers.

This elective feature was designed to keep the assessment from being a mandatory tax but to strongly discourage payers from paying as part of their providers' rates.  Providers felt that if they had to collect the 8 percent from competitive payers, the payers would just reduce the hospital payment rate as an offset.  So hospitals wanted payments made directly to the state in order to separate negotiation over rates and collection of the assessment.  New York also relied less heavily than Massachusetts on the provider assessment because there was fear that, without mandatory rate regulation, it could not be effectively passed through to payers.

Reformers thought that the Travelers style of add-on surcharge on hospitals would not work well under deregulated rates, and they hoped that giving payers a choice of payment method would help the state avoid an ERISA challenge.  The 24 percent surcharge noted above coincides with the commercial plan assessment of 11 percent on top of a 13 percent rate differential, which Travelers had upheld.19   More recently, the Supreme Court upheld a New York gross receipts tax on health facilities, which included health centers operated by self-insured plans,20 so the state has some legal precedent for its approach.


Recent Developments

Medicaid Demonstrations and Use of Pool Funds

All three states have recently undertaken initiatives to revise uncompensated care pools and expand coverage under Medicaid.  With these reforms, each state exhibits a different level of attentiveness to the financial situation of providers.

In Massachusetts, reform of the free care pool was part of a renewed effort to expand health insurance coverage in the state.  In 1994, Governor Weld proposed a coverage expansion as part of a Medicaid research and demonstration waiver (Section 1115).  The proposal would fold certain state-only funded programs, such as CommonHealth and the Medical Security Plan, into Medicaid in order to receive federal matching payments.  The added funds were to be used to expand coverage for children (up to 133 percent of the FPL) and long-term unemployed adults.  In addition, Governor Weld proposed using part of the pool to finance a new program, the Insurance Reimbursement Program (IRP), for low-income workers.

Under the IRP, employers would receive subsidies for providing health insurance to low-income workers.  The legislature modified the governor's proposal by extending coverage to more children (up to 200 percent of the FPL for children up to age 12), authorizing the IRP for smaller employers than originally proposed, and creating a supplemental payment to the two public hospitals.21   The latter program was designed to secure $70 million for the two hospitals to offset the reduction in their pool payments.  The funding is provided in exchange for designing a managed care program for low-income patients enrolled to receive care at the two hospitals.  Financing under the final compromise included a tobacco tax, additional federal funds through matching payments, and a federal match on the intergovernmental transfers from the two public hospitals.  With the addition of this new program, Massachusetts has the highest non-pool subsidies per uninsured person of the three states - $187, compared with $145 in New York and $124 in New Jersey (table 6).

In March 1997, New Jersey submitted a Section 1115 proposal to redirect a portion of pool funds for a managed charity care program, but without new sources of funding.  Given the overall shortfall in the pool, good management of charity seems unlikely to relieve the financial burden on hospitals.  Since that time, the passage of CHIP in the Balanced Budget Act of August 1997 has ensured additional state and federal funding for new insurance.  While only some of this will find its way to hospitals, from a patient standpoint, access to primary and preventive care should increase.

In pursuing its recent Medicaid demonstration, New York is primarily focusing on enrolling Medicaid beneficiaries in managed care, not eligibility expansion.  The eligibility expansion is very limited and consists of enrolling persons with coverage through Home Relief, the state medical assistance program, into Medicaid.22   Although the state is seeking to reduce expenditures through managed care, it will retain total pool spending and numerous non-pool subsidies for hospitals.  The intent is to move hospitals to greater efficiencies and, ultimately, lower costs, while retaining reimbursement for charity care.


Assessing Pool Finance

Using bad debt and charity care pools has advantages and disadvantages for states.  The main advantage is that pools appear to maintain access for the uninsured to safety net hospitals.  Redistribution of charity burdens through pools is one way of lessening some of the potential adverse effects of competition on a hospital's ability to serve the uninsured.  Increased competition in the hospital marketplace should lead to more efficient production of services, but it is difficult for hospitals to compete and provide large amounts of uncompensated care.  Uncompensated care imposes costs on hospitals for which there are no revenues; in a competitive environment, hospitals cannot pass these costs on to paying patients through higher charges.  Redistribution through pools allows market competition to take place while at the same time leveling the playing field between hospitals that serve large numbers of nonpaying patients and those that serve few.  Empirical analyses have documented improved access under rate setting, and personal interviews with multiple sources suggest that, in these states, indigent access to hospital care remains reasonable after deregulation.

Pools are also likely to be much less costly than universal coverage because they provide limited support to targeted hospitals that provide the most urgent charitable care to poor people.  Even beyond pooling for uncompensated acute care generally, pool-like mechanisms can be used to provide even more targeted support to hospitals providing care to certain patients - for instance, those with AIDS, mental illness, tuberculosis, substance abuse, or high-risk pregnancies, as in New Jersey.  While, in general, pool funding has been used to support hospital inpatient care, the states reviewed in this paper are also beginning to support nonhospital primary care and preventive services.

At the same time, there are a number of problems that pooling mechanisms have to overcome.  First, raising the funds is harder under a competitive regime than under a fully regulated all-payer system or even in an indemnity insurance-dominated hospital market.  States clearly have authority to assess hospitals for funds under any reading of ERISA constraints on their power, and federal law considers such assessments matchable by DSH funding under Medicaid.  Yet, whereas hospitals could once routinely pass through assessments, and by law under rate setting, today large private payers can avoid or reduce hospital markups by credibly threatening to move their business elsewhere in search of a better rate.

Funding was easier in the mid-1980s because most private payers were fully insured, their insurers were clearly subject to state insurance regulation, and ERISA challenges were rare.  Today, in contrast, hospital payment comes mainly from self-insured employee benefit plans and the state must avoid coming too close to direct state regulation of ERISA-protected employee benefit plans, although the U.S. Supreme Court appears to have given states new leeway.  In the past year, ERISA-conscious redesign of pool funding in Massachusetts and New York has sought to make payers bear a share of funding by strongly encouraging them to pay the state pool, without directly imposing a tax.  After market renegotiation of hospital contracts, however, these new assessments may yet be borne mainly by hospitals "downstream" from health plans in the flow of funding, rather than solely by plans' employment-groups customers "upstream."23

Second, it is hard to target pooled funds fairly and efficiently.  Consider, initially, the patient population to be helped.  Over time, states have become more likely to allocate funds in proportion to charity care rather than bad debt, and enforcement raises new issues.  States must define levels of patient or family income that make recipients of services a charity admission rather than a potential source of bad debt, but more investigation is needed to determine how well hospitals and state overseers can distinguish between the two.  It is difficult to verify patients' self-reported income information, and securing documentation adds expense.  Large administrative efforts also add time; before the 1997 revisions in Massachusetts, there was a six-year backlog for finalizing settlements (Special Commission 1997).

Pool funds can be distributed disproportionately to hospitals providing high levels of uncompensated care on the ground of greater need.  Special revenues can also be set aside for public hospitals, whose mission it is to be the provider of last resort.  Such targeting has increased in the three states, especially after deregulation.  While it is an understandable reaction to caps or cuts in funds, targeting of safety net hospitals also curbs access to non-safety net institutions and can be argued to be unfair treatment of nonprofit facilities that view indigent care as integral to their mission.  These states have all targeted at least some funding in such ways, but numerous changes have occurred, partly because of shifting political support for redistribution.  Similar difficulties attend the issue of whether pool revenues should be allocated only to hospitals or instead spread more widely to cover some primary care and preventive services.  In Massachusetts and New York, community health centers get some pool funds; in New Jersey, a demonstration was planned to have hospitals spend some pool monies on nonhospital charity care, in managed care fashion.  But all such distribution issues are difficult - and made more so by political campaigning in favor of certain hospitals.  None allows the trade-offs across types of care possible under insurance in the care of an individual patient.

A third problem with bad debt and charity care pools is moral hazard.  Pools intentionally seek to promote care to the poor; or, more accurately, to make access to hospital care neutral with respect to patient income and insurance status.  Pools are not meant to promote inefficiency, yet may as a side effect create incentives for profligacy, starting with defensible accounting shifts to increase the share of costs eligible for a pool.  Beyond that, hospitals may incur high uncompensated care costs not simply because of the poor patients they attract, but also because of their poor management practices.  Pool subsidies somewhat undercut the new competition's incentives to improve efficiency.  So pool administration, like rate-setting administration before it, has had to impose various screens or find other judgmental ways of avoiding inefficiency.  Again, such measures pose issues of administrative efficiency as well.

Fourth, subsidizing bad debt and charity care through pools may help protect some hospitals with poor quality standards that cannot attract paying patients.  Distributing the money directly to individuals through expanded insurance would allow individuals more directly to "vote with their feet," taking their money elsewhere and improving, or eliminating, poor hospitals in the process.  The pool and pool-like subsidy programs reduce payments to hospitals attracting fewer patients, but do so less directly and with a time lag.

The three states have all addressed these issues and have decided politically that the pros outweigh the cons.  A detailed study remains to be made.


Conclusion: Pools in Policy Context

How to provide a health care safety net for people without health insurance is a continuing, if not a growing, issue for government.  Insurance coverage continues to decrease despite numerous public initiatives to expand access to Medicaid and private coverage.  Most hospitals maintain some level of charity care, but some face far higher demand for charity than others and have higher bad debt as well.  Further, almost all hospitals have seen their ability to fund charity reduced during this era of excess capacity, vigorous selective contracting by managed care organizations, increased hospital price competition, and cuts in Medicare and Medicaid.  Competition promotes efficiency but also squeezes out cross-subsidies.

Policymakers in Massachusetts, New Jersey, and New York concluded that the ability of individual hospitals to finance charity care on their own falls below levels of socially perceived need.  Hence, the states reaffirmed the utility of the hospital uncompensated care pools they created in the 1980s, though more targeted in recipients, both patients and hospitals.

Originally begun under mandatory hospital rate setting, the pools served to assess all covered hospital payers and to reallocate funds to hospitals disproportionately serving the poor and uninsured.  The original goal was to make hospitals economically indifferent to the insurance status of potential patients.  All three states have continued their pools even after deregulating hospital rates and emphasizing increased insurance coverage to improve access and reduce uncompensated care.  In Massachusetts and New York, careful statutory design seeks to continue the earlier practice of making payers contribute to a state-run pool, to the full extent allowed under federal ERISA law.  New Jersey has no provider or payer assessment; for several years, its pool has had temporary funding from surpluses in the state unemployment compensation insurance fund, an indirect assessment on payers.  In 1997, the pools cost only $300 to $500 per uninsured person in these three states (table 5), and there is some evidence suggesting reasonable hospital access.

With regard to the twin social goals of promoting access to appropriate levels of care and maintaining cost-effectiveness, pool subsidies are probably inferior to more thoroughgoing support for health coverage, although the latter would likely prove more expensive overall because of its more comprehensive coverage.  The bottom line is that bad debt and charity pools are not a perfect solution.  There are many significant problems to be addressed, especially balancing uneven distributions of need with political resistance to redistribution and encouraging better management of the care provided, especially to repeat users of free care.

Pools can still be useful policy tools.  They may be a useful adjunct for states that have expanded coverage to the extent feasible, or they may be a helpful stop-gap for states that cannot promote coverage.  Pools can serve an important role in leveling the playing field among hospital competitors, making it possible for the system as a whole to gain the benefits of increased competition without sacrificing its ability to provide adequate amounts of uncompensated care to low-income individuals.


Notes

1. Massachusetts, Chapter 574 of the Acts of 1985; Weissman, Lukas, and Epstein 1992; Deloitte & Touche 1996; New Jersey Pub. Law 1986, chapter 204; Gaskin 1997.

2. In addition, all three states have supplemental payments to publicly owned hospitals, whereby the state Medicaid share is funded through local funds, transferred to the state, and matched with federal Medicaid funds.

3. Most of the funds ($265 million) go to public hospitals, although other hospitals with high indigent loads became eligible for adjustments in mid-1991.

4. ERISA is codified at 29 U.S.C. 1001 et seq. and broadly preempts "any and all state laws insofar as they . . . relate to any employee benefit plan," 29 U.S.C. at 1144(a). The decisions affecting New Jersey and New York are discussed later.

5. New York state lost at the trial level and also on appeal (unlike New Jersey), but was upheld by the U.S. Supreme Court in a landmark ERISA case.

6. Early on, New Jersey hospitals sometimes had trouble documenting charity care at the level of allowed amounts (as against bad debt of nonpoor patients), so the exact amount of pool funding delivered could be lower. By 1997, documented charity exceeded the pool amount.

7. A very similar fund, the Hospital Relief Fund - Mentally Ill/Developmental Disabilities, was funded at $15 million in 1993 and rose to about $18 million for 1997. In addition, the state-owned university hospital qualifies for a separate Medicaid disproportionate share payment.

8. Emergency services were defined quite broadly and included admissions originating in the emergency department.

9. In Massachusetts, hospital-affiliated community health centers were eligible for pool funds beginning in 1985; however, freestanding community health centers did not become eligible until 1991 (Massachusetts, Chapter 495 of the Acts of 1991).

10. In related developments, all three states have also funded substantial new insurance coverage that is expected to reduce charity care and bad debt by more than 10 percent.

11. As in New Jersey, but not in New York, Connecticut deregulated in response to an ERISA-based court decision (American Health Line 1994).

12. Metropolitan Life Insurance Co. v. Massachusetts, 471 U.S. 724 (1985).

13. As already mentioned, trial and circuit courts in New York had struck down the surcharges, whereas in New Jersey the circuit court had upheld rate setting. Compare Travelers Insurance Co. et al. v. New York, 813 F. Supp. 996 (SDNY 1993), affirmed 14 F. 3d 708, 718 (2d Cir. 1994) with United Wire, Metal and Machine Health and Welfare Fund v. Morristown Memorial Hosp., 995 F. 2d 1179, 1191 (3d Cir. 1993), cert. denied, 510 U.S. (1993). This difference in ERISA interpretation was one reason the U.S. Supreme Court agreed to hear the case, or accepted "certiorari"; see New York State Conference of Blue Cross and Blue Shield Plans v. Travelers Insurance Co., 514 U.S. 645 (1995). New York's tax on provider gross receipts also used for the pools was not directly challenged in this case, but parties perceived that it would be affected by the ruling on the surcharges to payers. The tax was upheld in the DeBuono case two years later.

14. Commercial insurers were billed an 11 percent surcharge above a rate already surcharged 13 percent above the baseline hospital diagnosis-related-group rate billed to Blue Cross, with the proceeds turned over to the state; see Travelers, above. In addition, HMOs were surcharged up to 9 percent of their aggregate monthly inpatient charges based on their enrollment of Medicaid patients.

15. DeBuono v. NYSA-ILA Medical and Clinical Services Fund, 520 U.S. 806 (1997).

16. Massachusetts P.L. 1997, Chapter 47, as amended by Chapter 170, codified at Mass. General Laws, chapter 118G. Payers' surcharges apply both to hospital payments and to ambulatory surgical centers' payments.

17. The statute provides that the ultimate responsibility for payment lies with the patient, not the payer acting on the patient's behalf, and that hospitals do the billing. Both provisions were thought to help ensure survival against ERISA attack. The precise surcharge amount is established by the state's Division of Health Care Finance and Policy, Executive Office of Health and Human Services, under implementing regulations 114.6 Code of Mass. Reg'ns 7.01-7.18.

18. Health Care Reform Act of 1996, Pub. Hlth Law, sect. 2807-j, -s, and -t.

19. New York State Conference of Blue Cross and Blue Shield Plans v. Travelers Insurance Co., 514 U.S. 645 (April 26, 1995). Blue Cross plans were not required to pay the surcharge.

20. DeBuono v. NYSA-ILA Medical and Clinical Services Fund, 520 U.S. 806 (1997).

21. The compromise was reached in two stages: Chapter 203 of the Acts of 1996, which included the children's coverage expansion and the tobacco tax, and Chapter 47 of the Acts of 1997, which included authorization for the IRP and reform of the free care pool. Chapter 203 also repealed the employer mandate, which had been repeatedly delayed.

22. This permits the state to receive federal matching payments for expenditures on this population. New York already claims federal Medicaid match for inpatient expenditures on the Home Relief population. Under the waiver, nonhospital expenditures also would be eligible for match.

23. Market forces now set what rates payers pay hospitals, and if hospital costs go down because the state lowers its assessment, one would expect competition to drive down prices to payers as well. Blue Cross and Blue Shield of Massachusetts, still a very large payer there, sought to reduce hospital payments in December 1997 by the amount of its new surcharges under the 1997 reform set to begin January 1, 1998. This seeming end-run around new legislation prompted threats of state action (Pham 1997a), and Blue Cross quickly backed down (Pham 1997b).


References

Altman, S., C. Brecher, M.G. Henderson, and K. Thorpe. 1989. Competition and Compassion: Conflicting Roles for Public Hospitals. Ann Arbor, Mich.: Health Administration Press.

American Health Line. 1994. "Connecticut: Hospital Regulation Comes to an End." American Health Line. Washington, D.C.: National Journal. http://nationaljournal.com/pubs/healthline April 4.

----.1997. "New Jersey: Whitman Signs Charity Care Funding Plan." American Health Line. Washington, D.C.: National Journal. http://nationaljournal.com/pubs/healthline December 22.

Arnould, R.J., R.F. Rich, and W.D. White, eds. 1993. Competitive Approaches to Health Care Reform. Washington, D.C.: The Urban Institute Press.

Baxter, R.J., and R.E. Mechanic. 1997. "The Status of Local Health Care Safety Nets." Health Affairs 16 (4, July-August): 215-22.

Biles, B., C.J. Schramm, and J.G. Atkinson. 1980. "Hospital Cost Inflation under State Rate-Setting Programs." New England Journal of Medicine 303 (12): 664-8.

Blumberg, L.J., and D.W. Liska. 1996. The Uninsured in the United States: A Status Report. Washington, D.C.: The Urban Institute Press.

Blumberg L.J., and L.M. Nichols. 1997. Health Insurance Market Reforms: What They Can and Cannot Do. Washington, D.C.: The Urban Institute.

Bovbjerg, R.R. 1992. "Reform of Financing for Health Coverage: What Can Reinsurance Accomplish?" Inquiry 29 (summer): 158-75.

Bovbjerg, R.R., and R.E. Curtis. 1987. "States Are Confronting the Adverse Side Effects of Health Competition." Business and Health 4 (5, March): 49-50.

Bovbjerg, R.R., and W.G. Kopit. 1986. "Coverage and Care for the Indigent: Public and Private Options." Indiana Law Review 19: 857-917.

Bovbjerg, R.R., A. Evans, J. Holahan, F. Ullman, and S. Flanagan. 1998. Health Policy for Low-Income People in New Jersey. Washington, D.C.: The Urban Institute.

Bruen, B.K., and F. Ullman, 1998. "Children's Health Insurance Programs: Where States Are, Where They Are Headed." Washington, D.C.: The Urban Institute.

Campbell, J.A. 1999. "Health Insurance Coverage: 1998." Washington, D.C.: U.S. Census Bureau. October 4.http://www.census.gov/prod/99pubs/p60-208.pdf(accessed 12/14/99).

Coughlin, T.A., L. Ku, and J. Holahan. 1994. Medicaid Since 1980: Costs, Coverage, and the Shifting Alliance between the Federal Government and the States. Washington, D.C.: The Urban Institute Press.

Deloitte & Touche, LLP. 1996. Safety Net at Risk: The Future of the Massachusetts Uncompensated Care Pool. Burlington: Massachusetts Hospital Association.

Eby, C.L., and D.R. Cohodes. 1985. "What Do We Know about Rate-Setting?" Journal of Health Politics, Policy and Law 10 (2): 299-327.

Employee Benefit Research Institute (EBRI). 1995. ERISA and Health Plans. Issue Brief No. 167. Washington, D.C.: EBRI.

Evans, A.E. 1997. "Provider Reimbursement in New Jersey." Washington, D.C.: Assessing the New Federalism, Urban Institute unpublished memo. August.

Fishman, L.E. 1997. "What Types of Hospitals Form the Safety Net?" Health Affairs 16 (4, July-August): 215-22.

Friedman, E. 1987. "Public Hospitals: Doing What Everyone Wants Done but Few Others Wish to Do." JAMA 257 (11):105-9.

Gaskin, D.J. 1997. "Altruism or Moral Hazard: The Impact of Hospital Uncompensated Care Pools." Journal of Health Economics 16: 397-416.

Hadley, J., and K. Swartz. 1989. "The Impacts on Hospitals' Costs between 1980 and 1984 of Hospital Rate Regulation, Competition, and Changes in Health Insurance Coverage." Inquiry 26 (1, spring): 35-47.

Harden, B. 1998. "Maternity Wars: Hospitals Lure Medicaid Moms with Luxury, Convenience." Washington Post, February 17, p. A1.

HANYS (Healthcare Association of New York State). 1996. The New York Health Care Reform Act of 1996. Albany, N.Y.: HANYS.

Holahan, J., and J. Kim. 2000. "Why Does the Number of Uninsured Continue to Grow?" Health Affairs (forthcoming).

Holahan, J., J. Wiener, and S. Wallin. 1998. Health Policy for the Low-Income Population: Major Findings from the Assessing the New Federalism Case Studies. Washington, D.C.: The Urban Institute.

Holahan, J., C. Winterbottom, and S. Rajan. 1995. "A Shifting Picture of Health Insurance Coverage." Health Affairs 14 (4, winter): 253-64.

Holahan, J., S. Zuckerman, A. Evans, and S. Rangarajan. 1998. "Medicaid Managed Care in Thirteen States." Health Affairs 17 (3): 43-63.

Holahan, J., R. Bovbjerg, A. Evans, J. Wiener, and S. Flanagan. 1997. Health Policy for Low-Income People in Massachusetts. Washington, D.C.: The Urban Institute.

Holahan, J., A. Evans, K. Liu, M. Sulvetta, K. Haslanger, and J. Cantor. 1997. Health Policy for Low-Income People in New York. Washington, D.C.: The Urban Institute.

IHPS (Institute for Health Policy Solutions). 1999. Directory of Consumer-Choice Health Purchasing Groups. Washington, D.C.: Institute for Health Policy Solutions. http://www.ihps.org.

Jackson, H. 1997. "Matters of State: Hospital Cash Flow Improves." Asbury Park Press, December 28. http://www.app.com.

Jain, S.C., and S. Hoyt. 1992. "Patient Dumping in the Federal Courts: Expanding EMTALA without Preempting State Malpractice Law." Law Medicine and Health Care 20 (3): 249-52.

Jensen, G.A., M.A. Morrisey, S. Gaffney, and D.K. Liston. 1997. "The New Dominance of Managed Care: Insurance Trends in the 1990s." Health Affairs 16 (1, January-February): 125-36.

Kronick, R. 1990. "Health Insurance for the Uninsured in New Jersey." Report to New Jersey Department of Health. University of California at San Diego, Department of Community Medicine (since changed to Family and Preventive Medicine). December 31.

Levit, K.R., H.C. Lazenby, B.R. Braden, and the National Health Accounts Team. 1998. "National Health Spending Trends in 1996." Health Affairs 17 (1): 35-51.

Liska, D., B. Bruen, A. Salganicoff, P. Long, and B. Kessler. 1997. Medicaid Expenditures and Beneficiaries: National and State Profiles and Trends, 1990-1995. Third Edition. Washington, D.C.: Kaiser Commission on the Future of Medicaid.

Mariner, W.K. 1996. "State Regulation of Managed Care and the Employee Retirement Income Security Act." New England Journal of Medicine 335 (26): 1986-90.

Marsteller, J.A., R.R. Bovbjerg, and L.M. Nichols. 1998. "State Policy Options for Nonprofit Conversion." Health Services Research 33 (5): 1495-1535.

McDonough, J.E. 1992. "Mass Retreat: The Demise of Massachusetts Hospital Rate Regulation." Journal of American Health Policy 2 (2, March/April): 40-44.

McDonough, J.E., C.L. Hager, and B. Rosman. 1997. "Health Care Reform Stages a Comeback in Massachusetts." New England Journal of Medicine 336 (2): 148.

National Hospital Rate-Setting Study. 1988. Final Report. Cambridge, Mass.: Abt Associates Inc.

Norton, S.A., and D.J. Lipson. 1998. Public Policy, Market Forces, and the Viability of Safety Net Providers. Washington, D.C.: The Urban Institute.

Pataki Task Force. Governor George E. Pataki's Ad Hoc Task Force on New York's Prospective Hospital Reimbursement Methodology (NYPHRM). 1995. Findings and Recommendations: New Directions for a Healthier New York Reform of the Health Care Financing System. Albany: State of New York, Department of Health. December 5.

Pauly, M.V. 1988. "A Primer on Competition in Medical Markets." In Health Care in America: The Political Economy of Hospitals and Health Insurance, edited by H.E. Frech and R. Zeckhauser (27-71). San Francisco: Pacific Research Institute for Public Policy. Pham, A. 1997a. "Blue Cross Threatened with Renewed Call for $200m Tax." Boston Globe, December 11.

----. 1997b. "Blue Cross Backs Down on Free Care: Insurer Drops Its Plan to Hit Hospitals for Payments to Cover Pool Obligation." Boston Globe, December 20.

Reeve, C. 1998. Still Me. New York: Random House.

Robinson, J.C., and H.S. Luft. 1988. "Competition, Regulation, and Hospital Costs, 1982 to 1986." JAMA 260 (18): 2676-81.

Rosco, M.D. 1990. "All Payer Rate Setting and the Provision of Hospital Care to the Uninsured." Journal of Health Politics, Policy and Law 15 (4): 815-31.

Schramm, C.J., S.C. Renn, and B. Biles. 1986. "Controlling Hospital Cost Inflation: New Perspectives on State Rate Setting." Health Affairs 5 (3, fall): 22-33.

Sloan, F.A., J.F. Blumstein, and J.M. Perrin, eds. 1986. Uncompensated Hospital Care: Rights and Responsibilities. Baltimore: The Johns Hopkins University Press.

Special Commission. 1997. Report of the Special Commission on Uncompensated Care. Boston: Special Commission (Massachusetts Senate bill 1691). February 3.

Spencer, C.S. 1998. "Do Uncompensated Care Pools Change the Distribution of Hospital Care to the Uninsured?" Journal of Health Politics, Policy and Law 23 (1): 53-73.

Sulvetta, M., and K. Swartz. 1986. The Uninsured and Uncompensated Care. Washington, D.C.: The Urban Institute.

Swartz, K. 1989. The Medically Uninsured: Special Focus on Workers, A Chartbook. Washington, D.C.: The Urban Institute and National Health Policy Forum. July.

Thorpe, K.E. 1987. "Does All-Payer Rate Setting Work? The Case of the New York Prospective Hospital Reimbursement Methodology." Journal of Health Politics, Policy and Law 12 (3): 391-408.

----. 1988. "Uncompensated Care Pools and Care to the Uninsured: Lessons from the New York Prospective Hospital Reimbursement Methodology." Inquiry 25 (3): 344-53.

Thorpe, K.E., and C. Spencer. 1991. "How Do Uncompensated Care Pools Affect the Level and Type of Care? Results from New York State." Journal of Health Politics, Policy and Law 16 (2, summer): 363-81.

Weissman, J.S., C.V.D. Lukas, and A.M. Epstein. 1992. "Bad Debt and Free Care in Massachusetts Hospitals." Health Affairs (summer): 148-61.

Wiggins, O. 1997. "Whitman Approves Surcharge of 40 Cents on Tobacco Products." Bergen Record, December 20. http://www.bergen.com/news/carefolo199712206.htm.


About the Authors

Randall R. Bovbjerg is a principal research associate in the Urban Institute's Health Policy Center. His main research interests are health coverage and financing; regulation, competition, and the appropriate role for government action; and the workings of law in action, notably for policy on medical injuries. Earlier work on changing state policy led him to coauthor one book on Medicaid and another on block grants, both from the Urban Institute Press.

Alison Evans Cuellar is a former research associate in the Urban Institute's Health Policy Center. She participated in several case studies within the Assessing the New Federalism project, including those for the states discussed in this paper. She also helped analyze Medicare budget issues. She is currently a doctoral student at the University of California at Berkeley, where her dissertation addresses competition in hospital markets.

John Holahan is director of the Health Policy Center at the Urban Institute. He has authored several publications on the Medicaid program. He has also published research on the effects of expanding Medicaid on the number of uninsured and the cost to federal and state governments. Other research interests include health system reform, changes in health insurance coverage, physician payment, and hospital cost containment.



Topics/Tags: | Health/Healthcare


Usage and reprints: Most publications may be downloaded free of charge from the web site and may be used and copies made for research, academic, policy or other non-commercial purposes. Proper attribution is required. Posting UI research papers on other websites is permitted subject to prior approval from the Urban Institute—contact publicaffairs@urban.org.

If you are unable to access or print the PDF document please contact us or call the Publications Office at (202) 261-5687.

Disclaimer: The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders. Copyright of the written materials contained within the Urban Institute website is owned or controlled by the Urban Institute.

Email this Page