Become a Patron!
Michael Campbell, Did I Do That? An Argument for Requiring Pennsylvania to Evaluate the Racial Impact of Medicaid Policy Decisions Prior to Implementation, 82 Temple Law Review 1163 (Spring-Summer 2010)(110 Footnotes)
In its 2003 report Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care, the Institute of Medicine Committee on Understanding and Eliminating Racial and Ethnic Disparities in Health Care (“the Committee”) noted that “even when insured, [racial and ethnic minorities] may face additional barriers to care due to other socioeconomic factors, such as high co-payments, geographic factors (e.g., the relative scarcity of healthcare providers and healthcare facilities in minority communities), and insufficient transportation.” The Committee characterized “access-related factors” as “likely the most significant barriers to equitable care,” and opined that they “must be addressed as an important first step toward eliminating healthcare disparities.” To this end, the Committee issued “Recommendation 5-7: Structure payment systems to ensure an adequate supply of services to minority patients, and limit provider incentives that may promote disparities.”
In Pennsylvania, Medicaid is a critical source of health insurance for people of color, far more so than for white persons. Currently, 38.7% of black or African American Pennsylvanians and 32.4% of Hispanics rely on Medicaid to pay their medical bills, compared to only 12% of white non-Hispanics. With the advent of national health care reform, Medicaid promises to take on an expanded role in opening doors to the health care system for people of color, by extending coverage to many who previously lacked insurance. But while Medicaid facilitates access to health care for those who might otherwise do without, some Medicaid policies fail to address and may even foster racial and ethnic disparities among its recipients. I first consider the impact on blacks or African Americans of some of Pennsylvania's major decisions regarding *1164 payment for long-term care services for the elderly over the past three decades. I next examine some recent changes to Pennsylvania's Medicaid program, and whom they impacted. Finally, I recommend that a racial impact statement accompany proposed policy decisions by the Medicaid agency and its major contractors.
II. Pennsylvania's Policy Decision to Limit the Growth of the Nursing Home Industry
In 1982, administrators of Pennsylvania's Medical Assistance (Medicaid) program determined that in order to save costs, they would need to slow the expansion of nursing homes in the state. Between fiscal years 1975-76 and 1982-83, the state's nursing facility costs had risen from 2.4% to 4.5% of the State General Fund Budget. Nursing facility care is an entitlement under federal Medicaid law, which means that the state must pay for the institutional care of anyone who meets financial and status requirements and can prove that they need the services of a nursing home, irrespective of the state or federal government's appropriation of funds. State officials operated under the belief that nursing home operators could find someone to fill virtually any new nursing facility and that Medicaid would eventually end up paying for most of the residents when their money ran out. If those officials could neither meet nor reduce the demand for nursing home care, they felt that perhaps they could limit costs by controlling the bed supply.
State officials decided to impose a moratorium on capital cost reimbursement for new nursing home beds. In July 1982, the Pennsylvania Department of Public Welfare (“DPW”), which administers the state's Medicaid program, promulgated regulations denying Medicaid payment for depreciation or interest (i.e., capital costs) on nursing home beds unless a Certificate of Need (“CON”) for the project was issued prior to September 1, 1982.
Pennsylvania law at the time required nursing home developers to obtain a CON from the Pennsylvania Department of Health before new beds could be licensed. The CON process operated as the only government-imposed safeguard against unnecessary facility expansion.
The moratorium made economic sense to state officials, since at the time, the Medicaid program was spending approximately $50 million per year just for nursing home depreciation and interest. Medicaid pays for at least part of the cost of two *1165 thirds of nursing facility residents in Pennsylvania, as in other states. The moratorium made far less sense to the poor, frail, and elderly of Philadelphia. It exacerbated financial incentives, already existent in the state's Medicaid reimbursement system, for nursing home operators to seek out a greater share of private-paying customers. This meant moving to the more affluent suburbs, where applicants for admission could likely afford to pay private rates for a longer period of time before “converting” to Medicaid.
The nursing home industry has maintained a historical love-hate relationship with Medicaid. Medicaid typically pays the lowest rate of any payer, and by federal law cannot pay nongovernment facilities more than Medicare would pay, which makes it an unattractive payment option from a facility's perspective. By contrast, facilities can charge privately paying residents whatever the market will bear. Medicare provides very limited coverage for long-term care, and few people carry long-term care insurance, leaving residents and their families along with Medicaid as the primary payment sources for most nursing home care. While it is in a nursing facility's financial interest to maximize private payments, the funds of many private-paying customers will run out if they live long enough. If they believe that a nursing home will throw them out when they can no longer pay private rates, they won't enter the nursing home in the first place. To assure an adequate supply of applicants, nursing facilities participate in Medicaid in overwhelming numbers (eighty-one percent of facilities statewide, accounting for ninety-three percent of beds in 1997), and two-thirds of residents in participating nursing homes have Medicaid as their payment source. Medicaid law protects nursing home residents who become impoverished by prohibiting facilities from discharging residents who convert to Medicaid when their funds run out. The law has been interpreted to further prohibit facilities from playing a shell game by decertifying Medicaid beds in order to claim that there is no bed available for a converting resident.
Allowing time for the impact of the moratorium to be felt (i.e., for expansion of facilities that obtained CON approval prior to September 1, 1982), it is revealing to compare changes to the nursing home bed supply in Philadelphia against changes in its most affluent suburban counties between 1985 and 1988. During this period, Montgomery, Chester, and Bucks counties in suburban Philadelphia--the three *1166 counties with the highest per capita income in the state --experienced increases in Medicaid certified nursing home beds of eighty-five, fifty-four, and forty beds respectively. Over the same span, Philadelphia lost 806 Medicaid beds.
The percentage of the white population over age sixty-five, compared to the total population over sixty-five, in the three suburban counties in 1985 was much higher than in Philadelphia. In Montgomery, Chester, and Bucks counties, the white population over age sixty-five in 1985 represented 96.4%, 94%, and 98.3%, respectively, of the total population over age sixty-five. In Philadelphia, the white population over age sixty-five was 72.4% of the total population of the same age. Thus, the county with the substantial non-white population lost resources, while counties with overwhelmingly white populations gained resources.
Today, Philadelphia stands in stark contrast to bordering Montgomery County. In 2008, there were 1,359 fewer Medicaid nursing home beds in Philadelphia County than in 1985. By contrast, Montgomery County had 2,433 more Medicaid beds than in 1985. Philadelphia, with an estimated population of 1,447,395 had 7,356 Medicaid beds. Montgomery County, with an estimated population of 778,048 had 6,535 Medicaid beds. Philadelphia thus had one Medicaid bed for every 196 persons, while Montgomery County had one bed for every 118 persons. The rate of Medicaid beds to adults age sixty-five or over receiving Medicaid jumped to one bed for every 1.3 persons in Montgomery County as compared with one bed for every 5.7 persons in Philadelphia.
*1167 The population of Montgomery County was 8.6% black according to the 2008 census data. Philadelphia's population was 44.8% black. A comparison of the quality of care delivered in facilities in the two counties is beyond the scope of this Essay. However, David Barton Smith et al. have recently demonstrated that nationally, predominantly black nursing facilities deliver poorer quality care.
III. Pennsylvania's Policy Decision to Expand Non-Institutional Long-Term Care Services
What occurred in Philadelphia and Montgomery counties following imposition of Pennsylvania's moratorium can be contrasted with what followed its policy decision more than a decade later to fund alternatives to nursing home care for the elderly under Medicaid. In 1995, the state chose to implement the “Pennsylvania Department of Aging Waiver” (“PDA Waiver”), pursuant to Section 1915(c) of the Social Security Act. Congress enacted Section 1915(c) in 1981, the year before Pennsylvania promulgated the moratorium.
Section 1915(c) waivers permit states to draw down federal matching funds for services not generally available under their Medicaid plans, and to offer these services in the home or other community settings to persons who would otherwise need *1168 institutional placement in a nursing facility. In short, waivers fund diversion projects, offered to select populations. Because they must demonstrate cost effectiveness as a precondition to gaining federal approval of a proposed waiver, states limit the menu of services to be offered, as well as the population to whom they will be offered. Federal waiver of the Medicaid statute's requirements of statewideness, comparability of care and services, and methodology for determining financial eligibility authorizes such flexibility. Consistent with federal law, states can target populations of a certain age, those living with a specific medical condition, or people residing in a certain part of the state. As contrasted with the open-ended Medicaid “entitlement” to nursing facility services, waivers also permit states to control expenditures by limiting the number of available “slots” (the equivalent of a nursing home bed). When compared with the capital cost moratorium, which is a blunt instrument designed to discourage the increasing bed supply, waivers enable the state to strategically plant and cap the supply of slots.
In November 1995, Pennsylvania initially implemented the PDA Waiver for individuals age sixty and over in Philadelphia County. Services provided through the PDA Waiver, but not otherwise available to Medicaid recipients, included attendant care, personal care services, companion services, respite care, home support, older adult living centers, environmental modifications, counseling, personal emergency response systems, and home delivered meals. In addition, the PDA Waiver paid for home health services, physician services, and transportation, to the extent not covered under the state plan. The State limited eligibility for the PDA Waiver to individuals who could be served at a cost not to exceed eighty percent of the statewide average Medicaid payment for nursing facility services, consistent with the federal requirement of cost effectiveness. Thus, by rule, it costs substantially less to serve an individual under the PDA Waiver than in a nursing facility.
The State expanded the original waiver, effective February 1, 1996, to twelve additional counties: Allegheny, Bradford, Cambria, Delaware, Fayette, Greene, *1169 Montgomery, Sullivan, Susquehanna, Tioga, Washington, and Westmoreland. Beginning in October 1998, Pennsylvania expanded its PDA Waiver statewide. As of November 30, 2009, it served 15,816 persons in all of Pennsylvania's sixty-seven counties, ranging from a high of 5,949 in Philadelphia County to a low of six in Montour County.
PDA Waiver slots more intensely serve counties with a higher percentage black population. Philadelphia, whose population is 44.8% black or African American (as contrasted with the statewide population, which is 10.8% black or African American) has 37.6% of the PDA Waiver slots. Nearly half of the PDA Waiver slots (49.4%) serve the three counties with the highest black or African American population, although those counties, Philadelphia, Allegheny and Delaware, have just more than a quarter (25.8%) of the total state population.
Predictably, the PDA Waiver serves a much higher percentage minority population than does the nursing home industry. Enrollees in the PDA Waiver are 26.33% black or African American. Non-minority whites comprise 64.23% of enrollees. By contrast, the racial composition of Pennsylvania Medicaid residents age sixty or over in nursing homes is twelve percent black or African American and eighty-five percent non-minority white.
Unfortunately, data is not available to determine the racial composition of the pool of individuals who qualify financially for Medicaid long-term care services. Several factors impede the inquiry: (1) Pennsylvania uses different income standards to determine nursing facility and PDA Waiver eligibility; (2) the nursing facility income limit is not a constant, but fluctuates according to the rate charged by the nursing home to which a person is admitted (i.e., if an individual's monthly income does not exceed the nursing home's monthly charge, she meets the income test); and (3) both *1170 programs impose a resource test (currently $8,000 of “countable” resources for anyone in the waiver and nursing facility residents with income at or below 300% of the federal Supplemental Security Income (“SSI”) rate) which introduces a hard-to-determine variable. However, given the fact that both the PDA Waiver and the Medicaid nursing facility benefit serve individuals of lower income, one can presume that the potential applicant pool is more heavily black or African American than the statewide elderly population as a whole. And since, as explained below, Medicaid eligibility limits for both nursing facility services and the PDA Waiver are higher than for the rest of the elderly Medicaid population, one can conclude that the pool is less heavily black or African American than the elderly population on Medicaid.
Pennsylvania's choices with respect to eligibility requirements for nursing facility services versus the PDA Waiver may also foster a racial disparity between the populations served. Pennsylvania sets the PDA Waiver income standard at three times the federal SSI payment rate, and the asset standard as countable resources below $8,000. The income ceiling was $24,264 annually for a single individual in 2009, based on a payment rate of $674 per month.
Pennsylvania applies the $8,000 asset standard for nursing facility services as well. However, Pennsylvania applies a much higher “medically needy” income rule to nursing home residents only. As noted above, in Pennsylvania a person qualifies for nursing facility services if he or she has a monthly income at or below the nursing facility monthly rate. Some states decline to offer this higher standard for nursing facility services, whereas other states extend the standard to Waiver services. Pennsylvania takes the liberal approach for institutional care and applies the more stringent test for non-institutional services. The current average monthly rate for a nursing home in Pennsylvania is $7,235.82. The annual ceiling is therefore $86,830, which is more than three and a half times the income ceiling for the Waiver. Thus Pennsylvania's nursing home program favors individuals with much higher income, allowing two-thirds of Pennsylvania nursing home residents to qualify for Medicaid.
While nursing homes serve a higher percentage of elderly white persons when compared with the PDA Waiver, it is important to also note the significant disparity in expenditure of resources. Recently, the Commonwealth has sought to shift the balance *1171 of services away from nursing home care and into non-institutional services. However, in fiscal year 2008-09 far more Pennsylvania Medicaid recipients over age 60 received their long-term care in nursing homes than through the PDA Waiver (72.6% vs. 26.6%), at a much higher average monthly cost to the state ($4,437 vs. $1,700). So the program that serves a disproportionate white population serves far more persons at a much higher per capita cost.
IV. Other Medicaid Policies Favor a Disproportionately White Population
Pennsylvania's long-term care reimbursement system offers clearly contrasting examples of how Medicaid policy decisions can facilitate or impede access to care for blacks or African Americans. But recent history reveals other instances in which choices by the Medicaid agency have benefited different racial and ethnic classes of recipients. Because race is so closely tied to economic status, decisions to expand Medicaid eligibility stand out, but more subtle decisions, such as whether to contract with “poor people only” managed care plans, can also impact access to care.
In the late 1980s and early 1990s, Congress expanded Medicaid eligibility for women and children, which resulted in increased enrollments nationally of almost seven million between 1989 and 1995. The federal expansion came in the form of a mandate with which states had to comply. In Pennsylvania, the expansion was termed “Healthy Beginnings.” Today, Healthy Beginnings extends Medicaid eligibility for pregnant women and children up to age one to families with income up to 185% of the federal poverty income guidelines (“FPIG” or “poverty level”). Children ages one through five qualify with family income of 133% of the poverty level, and children ages six through eighteen qualify at 100% of the poverty level. In 2009, an infant in a family of three qualified if the family income was at or below $2,823 per month, a one-*1172 year-old qualified if the monthly family income was at or below $2,030, and a six-year-old qualified at or below $1,526 per month.
As a program serving families of modest means, Healthy Beginnings predictably enrolls a significant percentage of black or African American women or children. As of May, 2009, between 21.2% and 24.4% of Pennsylvania's Healthy Beginnings enrollees were black or African American.
By contrast, in the mid 1990s, Pennsylvania became the only state in the country to extend coverage to children under age 18 with severe disabilities who did not qualify for Supplemental Security Income (and automatic Medicaid entitlement) because their parents' income was too high. Pennsylvania did so by disregarding the income of the parents. Other states extended Medicaid to these special needs children only if they otherwise qualified for nursing home services. Despite a 2005 change to the state welfare code that imposed premiums for children in families with income above 200% of the FPIG, the program is very expensive, and serves an overwhelmingly white population. During 2006, the total number of children enrolled at least sometime was 47,632, at a total cost (state and federal matching funds) of $140,260,140.91.
The six counties with the highest number of enrolled children were Allegheny, Montgomery, Bucks, York, Delaware, and Chester, respectively. Five of these (all but York) rank among the top six Pennsylvania counties in per capita income. Predictably, Medicaid enrollment of white children in this category of eligibility is between 93.8% and 94.1%. Black or African American children comprise between 4.3% and 5.1% of the enrolled population. As in its long-term care programs, Pennsylvania has established policies which steer funding to a disproportionately white population.
*1173 V. The Need for Racial Impact Statements
Policies that disproportionately affect one or more racial groups are not necessarily illegal or even wrong. One would be hard pressed to argue against extending health coverage to severely disabled children, or to children with severe disabilities. But it is naïve to believe that such decisions occur in a vacuum, without any potential impact on other programs, especially those that serve the poorest groups with the highest minority representation. In recent years, Pennsylvania has done a remarkable job of avoiding cuts in eligibility and services or the imposition of burdensome cost-sharing requirements on vulnerable populations during a time of fiscal strain. However, for several years, state officials have distributed a list of cost-containment strategies employed by other states as a regular feature at the monthly meetings of the Pennsylvania Medical Assistance Advisory Committee (“MAAC”). For example, in October 2005, committee members were informed that Missouri was reducing its adult eligibility limits from seventy-five percent of the federal poverty level to twenty-three percent, and eliminating some coverage for 90,000 adults. The state also advised members that Tennessee was cutting eligibility for 323,000 adults. No Medicaid policy discussion occurs without an eye on the budget, and the concern here is that the poorest people, such as adults in Missouri with income less than 75% of the poverty level, provide an easier target for cuts than more affluent (and whiter) groups.
What can be done to focus the attention of policymakers and the public on the potential racial impact of health policy decisions at a point in time when they can be reversed without harming innocent recipients? The courts do not presently offer a viable forum. The Supreme Court long ago held that in the absence of racial motivation, neither the Equal Protection Clause of the Fourteenth Amendment nor Title VI of the Civil Rights Act of 1964 requires states to treat the various categorical grant classes in welfare programs equally. In rejecting the plaintiff's “naked statistical argument” (much like the arguments made here) when they challenged Texas's disparate reduction in benefits to its Aid to Families with Dependent Children program, as compared with its treatment of other programs serving the elderly, blind, and those *1174 with disabilities, Justice Rehnquist noted in his opinion for the majority, “given the heterogeneity of the Nation's population, it would be only an infrequent coincidence that the racial composition of each grant class was identical to that of the others.” Although Medicaid policies have the potential to ignore or even exacerbate racial disparities in health care access and quality, evidence does not suggest that Pennsylvania policymakers are deliberately or even consciously steering scarce health care funds on the basis of race.
The Supreme Court recently dealt racial impact claims a possibly fatal blow when it held that private individuals do not have a right of action to enforce regulations promulgated under Title VI that proscribe activities having a disparate impact on racial groups. Now, when federal health programs are at issue, only the U.S. Department of Health and Human Services' Office for Civil Rights (“OCR”) may raise disparate impact claims in the courts. But the capacity, competence, and authority of OCR to set standards, ferret out violations, and compel reversal of state Medicaid policies that foster health care disparities, with or without litigation, has met scathing criticism. Singled out for special criticism has been the length of time--often three to six years--associated with complaint investigations.
Absent a change in federal law or massive changes at OCR, the legal system does not offer a timely and effective forum for identifying, much less resolving, the disparate racial impact of Medicaid policy decisions. But an issue as important as equal access to health care deserves an effort to at least raise awareness, within the bureaucracy and among the public, of the potential racial impact of proposed policies. I recommend requiring that prior to publication of proposed regulations, the Medicaid agency first assesses the racial impact of its proposed policies and then presents these findings publicly. A necessary second step would be to provide the opportunity for a response from the public to its findings.
Advocates for a fairer criminal justice system have advanced such a process, and both Iowa and Connecticut have enacted legislation requiring racial impact statements to accompany all legislative proposals that could lead to minority disparities in rates of incarceration. Iowa's law requires a minority impact statement to accompany *1175 any legislation related to a public offense, sentencing, or parole and probation procedures. Upon signing the law, the Governor stated,
This means when members of the General Assembly and Executive branch are considering legislation of this nature, we will now be able to do so, with a clearer understanding of its potential effects--positive and negative--on Iowa's minority communities. Just as Fiscal Impact Statements must follow any proposed legislation related to state expenditures, with my signature, Minority Impact Statements will serve as an essential tool for those in government--and the public--as we propose, develop, and debate policies for the future of our state.
Pennsylvania law already requires its administrative departments, including the Department of Public Welfare, to prepare a fiscal note for regulatory actions and administrative procedures. For each initiative, the agency must identify the funding source, the cost for the upcoming fiscal year, the projected cost for each of the five succeeding years, and the fiscal history of the program for which expenditures are being made, among other things. Imposing a like requirement on agencies engaged in making health care expenditures on behalf of the state would force the officials to evaluate the consequences of actions such as the moratorium, and would--as the Governor of Iowa pointed out--empower concerned members of the public to participate in the dialogue around the advisability of, as well as alternatives to, such policy initiatives.
The requirement of a racial impact statement in Pennsylvania would represent a modest expansion of provisions that currently exist in the state's Regulatory Review Act. The Act mandates that at the time it submits a proposed regulation for publication, an agency must submit a regulatory analysis form to the Independent Regulatory Review Commission (“IRRC”) and the standing committees of the Pennsylvania House and Senate with jurisdiction over the agency. The regulatory analysis form must include:
(10) An identification of the financial, economic and social impact of the regulation on individuals, business and labor communities and other public and private organizations and, when practicable, an evaluation of the benefits expected as a result of the regulation.
(11) A description of any special provisions which have been developed to meet the particular needs of affected groups and persons, including minorities, the elderly, small businesses and farmers.
*1176 The Act permits but does not require publication of this information. The Commission uses this information in its determination of whether the proposed regulation is in the public interest, which consideration must include “[t]he protection of the public health, safety and welfare.”
A logical extension of the current scheme would be to require that the submitting agency include a racial impact statement with all initial IRRC filings that propose changes to state-funded health care programs. Borrowing from the current law pertaining to fiscal notes, the racial impact statement could be required to identify the expected impact on minorities in both the near future and five years out. Such a provision would offer protection against policies, such as the moratorium, whose impact evolves over time. And expanding on current requirements that the regulatory analysis form identify the social impact on individuals and describe special provisions developed to meet the needs of minorities, the racial impact statement should identify both the positive and negative impact on the ability of minorities to access quality health care as compared with the impact on non-minority populations. Such changes would force policymakers, and hopefully the public, to consider not only whether a policy will likely help or hurt the minority population, but also how it will help or hurt the minority population as compared to the non-minority population. This would force consideration of not only the overall impact, but also the disparity.
Finally, the new rule should require, rather than permit, the publication of this information. It should reach policy decisions which do not take the form of regulations, but which are instead found, to an increasing degree, in contracts such as those with entities, like managed care plans, which act as surrogates for the state in running the Medicaid program.
Given the link between Title VI and the passage of Medicare and Medicaid in 1965, it is ironic that notions about who is more worthy of governmental help, or simply the pursuit of other worthy goals such as cost savings, can lead to policies that serve non-minority populations to a greater proportional degree than minority populations. The national goal of eliminating health care disparities requires that Pennsylvania make best efforts to anticipate such results in advance, and invite the public to join in a discussion about priorities and alternatives.
. Visiting Reuschlein Clinical Professor, Villanova University School of Law; former Executive Director, Pennsylvania Health Law Project.