Address correspondence to: Colleen L. Barry, Johns Hopkins Bloomberg School of Public Health, 624 North Broadway, Room 403, Baltimore, MD 21205 (email: ude.hpshj@yrrabc).
Copyright © 2010 Milbank Memorial FundContext: This article chronicles the political history of efforts by the U.S. Congress to enact a law requiring “parity” for mental health and addiction benefits and medical/surgical benefits in private health insurance. The goal of the Paul Wellstone and Pete Domenici Mental Health Parity and Addiction Equity (MHPAE) Act of 2008 is to eliminate differences in insurance coverage for behavioral health. Mental health and addiction treatment advocates have long viewed parity as a means of increasing fairness in the insurance market, whereas employers and insurers have opposed it because of concerns about its cost. The passage of this law is viewed as a legislative success by both consumer and provider advocates and the employer and insurance groups that fought against it for decades.
Methods: Twenty-nine structured interviews were conducted with key informants in the federal parity debate, including members of Congress and their staff; lobbyists for consumer, provider, employer, and insurance groups; and other key contacts. Historical documentation, academic research on the effects of parity regulations, and public comment letters submitted to the U.S. Departments of Labor, Health and Human Services, and Treasury before the release of federal guidance also were examined.
Findings: Three factors were instrumental to the passage of this law: the emergence of new evidence regarding the costs of parity, personal experience with mental illness and addiction, and the political strategies adopted by congressional champions in the Senate and House of Representatives.
Conclusions: Challenges to implementing the federal parity policy warrant further consideration. This law raises new questions about the future direction of federal policymaking on behavioral health.
Keywords: Parity, insurance, mental health, addictionOn October 3, 2008, President George W. Bush signed the Emergency Economic Stabilization Act of 2008 to bail out financial institutions as a step toward addressing the nation's financial crisis. As a provision of this legislation, the Paul Wellstone and Pete Domenici Mental Health Parity and Addiction Equity (MHPAE) Act (P.L. 110-343) was passed into law. The MHPAE aimed to create “parity” by eliminating historical differences in group health insurance coverage for mental health and substance abuse (MH/SA) benefits and medical/surgical benefits. The new law is expected to affect insurance coverage for 140 million people covered under employer-sponsored health plans and state and local government plans. Officials estimate that the new law will increase premiums by four-tenths of 1 percent, or $25.4 billion, over a ten-year period. Before passage of the MHPAE, coverage for behavioral health care often required a higher level of cost sharing (e.g., coinsurance of 50 percent compared with 20 percent for outpatient medical services) and special service limits (e.g., twenty outpatient visits and thirty inpatient days per year). In a floor statement, Representative Patrick Kennedy (D-RI), one of the chief architects of the new law, made the case for its passage on the grounds that “access to mental health services is one of the most important and most neglected civil rights issues facing the Nation. For too long, persons living with mental disorders have suffered from discriminatory treatment at all levels of society” (Congressional Record 2007).
In this article, we review the political history of efforts to enact comprehensive federal parity (see appendix). We identify three factors critical to passage of the law after years of political deadlock. First, the development of new research evidence regarding the costs of parity was instrumental in allaying concerns among employer and insurance groups and policymakers about its cost impact. Second, the willingness of members of Congress to cast the issue in terms of their own and family members’ experiences with mental illness and addiction was crucial. Third, passage of parity can be credited to the savvy political strategies adopted by the members of Congress championing the issue in the Senate and the House of Representatives. These legislators crafted two separate parity bills with the goal of gaining passage in each chamber. While the differences between the House and Senate bills created tensions among parity advocates at certain junctures, this approach helped overcome the barriers to adoption in each chamber and paved the way for final negotiations over a compromise bill. We conclude by discussing challenges to implementing the policy. Earlier in 2008, Congress enacted, and the president signed, legislation that mandated parity in Medicare coverage for “nervous and mental diseases.” The political history of that set of events, however, is beyond this article's scope.
Limits on insurance benefits date back to the inception of third-party payment for mental health services (Ridgely and Goldman 1989). Not until after World War II did insurance policies include mental health services, when insurers began covering some hospital psychiatric care (Goldman, Sharfstein, and Frank 1983). Before the period of deinstitutionalization, beginning in the late 1950s when most long-term care in psychiatric hospitals was replaced with community-based mental health care, there was little incentive for private insurers to cover services already paid for through the public sector (Frank, Koyangi, and McGuire 1997). But even though the proportion of private firms that include mental health and substance abuse coverage in their benefit packages has increased dramatically over the last thirty years, the National Compensation Survey conducted by the U.S. Bureau of Labor Statistics indicates that day and visit limits have become more stringent over time (U.S. Bureau of Labor Statistics 1982, 2007). One survey conducted before the MHPAE law was enacted found that 74 percent of workers in employer-sponsored health plans with mental health benefits were subject to an annual outpatient visit limit, 64 percent were subject to an inpatient day limit, and 22 percent had higher cost sharing for mental health benefits than for general medical benefits (Barry et al. 2003).
The MHPAE Act includes the following provisions:
Equal benefits: If an employer offers behavioral health coverage, all financial requirements (deductibles, copayments, and coinsurance) and treatment limits (number of inpatient days and outpatient visits) for MH/SA benefits must be equal to those for medical/surgical benefits.
Annual and lifetime dollar limits: The law expands on the Mental Health Parity Act of 1996, which prohibited the use of special annual and lifetime dollar limits for mental health benefits, to include addiction benefits.
Coverage not mandated: According to the law, employers are not required to provide MH/SA benefits.Group coverage: The law applies to all U.S. employers with more than fifty employees. The law does not extend to the individual insurance market.
Conditions covered: Services are covered for those mental health conditions and substance use disorders defined under the terms of a health plan. The law does not require that specific conditions be covered.
Protections for state parity laws: States may impose parity requirements on health plans that are more stringent than required under the federal law.
Out-of-network coverage: Health plans that provide out-of-network coverage for medical/surgical benefits must also provide equal out-of-network coverage for MH/SA benefits.
Monitoring: The U.S. Government Accountability Office (GAO) will conduct a study to analyze the specific rates, patterns, and trends in coverage and exclusion of specific MH/SA diagnoses by health plans and health insurance under the law.
Compliance and enforcement: The secretary of labor must submit a report to Congress concerning compliance with the law by 2012 and every two years afterward. The Departments of Labor, Health and Human Services, and Treasury must publish and disseminate information about the law's requirements and must coordinate their enforcement.
Cost exemption: If a health plan's total costs increase by 1 percent and are attributable to parity (2 percent in the first year after implementation), the plan can file for a one-year exemption from this law.
Other populations: The act applies to Medicare Advantage coverage offered through a group health plan, Medicaid managed care, the State Children's Health Insurance Program, and state and local government plans.
For this study, we conducted structured interviews with twenty-nine persons closely involved in the federal parity debate, including members of Congress and congressional staff; lobbyists for consumer, provider, business, and insurance groups; and other key contacts. 1 Most of these interviews were conducted by two- or three-author teams, providing an opportunity for the research team to compare observations and build reliability in reporting. We also reviewed historical documentation, academic research on the effects of parity, and public comment letters submitted to the U.S. Departments of Labor, Health and Human Services, and Treasury related to implementation of the new law.
Efforts to achieve benefit parity date back almost fifty years to President John F. Kennedy's administration. In 1961, President Kennedy called on the U.S. Civil Services Commission (the predecessor to the U.S. Office of Personnel Management) to require the Federal Employees Health Benefits Program (FEHBP), the health insurer for federal employees, to cover psychiatric illnesses at a level equivalent to general medical care (Hustead et al. 1985). Before this, health plans participating in the FEHBP had been replicating the prevailing private-sector pattern of limiting coverage for mental health care. By 1975, however, FEHBP health plans were permitted to scale back their mental health benefits. At that time, the Blue Cross Blue Shield High Option Plan was the only plan remaining that provided parity-level psychiatric coverage, and so for a number of years, it attracted higher-cost beneficiaries until 1981 when it was permitted to cut its coverage significantly (Hustead et al. 1985; Scheffler 1985; Watts, Scheffler, and Jewell 1986).
In the 1970s and 1980s, efforts to improve private insurance benefits were made primarily in the state legislatures, with the passage of mandated benefit laws establishing minimum benefit levels for alcoholism (38 states), drug abuse (25 states), and mental health (18 states) treatment (Laudicina, Loseleben, and Pardo 2000). The Massachusetts mandated benefit law, for example, required health insurers operating in the state to cover $500 of outpatient treatment annually, sixty days of inpatient care in a psychiatric hospital, and inpatient mental health care in general hospitals at the same level as for other illnesses (McGuire and Montgomery 1982).
Beginning in the 1990s, consumer advocates pressing for benefit parity at the state and federal levels began framing the issue more explicitly as an antidiscrimination measure. In 1992, Senators Pete Domenici (R-NM) and John Danforth (R-MO) introduced the first federal mental health parity legislation (S.2696) in Congress. Although no action was taken at that time, the issue reemerged in 1993/1994 in the context of efforts by President Bill Clinton's administration to reform the health care system. The basic package of health services proposed under the Clinton plan would have included thirty days of inpatient and residential care and thirty outpatient psychotherapy visits, with $25 in shared costs for managed care and a 50 percent shared outpatient cost for unmanaged care (Arons, Frank, and Goldman 1994). The second phase of the proposal would have required full integration of mental health services at parity with other medical services, beginning in 2001.
After the health care reform efforts under President Clinton failed, Senator Domenici teamed up in 1995 with Senator Paul Wellstone (D-MN) to introduce a relatively comprehensive parity bill (S.298) and attempted to attach it to the Health Insurance Portability and Accountability Act (HIPAA). At this time, the Congressional Budget Office (CBO) released an estimate that the bill would impose direct costs on the private sector of 4 percent of private health insurance premiums (CBO 1996). In April 1996, the Senate voted 68 to 30 to approve parity as an amendment to HIPAA legislation, but this provision was dropped during House–Senate conference negotiations.
In 1996, Senators Domenici and Wellstone introduced a scaled-back bill (S.2031) requiring mental health parity only for annual and lifetime dollar limits and succeeded in attaching it as an amendment to the Veterans Affairs, Housing and Urban Development (VA-HUD) appropriations legislation. The Mental Health Parity Act (MHPA) of 1996 (P.L.104-204) required group health plans with fifty or more employees that offered mental health benefits to apply the same lifetime and annual dollar limits to mental health coverage as those applied to coverage for medical/surgical benefits. This law did not apply to other kinds of mental health benefit limits, such as special annual day or visit limits and higher cost sharing, and it did not cover the treatment of substance use disorders. Research indicates, however, that health plans circumvented the law by tightening restrictions on the number of hospital days and outpatient visits for mental health services (U.S. General Accountability Office 2000). The provision for mental health parity on lifetime limits, however, did give patients some protection from bankruptcy, an important benefit of insurance coverage.
Because of its limited scope, the 1996 MHPA has been viewed as only a symbolic policy change, and long-time parity advocates were not satisfied. In our interview, Congressman Jim Ramstad (R-MN), one of the bill's sponsors in Congress, recalled that Senator Paul Wellstone expressed a mix of frustration and optimism with the 1996 law, noting that “we didn't even get half a loaf, we just got crumbs but it's a start.” Nonetheless, the 1996 MHPA is credited with heightening the profile of the parity issue and, with the encouragement of mental health advocacy groups, prompting state legislatures to step into the breach to experiment with more comprehensive parity laws (Gitterman, Sturm, and Scheffler 2001). By 2006, thirty-seven states had parity laws on their books. These laws vary substantially in the type of benefits covered, diagnoses included, population eligible, and direction regarding the use of managed care. Some state laws also are quite limited in scope. For example, South Carolina's law applies to public employees only. The more comprehensive state laws require equal cost sharing and prohibit the imposition of special inpatient day and outpatient visit limits. State laws also differ in the conditions covered, with some applying to only a subset of severe or “biologically based” disorders. 2 Finally, only a handful of states cover treatment for substance use disorders (Barry and Sindelar 2007).
We should note that the Employee Retirement Income Security Act (ERISA) of 1974 limits the reach of all state parity laws by exempting from state insurance mandates those firms that self-insure. The Kaiser Family Foundation estimated that between 33 and 50 percent of U.S. employees were in self-insured plans in 2000 and thus, in accordance with ERISA, would not be covered by state parity laws (Butler 2000). Using the Medical Expenditure Panel Survey Insurance Component (MEPS-IC), Buchmueller and colleagues (2007) estimated that even though 45 percent of all insured private-sector employees lived in a state with a strong parity law, the ERISA exemption reduced by half the number of employees actually covered by state parity laws, thereby substantially limiting their reach. For this reason, advocates argued that state parity laws were inadequate, since they did not cover those working for self-insured firms, and they have been pushing for enactment of a federal parity law that would apply to all privately insured individuals with group coverage.
Although efforts to pass a more comprehensive parity law repeatedly stalled in Congress between 1997 and 2006, there were several other important developments during this period. In 1999, at the first-ever White House Conference on Mental Health, President Clinton directed the Office of Personnel Management (OPM) to implement comprehensive MH/SA parity in the FEHBP, beginning in 2001 (OPM 1999). This directive affected the health insurance of 8.5 million federal employees, retirees, and their dependents. The OPM had already been moving in the direction of parity, having established a minimum mental health benefit for federal employees that modestly improved coverage for these services in 1994 and having shifted to parity for psychiatric medication management visits but not other services in 1999. The FEHBP parity directive constituted the most extensive effort of its kind, covering all diagnoses listed in the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders (DSM) and all aspects of in-network MH/SA benefits, including cost sharing, deductibles, and lifetime and annual dollar, day, and visit limits. The FEHBP also held political and symbolic importance as the insurer of members of Congress and their families.
Another milestone was reached in 2002 when President George W. Bush publicly stated his support for more equivalent mental health coverage while announcing the formation of the president's New Freedom Commission on Mental Health. 3 He made this endorsement at a job-training facility for persons with mental illness in New Mexico, the home state of Senator Domenici. In the past, Democratic Party leaders had supported parity, but this event put a Republican president on record in support of it, along with a powerful GOP congressional leader, thus depicting parity as a bipartisan policy aspiration. In 2003, a recommendation for parity appeared in the final report of the president's New Freedom Commission on Mental Health (2003) (over the objections of White House staffers, who felt that the New Mexico announcement had been sufficient), making parity an official recommendation of President Bush's commission.
After years of political impasse, a number of developments proved critical to the eventual enactment of the comprehensive federal parity law in 2008. In this section, we describe how the emergence of new evidence regarding the cost of parity, the personal experiences with mental illness and addiction among parity proponents in Congress, and the political strategies adopted in the Senate and the House of Representatives created the conditions for legislative passage.
Research evidence was instrumental in allaying business groups’ and policymakers’ concerns about the cost impact of parity. From the time that mental health coverage was first offered in the 1950s as part of major medical insurance, insurers worried that intensive and/or long-term psychotherapy would drive up health insurance premiums. Furthermore, they feared adverse selection—that those plans that offered better benefits would attract those people who were more likely to use services and thus place them at a price disadvantage with regard to premiums. As noted earlier, research evidence from the 1970s and 1980s was interpreted as supporting higher cost sharing, particularly for psychotherapy (Ellis and McGuire 1986; Horgan 1986; Manning et al. 1989; McGuire 1981; Taube, Kessler, and Burns 1986).
The 1990s began an era of invigorated research on the impact of expanding mental health insurance benefits. The new research evidence that emerged over this period challenged the prevailing view that expanding mental health benefits would drive up costs. Rather, new evidence suggested that the presence of managed care would likely make parity much more affordable. Managed care had fundamentally altered mental health care delivery by shifting from demand-side mechanisms (e.g., cost-sharing requirements, benefit limits) to supply-side mechanisms (e.g., utilization review) for controlling mental health spending. On average, the academic literature indicated that managed care reduced mental health costs by around 25 percent, with most of the savings resulting from decreases in use and spending for inpatient care (Bloom et al. 1998; Callahan et al. 1995; Christianson et al. 1995; Goldman, McCulloch, and Sturm 1998; Ma and McGuire 1998).
The National Institute of Mental Health (NIMH) was instrumental in making the initial investment in developing and disseminating these new research findings. Starting in 1991, under a process initiated by Senator Pete Domenici, Congress directed the NIMH to conduct a series of studies on expanded mental health benefits. Reports summarizing the current knowledge base on parity under managed care were prepared by the NIMH's National Advisory Mental Health Council (NAMHC) and delivered to the Congress (NAMHC 1993, 1997, 1998, 2000).
It took some time for this newer evidence to influence the policy debate over parity. During the Clinton administration's health care reform efforts of 1993/1994 and during the debate over the 1996 MHPA, CBO actuaries did not feel that the cost-controlling effects of managed care had been demonstrated sufficiently to use these newer cost assumptions in making their estimates.
Like the CBO, during the 1990s the actuarial firms commissioned by groups representing the business community and the insurance industry usually did not take into account the effects of managed care. These estimates, calculated as the expected change in total premium due to parity, ranged as high as a 12 percent increase in total premiums. For example, the ERISA Industry Committee and the Association of Private Pension and Welfare Plans sent out a bulletin in 1996 estimating an 8.4 to 11.4 percent increase in premiums due to parity (Watson Wyatt 1996). In response, mental health provider groups commissioned PriceWaterhouseCoopers to estimate and distribute their own estimates (Bachman 1996). Some of these estimates prepared for mental health advocacy groups did take into account managed care and projected increases in premiums as low as 1 percent. This so-called battle of the actuaries created among members of Congress the perception that the cost impact was unknown, which contributed to their failure to enact a comprehensive parity law in the mid-1990s.
Then two developments led to a reduction in actuarial estimates of the cost of parity. First, the NIMH commissioned the Hay Group to update the methodology used for conducting cost estimates that incorporated the effects of managed care. Second, the Robert Wood Johnson Foundation convened a workshop in May 2001 for actuaries, academics, and government policy experts to discuss methods for estimating the costs of parity for mental health with attention to the effects of managed care (RWJF 2001). Following this workshop, staff at the CBO who attended the workshop released a new analysis using the Hay Group's methodology to score parity legislation that reflected updated cost assumptions related to managed care (CBO 2002). The CBO estimated that premiums for group health insurance would increase by an average of 0.9 percent.
Subsequent empirical research supported the contention that parity under managed care would not substantially increase costs. Research from states’ experiences with parity under managed care arrangements found consistent evidence of little effect on spending but increased financial protection, which was reflected in reduced out-of-pocket expenditures (Bao and Sturm 2004; Barry and Busch 2007, 2008; Busch and Barry 2008; Rosenbach et al. 2003).
A unique opportunity to study the cost impact of parity in a large, national insurance program came in June 1999 when President Clinton directed the OPM to offer comprehensive parity mental health and substance abuse benefits for all in-network services offered by health plans in the FEHBP. Parity in the FEHBP was evaluated by a collaboration between the Department of Health and Human Services (DHHS) and the OPM, which operates the FEHBP. The Office of the Assistant Secretary for Planning and Evaluation (ASPE) within the DHHS drew up a contract for the parity evaluation in collaboration with several other agencies within the DHHS, including the NIMH, National Institute on Alcohol Abuse and Alcoholism, National Institute on Drug Abuse, Agency for Healthcare Research and Quality, and Centers for Medicare & Medicaid Services. These agencies oversaw the studies and provided guidance to the investigators. The results were posted on an ASPE website and were published in several peer-reviewed journals, including the New England Journal of Medicine and Pediatrics. Testimony was given before Congress. Results from the evaluation indicated that parity had little impact on spending within managed care plans but did lower out-of-pocket costs (Goldman et al. 2006). The post-parity change in MH/SA spending in the seven health parity plans studied was on a par with or below that of other large, privately insured populations that were not covered by the parity requirement. In five of these plans, the parity policy was associated with reductions in beneficiaries’ out-of-pocket spending. Similar effects were found in an examination of child dependents of FEHBP beneficiaries (Azrin et al. 2007).
All the employer groups we interviewed pointed out that this newer research evidence, together with their own experiences with benefit expansion under managed care, contributed to the evolution in their view that comprehensive parity would not break the bank. Likewise, all those policymakers and federal and congressional staff members we interviewed cited the importance of academic research evidence to perceptions about parity in Congress. Our interviews suggest that without defensible academic research on costs and a favorable score from the CBO, their views would likely not have shifted and that this proved essential to removing a barrier to advancing comprehensive parity legislation.
Research suggests that personal experience with mental illness matters in determining support for mental health policy. Using data from the 1996 General Social Survey, Jean McSween (2002) found that having a personal experience or a family member with personal experience with mental illness was associated with support for increased government spending for mental health. In the case of parity, personal and family members’ experiences with mental illness and substance use disorders among members of Congress were critical. Without exception, the central proponents of parity in Congress became involved in the issue through a personal connection. Congressmen Patrick Kennedy and Jim Ramstad, sponsors of the bill in the House of Representatives, attributed their leadership on parity to their own struggles with mental illness and/or addiction. In the Senate, Senators Domenici, Wellstone, and Edward Kennedy (D-MA) became involved with parity through the struggles of family members.
Congressman Ramstad connected his commitment to parity to the day in 1981 when he woke up in a jail cell in South Dakota after an alcohol-induced blackout. Congressman Kennedy remembered in the years before speaking publicly about this issue, being a young state senator and parking blocks away from his therapist's office to avoid having anyone find out he was seeking treatment for bipolar disorder. After a middle-of-the-night car crash on Capitol Hill in 2006, Congressman Kennedy publicly disclosed his decision to receive treatment at the Mayo Clinic for addiction to prescription medication. His personal experience with these conditions being treated as “second-class illnesses” motivated his interest in casting insurance parity in terms of civil rights. Congressmen Ramstad's and Kennedy's willingness to come forward and discuss their private struggles was viewed as particularly helpful in humanizing the issue for their congressional colleagues in the final phases of reform.
Our interviews largely attribute the inclusion of addiction treatment in the parity bills in 2007 to Congressmen Kennedy and Ramstad presenting the need for parity for addiction benefits in terms of their own personal experiences. From 1992 through 2006, substance abuse services were typically not included in parity legislation. For example, a stand-alone substance abuse bill, the Help Expand Access to Recovery and Treatment (HEART) Act of 2003, was introduced by Congressman Ramstad and Senator Norman Colman (R-MN) in 2003 with the understanding that once parity for mental health was enacted, expanding parity to addiction treatment might be achieved on an incremental timetable. Less sympathetic views among members of Congress were attributed in part to the ambivalence over whether to regard addiction problems as medical conditions to be treated or personal failings to be overcome by other means. In the words of a lobbyist for one of the consumer advocacy groups, “Something changed [related to congressional consideration of substance abuse parity] and that something was Patrick Kennedy.”
As the father of a child with schizophrenia, Senator Domenici connected his work on parity to his desire to respond to constituents who spoke of becoming bankrupt over the out-of-pocket expenses for the mental health–related treatment costs of family members. Senator Wellstone attributed his work on the issue to his brother. A statement honoring Senator Wellstone after his death in a plane crash in 2002 by the then Senate majority leader Tom Daschle described Wellstone's commitment to parity:
Like tens of millions of Americans, Paul Wellstone knew well the anguish that mental illness can cause families. Nearly 50 years ago, when his older brother Stephen was a freshman in college, he suffered a severe mental breakdown. Stephen Wellstone spent the next two years in mental hospitals. Eventually, he recovered and graduated from college—with honors—in three years. But it took his immigrant parents 20 years to pay off the bill from those two years. (Congressional Record 2004)
Another example of the personal dimension of the political debate over parity was the relationship between Senator Kennedy and his son, Congressman Kennedy. One person we interviewed recalled that on March 5, 2008, the day the House of Representatives passed the parity bill, Senator Kennedy came to the House chamber to observe the proceedings. After the vote, he walked over to shake hands with Congressman Ramstad and to give his son a hug, an indication of the personal significance the issue held for him. In the final days before its passage in the fall of 2008, with the retirement of both Senator Domenici and Congressman Ramstad and the illness of Senator Kennedy looming, the personal dimension of this policy issue was widely viewed by those interviewed as having been decisive.
Finally, passage of parity can be credited to the political strategies adopted by the champions of parity in the Senate and the House of Representatives, who crafted separate parity bills aimed at gaining passage in each chamber. Although the differences between the House and Senate bills created tensions among parity advocates at certain junctures, this approach proved effective in overcoming the barriers to adoption in each chamber and paving the way for final negotiations over a compromise bill.
In the Senate, it had become increasingly clear to Senators Domenici and Kennedy that the only viable approach to gaining passage would be to draft a bill that both parity advocates and long-time industry opponents could support. According to Senate rules, a single senator can halt floor action with a filibuster, which requires sixty votes to overcome. Unless a majority party holds the Senate by a wide margin, this structure tends to require some bipartisan support to pass a bill. In 2005, Senator Domenici called on his fellow Republican senator Michael Enzi of Wyoming to help overcome employer and insurer groups’ objections to parity. Senator Enzi chaired the Health, Education, Labor and Pensions (HELP) Committee, which had jurisdiction over parity legislation in the Senate. Senator Kennedy was the HELP Committee's ranking minority member. Senators Enzi and Kennedy invited business and insurance groups to meet privately to explore areas of compromise that might address their concerns. These groups respected both Senator Enzi, with whom they had worked closely on a range of business issues, and Senator Kennedy, who had a history of bipartisan leadership on major health legislation. MH/SA consumer and provider groups were not initially invited to join these talks.
A number of other factors brought business and insurance groups to the negotiating table in the Senate. As noted earlier, the views of employer and insurance groups had shifted over the years regarding the cost impact of parity. In addition, employers were increasingly aware of research on both the effectiveness of treatment and the productivity gains associated with improved mental health and had observed these effects in their own workforces. In the words of one business lobbyist, “Employers got it—they wanted to offer good mental health benefits and they realized that there are additional costs on the medical side when mental disorders are not treated.” Anticipating the likelihood of health care reform in the event of a Democratic presidential victory, industry groups also were motivated to come to a resolution on parity before the 2008 election. They reasoned that it would be more difficult to influence the shape of parity legislation if it were incorporated into the larger, more unpredictable health care reform debate.
Despite their willingness to negotiate, employer and insurance groups had several major concerns related to parity. First, employers worried that parity would open the door to health insurance mandates on businesses. Second, if they agreed to go along with parity, these firms would seek uniformity in its implementation, since many operated across state lines and had to contend with varying parity laws in different states. They argued that a federal parity law should establish a ceiling rather than a floor for regulating MH/SA benefits, and they sought federal preemption of state parity laws to allow them to comply with a single federal law. Third, they opposed the requirement in earlier parity bills and the FEHBP parity directive that the policy be applied to all medically necessary diagnoses listed in the DSM, instead favoring an approach that would leave to the health plans the decision of which diagnoses should be covered under a parity law. Previously, to combat this provision, employer and insurance groups had adopted a strategy attacking the legitimacy of the DSM as the basis for determining coverage. They called attention to various DSM diagnoses that arguably would be covered under parity, including jet-lag sleep disorder, phase-of-life problems, malingering, and religious or spiritual problems. This strategy proved effective. One mental health advocate we interviewed admitted that parity opponents had “Harry and Louise-d them,” referring to the insurance industry's television commercials critical of President Clinton's 1994 health plan that were widely credited with undermining support for reform. In a similar strategy to undermine support among members of Congress, industry groups used the DSM provision to suggest that parity would lead to much unnecessary care.
While meetings with employer and insurer groups were being held behind closed doors, nine consumer and provider groups organized as the “Fairness Coalition” 4 in support of parity received monthly updates from the HELP committee staff. The Fairness Coalition had formed in the aftermath of the failed 1993/1994 health reform efforts out of a concern that the consumer and provider umbrella group, the Mental Health Liaison Group, had been too large and diverse to articulate a single, concise message for addressing mental health issues during health reform under President Clinton.
By early 2006, as progress with business groups was gaining momentum, Fairness Coalition members were invited into these discussions with the goal of crafting a bill that both sides could support. These organizations engaged in two years of quiet negotiations, discussing specific language and bill wording, and they eventually hammered out a compromise. Most notably, members of the Fairness Coalition agreed to back down on the requirement that the policy apply to all medically necessary diagnoses listed in the DSM, and insurers made concessions on the issue of state preemption. Participants credit this outcome to the pragmatism of Senators Kennedy and Enzi and their staffs, who repeatedly asked each group whether a contentious provision was really necessary to have a good law.
In the House of Representatives, Representatives Kennedy and Ramstad took advantage of the window of opportunity provided by the 2006 election to advance a comprehensive parity bill. In November 2006, the Democratic Party gained control of the House for the first time in twelve years. Election results yielded a strong majority of 233 Democrats to 202 Republicans. A key role of the House leadership is determining which bills will be brought up for consideration on the House floor. Before the 2006 election, Speaker Dennis Hastert (R-IL) had refused to bring up the parity bill for a floor vote in deference to the objections of conservative members of Congress and the employer groups. The fundamental concern of the House leadership was that parity constituted a mandate on businesses (although none of the parity bills required employers to offer MH/SA coverage), that it would open the door to other types of mandates (e.g., affecting chiropractors, dieticians), and that it would result in higher mental health benefit costs. Even though the House parity legislation had more than enough cosponsors—both Democrats and Republicans—to gain the 218 votes needed for passage, floor consideration had been blocked by the House leadership.
This dynamic changed with the ascent of Nancy Pelosi (D-CA) to the Speaker's office after the 2006 election and her willingness to bring parity up for a vote. Congressmen Kennedy and Ramstad worked with mental health and substance use disorder advocacy groups to draft a broader bill than was being considered in the Senate. The groups working with the House bill sponsors were organized under the Parity NOW coalition, a counterpart to the Fairness Coalition, the group working with Senate bill sponsors. The Parity NOW coalition was made up of interest groups almost entirely different from those working to develop a Senate parity bill. Only two groups, Mental Health America and the American Psychiatric Association, joined in the deliberations over both the House and the Senate bills. The House bill applied to all medically necessary diagnoses listed in the DSM and to both in- and out-of-network coverage. The sponsors of the House bill were concerned that without an out-of-network parity benefit, health plans would establish “phantom networks” without enough participating providers to offer access to care. The House bill also required that the criteria for medical necessity determinations and coverage denials be made available to beneficiaries and include stringent protections for state parity laws. Like the Senate bill, the House version applied only to group coverage and was not a mandate, since it did not require employers to cover MH/SA benefits.
Employer and insurance groups opposed the House bill, opting instead to support the Senate version only. Most members of the Fairness Coalition (with the exception of Mental Health America and the American Psychiatric Association) did not support the House bill out of deference to the negotiation over the Senate bill with employer and insurers. The fact that some consumer and provider groups supported the Senate bill while others supported the House bill created some tensions in the parity advocacy community.
In February 2007, Senators Domenici, Enzi, and Kennedy introduced the Mental Health Parity Act of 2007 (S.558), the compromise that had been worked out in the Senate HELP Committee and that was supported by business and insurance groups. The HELP Committee marked up and approved the bill two days after it was introduced. In September 2007, the bill passed on the Senate floor by unanimous voice vote attached to unrelated legislation that delayed expiring tax breaks.
In the House of Representatives, after a series of fourteen field hearings around the country to rally support, Congressmen Kennedy and Ramstad introduced the Paul Wellstone Mental Health and Addiction Equity Act (H.R.1424) in March 2007. The three House committees with jurisdiction over the bill—Education and Labor, Ways and Means, and Energy and Commerce—all voted the bill out of committee between July and October 2007. In March 2008, the bill was passed on the House floor by 268 to 148, with 47 Republicans joining 221 Democrats to support it. 5 House Republicans made an unsuccessful attempt to substitute the Senate bill language in the final bill. A rally was held on the steps of the House of Representatives to celebrate the bill's passage, with all the House Democratic leaders on hand to show support.
After passage of legislation in both chambers, the House and Senate bill sponsors began informal negotiations to reconcile the differences between the bills. In the last days before fall recess in a presidential election year, timing was an issue. On the Senate side, there was pressure on the House sponsors of the bill to accept the Senate version. However, House members expressed willingness to start from scratch the following year rather than accept what, in their view, was a much weaker bill. Although this strategy had risks, these members were willing to gamble that the Democrats would do well in the 2008 election, thus giving them an opportunity to negotiate an even strong parity bill the following year if necessary. Given this possibility, employer and insurance groups were also somewhat more willing to compromise.
Over the summer, there was a breakthrough when the House sponsors decided to drop their provision requiring insurers to cover all conditions listed in the DSM and instead agreed to the Senate language. But they were concerned that allowing the health plans to determine which conditions would be covered would result in the systematic exclusion of certain diagnoses from the parity benefit. To address this, the compromise required that in the years following enactment, the GAO would report on specific coverage rates for all mental health conditions and substance use disorders; which diagnoses were most commonly covered or excluded; whether implementation of this law had affected trends in coverage or exclusion of diagnoses; and the impact of covering or excluding specific diagnoses on participants’ and enrollees’ health, their health care coverage, and the costs of delivering health care. In exchange, the Senate negotiators agreed to extend parity to out-of-network services and to additional language allowing the states to impose parity requirements on health plans that were more stringent than the federal parity requirement. Finally, the compromise bill allowed the health plans to file for an exemption from the requirement if parity led to an increase in total costs of 1 percent that could be attributed to parity (2 percent in the first year after implementation). But the exemption language was written in such a way as to make it less desirable for plans to use this option. Specifically, it required them to submit an actuary report, using claims data, on the cost increase after the plans had complied with the law for six months, and the exemption would apply for one plan year only.
By late June 2008, all parties had agreed to a compromise bill. But in July 2008, an effort to enact parity as a rider to legislation to extend tax breaks stalled. In late August and early September, action in Congress halted during the Democratic and Republican party conventions. At that point, many of the key stakeholders feared that the bill would not progress any further. A number of those interviewed emphasized the crucial role of House and Senate leaders in this last phase, and the willingness of Senator Kennedy to push the issue with Senate leaders on behalf of his son. Given that parity was not a priority for most members of Congress and that other major issues were under consideration in Congress, parity would most certainly have died without commitments by Senate Majority Leader Harry Reid (D-NV), Assistant Majority Leader Richard Durbin (D-IL), and Speaker of the House Nancy Pelosi to find a way for the bill to be voted on.
In late September, Congress was debating a $700-billion rescue of the financial industry. On September 29, the House of Representatives voted to reject the bailout package for the financial industry in defiance of President Bush and congressional leaders of both parties, who argued that the measure was crucial to stave off widespread financial collapse. Taking advantage of this window of opportunity, House and Senate leaders called up the parity bill for consideration. This development initially confused the parity bill's sponsors until Senate Majority Leader Reid explained that the leadership would use the parity bill as the vehicle to try to push through the bailout package. Parity advocates lobbied fifty-one members of the House who had cosponsored the parity bill but had voted against the financial bailout package to reconsider their vote. Thus, in an unexpected turn, the bailout was amended and passed under the House parity bill H.R.1424. The parity law was passed as a “sweetener” to this unpopular but momentous law. This example of “the tail wagging the dog” ended a twelve-year odyssey since passage of the 1996 law.
After its passage, the new federal parity law entered the rule-making phase, in which those federal agencies that have jurisdiction make administrative decisions about how to interpret the legislative language. This process shapes implementation of a new law. Political scientists have long observed a tendency of interest groups to refight policy battles at the rule-making stage (Balla 1998; Yackee 2005). In the case of the federal parity law, rule making created a somewhat contentious second bite at the apple for groups that had been temporarily united under an alliance to win passage of the law.
The first step in the rule-making process was a request for comments on the proposed regulations, which was posted in the Federal Register in April 2009 by the Departments of the Treasury, Health and Human Services, and Labor, the three regulatory agencies charged with implementing the new law. These agencies received more than four hundred comment letters. The Interim Final Rules were not released until February 2010, a month after the date when federal parity took effect for most plans. The delay in receiving guidance from the federal agencies posed challenges for firms and insurers charged with restructuring their benefit packages to comply with the new law. (Health plans were technically not required to comply with the Interim Final Rules until the start of new plan periods beginning after July 2010.) The release of final rules was still pending in late spring. The consequence is that employers might have to alter their benefit plans three times as more information is provided regarding how to comply with the new law.
Release of the Interim Final Rules helped clarify some areas of uncertainty related to interpretation of the statute. For example, the statute stated that financial requirements and treatment limits should be no more restrictive than the predominant financial requirements and treatment limitations applied to substantially all medical and surgical benefits covered by a health plan. The rules provide specific guidelines for how health plans should interpret the terms predominant (i.e., more than one-half of the medical/surgical benefits subject to the financial requirements or treatment limitations) and substantially all (i.e., at least two-thirds of the benefits in a classification). Regulators identified six broad classifications of benefits, noting that plans often varied financial requirements and treatment limits by whether treatment was provided in an inpatient, outpatient, or emergency setting and by whether the provider was in or out of network.
One area of particular contention was how to apply parity to deductibles. The comment letters reflected two opposing views. One was that health plans could have separate deductibles for medical/surgical benefits and for MH/SA benefits as long as the level of the two deductibles was the same. This was the view advanced by business and insurance groups, which argued that the costs of converting administrative systems to allow for unified deductibles would be high for health plans contracting with managed behavioral health organizations (MBHOs) to manage their behavioral benefits. The opposing view, of provider and consumer groups, was that medical/surgical and behavioral expenses together should satisfy a single deductible and that allowing separate deductibles “undermines a central goal of parity legislation, to affirm that mental health and substance use disorder benefits are integral components of comprehensive health care generally and should not be distinguished from general/surgical benefits” (Interim Final Rules, p. 5415). The argument was also made that since behavioral benefits typically comprised only 2 to 5 percent of a plan's costs, using identical but separate deductibles would impose a higher burden on MH/SA benefits. While acknowledging that the statute could be interpreted to support either position, the regulatory agencies decided that a combined deductible was “more consistent with the policy goals that led to enactment of the MHPAE.” To comply with this provision of the law, health plans and MBHOs are now developing integrated information technology systems to allow for a single deductible.
Another area of uncertainty related to how parity would affect the plan's management. Some consumer groups sought a broad interpretation of parity that would extend to benefit management so that, for example, utilization review practices would have to be the same for behavioral and medical/surgical benefits. This view was related to a long-standing concern that insurers would respond to parity by imposing stringent managed care practices on behavioral health, so that removing nominal benefit limits would not lead to more equitable coverage (Burnam and Escarce 1999). In contrast, insurers, including MBHOs, interpreted the language in the federal parity law as providing greater latitude in managing benefits. When the Interim Final Rules were released, they included a section on so-called nonquantitative treatment limitations to address this issue. Examples of nonquantitative treatment limitations are medical management standards such as medical necessity determinations; prescription drug formulary design; step therapy protocols requiring the use of less expensive therapies before a plan will cover more expensive therapies; standards for provider admission to participate in a network, including provider reimbursement; plan methods for determining usual, customary, and reasonable charges; and conditioning benefits on the completion of a course of treatment. The regulations prohibit the use of nonquantitative treatment limitations for MH/SA benefits unless the processes or standards used in applying these limits are comparable to or apply no more stringently than those used for medical/surgical benefits. The regulating agencies note that the regulations allow variations to the extent that recognized clinically appropriate standards of care may differ. For example, they note that the acute versus the chronic nature of a condition, the complexity of the treatment, and other factors may affect how the standards are applied in practice.
A third area of dispute was how parity applied to the continuum of care for treating MH/SA disorders. Some comment letters requested that health plans not be required under parity to cover specific treatments or treatment settings (e.g., counseling, nonhospital residential treatment, partial hospitalization, intensive outpatient treatment) if such treatments or treatment settings were not provided for medical/surgical conditions. Others argued that parity implied that the full scope of medically appropriate services would be available to treat these disorders, just as the full scope of services would be offered to treat medical/surgical conditions. Others interpreted the statute to require that health plans provide benefits for any evidence-based treatment. Finally, the Interim Final Rules recognized that not all treatments and treatment settings for MH/SA conditions corresponded to those for medical/surgical conditions, but they did not specifically address scope of services or continuum of care under federal parity. Rather, the agencies invited additional comments on whether and to what extent the federal law should address the scope of services or continuum of care. Further guidance on this point and in a few other areas (e.g., clarification of how employers can qualify for a cost exemption) was deferred until the release of the Final Rules, although it was uncertain when these final regulations would be released.
Earlier research suggested that the new parity law could offer financial protection to those with MH/SA conditions but have relatively little impact on total costs. However, since the provisions of the MHPAE Act differ in certain respects from the FEHBP parity directive and state parity laws, these experiences with parity provided only limited insight into the probable effects of the new law. For example, since the FEHBP parity directive applies only to in-network benefits and to all DSM diagnoses deemed medically necessary, the out-of-network parity benefit and the more flexible policy related to conditions covered under the federal parity law constitute “untested” provisions that warrant study. Likewise, the federal parity law differs from the FEHBP directive and most state laws (with the possible exception of the Oregon statute) in the extent to which it regulates how health plans manage the MH/SA benefit (e.g., nonquantitative treatment limits).
Employer and insurance groups have raised concerns about the regulations as being more expansive than the lawmakers intended. They contend that the nonquantitative treatment limit provision will hamstring insurers’ ability to manage the MH/SA benefit and warn that costs may be higher than predicted and compliance may be sufficiently burdensome to lead some employers to drop MH/SA coverage entirely. It is worth noting that this is the first major set of federal regulations (other than HIPAA) with which the MBHO industry has had to comply. Litigation is expected to resolve some of the important areas of continuing uncertainty related to implementing the new law. MBHOs have already filed a lawsuit charging that the departments failed to issue a proposed rule for public comment before issuing the Interim Final Rules. They contend that the issuance of the Interim Final Rules without a formal notice and comment period denied them their statutory right to participate in the rule-making process (Coalition for Parity 2010).
The passage of the parity law also raises questions about the future direction of federal policymaking on behavioral health. For many years, parity has been the signature mental health issue on Capitol Hill. By giving consumer and provider advocates a relatively high-profile issue to lobby for, the issue has raised the visibility of these interest groups among federal policymakers. It remains to be seen what the long-term impact of the federal law will be. In the short term, its passage was crucial to shaping mental health benefits under the landmark health care reform law passed in early 2010. Under that law, beginning in 2014, parity will be extended to individuals purchasing health coverage through state-based insurance exchanges. The core principle that behavioral health be treated equally with other medical conditions under health care reform and beyond is the fundamental achievement of the federal parity law.
This work was supported by a grant from the National Institute of Mental Health (R01 MH080797). Howard H. Goldman and Haiden A. Huskamp acknowledge support from the John D. and Catherine T. MacArthur Foundation. Colleen L. Barry acknowledges support from the National Institute on Drug Abuse (R01 DA026414).