stained2.jpg - 26.1 K Mental Health Parity
From Research to Policy
by Ardis Hanson

Does one create policy from research?
If so, this process begins, first with a need or issue, then with a question that describes it and breaks it down into components that require understanding and information. It continues with gathering data, then converting the data to information which addresses the solution for the problem. This information addresses the historical, political, social, and economic ramifications of the issue and the consequences, good, bad, or null, that come from implementing the policy.

Those of us interested in the intersection of social science and policy see the need for fundamental research that can be applied to answer the problems in today's society and in the future. Historically, government decision making is a process of advocacy and bargaining, that power and interests are at stake, and the social science researcher is a neutral and dispassionate person. However, social science research is not value-free. Research either explicitly or implicitly embodies a set of values by its focus on certain issues and exclusion of others, by the concepts it uses, and by its methodology. Ideally research can only be used effectively when researchers and decision makers have accepted the same set of values. It is up to the researcher to frame the results of their research into a product that the decision maker can understand, in a common language, not overly embellished with the jargon of the discipline or of academia.

How is policy decided?
For example, during the Florida Congressional sessions every year, only a few major policy decisions are considered. Such decisions involve fundamental revisions of existing policies, programs, or activities, such as the Baker Act; reorganisation of state agencies, such as the decision to reorganise HRS and implement the new Department of Health; and increases, decreases, and terminations of ongoing activities.

Major policy decisions also break from past practice and are, in many cases, considered experimental especially when it is a previously unexplored area. Predictions of the outcome of a major policy decision are often uncertain or range from good to worst case scenarios. Therefore decision makers invest time and resources - money and people - to reassess the facts, frame an issue itself, and to find and evaluate the broadest range of options. This requires credible and qualified judgement to support one's position. Decision makers evaluate the political strengths of the viewpoints, assign measures to the intangible factors that make a decision controversial, and arrive at a political victory or compromise that will result in feasible and good policy. James Buckley wrote in his If Men Were Angels (1975 p. 127), that

"...each member of the Congress is expected to cast an independent judgement on every matter that comes before the body as a whole. The Congress can only try to cope with the ever-expanding number and variety of federal activities by dividing ever thinner the amount of time its members can devote to any one."

Remember this when writing analysis for policymakers. Their time is too limited to review one hundred pages when fifty will suffice, and an executive summary of three to five pages is perfect.

What services do researchers provide to policymakers?
Research serves as many things. It can be an information base for decision makers. It can be translated into practical measures and action strategies. It can serve as a substitute or "holding pattern" for decision makers, generating reports on issues while the actual decision making is put on hold. Finally, research can be used to effectively support or trash an opinion held by a decision maker.

There are different orientations to research. Three orientations are clinical or scientific, academic, and advocacy. All three have their benefits and minuses. A person using a clinical approach gathers as comprehensive data, reviews it and makes a decision about the content of the data. He or she then looks at the external, e.g. political, social, economic, realities of the situation and tries to reconcile the data and the reality to a workable solution, often ending with a compromise. The person using the academic approach gathers data and makes it into a statement on the issue, oftentimes disregarding the political/socioeconomic consequences of the issue. The person using the advocacy approach is pushing for their model solution; their research contains all the information necessary to sway policy in their direction only.

How much is research used by decision makers?
It depends. If it is solely determined by the direct influence research has on influencing policy decisions, the answer is "not much". If we expand its definition to include using research concepts in formulating questions, setting agendas, and constructing images of policy, the answer could be "significant". Useful research for policy has two distinct features. First, it makes an intrinsic contribution to the work of an agency, i.e. it contributes ideas or information that advance the decision making process. Second, it can be incorporated by the official easily into their own "work", i.e. the decision making process.


A Case Study
When Senator Grant asked the de la Parte Institute to explore the ramifications of implementing parity for persons with mental illness, he sent the following questions:

  1. What are the experiences of other states that have developed and/or implemented parity legislation for mental illness?
  2. What have been the costs and service utilization experiences resulting from implementing parity for mental illnesses? What does the data tell us about managed care strategies and benefit designs affect the costs of establishing parity?
  3. Based upon the experiences of other states, other national data, and the information available from Florida, how should we proceed in developing mental health parity legislation?
The first question, alone, requires a fairly extensive review of state and national sources, not the easiest thing to find if you are used to traditional library skills for academic programs. Sources for this kind of question include the national and state newspapers, organisations dealing with policy analysis, legal resources, and state and federal government agency information.

Question two covers financial and insurance sources, as well as state government analyses (insurance committees, etc), business sources, health and human resources management, to name a few.

The last question requires Florida data, analysis of state and federal parity legislation (current and pending), copies of the actual legislation for wording, any analyses prepared by the committee staffs, advocacy briefs, etc. All of this is then translated into how the state should proceed if we decide to implement parity.

Where did we go? What did we find?
The literature review was broken into several areas. The first was the state and federal government information. Guides to state governments can net the phone numbers of the appropriate agencies and, with persistence, phone calls can be routed to the appropriate staff person. Remember, most staff are also as pressed for time as you are so be clear about the type of information you need.

Other resources on the states included the the National Association of Insurance Commissioners for their compendium of state laws on insurance topics, the Intergovernmental Health Policy Project for their state mandates, the American Managed Behavioral Healthcare Association, the Bazelon Center for Mental Health Law, the Coalition for Fairness in Mental Illness Coverage, the Congressional Budget Office (CBO), and SAMHSA were called for current documents. The state and national advocacy organisations, e.g. the Florida Alliance for the Mentally Ill (FAMI) and NAMI, were also more than willing to share their action alerts and parity packages. When you approach an organisation that has a vested interest in a topic, be sure to gather more than just their documents. The more the range of opinions, the better you can understand everyone's interest in the area. It is also essential that if you are writing as a consultant that your work should be as unbiased as possible and as neutral in your presentation.

Literature reviews on parity legislation were done as well - the first was a basic literature review across academic disciplines, the second was across news and legislative sources. The literature reviews were updated every two weeks during the writing of the parity paper document.

Federal parity legislation was tracked and arranged into a timeline, indicating key players, revisions to bills, supplemental congressional documents such as national reports on the effects of implementing parity, and the final version of the federal legislation. This required searching the national newspaper (using LEXIS/NEXIS), reading two seminal newsletters - Mental Health Weekly and Mental Health Report, contacting advocacy groups such as the National Alliance for the Mentally Ill (NAMI) and the National Mental Health Association, reading the Congressional Record, and two internet sites for congressional information - Thomas and Marvel.

Essential to this area was the acquisition of the major reports bandied about in the press on the financial consequences of implementing parity on a national level. Five major reports were mentioned in the national legislative sessions. In each state that had implemented parity or was in the process of implementation, a number of locally commissioned reports were cited within the legislative histories or news articles. All of these were also considered important background information.

Statistics were, and are always, a problem. For national statistics, SAMHSA and NIMH were used as prime resources. Other governmental agencies , on the national level, included the Social Security Administration (SSDI and SSI benefits) and the National Advisory Mental Health Council. State statistics for Florida came primarily from the ADM and AHCA offices in Tallahassee. The United States Statistical Abstract and the Florida Statistical Abstract were also used for baseline population information. Other statistics came from the Epidemiologic Catchment Area (ECA) study, advocacy groups, the Bazelon Center, the journal literature, recent analyses by Institute and other USF faculty.

There was also an economic analysis done by one of the authors, an economics professor at New College/USF Sarasota. His analysis was an evaluation of the benefits of implementing mental health parity only for persons with severe mental illness. His section alone needed the following:

  1. number of people who suffer from severe mental illness,
  2. number of people who receive inpatient and outpatient care (national and state),
  3. treatment efficacy rates,
  4. the social costs of treatment (national and state)
    1. including direct treatment costs
    2. related medical or assistance costs
    3. indirect costs
      1. Maintenance costs (costs of housing assistance, administrative costs of transfer payments,
      2. Legal (law enforcement and public safety costs)
      3. Productivity (lost productivity and productive capacity- absenteeism, suicide, premature death, etc.
In itself, this alone required a significant amount of time to find the numbers. Plugging them was the easy part (for an economist). Additionally, some of the information we needed to write this report was simply not available. Some of it had to be extrapolated from older data. Datasets didn’t have specific data we needed or had corrupted/incomplete data.

A quick question- what is the standard the state of Florida uses in determining the prevalence of severe mental illness in the state? The ECA prevalence rate for severe mental illness, 2.8%, is the accepted standard for the state of Florida as well as for national reports. Establishing the authoritative standards and definitions for state data and terminology is important. Sometimes it takes a number of phone calls to find the information and cite it accurately and correctly.

One other resource we used was the Internet. However, there were specific sites we were using which, in turn, led us to other sites for information. I could have spent days reading through 68,724 hits on "mental + health + parity". Instead I chose search engines that allowed me to narrow my search, bookmarked my sites carefully, and moved through them accordingly.

The following is a checklist I use for these kind of questions. I have found a checklist of this kind helps me to stay on target while keeping the big picture in mind.

  • What are the objectives of this research?
  • Exactly what issue is being addressed? Always focus on the issue most important to the decision maker.
  • What are the alternatives or options? Always keep in mind the goals of the decision maker.
  • What possible impacts will there be? Include positive impacts as well as negative impacts. Look for secondary or unintended results within each impact identified.
  • Put it into understandable and accurate terms.
  • Any numbers should be as accurate and current as you can get and cite where and when and any limitations with the data on the chart
  • Any information should be to the point.
  • If it is too complicated to explain, don't use it.
  • Check the facts. Again.
  • Who generated the numbers? Are they biased?
  • Never rely on a single source for data. Gather corroborating data.
  • Be sure your definitions match the definitions in the data and the definitions of the policymaker.
  • God, or the devil, is in the details.
  • Be the devil's advocate for everyone. Use opposing viewpoints to gather the facts to substantiate their case.
    and finally,
  • Give analysis, not decisions. That's not your job, it's theirs.


A more extensive review of this article, including a review of mental health as public policy, was published. The citation is: Hanson, A. From research to policy: a look from an information perspective at the process of developing a mental health parity report. Behavioral & Social Sciences Librarian, 1998; 17(1):1-20.


Institute Home PageTHE LOUIS DE LA PARTE
FLORIDA MENTAL HEALTH INSTITUTE
USF Home Page
@ Ardis Hanson, the de la Parte Institute, May 1997.