Milbank Memorial Fund

The purpose of this article is to honor a beloved friend and collaborator by giving readers access to his voice as he used it in articles we wrote together. Dan's conversational voice became his written voice when he remarked on the ironies of tax law and of the politics of making health and social policy. I write below about our goals, how we conducted research and wrote our articles, and what we learned. Then I tell a story about some influence our work has had. I conclude by describing work we left unfinished and urge others to continue it.Dan and I began to talk in 1956, when we were sophomores at Harvard College. For the next fifty-seven years we talked over meals in the Leverett House dining room, in rooms in Leverett after the libraries closed, first as roommates, then as neighbors, collaborators, husbands, and fathers. From the beginning we talked as friends.We talked most frequently about the politics of making and implementing public policy. This was our daily subject, for example, as we worked on our senior honors theses in 1958–59: Dan's on the constitutionality of the Smith Act of 1940, which made it a federal crime to teach or advocate violent overthrow of the government; mine on how, during the New Deal, the federal government hired unemployed writers to write guidebooks about the history and current circumstances of each state.As we worked on our theses, we told each other stories about our mentors, whom we admired deeply: V. O. Key Jr. for Dan, Frank Freidel and Arthur Schlesinger Jr. for me. We had not yet understood that these distinguished scholars were demonstrating three principles of effective teaching: love your discipline, express that love as engaged citizens, and take your students seriously.We continued to talk frequently, in Cambridge, New York, and Boston. Then, in the mid-1980s, we began to conduct research and write together. Dan described our subject as the “curious way we make health policy in the United States.” He explained what he meant by “curious” in one of our subsequent articles. “Officials of the Internal Revenue Service,” he wrote, “made health policy while convinced that they were not making policy at all.” He complained that “this method of making health policy serves the public interest badly” because “we all pay a price when substantive policy is made by people who agree that they are not equipped to do so and insist in spite of evidence that since they are not making policy there is no reason to hold them accountable for it.”The price all of us pay mattered a great deal to Dan. In this instance, for example, the federal officials who persuaded themselves that they were deciding what Dan called “an issue of technical tax law” were in fact, he continued, “accepting as decisive the hospital industry's view of health policy.”Our collaborative research and writing began in 1985 as a result of one of Dan's standard conversational gambits, a mock-innocent question. He asked me what I knew about cafeteria plans, a new tax-exempt employee benefit that had been made possible in 1978 by the enactment of section 125 of the Revenue Code. Not much, I replied. Dan then wondered why these plans, which the Office of Management and Budget and the Treasury had assured Congress would be revenue neutral, had added $20 billion to federal tax expenditures during the past year.We agreed to try to learn why the executive and legislative branches of the federal government had made this error and devised a way of working — methodology would be too pompous a word — that we continued in our subsequent projects. We relied mainly on our linguistic skills: Dan spoke tax; I spoke health policy. We chose one of these obscure languages for each interview we conducted in person or by telephone; interviews with officials of the Revenue Service, congressional staff, and lobbyists. During each interview, Dan or I would pause occasionally to give the other a running translation. We used the same technique when we read published documents and decoded them for each other.The only unpublished documents we used were letters responding to our queries. Our attempts to obtain documents that the Revenue Service is required to deposit in the National Archives had only helped to reduce the volume of paper saved by the federal government. We routinely requested such documents under the Freedom of Information Act. The Revenue Service routinely responded to our requests, as Dan wrote, “by first telling us that we were entitled to the documents and then regretting that the documents could not be found.”A few weeks after we completed our interviews and had translated and discussed published documents, Dan sent me a draft manuscript of approximately 150 double-spaced pages. We never discussed Dan's first drafts or even why we didn't talk about them. I interpreted this unusual absence of conversation as recognition that each of us was acting in our accustomed professional roles. Creating a huge manuscript that included every pertinent fact we had scrounged, every potentially relevant point of law, and every conceivable way to interpret fact and law was, I assumed, what Dan did as a professor of tax law.Another reason I never tested this interpretation was that Dan had informed me, speaking in a regretful tone, that I lacked understanding of the forensic criteria that academic lawyers use to judge one another's skills. The occasion for his comment was the only time I participated in reviewing a candidate's qualifications for tenure in the law faculty at Northeastern. It had not mattered to Dan that he and I agreed that the candidate was superbly qualified for promotion.A few weeks after I read the long manuscript Dan had sent, I sent him a revised draft of about thirty-five double-spaced pages. We never discussed why I made particular decisions about content and style. Dan may have assumed that I was simply behaving as a practitioner of the politics of making and implementing health policy who also writes articles and books. He would not have presumed to practice this art, just as I would not have presumed knowledge of the forensic criteria by which academic lawyers judge their colleagues.Dan edited my draft for accuracy about facts and law and added citations, footnotes, and appendixes. Then we submitted the article for publication. After we received reviews and made revisions and an editor had accepted the article, we made additional revisions and, with the editor's permission, submitted it to a law journal.Two editors' responses to the manuscript of our initial article on cafeteria plans revealed that we were doing unconventional research and analysis. The first editor to whom we sent it, of the Milbank Quarterly, a journal I would eventually serve as publisher, returned it a month later, covered with Post-its asking questions. He offered no overall judgment, not even an invitation to revise and resubmit. “Who is this Mr. Willis?” Dan asked, in a tone I by now recognized as forensic artistry.The second editor to whom we sent the manuscript was Larry Brown of this journal, who accepted it almost by return mail. Larry wrote that ours was the only paper he had received in several years as editor that was about politics, policy, and law. Having accepted the paper, Larry then sent it to external reviewers.A gratifying number of academic lawyers and researchers in the disciplines of the policy sciences have cited our articles. Dan was particularly pleased by comments from two eminent tax lawyers. Dan Halperin made the first of these comments in letters he wrote to us as we were finishing our first manuscript. He had been the lead author of section 125 as a policy maker at the Treasury in 1978. His initial response to our draft article was to insist that he had been making technical tax policy and that tax law “was a poor substitute for explicit social policy.” Thus spoke the future Stanley S. Surrey Professor at the Harvard Law School and the first holder of the Martin D. Ginsburg Chair in Taxation at Georgetown. Several weeks later, however, Dan wrote again to agree with us that “it is fair … to say that the Treasury took the policy of the tax law to encourage spending on health insurance and health care as a given. … Perhaps Treasury (and me) too often takes a tax preference as a given.” He also gave us permission to quote this correspondence in the published article.J. Mark Iwry made the second comment. In Barack Obama's administration, Mark is senior adviser to the secretary of the Treasury and deputy assistant secretary (tax policy) for retirement and health policy. He had been the Treasury Department's benefits tax counsel from 1995 to 2001. Several years later, Mark and I were panelists at a conference in Washington, DC, convened by the Employee Benefit Research Institute. I made a comment on the panel about the history of making health policy through tax law. Mark said, “Oh, you're THAT Fox.” During the next break, he told me that, when he was benefits tax counsel, he had required every lawyer on his staff to read Dan's and my articles.There are two projects we cannot finish. Dan and I had been following the interaction of politics, policy, and law in drafting section 501(r) of the Patient Protection and Affordable Care Act of 2010 (ACA), and in writing regulations to implement it. The purpose of the new section is to strengthen the accountability of nonprofit hospitals for providing benefits in the communities they serve.The passage and implementation of this new section of the tax code has, to date, followed patterns Dan and I had documented for other occasions on which tax lawyers have made health policy. For example, John McDonough, a former staff member of the Senate Health Committee, told us that staff of the Senate Finance and the Joint Tax Committees refused to discuss the history of section 501(r) with staff of the Senate Health Committee. A lawyer for a national health policy advocacy group told us that she had been unable to document discussions about drafting regulations to implement it, discussions that Congress had requested, between officials of the Revenue Service and of the Centers for Disease Control and Prevention. The draft regulations posted by the Revenue Service in April 2013, and its formal response to comments on them three months later, were silent about the existence of a body of expertise for measuring and prioritizing the health problems of communities.We left more important work unfinished because we ran out of time, or had other obligations, or had insufficient insight into the complexity of politics, policy, and law. Until I prepared the talk that preceded this article, I had forgotten that we had written about this unfinished work: twice. I close by quoting what we wrote — I hear both our voices in the prose — because I hope, and I know that Dan would have hoped, that others will get on with this work.One aspect of this unfinished business is to explain in detail why, since 1942, the Treasury not only has been “making health policy through tax law but seems to have had no doubt that this was its normal function.” We continued: “How did the Treasury first decide that this was proper? What is the history of its interaction with the people who make explicit health policy? How did the legislative clearance process operate to make health policy through tax law?” We directed these questions to “those who believe we cannot understand where we are without understanding how we got there.”The other piece of unfinished business resonates with current news. We described it in our articles on what we called the “semipreemption” of state lawmaking by the federal government in the Employee Retirement Income Security Act (ERISA) of 1974. Because these words are mostly Dan's, they are an appropriate conclusion for this article: “Do we as a nation want to continue to link health insurance with employment, or are we willing to move toward publicly provided benefits, even to those who are employed? If we stay with employment-linked insurance, do we want to preempt state regulation of employee benefits to enable the market to set the scope of benefits, or do we prefer national standards for health care coverage, or will we accept state experimentation?” We told our readers that, until American politics resolves these questions, “we will muddle along with a policy … created at a time of different national expectations to achieve different purposes.”I miss Dan.This article originated as a talk at a memorial for Daniel C. Schaffer organized by his colleagues on the faculty of the School of Law of Northeastern University after his untimely death in May 2013. Several people made this article possible. Sally Bould, Dan's widow, invited me to talk about Dan and me at the memorial event. Aaron A. Fox, my son, said that he attended the memorial in order to “represent the family.” He also intended to continue my talk if grief overcame me as it often did as I drafted it. Peter Jacobson read the article as a friend who practices law as a policy science. He commended it to Colleen M. Grogan, an editor who is expert at empathizing with authors.As I describe herein, the Journal of Health Politics, Policy and Law published three articles drawn from our research on how the federal government has made health policy through the politics of making policy to regulate employee benefits and tax-exempt organizations. After our articles appeared here, we published slightly modified versions of them in legal journals: (1) Daniel M. Fox and Daniel C. Schaffer, “Tax Policy as Social Policy: Cafeteria Plans, 1978–1985,” Journal of Health Politics, Policy and Law 12, no. 4 (1987): 609–64; revised and published as Daniel C. Schaffer and Daniel M. Fox, “Tax Law as Health Policy: A History of Cafeteria Plans, 1978–1985,” American Journal of Tax Policy 8, no. 2 (1987): 1–67; (2) Daniel M. Fox and Daniel C. Schaffer, “Health Policy and ERISA: Interest Groups and Semipreemption,” Journal of Health Politics, Policy and Law 14, no. 2 (1989): 239–60; revised and published as Daniel C. Schaffer and Daniel M. Fox, “Semi-Preemption in ERISA: Legislative Process and Health Policy,” American Journal of Tax Policy 9, no. 7 (1990): 47–69; and (3) Daniel M. Fox and Daniel C. Schaffer, “Tax Administration as Health Policy: Hospitals, the Internal Revenue Service, and the Courts,” Journal of Health Politics, Policy and Law 16, no. 2 (1991): 251–79; revised and published as Daniel C. Schaffer and Daniel M. Fox, “Tax Administration and Health Policy: The Tax Exemption of Hospitals, 1969–1990,” Tax Notes, October 21, 1991, 217–31.

For more than a century, policy in the United States has incentivized both expansion in the number and size of tax-exempt nonprofit organizations in the health sector and their commercialization. The implementation of the Affordable Care Act of 2010 (ACA) began yet another chapter in the history of this policy paradox.

This article explores the origin and persistence of the paradox using what many scholars call "interpretive social science." This methodology prioritizes history and contingency over formal theory and methods in order to present coherent and plausible narratives of events and explanations for them. These narratives are grounded in documents generated by participants in particular events, as well as conversations with them, observing them in action, and analysis of pertinent secondary sources. The methodology achieves validity and reliability by gathering information from multiple sources and making disciplined judgments about its coherence and correspondence with reality.

A paradox with deep historical roots persists as a result of consensus about its value for both population health and the revenue of individuals and organizations in the health sector. Participants in this consensus include leaders of governance who have disagreed about many other issues. The paradox persists because of assumptions about the burden of disease and how to address it, as well as about the effects of biomedical science that is translated into professional education, practice, and the organization of services for the prevention, diagnosis, treatment, and management of illness.

The policy paradox that has incentivized the growth and commercialization of nonprofits in the health sector since the late 19th century remains influential in health policy, especially for the allocation of resources. However, aspects of the implementation of the ACA may constrain some of the effects of the paradox.