Remembering Uwe

The healthcare world learned with great sadness this week of the passing of our friend, Uwe Reinhardt. I met Uwe in 1982 at the Federation of American Hospitals meeting in Las Vegas. Uwe opened the meeting by apologizing, in his disarming German accent, for not being his usual sharp self. He had, he said, skipped breakfast because his wife May had instructed him not to pay for anything in Las Vegas that he could get for free at home. This was vintage Reinhardt, innocent and knowing at the same time. That meeting was the beginning of a long and warm friendship.

Uwe would have been acutely uncomfortable with his colleagues referring to him as a “giant” in our field, because he was genuinely humble, and had not forgotten what it felt like growing up poor in devastated post-War Germany after WWII. And there was no sterner test of humility that occupying the James Madison Professorship of Political Economy at Princeton, just about as flossy an academic title as you will find.

For many years, Uwe taught a standing-room-only undergraduate course in Accounting at Princeton. The way he taught it was as a cleverly disguised course in moral philosophy. A main trope: “how would a ‘seasoned adult’ look at this problem?”  A ‘seasoned adult’ was someone who had lost his or her moral compass and sense of shame. So, where would a ‘seasoned adult’ book a bribe paid to a foreign official to obtain a contract, etc.? He was cleverly goading them not to lose their sense of outrage and their own moral compass, a tricky task without patronizing them.

Uwe had a devastating dry wit. There was the barest hint of vermouth in the Reinhardt martini.

You lived in fear of being placed on a program before him for what he might conceivably say about your talk. Please forgive the following, but I will never forget being on the program with him for the Board of Blue Cross Blue Shield of South Carolina about eight years ago. For some reason, our host had asked Uwe to talk about the current political landscape and me to talk about international health systems and what we could learn from other countries. This was a cruel thing to do, because Uwe had forgotten more about this topic than I know.

So I spent a month preparing this talk. I deconstructed the famous World Health Organization ranking of health systems, in which the US ranked 37th, behind Oman, Malta, San Marino, Switzerland etc., (largely by, effectively, double counting income inequality). And I found some great data it turned out Uwe hadn’t yet seen on WHY the US healthcare is so much more expensive than other countries (hint: ambulatory services dwarfed pharma pricing and the cost of our multi-payer system). As an almost 60-year-old “expert”, I was more nervous giving this talk than I was at my Ph.D. thesis orals at the University of Chicago decades earlier. To my immense relief, he LIKED it, and asked me to send him my slides.

Our last conversation, about a month ago, was about this very topic.   We were talking about the latest round of international cost comparisons, and he said “Our colleagues need to invent a new parlor game. This one is a complete waste of time.” Most of the countries we were unfavorably compared to, he went on, were smaller than Los Angeles County, and had been doing what they had been doing for generations. I riffed on how our health system WAS a country, bigger than Germany, sort of like a successful version of Afghanistan, replete with tribal conflicts, warlords, corruption, a bad communication system, language problems, etc. That reforming the US health system was a LOT harder and more dangerous than invading a middle eastern country like Iraq. He told me to be careful with that one.

He knew when not to be careful. During the 1980’s, he served on the Physician Payment Review Commission, which was eventually folded into MedPac. So he was a fixture on the medical society lecture circuit. On a panel in front of a bunch of physicians, he bridled at a physician who argued that cutting Medicare physician fees would damage the quality of care.

Uwe asked him whether he was going to leave a sponge in a patient after closing him up at surgery because his fee was reduced. That remark put Uwe on the American Medical Association’s “do not call” list for the better part of a decade.

Uwe had a soft heart, and if you wanted him to speak for you and could get to him, he would say “yes” and deeply discount his fee.   After some marital byplay, he eventually delegated the negotiations to his formidable wife, May. He met Tsung-Mei Cheng, a fellow immigrant, while studying graduate economics at Yale. May was the daughter of a Kuomintang (Nationalist Chinese) General who fled mainland China for Taiwan at the end of the Chinese Civil War. Photos of May photoshopped into Russian Commissar garb were a fixture in Uwe’s colorful presentations.

May is an improbable combination of steely, razor sharp and tough minded, but also sweet, gentle and loving. They were the most amazing couple. Delegating negotiation over speaking gigs to May was a classic economist’s move. Who better to place a value on a night away from home than his loving wife and fellow economist? She drove a very hard bargain. Later in his career, Uwe made numerous trips to Asia, “holding May’s coat” as he put it. He took immense pride in May’s role in advising Pacific Rim governments like Singapore and Taiwan on health financing, and eventually mainland China, on their financing reforms.

I talked to Uwe every few weeks or so, for decades.   Whenever something happened in the world that upset me or I just could not wrap my head around, I would call him up and ask him how he felt about it. He had a calming influence on me. I do remember him remarking on what a bad sign it was that so few members of the incoming administration of George W. Bush had passports, and that they could do a lot of harm without knowing it. That was a prescient forecast. The only time I can remember in 35 years where Uwe boiled over into outraged, white hot anger was over the War in Iraq, which he felt put his son, a Marine captain, at needless risk. His son was nearly killed by an IED on a subsequent tour of Afghanistan.

I remember Uwe saying once that our colleagues in health policy weren’t going to learn so much by travelling abroad, but that they would return perhaps feeling really lucky about how much talent and resources they had to work with. Our job, he said to me more than once, was to “keep them honest”. He never lost his moral compass, but also never mounted a high horse about it. Despite its manifest flaws, he loved his adopted country. It is VERY hard to imagine this world, and this field, without Uwe in it.

Categories: Uncategorized

3 replies »

  1. Thanks Jeff for this poignant piece. Uwe was indeed one-of-a-kind and the humor he infused into the rather dry subjects of health insurance and health policy kept many of us chugging along amid the chaos and more than occasional disappointment.

    Uwe’s last published paper–as far as I know—was in JAMA on Nov 2. It was a commentary on a proposed scheme (from Zeke Emmanuel and others, to measure “affordability” in healthcare. Herewith the entire text:

    “Among the much-repeated words woven into the US debate on national health policy are “affordability” and “sustainability.” Indeed, this debate is not confined to the United States. Remarkably, no one knows what these words actually mean at the practical level. Is there any economist or other expert, for example, who could be sure what percentage of the gross domestic product (GDP) the United States can “afford” to spend on health care, or what level of spending on Medicare is “sustainable”?
    To illustrate, according to the latest report from the Boards of Trustees of the Federal Hospital Insurance and Federal Supplementary Medical Insurance Trust Funds, spending on the federal Medicare program currently is 3.6% of a GDP per capita of about $58 000. That leaves $55 912 of non-Medicare GDP for all other spending. The trustees project that by 2050, Medicare will account for 6% of GDP. Assuming a low future rate of growth of only 1% per year for real GDP per capita, spending on Medicare will be 6% of a projected inflation-adjusted GDP per capita of $80 544 (with the 6% spending leaving the contemporaries living in 2050 a non-Medicare GDP per capita of $75 500). That amount is 35% higher than non-Medicare GDP per capita today. So will a spending level of 6% of GDP per capita in 2050 be affordable? Is Medicare sustainable?
    Undaunted by these definitional difficulties, in this issue of JAMA, Emanuel and colleagues, propose a new Affordability Index defined as “dividing the mean cost of employer-sponsored family health insurance by median household income.” Household income in this calculation represents household income from any source minus all taxes paid by households plus transfer payments, such as Social Security payments received by households.
    As the authors note, the data for this index are readily available from income surveys regularly published by the US Census Bureau. The numerators are from the annual surveys of employer-sponsored insurance (ESI) conducted by the Kaiser Family Foundation. The denominators are from annual surveys of household income published by the US Census Bureau.
    This proposed new index is closely related to information already routinely made available by the Kaiser Family Foundation surveys. That information is reported in a graph that presents, in index form, the time path over the past several decades of (1) total ESI premiums, including the employers’ and the employees’ contributions to that premium; (2) the premium contributions made by employees; (3) a wage index; and (4) an index of general price inflation measured by the Consumer Price Index. That graphical display conveys a good sense of the changing burden health insurance premiums impose on household income.
    The proposed Affordability Index also is closely related to the Milliman Medical Index (MMI), which is based on a large data bank of ESI-covered US families. The data bank is maintained by the employer-benefit consulting firm Milliman. The MMI shows, in dollars, the total cost of health care for a typical US family of 4 covered by a Preferred Provider Health Insurance Plan under ESI. Total costs in the index include the employee’s direct contribution toward his or her family’s insurance coverage, plus the family’s out-of-pocket health spending. In 2017, the estimated total cost is $26 944, of which employers paid $15 259 and employees, $11 685.
    As Emanuel et al suggest, the estimated out-of-pocket spending in the MMI is more than twice as high as the estimate reported in data series maintained by the Medical Expenditure Panel Survey (MEPS) of the Agency for Healthcare Research and Quality and another estimate published by the US Bureau of Labor Statistics. But the MMI represents a family of 4, whereas the other 2 indexes represent an average family of only 2.5 persons.
    In short, there already exist well-known sources of information on health spending per family that can easily be related to family income. The virtue of the Affordability Index proposed by the authors is that it is compact and easy to calculate. As such, it is a welcome new potential tool in the armamentarium of health statistics.
    Based on 2016 data, the Affordability Index proposed by Emanuel et al would be 30.7%. What does that number really mean? Was the spending it represented “affordable”? If not, how could the United States spend that much nevertheless?
    The affordability of health spending can be measured at 2 distinct levels: (1) at the aggregate, macroeconomic level based on national mean or median values and (2) at the microeconomic level of health spending by individual families in relation to their “discretionary disposable income,” ideally defined as disposable income after taxes minus essential expenditures on food, housing, clothing, transportation to and from work, and other essentials. (It is not the definition of income used by Emanuel et al.)
    The proposed Affordability Index is aimed at the aggregate level, as it is based on the mean cost of health insurance coverage per family divided by median family income. This index has some meaning to health economists, who believe, on the basis of theory and some limited empirical evidence, that the cost of the contributions employers make to their employees’ health insurance is recovered by employers out of the cash take-home wages they pay their employees.,6 Based on that theory, the median income figure used by the authors already has the mean premium cost for health insurance deducted from it, which leads to an exaggeration of the burden of health insurance the authors seek to measure.
    However, economists do not know exactly how an employer’s total outlay for the employees’ health insurance is reflected in lower take-home pay for particular individual employees. So to tell a particular employee that the Affordability Index in 2016 was 30.7% does not really tell that employee much about how their employer adjusted wages for health care expenses because it is unlikely that the employee’s take-home pay was reduced by 30.7% owing to health spending. Furthermore, many, perhaps most, employees probably do not believe the economists’ story and assume instead that the owners of their companies absorb that part of the premium paid by employers.
    What has long been needed in the United States is a regular, user-friendly annual report on the burden that total health spending by individual families imposes on their family budgets. For example, some individuals would like to know how many US families actually went bankrupt over health spending or were teetering on bankruptcy. Such a report could be stratified by age, income, race/ethnicity, geographic location, and other factors.
    In place of a written report, an interactive information system could be used that makes it easy for individuals to relate health-spending data to the income of households of different types. The Kaiser Family Foundation now routinely provides interactive maps showing the influence of health reform proposals on different types of families.
    The MEPS, conducted regularly by the Agency for Healthcare Research and Quality of the US Department of Health and Human Services, already contains most of the survey data required for such reports. The survey “gathers data on individuals’ health status, health conditions, medical events, charges and payments, employment, health insurance coverage, income, access to and satisfaction with care…” If additional data were needed, they could most likely be accommodated in future MEPS surveys.
    The desired reports or interactive information systems could be produced by the Agency for Healthcare Research and Quality or the Kaiser Family Foundation. But because the MEPS database is an open-access file, it could be produced by any organization capable of producing such analyses. All that is needed would be the fashioning from the data regular user-friendly reports or interactive information systems that would make them prominent for the media and the general public. Most importantly, such a data system could help constrain the folklore on which health policy so often is conducted and provide to legislators robust data about the fragile US health insurance system over which they preside.

    Published Online: November 2, 2017. doi:10.1001/jama.2017.16187

  2. Uwe Reinhardt was a giant in his field. I heard him speak several times at conferences and he always dazzled with insights and wit. I also read his NYT blogposts which were also terrific. What a great man! He will be missed.

  3. Thanks for this Jeff. You rounded out the man for us. I always loved listening to his talks. We are all better because of Uwe.