Part II "Structuring Information Felicitously"

A Closer Look at an Arnold Ventures Funded Study

In the first part of our look at Arnold Ventures, we explored its business model and generous support of elite University health policy experts to further an ambitious health policy agenda. In this second part, we will explore some of the questions raised by Arnold’s aggressive approach.

Zack Cooper is an Associate Professor of Economics and Health Policy at Yale University*. He is the academic investigator at the heart of the so-called the 1% Solution, an Arnold Ventures funded project which encompasses most of its health policy agenda. The core idea of the “1% solution” is that while comprehensive health reform (e.g. “Medicare for All”) may not be achievable, pursuit of a bevy of policy goals with smaller price tags could generate savings that could be reinvested in policy improvements.

Cooper was the object of unwanted press scrutiny for receiving extensive sub rosa funding from United Healthcare for research work and writing instrumental in the enactment of the No Surprises Act in 2021, which was aimed at controlling out-of-network health insurance billing. United was expected to be the largest single beneficiary of this legislation. (The biggest “surprise” emerging from the No Surprises Act was that providers are winning 80% or more of the independent mediations of these disputes, suggesting that it was health insurers, not providers, who were gouging the public).

According to Arnold’s 990s, Cooper’s Yale policy shop, the Tobin Center for Economic Policy, received over $5 million from 2018 to 2024. Of this amount, $700 thousand funded the 1% Project itself, including more than a dozen papers by academic colleagues on topics ranging from surprise billing to PBM reforms to site neutral outpatient payment to hospital market concentration.

As part of this project, Cooper and a University of Chicago colleague, Zarek Brot-Goldberg, published a paper in early 2024 of the economic impact of hospital mergers: “Is There Too little Anti-trust Enforcement in the Hospital Sector?” which found that 20% of hospital mergers had an adverse economic impact on their communities. The alternative off-message headline, “80% of hospital mergers had no adverse economic on their communities” never surfaced.

However, a follow on piece got wide circulation thanks to a June, 2024 Wall Street Journal article, which exposed it to millions of readers without any reference to Arnold Ventures funding. The paper, which featured an astonishingly complex multivariate econometric model, was originally published by the National Bureau of Economic Research (NBER is also funded by Arnold Ventures). This paper linked hospital mergers to widespread layoffs in the communities where the mergers took place and a subsequent wave of suicides and drug overdoses (!).

According to this study, the 307 hospital mergers Brot Goldberg and Cooper analyzed from 2010 to 2015 resulted in hospital rate increases to commercial insurers of 1.2%. This sad exercise of supposed “monopoly power” (why not a 20-30% increase if you “owned” the market?) was not sufficient even cover a merger’s transaction costs (legal, accounting and bankers fees, consulting services, etc. typically amount to 3-5% of the revenues of the merged enterprise) , let alone generate free cash flow for the merged entity.

The hospital merger-related increases authors found were adding to a health benefit expense that was about 9% of total employment cost in their employer sample. A 1.2% premium increase in an expense that was 9% of payroll raised employers’ compensation costs by less than a tenth of a percent.

Yet, according to their model, this tenth-of-a-percent increase in compensation cost somehow triggered a wave of layoffs in the community where the mergers took place. The mergers in the study took place during a time period (2010-2015) immediately following the Great Recession of 2008, during which layoffs soared nationwide. Unemployment did not return to pre-crash levels until late 2017!

Econometric models such as the one in this study do not establish the direction of causation. Rather they infer it from correlating supposedly independent factors. The study did not control for the impact of the catastrophic economic downturn on the affected (mostly urban) communities or account for the potential role of the Great Recession’s in causing the hospital mergers. Rather, by association, the study chose to blame the victims. There was no control group of employers in communities that did not experience a hospital merger or did not have a hospital at all.

Authors also did not control for the other ways that employers typically respond to employment cost increases, such as raising their prices, reducing costs other than payroll, or most importantly, increasing patient cost sharing. Employees with high deductible health plans increased six-fold in the aftermath of the Great Recession according to KFF .

Straining credulity to the max, the merger-induced layoffs were extrapolated to have caused over 10 thousand deaths of despair (suicides and drug overdoses) nationally in communities where hospital mergers took place. No effort was made to control for other potential causal factors of those deaths - the arrival of fentanyl in the community, business closures, big increases in patient cost sharing and household financial trauma stemming from the recession.

Another crucial missing control group: communities where, instead of merging with an out-of-town health system, the hospital simply closed. Hospitals are often the largest employers in their communities. The layoffs resulting from a hospital closure, both of hospital employees and suppliers/contractors, would have dwarfed any layoffs that “resulted” from keeping the hospital open. The lack of hospital access would also almost certainly have had measurable effects on the mortality rate of the surrounding community.

As a sociologist and management consultant who spent more than forty years trying to help hospitals remain independent, I can say that only economists with an agenda could have constructed this garish, neon-lit throughline from hospital mergers to layoffs to suicides and drug overdoses. Economist Uwe Reinhardt had a term for statistical manipulations of this kind. He called them “siffing”, which stands for “structuring information felicitously”.

What Arnold Ventures did with this study was fund a headline: “A New Study Found Hospital Mergers Caused a Wave of Layoffs and Deaths in their Communities”. That is more than four dangerous words. We have found numerous examples of distorted findings in other Arnold funded studies.

With a foreign policy-oriented Republican in the White House and a Republican controlled Congress, an Arnold Ventures health policy agenda heavy on price controls and tighter government regulation seems unlikely to be implemented in the next couple of years. However, with health costs on the rise, and three more years of academic studies from elite University faculty flooding the zone, Arnold’s health policy agenda will be front and center for the next Democratic Congress or White House.

Arnold’s patient discipline, combined with his billions and sophisticated political action committee efforts, will unleash a fresh wave of technocratic policy solutions on our healthcare system, physicians and patients alike. Whether actual evidence supports this agenda won’t matter all that much!

Author gratefully acknowledges the financial help of the Federation of American Hospitals, in analyzing the paper discussed above. (You can read a more detailed analysis of this study here).