We propose to investigate the profitability of the hospital industry in California. Does the industry exhibit scale effects with respect to system size? Does the trend towards consolidation in the hospital market lead to profits for the owners, and higher prices for consumers? We fail to find strong evidence for this but more work is needed to make any definite conclusion.
In 1960, the United States spent 5% of GDP on health care. That grew to 10% by 1985, and 15% by 2003. In 2019, the United States spent 16.9% of GDP on healthcare, the highest share of output in the OECD, and roughly twice the per capita expenditure of the OECD average. $1.2T of the $3.8T expenditure, or about a third, went to hospital care.
With the cost of health care eating up a growing share of economic output, the market behavior of significant actors in the health care industry has drawn increasing criticism. One of these trends has been the consolidation of hospitals into larger and larger systems, drawing criticism. In 2020, the State of California reached a settlement with punitive measures against Sutter Health, alleging that Sutter Health consolidated services in order to drive up prices, rather than improve efficiency.
- give quantitative support
- use Cournot analogy here
- how to put citation references in markdown?
Economics defines economic rent as profit above the level that a competitive marketplace would allow. In a free market, if profits are sufficiently high, new firms will enter that market, increasing the supply and driving the price down to the competitive equilibrium. However if one or a few firms are able to keep new competitors from entering, and are able to collude with one another to raise prices, they can extract rent.
But actually, it's a little more complicated. Market concentration is necessary for rent-seeking, but not sufficient. That is, just because a few hospital systems dominate the market, other forces might prevent them from setting prices. If buyers of the output are also able to exert market power, or if sellers of labor are able to extract higher wages, these forces could reduce or eliminate hospital profits.
We want to see if we can find evidence that this is happening in the hospital industry in California. Are hospitals earning economic rents? Are they earning returns on investment that are higher than the rate of return on capital invested in the wider economy?
In a 1996 paper for the FTC, Simpson and Shin find that nonprofit hospitals do set prices higher when they have more market power. They test two definitions of market power -- HHI and distance to the closest competitor.
Based on the current state of our analysis, hospitals do not have significant power to set prices, as measured by the Lerner Index. The Lerner Index is given by
- can Github gist markdown handle LaTeX?
An alternative way to measure profitability is to look at the return on invested capital. But to do that we would need the complete balance sheet for each system. The individual facility balance sheet data reported to OSHPD does not include system-level cash, so the records are incomplete. So that analysis would have to be done at a later time.
We've already stated that an increase in market power -- the power to set prices -- leads to profits. But there is another factor that helps determine profitability in hospitals: the commercial share of net patient revenue.
- immediately introduce Tirole's formulation of HHI/elasticity of demand = p - mc
- use calculated HHI, price, marginal cost per adjusted discharge to calculate expected price-elasticity of demand, and comment
Hospitals are paid largely by three sources: commercial or private insurance, Medicare, and Medicaid (called Medi-Cal in the California). Casually we might say that the reimbursement rate of commercial payers is 40% of charges, while Medicare pays slightly more than 20% of charges, and Medicaid pays less than 10% of charges. The demographics of the patient population in the area surrounding the hospital will determine the share of the revenue coming from commercial payers. Therefore profitability will to some extent depend on the demographics of the patient population independently of the market power the hospital is able to exert.
- need a citation for this -- or can we calculate using OSHPD, CMS, CHHS?
- maybe get away from talking about economic rent as a concept, and just use accounting profit, unless Sutter lawsuit actually invokes it
In California, the state government -- via OSHPD and CHHS -- requires hospitals to report their finances, along with other statistics on a quarterly basis. They then make that information public. Likewise, at the federal level, CMS makes a lot of financial data regarding the hospital services it pays for public.
While the data is messy, and combining it is convoluted and requires significant subject matter expertise, it is possible to paint an accurate picture of the state of the health care provider sector using this information. We are presently engaged in this project in order to get a clear and accurate picture of the market and answer our central question.
Using HHI and CR4 as measures, we find that for most of California -- with the possible exception of LA, Orange County, and East LA, the hospital industry is oligopolistic bordering on monopolistic. But the two measures don't completely agree. The 4-firm concentration ratio (CR4) probably does a better job of measuring the degree of competition here.
But it is also unclear whether we should be measuring things by net revenue, licensed beds, discharges, or adjusted discharges. We should really look at all three. This demands further attention and we will return to it in the future.
We find that the hospital industry does exhibit significant concentration. However it is unclear whether they've been able to raise prices over costs sufficiently high to generate economic rent.
The question now remains, why might hospitals, even if they exert considerable control over the market, not be able to extract rents? The most likely answer is that there are other economics forces counteracting the market power hospitals have. Let's break this down.
Hospitals revenue comes from commercials payers, CMS, and Medicaid. The commercial heath insurance market is fairly concentrated as well. Furthermore these insurers serve as an intermediary between employers/patients, physicians, and hospitals, clinics, pharmacies, and surgery centers. They therefore often have more information than hospitals do about the actual willingness to pay or willingness accept of other actors in the market. They are able to exert pricing pressure on hospitals through contract negotiations.
- try to get detail on HHI of commercial payers
- should I develop this information argument or is it a distraction
The other payers -- Medicare and Medicaid -- exert even more market power. In fact, according to MedPac, efficient hospitals run a 2% loss on their Medicare patients. And they definitely lose money on their Medicaid population. In order to turn a profit, hospitals have to extract sufficient revenue from their commercial patients to subsidize the losses they take from government payers.
On the expense side, the largest expense category for any hospital is labor: salaries, wages, and benefits. And labor is unionized. Just a handful of unions dominate the nursing and tech skill categories that hospitals depend on.
- add detail
Economic theory offers a couple other considerations here: efficiency wages and expenses as a barrier to entry.
- can we compare average daily census to bed capacity to estimate excess capacity
- if we can, do mergers actually decrease excess capacity?
Efficiency wages are wages paid above the market clearing wage in order to attract the highest quality labor and reduce turnover. Henry Ford famously paid a $5 daily wage when the market wage for similar work was $2, and claimed it increased profits by cutting overall costs. But we don't see any evidence for this happening presently in the nursing and tech labor markets.
Another reason for hospitals to pay higher wages is to raise a barrier to entry or drive out competitors. If we saw an increase in wages trending along with an increase in hospital consolidation, we might suspect the two were related.
However, Prager and Schmitt (2021 cite) have found that hospitals have actually suppressed wages for nurses and techs -- exhibiting wage-setting power, like a monopsony or single buyer in a market. So rather than strategically raising nursing salaries to block competitors, hospitals are using whatever muscle they have to reduce wages.
But if hospitals are suppressing the cost of labor successfully, why isn't that leading to higher profits? That seems to leave the payers -- that is, commercial payers such as Blue Cross, United Health, along with government payers Medicare and Medicaid responsible for negating the opportunity for hospitals to extract economic rents.
Yet, as earlier mentioned, the California AG alleges that Sutter Health has exhibited price-setting power with commercial payers such as Blue Cross.
Point: Market power on the sell side and the buy side. Can market power of payers (especially the government) counteract the market power of health systems?
In a Cournot market with dispersed buyers, the HHI divided by the price-elasticity of demand equals the price-cost margin (Hendricks and McAfee 2007). The DOJ guidelines actually use capacity, not market share of sales, to calculate HHI. In the context of hospitals, this would mean using hospital beds rather than discharges to calculated the HHI on the sell side of hospital services. However the logic of Cournot would use discharges, the quantity units, or perhaps adjusted discharges, to account for the outpatient side of the business. The hospital market is also vertically separated, that is, there are no hospital systems that provide insurance or vice-versa.
Lee and Fong (2013) develop a Markov-perfect dynamic network model of bilateral oligopoly with an application to negotiations between hospital systems and commercial insurance payers. Their model allows payers to renegotiate contracts following a hospital merger. As they frame the issue,
Consolidation in the US healthcare delivery system has increased dramatically in recent years, and anti-trust regulators have become increasingly concerned with market concentration leading to decreased consumer welfare. However, regulators have historically had difficulty challenging mergers.
A key question in such anti-trust analyses is whether the potential benefits of consolidation, such as increased efficiency, reduced excess capacity, lower transactional frictions, and higher risk tolerance outweigh the potential costs of increased market power.
They find in simulations that payers can extract the value of increased efficiency when hospitals merge, preventing hospital profits from increasing. Alternatively depending on simulation parameters and luck, hospital mergers can decrease total profit if insurers can no longer steer patients to the lower-cost hospital, reducing system efficiency, hurting the merged provider as well as the payer. While such results are only simulations based on applied theory with a complex setup, and are not intended to accurately mimic how hospital-payer negotiations work in practice, they nevertheless support our intuition that the interaction of oligopolies can have unpredictable effects.
Moriya, Vogt, and Gaynor (2010) build a 3-year panel of eleven million privately-insured Americans and look at changes in prices as consolidation occurs among hospitals and among insurers. They reject the hypothesis that hospital mergers raise prices, though the estimated model coefficients are positive. However they find that mergers of insurers decrease prices. Using estimated model parameters, they show that a hypothetical merger of two of five equally-sized insurers in a market would decrease prices 6.7%.
Indeed what I've got is a set of fuzzy hypotheses. But to answer your question I think there are forces acting on both the hospitals and the payers to consolidate -- possibly govt. And if the idea that hospitals and payers are locked in a bilateral oligopoly situation with each other, it's unclear whether one can extract surplus from the other. I might be able to get data on consolidation in the insurance payers, and I might have to.
Hospitals are consolidating to increase profit -- but that profit could be siphoned off by payers. Or the opposite might happen. The California case against Sutter was brought by the unions, who alleged Sutter's prices were hurting their members, by driving up the cost of insurance. But that's a tricky argument to make, if the harm to the consumer passed through another agent. I need to read the case carefully and to understand the story better.
Another thing to notice here is that the unions are also playing a game. Nursing wages are the single largest expense for a hospital, and hospitals are regularly engaged in negotiations with the unions. The tactics can get pretty nasty. But it's not clear that it would be in the union's best interest to hurt Sutter. Unions can only extract wage increases from management if there are profits to extract. On the other hand, Sutter is now in a weak position all around. They tried and failed to use their market power to set prices. Now they need to find ways to control costs, or improve quality under a regime of lower prices.
Further work on the data is needed before we can estimate the relationship between profit and market share and commercial payer mix using panel regression (random effects).
- https://aspe.hhs.gov/basic-report/effects-health-care-spending-us-economy
- https://www.oecd.org/health/health-systems/health-at-a-glance-19991312.htm
- https://www.ftc.gov/reports/do-nonprofit-hospitals-exercise-market-power
- https://www.ama-assn.org/about/research/trends-health-care-spending
- https://www.healthcarefinancenews.com/news/efficient-hospitals-operate-2-margins-medicare-payments-medpac-reports
- https://www.aeaweb.org/articles?id=10.1257/aer.20190690
- https://www.healthaffairs.org/doi/full/10.1377/hlthaff.2015.1193
- https://www.healthaffairs.org/do/10.1377/hblog20140612.039598/full/
- https://www.mcafee.cc/Papers/PDF/Concentration.pdf
- http://economics.yale.edu/sites/default/files/lee_mpnenetworkformation.pdf
- https://heinonline.org/HOL/LandingPage?handle=hein.journals/hecpol5&div=37&id=&page=