By Jillian Chown, David Dranove, Craig Garthwaite, and Jordan Keener
National Bureau of Economic Research, July 2019
Perhaps more than any other sector of the economy, healthcare depends on government resources. As a result, many healthcare systems rely on the use of government monopsony power to decrease spending. The United States is a notable exception, where prices in large portions of the healthcare sector are set without government involvement. In this paper we examine the economic implications of a greater use of monopsony power in the United States. We present a model of monopsony power and test its predictions using price differences between the United States and Canada – a country that represents an example of a “Medicare for All” style system. Overall, we find that wage differences for medical providers across the two countries are primarily driven by the broader labor market while price difference for prescription drugs are more directly the result of buyer power. We discuss theoretical reasons why a Canadian monopsonist may be more willing to exploit its buyer power over prescription drugs rather than provider wages and why a U.S. monopsonist might not be willing to do the same.
From the Introductory Comments
Health care spending varies markedly across countries. It is widely accepted that at least a portion of spending differences results from the willingness and ability of governments to exercise monopsony power when purchasing healthcare goods and services. Broadly speaking, governments that finance healthcare services often use their buyer power to push down the prices paid for healthcare inputs such as pharmaceutical products, medical devices, and the wages of medical providers. Differences in the willingness and ability to exercise buyer power certainly contributes to the considerable healthcare price and wage dispersion across countries. However, the degree to which these price differences across countries are the result of buyer power as opposed to other economic forces is an open question that we explore in this paper.
At an aggregate level, the United States stands at an extreme of a relatively parsimonious government role in the financing of healthcare services, a limited use of government buyer power, and a relatively large percentage of its economic activity occurring in the health sector. Concerned about the rising cost of U.S. healthcare, many policymakers have proposed that the federal government make greater use of its buyer power. Proposals include various “Medicare-for-All” policies that would expand existing public insurance systems to cover more than simply the elderly, poor and disabled, as well as more limited proposals, such as requiring Medicare to use its size to negotiate lower pharmaceutical prices. The economic effects of a greater use of monopsony power obviously depend on the specifics of the proposed policy reform. While proponents have offered a variety of “Medicare-for-all” alternatives, we will use the term in this paper to refer to a truly single payer system where no private insurance options exist. Under this option, which has been proposed by numerous policymakers, the federal government would be the single purchasing agent for all Americans.
The potential savings from implementing such a single payer system would primarily come from two sources: (1) decreased administrative costs resulting from a single firm providing health insurance and (2) the exercise of buyer power to reduce input prices. We will briefly discuss the former before moving on to the latter, which is the main focus of this paper.
Centralizing insurance in a single entity would certainly eliminate some duplicative administrative functions. In addition, a single insurer that lacked competitive pressures would likely have lower marketing expenses.
While administrative expenditures are a potential source of savings, a single payer aiming to meaningfully reduce spending would, by necessity, follow the proverbial “Willie Sutton” rule of health spending and target the prices paid to healthcare inputs such as medical providers and pharmaceuticals. These input prices represent the bulk of U.S. healthcare spending with labor costs comprising approximately 60 percent (Kocher and Sahni, 2011) and pharmaceuticals accounting for an additional 15 percent. A government payer could cut both of these prices either by directly reducing the fees paid to physicians, other medical personnel, and pharmaceutical manufacturers or by reducing payments to hospitals and other institutions, which would then have to reduce their costs (i.e., cut wages and reduce other medical inputs) to remain financially viable.
Proponents of an expanded use of government buyer power frequently cite the experiences of other developed countries that have both far more extensive government involvement in healthcare and lower health spending. Within these comparisons, the Canadian system is perhaps the most similar to Medicare-for-all. In Canada, provincial governments offer health insurance and finance care delivery under rules established by the government, but medical services are mostly provided by private providers in privately owned and operated facilities. The similarity of the Canadian system to many U.S. proposals for Medicare-for-all can perhaps be most clearly seen in Senator Bernie Sanders (I-VT) explicit statements that the Canadian system is a model for his widely discussed universal coverage proposal (Kliff 2017). Thus, the decisions of Canadian insurers provide perhaps the best evidence of the potential optimal savings available from the creation of a single monopsonist in the United States.
In this paper, we present a simple theoretical model to predict the optimal decisions of a healthcare monopsonist such as Canada. We then use price data across the U.S. and Canada to empirically test the predictions of this model to help understand the scope of savings optimally available to a similar monopsonist in the United States.
The predictions of the model suggest the optimal exercise of market power will differ based on both the relative size of the monopsonist in the relevant market and the resulting elasticity of supply it faces. We test the predictions of this model using U.S. and Canadian price data for two key inputs of healthcare services: prescription drugs and medical providers.
Given our model predicts a different response based on the monopsonist’s relative market size (and therefore the respective elasticity of supply), it predicts different exercises of market power across these two input markets. Specifically, our model predicts the Canadian monopsonist would exercise more of its market power for pharmaceuticals than for labor.
We find evidence that the Canadian monopsonist does exercise buyer power with respect to drugs. Using highly detailed pricing data that has not been previously studied for this purpose, we estimate that for a consistent basket of pharmaceutical goods, Canadian consumers pay 54 percent less than Americans.
Our analysis finds that Canadian drug prices are 54 percent lower – a difference that is far greater than for other goods and services. We interpret this as evidence that Canadian provincial governments are willing and able to exercise monopsony power over drugs to a greater degree than in the labor market – which is consistent with our monopsony model’s predictions.
The fact that Canadian wages for highly skilled employees show limited evidence of an exercise of buyer power while prescription drug prices indicate a large role for the public payer’s negotiations suggests that Canadian policymakers carefully consider the potential consequences of exploiting monopsony power in healthcare markets. For products where Canadian consumers represent a relatively small part of the market, such as prescription drugs, policymakers have the freedom to exploit buyer power without meaningful consequences – and they do. In contrast, in labor markets, Canadian policymakers must contend with the fact that if they use their negotiating power to push down input prices it could have large and potentially harmful short and long term consequences for healthcare labor supply.
From the Conclusions
In this paper, we examine the role of monopsony power in healthcare markets by comparing prices and quantities of healthcare inputs in Canada and the United States. We provide a theoretical framework and empirical evidence for the optimal behavior of a monopsonist. We focus on two critical inputs to the healthcare production function that illustrate key features of the theoretical model – labor and prescription drugs. In line with the theoretical expectations we outline, we find that the Canadian monopsonist is more willing to exploit its market power over prescription drugs than provider wages.
Examining a market basket of prescription drugs, we find that prices in Canada are 54 percent lower than in the U.S. This price differential is much larger than the general cost-of-living differences between the two countries, which we interpret as evidence that the Canadian monopsonist is willing to use its power to lower the cost of this particular medical input. This is in line with our theoretical expectations: because Canada represents a small share of the global drug market, policymakers are able to exploit their buyer power without incurring broader consequences, such as reducing innovation.
In contrast, we find that the differential in provider wages between the U.S. and Canada can largely be explained by differences in the prevailing labor market conditions, implying that the Canadian monopsonist is not willing to exercise the full scope of its buyer power for this input. In particular, we find that Canadian medical workers earn about 24 percent less than their American counterparts. If this was solely the result of profligate U.S. health spending, we should expect that U.S. providers would not only be paid more than their Canadian counterparts but also relatively more than other highly skilled professionals in the United States. This is not the case. We primarily attribute differences in wages for U.S. medical providers to the structure of the broader labor market; i.e., U.S. providers are paid more than in Canada because all highly skilled U.S. professionals are paid more than in Canada. In particular, high skilled healthcare workers in Canada make about 26 percent less than similar workers in the United States, while high skilled non-healthcare workers make 22 percent less. Thus, the vast majority of the 26 percent wage differential likely reflects overall labor market conditions. The remaining 4 percentage point differential between healthcare and other professional wages may reflect a variety of factors—potentially including inefficiencies in the U.S. system that provide excess wages to healthcare providers or a choice by the Canadian monopsonist to exploit some of its market power.
Our findings are consistent with the idea that the Canadian monopsonist believes that a substantial reduction in the wages of physicians and other healthcare workers could cause a reallocation of talent across the economy. This is true even conditional on similar individuals receiving medical training. These individuals have a number of opportunities beyond clinical practice, and there are some reports physicians are increasingly choosing these options (Mostue 2017). Perhaps more concerning would be a reallocation of talent out of medical training altogether. For example, if the U.S. took steps so that its physicians were paid the same as Canadians, this would move physicians from the second highest paid profession in the U.S. to the eighth highest paid profession — situated just above computer engineers, mathematicians and biomedical engineers. This change in relative pay could push many talented young individuals considering medicine into entirely different fields. Given this concern, it should perhaps not be surprising that even the Canadian monopsonist does not appear to have chosen to meaningfully exercise its wage setting power. In Canada, as in the United States, physicians and surgeons are among the highest paid employees in society.
Taken together, our findings should temper the expectations of anyone who believes that Medicare-for-all will drive U.S. health spending to Canadian levels. Most health spending goes toward labor expenses. Returning to the Willie Sutton analogy, while labor is “where the money is,” a U.S. monopsonist might find, just as in Canada, that it is not worth “robbing the bank” by lowering labor expenses, as this would drive highly skilled workers into other fields. And while the U.S. monopsonist could drive down drug prices to levels at or below those in Canada, it might be tempered by concerns about disruptions to long term drug supply, a problem not shared by the much smaller Canadian monopsonist.
By Don McCanne, M.D.
In the debate over the proposal for a single payer model of Medicare for All, the claim is often made that giving the government that much power to set prices (functioning as a monopsony – a single purchaser of health care) would drive physicians and hospitals into insolvency thus decreasing access to health care. This NBER paper is important because it shows what is actually happening in Canada under its single payer model of health care financing and what might happen in the United States under a single payer Medicare for All monopsony.
Many studies have shown that much of the administrative waste that characterizes the health care financing system in the United States would be recoverable under a single payer system. But a single payer would also have the monopsonistic power to drive down prices, potentially saving even more funds. So how has Canada, with its Medicare for All-like system, utilized its monopsonistic power?
Canada has driven down drug prices so that they are 54 percent lower than in the United States. They have been able to do that since their drug market is only about one-tenth of that in the United States, thus there is no fear that lower profit margins would have any impact in suppressing new drug innovation. This fear of suppressing innovation has been expressed repeatedly in the United States considering our very large drug market, but other factors not addressed here suggest that the costs of drug innovation do not justify the price gouging that is taking place in the U.S.
What is particularly striking in this paper is that Canada is not using its monopsonistic power to drive down wages and salaries in the health care industry to levels that are comparable to those in our traditional Medicare program. Rather they are pricing health care wages based on prevailing labor market conditions.
In the United States, Medicare has exercised its monopsonistic power to drive prices down to levels that some claim to be below costs. It is the private insurance sector that has been complicit with driving prices up, and Medicare has been able to offset some of the excessive pricing through its monopsonistic leverage.
Private insurers have at least two incentives to drive up prices. Revenue in excess of medical losses (costs of patient care) are retained by the insurer to fund their own administrative costs and to provide profits. Their primary product is administration and the more administrative services they can sell, the better for them. So private insurers contribute directly to high health care costs in the U.S. both by driving up prices and by generating ever more administrative waste.
But in Canada, the public stewards are pricing health care at a level that ensures an adequate health care labor supply yet not wasting taxpayer funds on excessive prices. Further, through the administrative simplicity of the single payer model, they are not wasting financial resources that are better used for patient care,
In the United States, we are now seeing a softening of the demand to enact single payer Medicare for All while at the same time we are seeing the manipulation of public opinion in support of ensuring the preservation of the option to choose private insurance plans. This is backwards. The private insurers are in the business to increase their revenues (waste our funds) whereas publicly financed and administered systems are public service systems that have a mission to provide us with the greatest value in health care. They can do that with a monopsony that gets the prices right by balancing legitimate costs with patient needs. The private insurers will not do that for us.
Stay informed! Visit www.pnhp.org/qotd to sign up for daily email updates.