Pharmacy benefit managers (PBMs) are the under-discussed market participants who manage prescription drug insurance for the vast majority of Americans. PBMs claim to be a lone bulwark against the rapacious pricing decisions of pharmaceutical firms, but is recent consolidation in the PBM market also a driver of high prices for prescription drugs? 

Controversy over the rising cost of pharmaceuticals has led to calls from both sides of the political aisle for government interventions, causing members of the once-cozy pharmaceutical sector to pass the blame for high prices like a proverbial hot potato. This phenomenon was on display following the pricing controversy over Mylan’s EpiPen. CEO Heather Bresch placed the blame for high list prices squarely on “middlemen” such as pharmacy benefit managers (PBMs)—the under-discussed market participants who manage prescription drug insurance for the vast majority of Americans. PBMs responded to this criticism by asserting that they, in fact, stand as a lone bulwark against the rapacious pricing decisions of pharmaceutical firms. This infighting over the cost of pharmaceuticals likely plays a role in the recent reorganization of payer-PBM relationships including the decision of Anthem to end its contract with Express Scripts and of CVS to attempt to purchase Aetna. 

To be sure, some of the high prices for pharmaceuticals in the United States are the result of a legitimate attempt to provide the appropriate incentives for innovation. However, another contributing factor to the current high prices is a dysfunctional market that rewards both PBMs and pharmaceutical manufacturers if prices rise. No one wants pharmaceutical innovation to slow, so attention in health policy must focus on understanding the nature of price competition in the pharmaceutical market—how it works, why it doesn’t work, and how to fix it. The answer to high prices isn’t broad price regulation, but restoring the intended level of competition to a market characterized by a dangerous combination of PBM consolidation and opaque pricing.

Economists often describe the adverse consequences that result from consolidation of market power in the hands of either sellers or buyers. In the case of the pharmaceutical market, we argue that the recent consolidation of the middlemen market, i.e., the PBMs, combined with opaque pricing is one cause of higher prices. When only a few PBMs exist, it is all too easy for them to stop functioning as brokers that increase market efficiency, and start looking for win-win arrangements in which consumers are the ultimate losers. The first step in disrupting this dynamic is understanding it.

Net prices for on-patent drugs are determined as follows. A final payer (e.g., an employer or insurer) contracts with a PBM to manage its pharmacy claims, design formularies, and negotiate the prices paid to pharmaceutical manufacturers. Manufacturers set an initial list price for their products. PBMs then negotiate with manufacturers over a rebate (i.e., discount) from the publicly known list price. A PBM can stimulate price competition where an individual consumer cannot because the PBM acts on behalf of a group of consumers and can shift those consumers between competing, substitute drugs. If there are two good blood pressure medications, the PBM can place the one that offers more favorable terms on the lower cost tier and encourage consumption by its enrollees. The higher-cost competitor then loses market share, which induces both drug manufacturers to offer discounts. If PBMs can credibly shift many consumers across rival products, they have more bargaining power and can earn larger rebates. To “move share,” PBMs invest in tools like differential out-of-pocket payments, prior authorization, and step therapy.

“When only a few PBMs exist, it is all too easy for them to stop functioning as brokers that increase market efficiency, and start looking for win-win arrangements in which consumers are the ultimate losers.”

For many good reasons, the magnitude of the rebates made by manufacturers to PBMs is kept confidential. This lack of transparency is not inherently problematic and in fact has many benefits. However, the tradeoff for these benefits is that payers cannot fully observe what is at stake in their PBM negotiations.

Even when rebates are confidential, an important question to ask is who ultimately gets the rebate: the PBM or the final payer (i.e., the employer)? The contract between the employer or insurer and the PBM is complex. The contract may allow for sharing of the rebates between the PBM and final payer, per-member-per-month charges, administrative fees, consulting fees, and a proportion of the list price of the medication that the final payer pays. The share of the rebate kept by the PBM can take the form of a flat fee per prescription or a percentage of the rebate paid. The ultimate terms of this contract depend on both the relative bargaining power of payers and PBMs and the information each has about the amount of economic surplus that is available.

In a well-functioning competitive market, the combination of this contract structure and lack of rebate transparency shouldn’t matter to the final price. In such a market, PBMs would compete by offering payers a package of total costs and quality and payers would pick the PBM with the most attractive options. In an attempt to win the payer contract, a PBM in a competitive market would compete with lower prices until the total package of payments just covers its costs. In other words, effectively, the entire rebate would be transferred to the payer. However, any lack of competition in a more concentrated market decreases the incentive to lower price. Raising price by increasing the per-member-per-month fee is transparent and salient for consumers. But if the PBM chooses a contract type that allows it to keep a share of the rebate, and then goes on to create contracts with manufacturers featuring higher prices and larger rebates, net prices rise in a way that is hard for the final payer to see. Note that a spate of recent PBM mergers has resulted in the top three firms controlling nearly 80 percent of all prescriptions.

Contracts with shrouded rebates in a concentrated market can result in higher manufacturer prices and increased profits for PBMs. This situation is illustrated by the following simple example. Suppose the manufacturer raises its list price by $10 and its rebate by $9. The result is a $1 higher net price so the manufacturer is better off. If a lack of competition allows a PBM to return $8 to the payer instead of the full $9, the PBM is better off by $1 also. The PBM has little reason to bargain with manufacturers to keep prices from increasing in the first place; indeed their incentive is to encourage higher prices and higher rebates. Meanwhile, the payer’s drug costs increase by $2. In a competitive PBM market, we would expect another PBM to approach the payer and offer to give it $8.50 of the rebate, and another would approach offering $8.75, and eventually all of the rebate would be returned to the payers.

In a truly competitive market we would expect the PBM to attempt to win the payer’s contract by also negotiating with the manufacturer to increase the rebate to $10 and return that full rebate to the payer. These patterns, however, depend on the existence of multiple PBMs actively competing for the payer’s business—something unlikely to emerge in a market with few PBM participants and large barriers to entry, including barriers like contract terms created by the PBM. The situation becomes even less competitive should the small number of existing PBMs begin to realize that such price competition would decrease everyone’s profits.

In private, many in the industry acknowledge the existence of these perverse incentives. However, there is little systematic evidence of their impact on pharmaceutical prices. We took a step towards understanding these relationships by investigating whether PBM stock prices rise when pharmaceutical prices are expected to increase. If the PBM market were competitive, changes in beliefs about future drug prices should cause the stock prices of manufacturers (who benefit from high prices) and PBMs (for whom high prices mean increased costs) to move in opposite directions. However, if both types of firms benefit from higher prices these stock prices should move in the same direction.

The first event we examined was the election of Donald Trump. Prior to the election, nearly all pundits and polls predicted an easy victory for Hillary Clinton. As a result, stock prices just prior to the election were based on an expectation of four years of Democratic policies such as the use government power to decrease drug prices. Trump’s surprise election provided a clear shock to this belief. Figure 1 shows the results of an event-study analysis estimating the effect of the election on the stock prices for Express Scripts (the largest and only publicly traded firm that is solely a PBM) and a set of pharmaceutical manufacturers.((Details of the methodology for this event study analysis are contained in the supplementary appendix at this link.)) These estimates show that the election caused stock prices to meaningfully increase for all of these market participants.

Figure 1: Stock Price Change Following Presidential Election. The estimates in figure 1 are the estimated abnormal returns in the stock markets for the firms (or groups of firms) from an event-study analysis of the election of Donald J. Trump to the office of President of the United States. The estimated coefficients in the figure represent the change in the stock that occurred after accounting for normal market fluctuations occurring over that time period. Estimates are statistically significant at a p-value of 0.05. More details on the empirical strategy are contained in the supplementary appendix.

We next examined president-elect Trump’s public statements in a post-election Time magazine interview in which he said that he would break from traditional Republican orthodoxy and implement policies to decrease drug prices. These statements received wide press coverage. Figure 2 contains similar estimates to figure 1 for stock prices before and after these public statements. In this case, firms’ stock prices simultaneously decrease. Other health care stocks, such as those of for-profit payers, did not track differently from the rest of the market following this news of lower expected drug prices.

Figure 2: Stock Price Change Following Time Magazine Interview. The estimates in Figure 2 are the estimated abnormal returns in the stock markets for the firms (or groups of firms) from an event-study analysis of President-Election Donald J. Trump’s interview with Time magazine in which he suggested he would support policies to decrease drug prices. The estimated coefficients in the Figure represent the change in the stock that occurred after accounting for normal market fluctuations occurring over that time period. Estimates for pharmaceutical, biotechnology, and Express Scripts are significant at a p-value of 0.10. More details on the empirical strategy are contained in the supplementary appendix.

Taken together, these results suggest that perverse market incentives have created a common interest in high prices for sellers and the initial buyers of drugs. Changes are needed to return the market to something that approximates competitive fundamentals. Clearly, unwinding PBM mergers is very costly. As an alternative, more insurers could take steps similar Anthem’s recent actions and break off relationships with PBMs that they believe are capturing too high a share of rebates. Anthem is poised to follow the strategy of insurers like United and Humana who have vertically integrated into the PBM function, perhaps in order to address the problem of the PBM’s incentives. Similarly, the recent attempt by CVS to purchase Aetna could eliminate many of these incentives for drugs purchased by Aetna’s clients. 

When the medical insurer carries out the pharmaceutical insurance function two efficiencies are generated. First, the PBM now internalizes the effects that medications have on health and charges lower co-pays for drugs that keep people out of the hospital. Secondly, the incentive to keep drug prices down is stronger. However, remember that final payers will only benefit from those lower pharmaceutical costs if there is a competitive market in medical insurance, which may not always be present.

Another alternative often proposed by policy makers is requiring public transparency of all rebates. However, the economics literature predicts this will lead to higher, not lower, prices.((Pharmaceutical firms are less likely to give large discounts to a PBM if that rebate could be seen and used against them in subsequent negotiations with other buyers. If rebates are public, manufacturers anticipate this eventual pressure and offer smaller, more uniform rebates that PBMs cannot negotiate—and ultimately consumers pay higher prices. For these reasons, public discounts could well raise net prices. In addition, public rebates make it easier for manufacturers to tacitly collude or engage in other behaviors designed to maintain high prices.)) A broadly feasible third approach would avoid these negative consequences and restore PBMs’ role as an agent for purchasers. Our proposal: all negotiated payments should be initially transferred in full from the manufacturer to the final payer. This would include rebates, and also any administrative fees or other payments from manufacturers that are increasingly common and unobservable to final payers. This change could be accomplished either by private contract or government action. Payers who see their entire rebate will know the true aggregate net price of the drugs they are buying.

Our policy would eliminate information asymmetries regarding rebates in negotiations between PBMs and insurers. We realize there may be a concern that not allowing PBMs to initially keep a portion of the rebate might diminish incentives to negotiate discounts. However, under our proposal payers can still efficiently negotiate with PBMs regarding subsequent transfers of the rebate if they desire. The final payer would simply pay the PBM according to the contract after it had received all the rebates. Negotiations are more likely to reach competitive outcomes when both participants have full knowledge of the scope of payments. In particular, the additional information will allow payers to more effectively push back against the current preferences of both manufacturers and PBMs for higher prices. Finally, more attention to the problem of PBM incentives and poor performance might create demand from employers for better service from new entrants.

Supplementary Appendix on Data and Methodology

Craig Garthwaite is an Associate Professor of Strategy and the Director of Kellogg School of Management’s Health Enterprise Management Program (HEMA) at Northwestern University. Fiona M. Scott Morton is the Theodore Nierenberg Professor of Economics at the Yale University School of Management where she has been on the faculty since 1999.

Disclaimer: The ProMarket blog is dedicated to discussing how competition tends to be subverted by special interests. The posts represent the opinions of their writers, not those of the University of Chicago, the Booth School of Business, or its faculty. For more information, please visit ProMarket Blog Policy.