Demand Inertia and the Hidden Impact of Pharmacy Benefit Managers
Research by Josh Feng1 and Luca Maini2
1University of Utah, 2Harvard Medical School
What is the best approach to controlling prescription drug prices? Branded drug manufacturers enjoy significant pricing power in U.S. markets, thanks to patent protection, price-insensitive demand, and consumer inertia. To compensate for this imbalance in market power, US drug prices are often negotiated by intermediaries called Pharmacy Benefit Managers (PBMs), large entities that design drug benefits on behalf of almost 80 percent of all insurance enrollees across the commercial market (e.g., employers, unions, and large insurers), Medicare Part D, and the ACA health exchanges. PBMs claim that they help control spending and improve patient welfare, but critics argue that they have done very little to slow the rapid price growth of branded drugs. To date, empirical research on PBMs is scarce, and the role these intermediaries play in moderating drug prices remains poorly understood.
In this article, we shed light on the role of PBMs and argue that their effect on branded drug prices is masked by the presence of inertia in drug demand. Inertia has two theoretical effects. First, it means that manufacturers want to increase prices over time, as their focus gradually shifts from building up market share to extracting higher markups from existing customers. Second, it limits the impact of later entrants by putting them at a disadvantage when competing for existing patients. As a result, inertia reduces competition and conceals the true impact of PBMs, making them appear ineffective at controlling drug prices.
Left: impact of PBMs on manufacturer profits; right: counterfactual price series without demand inertia
We used MATLAB on CHPC to estimate the dynamic game. It helped because in each year, we have to compute value functions for a large number of states. Having access to many nodes greatly reduced the time to compute one iteration of the game, which allowed for faster estimation.
Attribution: This content was provided by the researchers and edited by staff at the CHPC for style.