Demand Elasticities for Pharmaceuticals: Evidence from a Payment Update in Medicare Part D

Medicare Part D implements a prescription drug insurance benefit for twenty million elderly and disabled Americans using Federal payments to private insurers.  This research analyzes the primary component of those Federal payments and in particular how firm behavior and beneficiary utilization respond when payments are recalibrated to reflect background trends in medical costs due to technological change.

The majority of subsidies to insurers in Part D are distributed through a system of diagnosis-specific payments (“risk adjustment”).  The diagnosis-specific payments in Part D aim to equalize insurer incentives across the healthy and the sick by paying insurers the diagnosis-specific marginal cost of treating each enrollee’s illnesses.  For example, an average-premium plan in 2010 enrolling a 66 year old man whose medical claims reflect Infectious Diseases would receive a diagnosis-specific payment of $454. If a similar enrollee’s medical claims instead reflect HIV/AIDS, the plan would receive $2541.  In principle, plans are equally willing to enroll both men because the diagnosis-specific payments offset the higher expected cost of the HIV/AIDS patient.

However, the levels of the diagnosis-specific payments were calibrated using data from the early 2000s and then left in place through 2010, despite new drug entry and the onset of generic competition (technological change) raising or lowering the costs of treating certain diagnoses.  In 2011, the payment system was updated to again set payments equal to associated treatment costs.  For the two men discussed previously, their insurer in 2011 now receives $560 for the Infectious Disease patient and $2141 for the HIV/AIDS patient.

According to the theoretical model of Frank, Glazer, and McGuire (2000), insurers’ benefit designs – whether a drug is covered and its copay – should respond to the change in incentives provided by this updating.  In particular, insurers should reduce benefit design generosity for diagnoses receiving negative payment updates (to deter individuals made less profitable by the update), and increase benefit design generosity for diagnoses receiving positive payment updates (to attract individuals made more profitable by the update).  We use the Part D formulary files for 2011 to test the prediction that Part D plans increased coverage and reduced copay for drugs treating diagnoses whose payments rose, and conversely decreased coverage and increased copay for drugs treating diagnoses whose payments fell.  Preliminary evidence is consistent with this hypothesis.

Simple demand theory implies that if enrollees face a decline in benefit design generosity for certain diagnoses in 2011, they will respond by (weakly) reducing utilization.  In general, straightforward applications of demand theory face a significant challenge in the endogeneity of price.  The Part D payment system revision, however, provides a plausibly exogenous shifter in Part D insurers’ input prices.  Using a two-year panel of drug claims, we test the enrollee-level change in each drug’s utilization (measured by days supplied in Part D claims) between 2010 and 2011.  We recover own-price demand elasticity (under an assumption of no intermolecular drug substitution) or cross-price demand elasticities (under a functional form assumption for drug demand within a pharmacologic class).