Working Paper: NBER ID: w21045
Authors: Michael Sinkinson; Amanda Starc
Abstract: We measure the impact of direct-to-consumer television advertising (DTCA) by drug manufacturers. Our identification strategy exploits shocks to local advertising markets generated by idiosyncrasies of the political advertising cycle as well as a regulatory intervention affecting a single product. We find that a 10% increase in the number of a firm's ads leads to a 0.76% increase in revenue, while the same increase in rival advertising leads to a 0.55% decrease in firm revenue. Results also indicate that a 10% increase in category advertising produces a 0.2% revenue increase for non-advertised drugs. Both the business-stealing and spillover effects would not be detected through OLS. Decomposition using micro data confirms that the effect is due mostly to new customers as opposed to switching among current customers. Simulations show that an outright ban on DTCA would have modest effects on the sales of advertised drugs as well as on non-advertised drugs.
Keywords: Direct-to-Consumer Advertising; Pharmaceuticals; Advertising Effects
JEL Codes: I11; L1
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Ban on DTCA (M38) | Sales for advertised drugs (L42) |
| Ban on DTCA (M38) | Sales for non-advertised drugs (D49) |
| Firm's advertising (M37) | Revenue (H29) |
| Rival advertising (L49) | Firm's revenue (D21) |
| Category advertising (M37) | Sales for non-advertised drugs (D49) |