Working Paper: NBER ID: w16854
Authors: Roger Bate; Ginger Zhe Jin; Aparna Mathur
Abstract: Focusing on 8 drug types on the WHO-approved medicine list, we constructed an original dataset of 899 drug samples from 17 low- and median-income countries and tested them for visual appearance, disintegration, and analyzed their ingredients by chromatography and spectrometry. Fifteen percent of the samples fail at least one test and can be considered substandard. After controlling for local factors, we find that failing drugs are priced 13.6-18.7% lower than non-failing drugs but the signaling effect of price is far from complete, especially for non-innovator brands. The look of the pharmacy, as assessed by our covert shoppers, is weakly correlated with the results of quality tests. These findings suggest that consumers are likely to suspect low quality from market price, non-innovator brand and the look of the pharmacy, but none of these signals can perfectly identify substandard and counterfeit drugs. Indeed, many cheaper non-innovator products pass all quality tests, and are genuine generic drugs.
Keywords: poor-quality drugs; price signaling; public health; substandard medicines
JEL Codes: D18; D4; I11; I18; L15
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| lower prices (P22) | higher likelihood of poor quality (L15) |
| lower literacy rates (I24) | larger price discount for failing drugs (L42) |
| look of pharmacy (L65) | quality test results (Y10) |
| local regulations (R48) | drug quality outcomes (L15) |
| income levels (J31) | drug quality outcomes (L15) |
| literacy rates (I24) | drug quality outcomes (L15) |
| presence of product registration (L15) | better drug quality (L15) |
| presence of innovator brands (O35) | better drug quality (L15) |
| higher tariffs and duties (F13) | lower quality (L15) |