Direct-to-Consumer Advertising of Pharmaceuticals (Figures)

Figures Figure 1: Political Ad Levels, January-June 2008 Notes: The above maps show a circle for each DMA in the USA. The diameter of each circle is proportional to the number of political ads aired in that market, in that month, for all races (Presidential, Senatorial, House, Gubernatorial). The first row are January and February; second row are March and April, and third row are June and July. Figure 2: National Pharmaceutical Ad Levels for Statins Notes: The above graphic plots national advertising spots from the Kantar data. Data spans January 2007-November 2008. Figure 3: National Pharmaceutical Ad Levels for Statins Notes: The above graphic plots the maximum of local advertising spots across DMAs from the Kantar data. Data spans January 2007-November 2008. The axes are the same as the previous figure. Figure 4: Political Ads Displace Local Drug Ads, Binned Scatter plot Notes: The above plots bins of observations from July 2007 to November 2008 at the market-month level after residualizing by market and drug-year-month fixed effects, and adding back the sample mean. The plot uses local drug ads, although the plot that also includes national is identical due to the drug-year-month fixed effects. Twenty bins are used. The fitted line is based on a regression of all underlying data, not only the binned values. Figure 5: Effect of Primary Timing on Non-Advertised Statins Note: The above plots estimated coefficients for timing dummies relative to a market’s primary month. The dependent variable is the (one-month) change in market share, defined as the percentage of the population taking a non-advertised statin. Figure 6: Effect of Primary Timing on Crestor and Lipitor Note: The above plots estimated coefficients for timing dummies relative to a market’s primary month. The dependent variable is the (one-month) change in market share, defined as the percentage of the population taking Lipitor or Crestor. Figure 7: Simulation Results: Eliminating DTCA Note: The above plots are histograms of the change in quantity for each drug (or drug group) from bootstrapped simulations that eliminate DTCA from the market over the sample period. See section 5 for an extended discussion of the methodology. Tables Table 1: Summary Statistics Drug Number of Markets Number of Months Advertised Statins 186 17 4 Political Ads Average 774 Standard Deviation 1,897 Minimum 0 Maximum 22,636 Drug Usage (Truven Analysis Data set) Average Branded Share 0.829% Range, Branded Share (0.000%, 4.71%) Average Generic Share 3.05% Range, Generic Share (0.000%, 7.62%) Drug Ads Conditional Mean of Local Ads by Drug 9.67 Range, Local Ads (0, 105) Conditional Mean of National Ads by Drug 45.31 Range, National Ads (0, 145) Notes: Unit of observation is the market-month-product combination. Data (source: left panel, Wisconsin and Kantar; right panel, Truven) span 17 months from July 2007 to November 2008. Averages in top right panel are over the entire population. Means in bottom right panel condition on advertising. Table 2: Covariate Balance, Part D Data Average Age % Female % White % Black % Hispanic Mortality Rate % Low Income Subsidy Below Median Markets 71.109 0.5489 0.8536 0.0849 0.0147 0.0423 0.6874 Above Median Markets 71.309 0.5519 0.8727 0.0933 0.0088 0.0425 0.6657 Difference −0.1994 −0.0030 −0.0190 −0.0083 0.0058 −0.0002 0.0217∗∗ Notes: We split the Part D beneficiary summary sample into two groups. We take the sum of political advertising over the 2008 calendar year and compare demographics for markets above and below the median. Statistical significance at the 10%, 5% and 1% levels are denoted by ∗ , ∗∗ , and ∗∗∗ . Table 3: Political Ads Displace Drug Ads Model: Log(Political Ads in 1000s + 1) Controls: Market FEs Year-Month FEs Drug FEs Drug-Year-Month FEs N R2 OLS −0.1825∗∗∗ (0.0129) X Dependent Variable: Log(Local Statin Ads + 1) OLS OLS Tobit Tobit ∗∗∗ ∗∗∗ ∗∗∗ −0.1063 −0.1063 −0.7729 −0.2329∗∗∗ (0.0143) (0.0143) (0.0494) (0.0581) X X X X 14, 280 0.267 14, 280 0.357 X X X X 14, 280 0.477 X X X X X 14, 280 0.213 14, 280 0.348 Notes: Unit of observation is the market-month-product combination. Data (source: Wisconsin and Kantar) span 17 months from July 2007 to November 2008 across 210 markets and include Crestor, Lipitor, Vytorin, and Zetia. OLS and Tobit standard errors clustered at the market-month level. The results with Drug-Year-Month fixed effects are identical if local plus national statin ads are used as the dependent variable. Statistical significance at the 10%, 5% and 1% levels are denoted by ∗ , ∗∗ , and ∗∗∗ . Reported R2 is adjusted for OLS, pseudo for Tobit. Tobit −0.1940∗∗∗ (0.0103) X X X X 14, 280 0.519 Table 4: Difference-in-Difference Estimates Ptm It Ptm ∗ It Crestor ∗ Ptm Crestor ∗ It Crestor ∗ Ptm ∗ It Products Fixed Effects Clustering N R2 Dependent Variable: Log(Days Supply) (1) (2) (3) ∗∗∗ ∗∗∗ −0.0291 −0.0955 −0.0705∗∗∗ (0.0106) (0.0137) (0.0174) ∗∗∗ −0.0873 −0.1051∗∗∗ (0.0160) (0.0187) ∗∗∗ 0.0979 0.1220∗∗∗ (0.0188) (0.0275) 0.0621∗∗ (0.0259) 0.1291∗∗∗ (0.0349) −0.1459∗∗ (0.0565) Crestor, Lipitor Lipitor Crestor, Lipitor Product, DMA DMA Product, DMA DMA DMA DMA 5, 914 3, 126 5, 914 0.834 0.907 0.835 Notes: Unit of observation is the market-month-product combination. Data (source: Truven) span 17 months from July 2007 to November 2008. Ptm is an indicator for any political ads in market m in month t. It is an indicator for the regulatory action months that prevented Lipitor from advertising. Statistical significant at the 10%, 5%, and 1% levels are denoted by *, **, and *** respectively. Table 5: OLS Revenue Regressions for Advertised Drugs Own Ads Rival Ads Dependent Variable: Log(Revenue per Insured), Two Month Trailing Average 0.0239∗∗∗ 0.0592∗∗∗ 0.0067∗∗∗ (0.0008) (0.0015) (0.0007) ∗ ∗∗∗ 0.0017 −0.0268 0.0037∗∗∗ (0.0009) (0.0046) (0.0008) Controls: Market FEs Drug FEs Year FEs Drug-Year FEs Drug*Reg. Action FE Drug FE*Time Trend Drug FE*Year-Month FE Clustering N R2 X X X X X X X X X 0.3168∗∗∗ (0.0835) −0.1792∗∗∗ (0.0589) X X X X DMA 11, 465 0.843 DMA 11, 465 0.844 DMA 11, 465 0.847 X DMA 11,465 0.852 Notes: Unit of observation is the market-month-product combination. Data (source: Truven) span 17 months from July 2007 to November 2008 and include Crestor, Lipitor, Vytorin, and Zetia. “Own Ads” and “Rival Ads” are constructed as log(1+X). “Two Month Trailing Average” indicates that the independent variables are constructed as the average of advertising during the revenue month and the month before. Number of observations is smaller than Table 3 because these specifications drop DMAs with no local advertising. This primarily affects small DMAs. “Reg. Action” refers to an indicators for before, during, and after the months in which Lipitor was prevented from advertising. Statistical significant at the 10%, 5%, and 1% levels are denoted by *, **, and *** respectively. Standard errors are clustered at the market level. Table 6: IV Revenue Regressions for Advertised Drugs Dependent Variable: Log(Revenue per Insured), Two Month Trailing Average Panel A: Second Stage Estimates Own Ads 0.1470∗∗∗ 0.1395∗∗ 0.0761∗∗∗ (0.0107) (0.0557) (0.0216) ∗∗∗ ∗∗ −0.1140 −0.1155 −0.0547∗∗∗ Rival Ads (0.0086) (0.0594) (0.0180) Panel B: First Stage Estimates (Excluded Instruments) Own Rival Own Rival Own Rival ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ Pol −0.2099 0.0463 −0.2602 −0.1573 0.1254 0.1427∗∗∗ (0.0465) (0.0198) (0.0530) (0.0215) (0.0447) (0.0256) 2 ∗∗∗ ∗∗∗ ∗∗∗ Pol 0.0218 0.0002 0.0248 0.0108 −0.0079 −0.0083∗ (0.0070) (0.0037) (0.0060) (0.0023) (0.0062) (0.0044) ∗∗ ∗∗∗ ∗∗∗ 3 −0.0006 −0.0001 −0.0007 −0.0002 0.0001 0.0001 Pol (0.0003) (0.0001) (0.0002) (0.0001) (0.0002) (0.0002) Reg. Action −1.6043∗∗∗ −1.3562∗∗∗ −1.3412∗∗∗ −1.2869∗∗∗ (0.0966) (0.0705) (0.0977) (0.0735) ∗∗∗ ∗∗∗ Reg. Action· −0.0979 −0.7648 0.3948 0.0103 −0.0127 −0.7374∗∗∗ .....Pol (0.1049) (0.0813) (0.0625) (0.0390) (0.1167) (0.0895) ∗ ∗∗∗ ∗∗∗ ∗∗∗ 0.0749 0.2276 −0.0449 0.0353 −0.0248 0.1977∗∗∗ Reg. Action· 2 (0.0388) (0.0328) (0.0122) (0.0109) (0.0465) (0.0365) .....Pol ∗ ∗∗∗ ∗∗∗ ∗∗ Reg. Action· −0.0065 −0.0170 0.0019 −0.0017 0.0023 −0.0143∗∗∗ .....Pol 3 (0.0035) (0.0030) (0.0008) (0.0007) (0.0043) (0.0033) Fixed Effects Product-Year Product-Reg.Action Product-Time Trend Partial F-Stat 1455.9 2979.3 51.6 28.9 722.9 1679.0 Notes: Unit of observation is the market-month-product combination. Number of observations is 11,465 in all specifications. Data (source: Truven) span 17 months from July 2007 to November 2008 and include Crestor, Lipitor, Vytorin, and Zetia. “Own Ads” and “Rival Ads” are constructed as log(1+X). All specifications are a “Two Month Trailing Average,” which indicates that the independent variables are constructed as the average of advertising during the revenue month and the month before. “Pol” is the number of political ads in a market-month, in thousands. “Regulatory Action” is a dummy variable for April-August 2008, when congressional action forced Lipitor to stop advertising. F-statistics are for excluded instruments only. Statistical significant at the 10%, 5%, and 1% levels are denoted by *, **, and *** respectively. Standard errors are clustered at the market level. All specifications include market, year, and drug fixed effects. Table 7: Timing Assumption Sensitivity Advertising Timing: Own Ads Rival Ads Controls: Market FEs Year FEs Drug FEs Drug FE*Time Trend Clustering N R2 Dependent Variable: Log(Revenue per Insured) This Month Two Month Three Month One Month Lag 0.0961∗∗∗ 0.0761∗∗∗ 0.0568∗∗∗ 0.0613∗∗ (0.0331) (0.0216) (0.0165) (0.0276) ∗∗ ∗∗∗ ∗∗∗ −0.0608 −0.0547 −0.0539 −0.0392∗ (0.0243) (0.0180) (0.0196) (0.0233) X X X X DMA 11, 466 0.807 X X X X DMA 11, 465 0.824 X X X X DMA 11, 465 0.843 X X X X DMA 11, 465 0.832 Notes: Unit of observation is the market-month-product combination. Data (source: Truven) span 17 months from July 2007 to November 2008 and include Crestor, Lipitor, Vytorin, and Zetia. “Own Ads” and “Rival Ads” are constructed as log(1+X). “This Month” indicates contemporaneous advertising, while longer time spans indicates that the independent variables are constructed as the average of advertising during the revenue month and the months before. First stage excluded instruments are political advertising, its square and cube, a dummy that takes on a one during April 2008-August 2008 (regulatory action dummy), and the interactions of the political variables and the regulatory action dummy. Statistical significant at the 10%, 5%, and 1% levels are denoted by *, **, and *** respectively. Standard errors are clustered at the market level. Table 8: Spillovers for Non-Advertised Drugs Dependent Variable: Model: Rival Ads Controls: Market FEs Time Trends Clustering N R2 Log(Revenue per Insured), Non-Advertised Drugs OLS IV 0.0018 0.0188∗∗ (0.0021) (0.0093) X X DMA 3, 112 0.880 X X DMA 3, 112 0.883 Notes: Unit of observation is the market-month combination. Data (source: Truven) span 17 months from July 2007 to November 2008 and include all non-advertised drugs. “Rival Ads” is constructed as log(1+X). All specifications have a “Two Month Trailing Average” as the independent variable, indicating that the “Rival Ads” is constructed as the average of advertising during the revenue month and the month before. First stage excluded instruments are political advertising, its square and cube. Statistical significant at the 10%, 5%, and 1% levels are denoted by *, **, and *** respectively. Table 9: IV Revenue Regressions for Advertised Drugs, Part D Data Model: Own Ads Rival Ads Controls: Market FEs Year FEs Drug FEs Drug FE*Time Trend Clustering N R2 All Prescriptions OLS IV 0.00835∗∗∗ 0.0610∗∗∗ (0.000709) (0.0180) 0.00210∗∗ −0.0369∗∗∗ (0.000830) (0.0132) X X X X DMA 11, 466 0.878 X X X X DMA 11, 466 0.863 New Prescriptions OLS IV 0.0476∗∗∗ 0.349∗∗∗ (0.00373) (0.0728) 0.0516∗∗∗ −0.200∗∗∗ (0.00549) (0.0630) X X X X DMA 10, 743 0.595 X X X X DMA 10, 743 0.325 Notes: Data created by restricting Medicare Part D event data to either all Crestor, Lipitor, Vytorin, and Zetia fills or the first prescriptions by beneficiary of those drugs and collapsing to the market-month-product level. Unit of observation is the market-month-product combination. Data span 17 months from July 2007 to November 2008. “Own Ads” and “Rival Ads” are constructed as log(1+X). All dependent variables are the “Two Month Trailing Average.” First stage excluded instruments are political advertising, its square and cube, a dummy that takes on a one during April 2008-August 2008 (regulatory action dummy), and the interactions of the political variables and the regulatory action dummy. Statistical significant at the 10%, 5%, and 1% levels are denoted by *, **, and *** respectively. In all specifications, we dropped markets with less than 30 days supply of new prescriptions, which reduces the sample size for new prescriptions. Table 10: Revenue Simulations Panel A: Simulations for Advertised Drugs % Change in Revenue: Crestor Lipitor (1) Eliminating Business-Stealing 0.2108 0.2350 Confidence Interval (0.0046, 0.4171) (0.0081, 0.4619) (2) Eliminate Political Ads 0.0166 0.0131 Confidence Interval (−0.0035, 0.0368) (−0.0019, 0.0282) Panel B: Simulations for Non-Advertised Drugs % Change in Revenue: Unadvertised Drugs (3) Eliminate Spillovers −0.0808 Confidence Interval (−0.1362, −0.0254) Notes: Estimates from “Two Month Trailing Average” and drug-specific time trend specifications are used in all simulations. Simulation (1) sets the coefficient on rival advertising in column 3 of Panel A of Table 6 equal to zero. Simulation (2) estimates the number of drug ads in the absence of political ads, and then estimates sales at those levels of advertising. In Panel B, (3) sets the coefficient on rival advertising in column 2 of Table 8 equal to zero. Estimates are bootstrapped by re-sampling the data set, re-estimating the primary specifications, and re-computing the counterfactual exercise on the observed data. We use 100 bootstrap replications and report the 2.5%-97.5% confidence interval as well as then mean. Table 11: Quantity Simulations % Change in Quantity Ban All Advertising Confidence Interval Crestor Lipitor Unadvertised −0.0007 −0.0197 −0.0356 (−0.0301, 0.0360) (−0.0540, 0.0104) (−0.0917, 0.0228) Notes: Estimates from “Two Month Trailing Average” and drug-specific time trend specifications are used in all simulations. The dependent variable is the log of the market share of a product. The simulation sets the coefficient on own and rival advertising equal to zero. Estimates are bootstrapped by re-sampling the data set, re-estimating the primary specifications, and re-computing the counterfactual exercise on the observed data. We use 100 bootstrap replications and report the 2.5%-97.5% confidence interval as well as then mean. Figure 7 shows the distributions of the simulated outcomes.