Spillovers and Other Externalities in Pharmaceutical Marketing - Innovation and Marketing in the Pharmaceutical Industry - page 680

Table 23.3 Regression results

Dependent variable
USA Canada

Log(category
Rx total)

Log(category
Rx totals)

Independent variable Model 1 Model 2

Intercept 14.259214*** 12.070408***

Quarterly sales dummy 1 (SD1) 0.045223** 0.045618*

Quarterly sales dummy 2 (SD2) −0.108323*** −0.018122

Quarterly sales dummy 3 (SD3) −0.085629*** −0.047146*

Share-weighted country-specific price −0.001767 −0.003712

Goodwill from country-specific detailing 0.000017*** 0.000054***

Goodwill from country-specific OME 0.000058** −0.000089

Goodwill from combined US DTCA
(TV + magazines)

0.000034*** 0.000029***

R -square 0.9040 0.8520

N 66 66

* p < 0.10; ** p < 0.05; *** p < 0.01

23.3.3 Results and Discussion

We begin with the results for the parameters of the category demand function.
Subsequently, short-run and long-run elasticities for detailing and DTCA, and single-
and multi-period return-on-investment (ROI) measures for detailing and DTCA are
discussed, in that order.

23.3.3.1 Parameters of the Category Level Demand

Table 23.3 shows the SURE parameter estimates for the category sales models
(we estimate ( 23.1 ) for each of the countries). We refer to the US results as Model
1 and the corresponding results for Canada as Model 2. The models fit the data well,
as judged by the high R 2 values. We find that after controlling for the seasonal
effects, the goodwill measures associated with detailing and DTCA have statisti-
cally significant coefficients and the correct (positive) sign for both countries.
The significant DTCA terms in the Canada regression (Model 2) showcases the
spillover effect. Price is not statistically significant, but has the correct (negative)
sign. OME has the right sign and is statistically significant in the USA. However, it
does not have a statistically significant effect in Canada. At least two of the three
seasonal dummies have significant effects, reflecting seasonality in the category.

23.3.3.2 Short-Run and Long-Run Elasticities

We report the short and long-run elasticities for DTCA and detailing in Table 23.4 (the
long-run elasticity is computed as in previous studies; see e.g., Narayanan et al. 2004 ).