This is the title of a new USC white paper by Darren Films, Karen Van Nuies, Daryas Lakdavalla and Dana Goldman, “How much does the revenue increase the new drug development?”
What is the elasticity of innovation?
It measures a percentage change in innovation – using the flow of approval of new drugs, or steps 1, 2, or 3 – due to the percentage change in revenue, is usually expected.
Future revenue.
In practice, it is a change in profits that matters, but future revenue is much more observable and estimated than future profits. Thus, the author focuses on the elasticity of innovation in relation to revenue rather than profits.
How much does revenue in future affect the possibility of new drug development?
It has been concluded from all studies that the elasticity is positive – IE, low revenue leads to low R&D – but estimates vary widely. However, we argue that a specific long -lasting elasticity associated with American revenue is within a range of 0.25 to 1.5, which means that for every 10% decrease in the expected revenue, we can expect 2.5% to 15% less drug innovation.
What is variability in these estimates?
An important question is why these estimates have such a big range? Certainly various study design cases (see below). Authors also claim that factors such as “Time Horizon’s study, price change size, cost of drug growth, obstruction for price-based pricing, and other market factors” affect all innovation’s magnitude.
Which method is used in literature to estimate the elasticity of innovation?
- Cross sectional: Exploitation of revenue variation in medical classes (or some other units of analysis) to estimate elasticity. For example, they can compare the “high-revenue” vs “low-revenue” classes to estimate elasticity [Examples: Lichtenberg (2005) and Civan and Maloney (2009)],
- Overall time chain: Exploitation of variation in industry-level revenue over time [Example: Giaccotto, Santerre and Vernon (2005)]
- Panel data approach: Include frequent differences in drug-class “fixed effects” and hard-to-map and class characteristics. In short, this approach approach focuses on inner-class revenue change, which is the driver of inner-class innovation changes. These analyzes usually require the use of “natural experiments” that causes a change in revenue in various areas of the market. Examples of natural experiments include future demographic changes or arrival of Medicare Part D. [Examples: Acemoglu and Linn (2004); Dubois et al. (2015); Blume-Kohout and Sood (2013)]
- Parameter computational model (aka structural model): The purpose of the firms specify the characteristics of the tasks, strategy sets and business environment, and when many firms are included in the model, the model usually requires that market balance. The parameters are chosen to match the real world (eg, average R&D expenses) and in such a way that models outputs also match the actual world consequences (eg, average flow of new drugs). [Examples: Abbott and Vernon (2007); Filson (2012); Adams (2021)]
Authors argue that panel approaches and parameter computational models are preferred.
To study with a favorite panel or computation approach, in which individual elasticity of innovation estimates arrived?
Authors have a good table that briefly presents the conclusions that I glued down.
Great work by my colleagues in USC! I definitely encourage you to read the full article.
