Economics - General Pharmaceuticals

Why R&D expenses rising and productivity falling could be a good thing

A few facts:

  • Pharma is increasingly a key driver of US R&D. Pharmaceutical firms share of R&D increased from <10% in the early 1970s to 24% in 2018 (among firms in Compustat data). Moreover, 5 out of top 10 largest R&D spenders in Compustat in 2010 were pharmaceutical firms.
  • A majority of drugs are targeted at the elderly. This is not surprising as the elderly are often have multiple comorbidities requiring medication. Between 1995 and 2013, drugs targeting indications with a majority of elderly individuals represented over half new drug development entry for preclinical projects.
  • In the US economy at large, R&D spending is rising, but productivity is not.

Based on these facts, an NBER working paper by Benmelechet al. (2021) aims to solve the conundrum posed by fact #3. The authors write:

…we have shown that a significant share of intangible investment, specifically R&D expenditure, is geared towards treating medical conditions afflicting the elderly. To the extent that these seniors remain out of the labor force, their increased quality of life and life expectancy will not directly increase labor supply and output (via a supply mechanism). Hence, output metrics may not benefit from this type of R&D investment.

While these R&D investments may not increase the productivity of the economy, these investments are likely still well worth the money. As the authors conclude in their abstract:

Increased life expectancy and quality of life among the elderly increases welfare but may not be reflected in estimates of total factor productivity

Methods detail:

The first fact listed above the authors arrive at through Compustat data. The second fact the authors get by using Cortellis Investigational Drugs and Medical Expenditure Panel Survey (MEPS) data. The Cortellis data is used to identify the drug pipeline and relevant indications. Then the authors map the indications to ICD-9 diagnosis codes and then map ICD9 to Clinical Classification Codes (CCC) available in MEPS data. The authors then classify a drug as elderly if the majority of patients in the CCC in MEPS were aged 65 years or older.


Leave a Reply

Your email address will not be published. Required fields are marked *