Wow. I last used the BRDPI estimate of inflation in the cost of biomedical research to illustrate how the full modular grant ($250K direct) had not changed and therefore purchasing power had eroded.

Jeff Mervis at Science has a blockbuster observation.

To remind you, the BRDPI is this:

The annual change in the Biomedical Research and Development Price Index (BRDPI) indicates how much the NIH budget must change to maintain purchasing power. The BRDPI was developed and is updated annually by the Bureau of Economic Analysis (BEA), Department of Commerce under an interagency agreement with the NIH.

That link also leads you to the data tables where you find the annual rates stretching back to the 50s.

The Mervis article highlights the historical low for FY2012 and reminds us of the cut in the salary cap (maximum amount of an Investigator’s salary that can be charged to NIH grants)

Congress passed a spending bill in December 2011 that lowered the salary ceiling for investigators on a standard NIH grant from $199,700 to $179,700.

and concludes with a caution:

NIH enjoys strong support in Congress, and the realization that biomedical inflation largely tracks salary trends, not the sticker price of essential lab equipment and supplies, is unlikely to have a major impact on policy debates. Still, it may behoove biomedical lobbyists to think twice before citing the cost of high-tech science as a rationale for pumping up NIH’s budget.

Yeah, I hear that. Let’s peer a little closer though.

The NIH Office of the Budget January 2015 overview [PDF] futher anticipates that salary cap is a major driver of the inflation index.

The modest BRDPI growth rate of 2.0 percent for FY 2014 reflects the effect of the NIH extramural investigator salary limitation (“cap”) of $181,500 and an increase on salaries of Federal civilian employees of 0.75 percent for that fiscal year.
The projected 2.2 percent growth for FY 2015 assumes a one percent increase for Federal salaries starting in January 2015, as well as an increase on the extramural investigator salary cap to $183,300.

OK, let’s route ourselves back to the NIH Office of the Budget report from January 2013 [PDF] which indeed draws an explicit link.

The modest BRDPI growth rate of 1.4 percent for FY 2012 reflects the effect of the reduction of the NIH extramural investigator salary limitation (“cap”) from $199,700 to $179,700 for that year and the continued freeze on salaries of Federal civilian employees.

However, it also goes on to note other contributions:

The BRDPI growth rate was also adjusted for the growth of stipends and related expenses on fellowships and training awards. In addition, the FY 2012 BRDPI growth is lower than the growth for FY 2011 because the rate of growth of prices for several input categories slowed down in FY 2012 compared with the growth during FY 2011. For inside NIH activities, the categories with slower growth in prices include travel, transportation, printing and reproduction, ADP and other IT services, instruments and apparatus, laboratory supplies, office supplies, utilities, repairs and alterations of facilities, compensation rates for consultants and support contracts. For extramural activities, fringe benefits, travel, equipment, supplies, patient care alterations and indirect costs each showed slower price growth during FY 2012 compared with FY 2011.

Wait. Sooooo, everything contributes a little bit? This seems out of step with Mervis’ column. Wait, wait….the 2013 overview continues….

Primarily because of the freeze on Federal civilian employee salaries and the cap on compensation of extramural investigators, the rate of growth of the BRDPI during the years FY 2011 through FY 2013 has been relatively low compared with its historical relationship with general inflation as represented by the growth of the GDP Price Index.

Primarily. So that circles us right back to the reduction in the extramural cap and elimination of Federal civilian salary raises. [You might ask why the Federal civilians did not also suffer salary reductions, merely freezes, eh?] But if we take this as a valid and intended connotation then it would seem Mervis has it right. The salary issues are huge.

I wonder why they didn’t just find the dollar figure. How many Investigators funded by the NIH in a given year are paying up to the cap? Multiply that by you favored reduction or increase and boom, you can translate that into new R01s.

The annual Salary Cap numbers can be found here. It seemed to steadily increase from 2005-2010, including that 2008-2009 interval that produced the most immediately prior reduction in the BRDPI. So why didn’t salary cap drive the BRDPI that year?