If you look around a bit on the NIH funding data at RePORT, you will find the following definitions.
Research Project Grants: Defined as R00, R01, R03, R15, R21, R22, R23, R29, R33, R34, R35, R36, R37, R55, R56, RC1, RC2, RC3, RC4, RF1, RL1, RL2, RL5, RL9, P01, P42, PN1, UA5, UC1, UC2, UC3, UC4, UC7, UF1, UH2, UH3, UH5, UM1, U01, U19, U34, DP1, DP2, DP3, DP4, and DP5 . Research projects were first coded to NLM in fiscal year 2007.
R01-Equivalent Grants: Defined as activity codes R01, R29 and R37.
The R29 was the FIRST award program and the R37 is MERIT, generally an extension of the noncompeting interval for a continuation R01 that scored really well. So...basically these are all R01s.
A post from Steven Salzberg begs to "Please save the unsolicited R01s" which includes this graph sourced from FASEB.
Making the same leap of considering these the "real" investigator initiated awards, we can see that the number of new awards in the past two Fiscal Years is lower than it has been since 95-96, *prior* to the doubling.
Everytime the NIH officialdom chooses to respond to criticism and concern about how their latest initiative will hurt the traditional strength (investigator initiated R01 equivalents) they try to claim that these are not paying the price. In various ways and with various incomplete analyses they try to give the impression that despite the invention of RC this and DP that, the failure to dismantle boondoggle Ps and the increased use of U-mechs...that the R01 remains sacred.
This graph gives you a retort.