Update and Revision History
This page compiles updates and revisions to the data as of February 15, 2019.
Errata: On February 15, 2019, the data product Agricultural Research Funding in the Public and Private Sectors was reposted to correct an issue with how the data series Total private food and agricultural research was adjusted to avoid double counting in combining private agricultural input research data with private food research data. In the previous version, the adjustment for double counting in computing the total was based on a slightly older estimate of agricultural input research; the component data for private agricultural input research and private food research were correct and current (and are unchanged). The changes in the series Total private food and agricultural research for any given year range from -1.4 percent to 1.4 percent, with a mean change over the entire period (1970-2014) of 0.01 percent.
Since the last update of the data Agricultural Research Funding in the Public and Private Sectors (March 26, 2012), the following changes have been made.
Public Sector Research Funding Series:
- In previous versions of the data, most USDA research estimates were made from National Science Foundation (NSF) information, with USDA, National Institute of Food and Agriculture (NIFA)-administered funding estimated using Current Research Information System (CRIS). In recent years, not all NIFA-administered funds are readily available from the CRIS source. The February 23, 2018, release thus also estimates NIFA-administered funds using NSF data. The current method for estimating NIFA funding was applied to earlier years, to make estimates comparable over time.
- NSF data for USDA agencies other than NIFA have now been adjusted to estimate intramural research only. This eliminates conceptual double counting of non-NIFA funding to State-level institutions from other USDA agencies. This funding is also captured in the State-level estimates based on CRIS.
- NIFA funding has been adjusted to eliminate the relatively small funding amounts for research in the Small Business Innovation Research (SBIR) program. Although administered by NIFA, these funds go to small private firms.
- State-level research expenditures have been reduced by $300 million in each of the years 2006, 2007, and 2008. This adjustment has been made to address overreporting by a single institution, the University of Minnesota Veterinary School. Some of the features of this overreporting have been documented by Alston et al. (2010), Persistence Pays: U.S. Agricultural Productivity Growth and the Benefits from Public R&D Spending, pp. 231-232.
The largest differences in the public sector series between previous versions and the current update are related to the "Minnesota correction" in 2006-08. Actual differences for 2008 are smaller than the differences for 2006 and 2007, as the previous version relied on preliminary NSF estimates for 2008 and 2009, which have since been revised.
Private Sector Research Funding Series:
- Private sector research is now identified by two components, private R&D in agricultural inputs and private R&D in food manufacturing. Private R&D in agricultural inputs is more likely to have impacts on agricultural productivity. Except in some vertically integrated sectors like poultry, food R&D has limited relevance for production agriculture. This distinction allows a better understanding of recent trends in private R&D. However, given available data sources, one component of agricultural inputs research—animal nutrition—is also part of the sampling universe for NSF estimates of food R&D. Thus, the total private R&D estimate is slightly less than the sum of agricultural inputs and food manufacturing research, as animal nutrition research has been counted only once in the total. See Documentation for details.
- Some estimates of private research investment for particular firms have been revised as more recent information has become available, and this changes estimated totals somewhat, compared to the previous estimates.
The R&D deflator has changed slightly, primarily due to minor revisions in the Bureau of Economic Analysis (BEA) deflators for two of the three components used to construct this deflator.