No R&D internal link—no link between patenting and R&D

Filippo, Northwestern University Kellogg School of Management; Timothy, Boston University Questrom School of Business and NBER. “Patent policy and American innovation after eBay: An empirical examination.” Volume 48, Issue 5, June 2019, Pages 1271-1281. LJS

Fig. 3 plots the NSF estimates of total R&D spending from 1990¶ through 2013 along with the cumulative R&D reported by U.S. public¶ companies in the Compustat database from 1999 through 2013.24 There¶ are three main take-aways. First, Compustat captures the majority of¶ U.S. business R&D spending– 87 percent, on average, for years in this¶ figure. Second, both time-series exhibit a sharp drop in R&D spending¶ in 2009, coinciding with the Great Recession. This drop is larger in the¶ Compustat data, but the recovery begins sooner and is more dramatic¶ for that series. Finally, there is no sudden change in the R&D investment¶ growth trend around 2006. Overall spending accelerated during the¶ boom years between 2006 and 2009 and fell during the recession. Even¶ after the end of the recession, we see R&D spending going back to the¶ same trend it was experiencing before the eBay decision. Overall, the R&D time-series data provide no evidence that changes in patent-law led¶ to any dramatic changes in R&D spending.¶ With the Compustat data, we can also examine industry-level trends¶ in R&D investment before and after eBay. Fig. 4 uses S&P industry¶ sector codes to group companies into three broad sectors, as above,¶ corresponding to Computers, Life Sciences and Other industries.25 Interestingly, for the public companies in these data, between 1999 and¶ 2013 total R&D spending grew faster outside of the Computer and Life¶ Science sectors. Computer-related R&D fell in the years following the¶ Internet bubble, and then grew steadily. For this industry, the progress¶ was more-or-less parallel to growth in Life Sciences R&D, although the¶ latter sector experienced the recession more sharply. To us, the broad¶ industry-level R&D trends suggest that large changes in aggregate R&D¶ spending are driven by changes in “demand” (specifically, the macro¶ economic environment) and shocks to financial conditions, rather than¶ changes in patent policy that occurred in 2006.¶ As a last effort to find a connection between changes in patent law¶ and R&D spending, we estimate a series of firm-level difference-in¶ difference regressions using the Compustat data. Our basic specification¶ will be¶ ln(RDit) = αi + λt + β1 PostEbayICTit + β2 ln(Assetsit) + εit¶ where the outcome (RDit) is annual R&D expenditures; αi are a set of¶ firm fixed effects that capture time-invariant unobserved factors that¶ drive R&D spending; λt are a set of year effects that capture changes in¶ the overall economic environment, including (importantly) the great¶ recession; and the main explanatory variable is Post Ebay ICTit, an indicator that equals one for firms in the ICT sector after 2006. By taking¶ the natural log of the outcome, we can interpret β1 as the percentage¶ change in computer-related R&D relative to the “control” industries¶ following the eBay decision. Our sample includes all firms that report R&D expenditures for every year they appear in the Compustat data, and¶ we include the log of total assets as a time-varying control for firm size.26¶ The basic idea behind this regression is that if changes in patent-law¶ (including but not limited to the eBay decision) disproportionately¶ impact the ICT sector, and if there is a strong link between patent¶ strength and R&D investment, then we should observe a decline in R&D¶ spending for ICT firms relative to firms in other sectors. That is, a strong¶ link between patents and R&D would lead to estimates of β1 less than¶ zero.¶ The assumption that the eBay decision had a disproportionate impact on ICT firms is consistent with Seaman’s (2016) findings on post¶ eBay patent injunctions in software and pharmaceuticals, as described¶ above. Prior research also shows that patent assertion by Non¶ Practicing Entities (NPEs) is concentrated in the ICT sector, and that¶ NPEs are unlikely to obtain injunctive relief (Allison et al., 2018). Finally, because ICT patents often cover a particular component within a¶ larger system, whereas patents in other industries (e.g. pharmaceuticals) may be more “fundamental” to the final product, we expect the¶ balance of hardships factor to favor injunctive relief outside the ICT¶ sector.¶ While we believe that the assumption of disproportionate impact on¶ ICT is a reasonable one, and consistent with most of the discussion on¶ the law change, different experts may disagree on the definition of a¶ control group. For instance, one could argue that some areas of Life¶ Science were also strongly affected by the law change. Because our¶ Compustat data end in 2013, the impact of the Mayo and Myriad decisions on the analysis should be minimal. Nevertheless, to guard¶ against this problem, we will conduct our analysis using Life Sciences¶ and “Other” industries as separate control groups. Our analyses will¶ therefore use three different control samples, where in each case the¶ treatment is defined as the ICT-sector and the control group changes¶ column-by-column. In the first column we compare ICT to all non-ICT¶ firms. In the second column the control group contains only life-science¶ companies, and in the third column the control group contains only¶ companies that are in neither ICT nor life-science industries. Comparing¶ the coefficients across the different specifications should assuage concerns that the choice of control group is driving the results.¶ It is important to note that we do not interpret these estimates as the¶ causal effect of eBay on R&D investment. In particular, our analysis may¶ also partially capture unobserved changes in the business environment¶ that interact with the effect of the new rules. However, if eBay led to a¶ significant decline in the expected returns to U.S. R&D investment, we¶ should be able to detect some changes across the different groups in our¶ model.¶ We estimate these models in two different specifications: ordinary¶ least squares regression with a logged outcome, and an exponential (or¶ Poisson) model of R&D in levels. The results of our regression analysis¶ are presented in Table 1.¶ Across all of the regression models, we can never reject the null¶ hypothesis that there was no change in ICT-sector R&D spending re¶ lative to other industries after 2006. Compared to all other firms, post¶ 2006 R&D investment in the computer sector declined by 2%, with a¶ standard error of 3%. Comparing estimates in the second and third¶ columns, it appears that the small relative decline in computer-related¶ R&D is driven by firms outside the life sciences sector, as suggested by¶ Fig. 4. Estimates based on the Poisson specification are very similar, but¶ exhibit less precision. In the Appendix, we show that the results in¶ Table 1 are robust to small changes in the sample and specification.27¶ Overall, the results in Table 1 imply that if one wishes to maintain that¶ eBay had a large negative impact on U.S. R&D investment incentives, it¶ is important to think about what else happened around the same time¶ period to increase R&D incentives by a similar amount.¶ Before moving on, it is interesting to compare the coefficient on the¶ Post Ebay ICT “policy” variable in Table 1 to the coefficient on log(As¶ sets). Because these regressions include firm fixed-effects, the latter¶ coefficient measures the association between firm-growth and R&D. It is¶ well known that growing firms increase their R&D spending, and our¶ estimates suggest that after controlling for time-invariant factors, a 10¶ percent increase in assets is associated with a 5 percent increase in R&D¶ spending. What is notable about these estimates is that they are large¶ and precise. Comparing the magnitude and precision of the two coefficients in Table 1 suggest that even hotly debated changes in patent¶ policy can have a minor impact on individual firms’ R&D investment¶ decisions compared to other firm-level factors that drive growth.