Changes
Jump to navigation
Jump to search
Noel Schankerman (2006) - Strategic Patenting And Software Innovation (view source)
Revision as of 18:48, 1 April 2013
, 18:48, 1 April 2013no edit summary
*This page is referenced in the [[Patent Thicket Literature Review]]
*This page is listed on the [[PTLR Core Papers]] page
Strategic patenting is widely believed to raise the costs of innovating, especially in industries characterised by cumulative innovation. This paper studies the effects of strategic patenting on R&D, patenting and market value in the computer software industry. We focus on two key aspects: patent portfolio size which affects bargaining power in patent disputes, and the fragmentation of patent rights (.patent thickets.) which increases the transaction costs of enforcement. We develop a model that incorporates both effects, together with R&D spillovers. Using panel data for the period 1980-99, we find evidence that both strategic patenting and R&D spillovers strongly affect innovation and market value of software firms.
==Review==
===Measures of Patent Thicket===
Patent thickets are measured by a concentration index that sums the shares of a firm's backward patent citations to the top 4 most-cited assignee firms in a given year.
*An 8-firm measure and Herfindahl index of concentration were also calculated to test robustness of results.
===Sample===
Panel data on 121 publicly traded U.S. firms in the software industry with at least 2 consecutive years of data from 1980-1999.
*Compustat is source of firm financial information, including R&D.
*USPTO is the source of information on the 29,363 patents held by firms in the sample (of which 12,507 are "software" patents), including backward citations, patent classifications and assignees.
*Firms are included when at least 45% of their patents are in Patent Office in International Patent Classifcation G06F "Electrical Digital Data Processing".
**About two-thirds of the firms are in SIC code 7372 "Prepackaged software".
===Results===
Market value:
*Market value is higher for firms whose patents are concentrated among fewer assignees [a less dense patent thicket] (5% point increase in the firm's citations shares held by the top-4-firms raises market value by 1.7%), but lower for firms in industries with a higher patenting propensity (10% increase in rivals' patenting propensity reduces firm value by 1.3%)
:''"These two findings strongly support the model's predictions about strategic patenting - there is evidence both that patent portfolio size (bargaining power) and transaction costs associated with the fragmentation of property rights affect the market value of [the] firm."''
*In the post-1994 period when U.S. courts became more pro-patent, the effects of patent concentration (less dense patent thickets) and rival firms' patenting propensity are highly significant, but insignificant in the earlier period.
:''"In summary, we conclude that the change in patent regime was associated with a sharp increase in the importance (as measured by the coeficients) of the strategic patenting variables. At the same time, despite a large increase in the level of patenting during this later period, we do not
find a sharp reduction in the impact of patents on market value. Evidently, whatever diminishing returns that was associated with the intensification of software patenting appears to have been largely countervailed by the increased value from the strengthening of software patent protection."'
*A 10% increase in technology spillovers is associated with a 1.7% increase in firm market value.
Patent stocks:
*Firms do less patenting, conditional on R&D, when they face technology rivals with higher patent propensities, (a 10% increase in patenting propensity is associated with a 4.5% reduction in patenting by a firm).
*Greater concentration of citations (less dense patent thicket) is associated with a reduction in firm patenting (a 5% increase in citations concentration reduces patenting by 12.8%) and the effect is unchanged between pre- and post-1994 periods when courts decisions became more pro-patent.
*A 10% increase in technology spillover, increases patenting bny 6.4%.
:''"Overall, the empirical results support the hypothesis that both technology spillovers and strategic patenting variables affect the decision to patent."
R&D:
*R&D is not significantly related to patenting propensity of rivals.
*R&D is significantly reduced by citation concentration (less dense patent thickets); in models with firm fixed effects, a 5% increase in citations concentrations is associated with a reduction in R&D by 1.4% (though no longer significant).
:''"[T]here is a smaller gain from having a larger patent portfolio when patent rights are more concentrated among rival firms. This finding is consistent with our expectations, since tacit forms of cooperation are more likely to develop in such cases and these make large patent portfolios less important as threats to resolve disputes."''
*Coefficients on time dummies show no significant differences, therefore:
:''"This finding suggests that the expansion of patentability over software during the 1980s and early 1990s was not associated with any major changes in R&D investment by these software firms as of the end of our sample period."
===Social Welfare Consequences===
:''"[W]e find that patenting by technology rivals reduces the firm's R&D investment, patenting and market value. Third, greater concentration (less frag-mentation) of patent rights among rivals reduces both R&D and patenting by the firm reflecting less need to have an arsenal of patents to resolve disputes when there are fewer players, but it increases market value because transaction costs are lower."
===Dependent Variable and Model===
Three dependent variables are analyzed:
*Tobin's Q (the market value of the firm divided by tangible assets) is analyzed using an OLS model with Newey-West standard errors robust to heteroskedasticity and first-order serial correlation.
*The firm's patent stock is analyzed using a negative binomial model.
*The firm's R&D investment is analyzed using an OLS model with Newey-West standard errors robust to heteroskedasticity and first-order serial correlation.
For all dependent variables the models account for:
*Patent thickets (lagged), measured by a concentration index that sums the shares of the firm's backward patent citations to the top 4 most cited firms in a given year;
*R&D spillover (lagged), measured by a weighted sum of R&D stock of other firms(depreciated at 15%), where weights are the technological proximity of other firms to the firm being observed (based on the uncentered correlation coefficient between the observed firm's backward citations to other firms' patents across patent classes);
*The patenting propensity of other firms (lagged), measured by a weighted sum of other firms' patent stocks divided by their R&D stock, where weights are the technology proximity of the observed firm to these other firms (based on an uncentered correlation coefficient between the observed firm's backward citations to other firms' patents across patent classes);
*Technological opportunity (two lags), measured by total patenting in a technology class weighted by the firm's closeness to each technology class reflected by backward citations;
*Indicators for the firm's industry (4-digit SIC) and year.
The market value equation additionally controls for:
*The firm's stock of R&D relative to tangible assets (lagged) with 4 higher order terms;
*The firm's patent stock over tangible assets (lagged) with 4 higher order terms;
*The log of firm sales (two lags).
The patenting stock equation additionally controls for:
*The log of the firm's R&D stock (lagged);
*Log of pre-sample patents;
*Overdispersion alpha to account for variance dispersion that distinguishes the negative binomial from the Poisson model.
The model of the firm's R&D investmet additionally controls for:
*The log of the firm's R&D (lagged);
*The log of firm sales (two lags).
