Ziedonis (2004) - Dont Fence Me In

• This page is referenced in BPP Field Exam Papers

Reference

• Ziedonis, R.H. (2004), "Don't fence me in: Fragmented markets for technology and the patent acquisition strategies of firms", Management Science. pdf

Abstract

How do firms avoid being "fenced in" by owners of patented technologies used, perhaps unknowingly, in the design or manufacture of their products? This paper examines the conditions under which firms expand their own portfolios of patents in response to potential hold-up problems in markets for technology. Combining insights from transactions cost theory with recent scholarship on intellectual property and its exchange, I predict firms will patent more aggressively than otherwise expected when markets for technology are highly fragmented (i.e., ownership rights to external technologies are widely distributed); this effect should be more pronounced for firms with large investments in technology-specific assets and under a strong legal appropriability regime. Although these characteristics of firms and their external environments have been highlighted in the theoretical literature, prior research has not explored the extent to which such factors interact to shape the patenting behavior of firms. To empirically test these hypotheses, I develop a citations-based "fragmentation index" and estimate the determinants of patenting for 67 U.S. semiconductor firms between 1980 and 1994. Accumulating exclusionary rights of their own may enable firms to safeguard their investments in new technologies while foregoing some of the costs and delays associated with ex ante contracting.

Background Facts

• There has been an 'unprecendented' surge in patenting in the US (1.5 million filed 1990-2003 (assumed), 170,000 in 20001 alone)
• The literature usually assumes that firms have the right to make, use or sell the technologies they internalize, but there are challenges to assembling the rights to outside technologies.
• The US spent $1b in lawsuits over 1991 patents, about a 1/3rd of the R&D in that year
• There was a "propatent" shift in IP protection in the 1980's - the 1982 creation of the Court of Appeals for the Federal Circuit. The CAFC was propatent:
  • It expended the boundaries of claims through its interpretation of scope
  • It raised evidentiary standards for challenges to validity
  • It was more willing to grant preliminary injunctions
  • It issued 'propatent' rulings

Survey evidence suggest that primary reasons firms patents are:

1. to prevent rivals from patenting
2. to use patents in negotiations with owners of outside technologies
3. to deter infringement lawsuits

Note that to capture the economic value of an innovation is not mentioned. Instead, patents are used for trading - to gain favorable access to outside technologies or to reduce expected outflows in licensing.

Theory Development

Optimal patent policy design is based upon the notion that strong rights will induce sufficient levels of R&D investment (for the first inventor in a cummulative chain). But patents:

• Are inherently difficult to value
• Have blurry boundaries that are difficult to demarcate
• Are such that parties in the "cummulative chain" may find it hard to identify existing underlying IP

Two theoretical problems include:

1. The costs associated with being held up
2. The diffuse entitlements problem - underlying IP might be held by a large number of disparate parties

Hold-up in IP Markets

Hold-up occurs when one party is able to expropriate rents from another. As per Williamson, hold-up requires relational-specific investments (and incomplete contracts, which are implicit here). Transaction cost theory predicts that firms will either internalize the problem (i.e. make rather than buy) or underinvest when the risk of expropriation is high.

With patents there is an additional twist (due to their public goods like nature). A patent is a right to exclude - it does not guarantee the owner the right to use if use infringes on the rights of others. Firms must consider the price they expect to pay to use technologies patented by others and so should try to improve their ex-post bargaining positions.

Ex ante a firm could try to invent around a patent to alleviate hold-up. This is only possible with knowledge of the hindering patent(s) ex-ante, which can not be gauranteed. If the hindering patent is discovered after the firm has made a relational-specific investment, such as in technology to implement the patent that is difficult to redeploy, the firm can be held up. Ex ante soltuions are less feasible when a large number of exclusionary rights exist, and perhaps more importantly, when these rights are widely held.

Here the anti-commons therory - that a large number of exclusionary rights leads to underinvestment - is supplemented by the notion of the dispersion of rights, not just the number. Assuming that the Coasian setting of zero transaction costs does not hold (and it shouldn't!), the nature of the dispersion of the rights will matter.

With concentrated rights, that is rights held by only one or a small number of holders, the firm may:

• Contact the owner and negotiate (alliance, joint venture or acquisition, or licensing etc).
• Attempt to invent around the rights
  • The rights are more easily identified because they are concentrated (this is Ed's addition)
  • Infringement is more likely to be detected
  • The most valuable patents are easier to identify
  • The per-patent licensing cost should be reduced (only license the valuable ones, and through a single agreement)

However, if the rights are fragmented:

• Ex ante solutions may be infeasible
  • The costs of licensing would be higher, as would any delays - many agreements and a need to identify the value of each and every patent (in part to determine the likelyhood of exclusion ex-post)

Institutions such as patent pools, cross-licensing agreements, and collective rights organizations will arise (have arisen) to bundle patents and reduce transaction costs. Acquisition(s) may be a valid strategy but this involves direct costs (of the acquisition) and indirect costs (unrealized gains from trade with specialized firms - though this appears a little dubious).

One mechanism to mitigate hazards is to amass a larger patent portfolio - this improves ex-post bargaining by creating an "exchange of hostages" (as per Williamson). The threat of a reciprocal suit may ameliorate rent seeking behaviour.

Hypotheses

Hypothesis 1: The more fragmented the external technology market, the more aggressively firms will patent

An aggressive strategy is particularly important when:

• External technology markets are fragmented
• The anticipated cost of being held up is large.

Hypothesis 2: The effect of fragmentated external rights on incentives to patents will be more pronounced among capital-intensive firms.

The paper uses capital-intensitivity as a proxy for the relational specific investments that can be held up. There is considerable rationale in the paper why this might be true particularly in the semi-conductor industry (which is the subject of the data sample).

Hypothesis 3a: The effect of fragmented external rights on incentives will be stronger following the propatent shift in the US legal environment

Hypothesis 3b: The interaction effect between fragmented rights and capital-intensity will be greater in magnitude following the propatent shift in the US legal environment.

Measures

The paper constructs a fragmentation measure as follows (time subscripts are ignored):

[math]FRAG_i = 1 - \sum_{j=1}^J \left ( \frac{NBCITES_{ij}}{NBCITES_i}\right )^2\, \quad i \ne j\,[/math]

This measure is then corrected for firms with few patents:

[math]\hat{F}_i = \left ( \frac{NBCITES_i}{NBCITES_{i-1}}\right ) \cdot FRAG_i\,[/math]

where [math]NBCITES_i\,[/math] is the total number of citations to a patent assigned to a firm (on an annual basis)

There are two types of error inherenent in such a measure:

• Type 1: a citation is observed but there is no risk of infringement or no need to license
• Type 2: There ia risk of infrigement or a potential need to license but a citation is not observed.

Capital intensity is measured as the deflated book-value of a firm's property, plant and equipment, normalized by the number of employees. Note that firm size is included as a covariate and is the number of employees (logged). The era in which the "new power of patents" was known, was defined to be 1985 onwards, after a series of important court cases.

Results

The paper runs a base model (without the key explanatory variables) first, to get a baseline propensity to patent.

Firms that have a capital intensive one standard deviation above the mean patent five times more aggressively in response to average levels of fragmentation, even after controlling for R&D and firm size. More importantly, capital-intensive firms do not patent more intensively unless they build upon fragmented technology pools.

Increases in technological opportunity are discounted as an explantion - patenting in non-US jurisdictions is used as a measure. Possibly higher capital-intensive firms with higher fragmentation are also engaged in a broader range of alliances (e.g. R&D agreements with other firms, universities, etc), which increases the efficiency of their R&D. However, this is discounted in unreported results that interact R&D with fragmentation. Likewise some firms may just be better - fixed and random effects models do not support this hypothesis.

Finally, it does appear that the propatenting environment had the hypothesized effect: stricter IP regimes increased the hold-up problem.