A Comparative Analysis of Patent Assertion Entities in Markets for Intellectual Property Rights

OrganizationScienceVol. 27, No. 1, January–February 2016, pp. 2–17ISSN 1047-7039 (print) � ISSN 1526-5455 (online) http://dx.doi.org/10.1287/orsc.2015.1021

© 2016 INFORMS

A Comparative Analysis of Patent Assertion Entities inMarkets for Intellectual Property Rights

H. Kevin Steensma, Mukund ChariFoster School of Business, University of Washington, Seattle, Washington 98195

{[email protected], [email protected]}

Ralph HeidlLundquist College of Business, University of Oregon, Eugene, Oregon 97403, [email protected]

The patent assertion entity is a relatively new organizational form that neither invents nor commercializes products,but acts as a distributor of intellectual property rights between inventors and commercializing entities. We combine

measurement and governance branches of transaction cost theory to compare the efficiency of market intermediation bypatent assertion entities to that of bilateral licensing agreements, patent pools, and firm integration. We consider the levelof complementarity between patents and the breadth of their commercial applications to develop four general intellectualproperty configurations that depict distinct relationships between patent supply and patent demand. The costs and benefitsof the various governance alternatives are then weighed for each configuration to identify when each alternative is likelyto be most efficient. Our analysis suggests that patent assertion entities are most efficient in allocating intellectual propertyrights when there is substantial patent complementarity such that value is created through patent bundling, and these bundlesare applicable across a broad range of product lines such that the costs of measuring infringement and its damages aresubstantial. We consider how the imperfections of patents as contracts between inventors and society in conjunction withrapid technological evolution contribute to the growth of patent assertion entities. This analysis provides some guidancefor managers on how to appropriate value from intellectual property.

Keywords : technology and innovation management; patents and intellectual property rights; transaction costHistory : Published online in Articles in Advance December 10, 2015.

IntroductionPatent protection and efficient markets for allocatingintellectual property rights provide both an incentiveto innovate and a means for matching inventions withopportunities for their commercialization (Arora et al.2008). Well-functioning patent markets allow firms tospecialize in either discovery or commercialization andform bilateral licensing agreements (Arora et al. 2004).Other types of governance, such as patent pools, areparticularly efficient in bundling patents to facilitate thecommercialization pursuits of pool members (Shapiro2001). While bilateral licensing agreements and patentpools have a long history, a relatively recent organiza-tional form used to allocate intellectual property rights isthe patent assertion entity (PAE). Pejoratively referred toas patent trolls, PAEs neither invent nor commercializeintellectual property, but rather act as independent dis-tributors, purchasing patents from inventors and licens-ing the rights to commercializing entities. The evolutionof PAEs has been remarkable in terms of their speed ofascendance and influence on business practices. In 2011,PAEs generated licensing revenues of over $29 billion(Bessen and Meurer 2012).

Although the growing influence of PAEs is indis-putable, there is much debate as to whether they are ben-eficial or exploitative. Some argue they destroy value

by enforcing frivolous patents and hampering legitimatecommercialization (Blumberg and Sydell 2011, Reitziget al. 2007). Others contend they incentivize innova-tion by offering inventors remuneration they could notobtain on their own (McDonough 2006). Irrespectiveof such debate, the growth of PAEs suggests that theyfill a void left by other governance alternatives suchas bilateral licensing agreements between inventors andcommercializing entities, patent pools, and fully inte-grated ownership of patents and commercializing assets.Nonetheless, questions remain. When are PAEs likely toplay a role in markets for intellectual property rights?Why has this relatively new type of organization evolved?What has precipitated their growth? In an attempt toprovide some perspective on PAEs, we meld the mea-surement and governance branches of transaction costtheory to identify measurement challenges specific tointellectual property that lead to contractual difficulties,and compare how well various governance alternativesresolve these challenges. Although we acknowledge thetraditional influence of asset specificity on governanceefficiency, accurately measuring the value and use ofintellectual property assets which can neither be seen nortouched can be particularly challenging and a primarysource of transaction costs when contracting for such

2

Dow

nloa

ded

from

info

rms.

org

by [

205.

175.

118.

86]

on 2

9 A

pril

2016

, at 1

2:29

. Fo

r pe

rson

al u

se o

nly,

all

righ

ts r

eser

ved.

mailto:[email protected]



Steensma, Chari, and Heidl: A Comparative Analysis of Patent Assertion EntitiesOrganization Science 27(1), pp. 2–17, © 2016 INFORMS 3

assets. Although others have explored the allocation ofintellectual property rights through market transactionssuch as bilateral licensing (Arora and Ceccagnoli 2006)and hybrid structures such as patent pools (Lerner et al.2007, Shapiro 2001), a comprehensive assessment of thebroader set of alternatives including PAE intermediationmay provide insight as to when they are likely to be com-paratively efficient and why they are playing a larger rolein patent markets.

Our analysis takes into account two forms of assetspecialization specific to patents, which influence trans-action costs and governance efficiency. One form occursbetween patents themselves (patent complementarity),and the second occurs between patents and other com-plementary assets used for commercialization (patentspecialization). From this, we derive four general intel-lectual property configurations for commercializationthat depict distinct relationships between the supply ofexisting patents and their demand. We consider the coreactivities of patent deployment, enforcement, and com-mercialization, and reflect on when it is likely to be mostefficient for patent ownership to be held by inventors,commercializing entities, or PAEs. In doing so, we takeinto account how alternative governance arrangementsinfluence transaction and production costs.

With some exceptions (Argyres and Liebeskind 2002,Jacobides 2005), the single transaction has been thetypical unit of analysis when using a transaction costframework. We adopt a system view and consider theincentives of a broader set of potential participants(inventors, commercializing entities, PAEs) attemptingto appropriate value from intellectual property. Ouranalysis suggests that a lack of specialization betweenpatents and complementary commercialization assets inconjunction with patent complementarity creates mea-surement challenges and generates costs that cannot beeasily economized through firm integration or hybridarrangements such as patent pools. In these instances,PAE intermediation can both alter how rents from intel-lectual property are spread across inventors and com-mercializing entities, and create an efficient system ofintellectual property allocation.

Throughout our inquiry, we assume that all entitiesoperate within a reasonably strong and consistent appro-priability regime in terms of patent enforcement, andother than possibly infringing on the property rights ofothers, conduct themselves within the rules of the game.Once we establish conditions where PAE intermediationis likely more efficient than the alternatives, we con-sider how the deficiencies in the institutional processby which patents are granted, in conjunction with rapidtechnological evolution leads to such conditions. Patentsare imperfect and incomplete contracts between inven-tors and society; PAEs can be viewed as either exploitingthese imperfections (destroying social welfare) or less-ening their influence (preserving social welfare).

We begin laying the groundwork for our analysis bydescribing property rights, transaction costs, and pro-duction specialization within the context of intellectualproperty.

Property Rights, Transaction Costs andProduction SpecializationFormalized property rights are imperative to efficientlycreate value from combining assets (Barzel 1997). If forexample there were no property rights associated witha particular piece of land, it would fall into the publicdomain and be subject to overuse. Only when owner-ship of the land and its attributes are defined can a landowner efficiently exchange with other input providers(e.g., labor) to combine assets and create valuable outputsuch as crops.

Nonetheless, even with property rights in place,exchange can be costly when the attributes of the inputsare difficult for those other than their owners to accu-rately measure. The effort put forth by transacting par-ties to evaluate the inputs of their counterparts reducesany value created from their combination. Such transac-tion costs may be minimized by designing contracts thatgive residual value of the output created from the com-bination of inputs to the provider whose input is bothconsequential and costly for transacting counterparts tomeasure (Barzel 1997). For example, when the effortsof labor are difficult to measure and land quality is rel-atively easy to evaluate, a rental agreement that grantsa fixed amount to the land owner and the residual valueof the crop output to a laborer will dissuade the laborerfrom shirking. By reducing the need for the land ownerto measure (at high cost) the laborer’s effort, the valuefrom combining land and labor to produce crops is opti-mized. In contrast, land quality may be misrepresentedby self-interested landowners and difficult for others tomeasure independently. If the efforts of labor are rel-atively easier for the land owner to measure, a wagecontract that grants a fixed payment to a laborer withthe residual value of the crop accruing to the landownermay be most efficient. Under these terms, laborers willfeel less compelled to expend costly time and energyevaluating land quality, costs that detract from the valueof combining land and labor.

How contracts are designed influence not only trans-action costs, but also production costs and the potentialto realize economies from production specialization. Ifinput providers can easily measure and evaluate theinputs of their counterparts, each provider could thenspecialize in their respective activity and generate pro-duction efficiencies by operating at an efficient scale.For example, landowners may specialize in managingtheir land holdings and deploy them through rentalagreements, whereas their tenants specialize in crop pro-duction. If inputs cannot be easily measured by others,

Dow

nloa

ded

from

info

rms.

org

by [

205.

175.

118.

86]

on 2

9 A

pril

2016

, at 1

2:29

. Fo

r pe

rson

al u

se o

nly,

all

righ

ts r

eser

ved.

Steensma, Chari, and Heidl: A Comparative Analysis of Patent Assertion Entities4 Organization Science 27(1), pp. 2–17, © 2016 INFORMS

the cost efficiencies from production specialization maybe outweighed by the costs of input providers having tovigilantly evaluate the complementary inputs of others.In essence, the activities of land management and cropproduction are integrated within one economic entity ifthe sum of management and production costs within thatentity are less than the sum of transaction and productioncosts that occur with a rental arrangement (Williamson1979). Management costs arise from having to incen-tivize and measure employee performance, and resolvedisputes through fiat. Transaction costs arise from firmshaving to identify exchange partners, measure the valueof their respective inputs to determine equitable prices,and resolve contractual disputes through a court of law.

In an ideal world where it is costless for input pro-viders to evaluate the inputs of their counterparts, eco-nomic entities specialize in a narrow activity and relyon market exchanges to secure inputs and to sell output.Multiple independent entities could devote their effortstoward a given solitary activity, conduct it at the optimalscale, and fully realize all efficiencies from productionspecialization. Market forces alleviate the challenges ofpricing inputs; the flexibility to switch exchange part-ners ensures that downstream firms receive quality input atcompetitive prices. When there are few transaction costs,organizing transactions within the firm where multipleactivities are coordinated by internal management is inef-ficient because of the inherent benefits of market exchangeand production specialization (Williamson 1991).

Such logic also applies when combining intellectualproperty and commercialization assets. By drawing legalboundaries around intellectual property and establish-ing ownership rights, patents place value that wouldotherwise be in the public domain into private con-trol. Whether intellectual property is patented determinesin part the type of governance that is most efficientfor combining intellectual property with complemen-tary commercialization assets. If for example, intellec-tual property is secured solely through trade secrets,providers of such intellectual property may be reluc-tant to disclose its innermost details for fear of beingtaken advantage of and losing the opportunity to fullyappropriate value from such intellectual property (Ganset al. 2002, Merges 1994). Similar to prospective ten-ants trying to measure the quality of land from ownerswho are not forthcoming regarding its attributes, it isdifficult for prospective providers of commercializationinputs to evaluate the quality of intellectual property ifits details are not adequately disclosed by its providers.In these cases, the most efficient means of combiningsuch intellectual property with complementary inputs isoften through fixed payment wage contracts whereby theproviders of intellectual property and commercializationassets are integrated within the same organization.

By codifying intellectual property and establishingownership, patents facilitate disclosure, making it eas-ier for prospective providers of complementary inputs to

evaluate the intellectual property contained within them.In such cases, arms-length licensing agreements may bemore efficient than wage contracts. Analogous to rentalagreements between landowners and tenants, licensingagreements occur between patent owners and commer-cializing entities which, for a fixed fee, are granted themajority of residual rights to the value of the combina-tion of assets.

Notably, by alleviating transaction costs caused bymeasurement challenges, patents enable firms to spe-cialize in either invention or commercialization. Suchspecialization not only leads to scale efficiencies in pro-duction, but also facilitates the use of refined and targetedincentives. Such incentives are particularly consequen-tial for invention; it is often difficult for integrated firmsthat invent and commercialize to maintain dual incen-tives to motivate both their inventors and those involvedin commercialization activities (Milgrom and Roberts1990). The political sway of those directly involved withcommercialization often gives way to uniform incentivesthat maintain the primacy of commercialization person-nel over inventors. Firms that specialize in discovery canprovide university-like environments supportive of inven-tor productivity that large integrated firms, frequently rifewith competing interests and internal politics, are unableto duplicate (Argyres and Liebeskind 2002).

Transactional Frictions in Patent MarketsEstablishing ownership rights to intellectual propertythrough patenting can encourage its disclosure, dimin-ish the transaction costs of combining intellectual prop-erty with commercialization assets, and enable firmsto attain production efficiencies through specialization.Nonetheless, patentable discovery and commercializa-tion activities are often integrated within one economicentity, suggesting that measurement challenges persistdespite the clarity and protection provided by patents.Dissecting the invention-commercialization value chain,and taking into account intermediary activities associ-ated with owning patents, provide further insight intopotential impediments to forging bilateral licensing con-tracts directly between those that discover patentableinventions and those that commercialize them. Deploy-ing patents entails managing how best to appropriatevalue from them through their legitimate combinationswith commercialization assets. Enforcing patents entailsmonitoring for their illegitimate combinations with com-mercialization assets (i.e., infringement) and all associ-ated litigation. Patent ownership confers discretion overtheir deployment and enforcement.1 Similar to the activ-ities of invention and commercialization, costs associ-ated with patent deployment and enforcement can beeconomized through specialization and scale. The extentto which measurement challenges arise when deploy-ing and enforcing patents depend on their relationshipto other patents and complementary commercializationassets.

Dow

nloa

ded

from

info

rms.

org

by [

205.

175.

118.

86]

on 2

9 A

pril

2016

, at 1

2:29

. Fo

r pe

rson

al u

se o

nly,

all

righ

ts r

eser

ved.


Patent ComplementarityAlthough securing ownership over intellectual propertythrough patenting enhances disclosure of the intellectualproperty contained in the patent, measuring the valueof any one patent remains challenging. By design, apatent grants ownership over a relatively unique pieceof intellectual property; there are few points of com-parison for which to benchmark its value. Unlike com-modity products where highly liquid markets signal theirfair value, precise pricing signals rarely exist for patents(Hagiu and Yoffie 2013). These measurement challengesare further exacerbated when the value of a givenpatent depends on whether it is combined with otherpatents. We refer to highly cospecialized relationshipsamong patents as patent complementarity. Commercial-izing products without rights to all relevant patents in thebundle can lead to claims of infringement by the ownersof these complementary patents. Each individual patentholds little independent worth; only the full bundle ofcomplementary patents has value.

The effort that prospective providers of commercial-ization assets need to expend to accurately evaluate abundle of complementary patents may be considerable.Simply identifying the full portfolio of patents neededfor commercialization pursuits may be difficult, partic-ularly when patent owners are geographically dispersed(Arora et al. 2004). Even if all relevant patents can beidentified, transacting with each owner can be costly.Because the marginal contribution of each patent to thebundle is difficult to measure, individual owners aremore likely to haggle over the price for their respec-tive patents; coordination and negotiation costs are exac-erbated (Hagiu and Yoffie 2013). When two or morepatents owned by independent price-setting patent own-ers are needed to create a product, each owner can applytheir own monopolistic markup, inflating the costs ofattaining the needed rights (Gilbert 2004).

In general, patent complementarity can make it dif-ficult for providers of commercialization assets toaccurately value individual patents. Such measurementchallenges and associated transaction costs are likely tohamper the efficiency of bilateral licensing agreementsbetween inventors and commercializing entities, givingrise to alternative arrangements that are more efficient.

Patent SpecializationRegardless of whether patents create value indepen-dently or in conjunction with complementary patents,they vary in their commercial breadth. Highly special-ized patents have a limited number of commercial appli-cations. For example, a patent covering the use of amolecule to treat heart disease cannot likely be used forany other purpose. Other patents or patent bundles canbe applied more broadly, such as those associated withmicroelectronic technologies (Arora et al. 2004). These

patents can be combined with a variety of componentryto create a diverse set of products in multiple industries.

Owning patented intellectual property is profitableonly if its use is effectively measured and monetized.Deliberate or unintentional infringement of patentedinventions by commercializing entities leads to sub-stantial economic losses for inventors and their non-infringing licensees (Eisenberg 2011). One source ofinfringement is when current licensees adapt their exist-ing product lines or develop new ones, and use licensedintellectual property beyond the scope of their licens-ing agreements. To restore the legitimacy of their use,original licensing agreements often need to be renego-tiated. Monitoring the behavior of licensees is neces-sary to limit infringement and losing value to the publicdomain. When the cost of measuring such infringe-ment and renegotiating licensing agreements are sub-stantial, the value from combining patented intellectualproperty with commercialization inputs through licens-ing agreements shrinks. Alternative governance arrange-ments may become more efficient.

However, acts of infringement are not limited to cur-rent licensees. In contrast to physical assets, intellectualproperty is nonrivalrous in that many individuals may usethe same intellectual property unbeknownst to its right-ful owners or other legitimate users (Romer 1990). Thus,measuring infringement entails monitoring not only thosecontractually permitted to use specific patents in a limitedway, but also those that have no contractual permissionwhatsoever. To the degree that the patent in questionis broadly applicable to a wide swathe of industriesand products, measuring the full extent of infringementbecomes costly because of the sheer number of potentialinfringers. The expense of accurately measuring infringe-ment when there is limited patent specialization is fur-ther heightened when patent bundles are involved. In theabsence of one-to-one correspondence between patentand product, assessing the range of patent infringementmay require specialized personnel to deconstruct a broadarray of complex products such as computers or telecomequipment (Ludlow et al. 2008).

Even after infringement is detected, enforcing patentrights and pursuing equitable compensation can beexpensive. Measuring the value of the infringed patentin relation to other patents and commercialization assetsused to develop the infringing product, and deter-mining damages can be difficult for patent owners.Although they can threaten legal recourse, the high costof litigation with uncertain outcomes combined withlimited resources of many inventors reduces the credi-bility of such threats. As a result, many patent ownerseither abandon all hope of patent protection (Rosenkopfand Nerkar 2001), tolerate high levels of infringement(Cohen et al. 2000), or sell their patents to firms thatcan efficiently enforce them (Galasso et al. 2013).

Dow

nloa

ded

from

info

rms.

org

by [

205.

175.

118.

86]

on 2

9 A

pril

2016

, at 1

2:29

. Fo

r pe

rson

al u

se o

nly,

all

righ

ts r

eser

ved.


Table 1 Intellectual Property Configurations for Commercialization

Patent Patent Supply of Commercial demandcomplementarity specialization patents for patents Examples

Low High Drug molecules

Low Low Aerosol spray

High High Insect/weed resistant crops

High Low Field programmable gate arrays

Note. � , patent; •, commercial application.

In general, the lack of patent specialization increasesthe cost of measuring infringement and assessing itsdamages. As with most activity costs, these enforcementcosts can be economized through scale efficiencies inoperations.

Intellectual Property Configurations forCommercializationBy taking into account both patent complementarity andpatent specialization, we derive four intellectual prop-erty configurations (depicted in Table 1) that delineatethe relationship between the supply of patents and theirdemand in terms of commercialization opportunities.

Low Patent Complementarity/High PatentSpecializationIn the simplest type of configuration, patent complemen-tarity is low where the value of a patent is not based onbundling opportunities with other patents, whereas patentspecialization is high where a patent’s potential value islimited to a narrow set of commercialization processes. Inessence, the supply-demand relationship involves unbun-dled patents with a narrow commercialization corridor.Patented molecules used for drug development typi-cally fall under this configuration; the applicability of amolecule used to treat a specific ailment is clearly lim-ited. In addition, the single patent/molecule stands on itsown in terms of its value in commercialization.

For example, a therapeutic drug for the treatmentof AIDS protected by U.S. Patent No. 7,514,085 (i.e.,immune modulating compounds from fungi) is based onan extract of a certain type of mushroom. This drug wasdeveloped specifically to control the growth of the AIDSvirus and normalize the immune response of a patient.The patent can be deployed only in the production of aclass of drugs that have only one use—treating AIDS.The chemical industry yields several similar single-useexamples.

Low Patent Complementarity/Low PatentSpecializationAnother intellectual property configuration occurs whenthere is neither patent complementarity nor patent spe-cialization; there is little in the way of bundling of intel-lectual property, and the knowledge contained in thoseassets can be used in a broad array of commercializationprocesses.

A patent related to the method and means for atom-izing or distributing liquid or semiliquid materials illus-trates the idea of this configuration. Otherwise known asthe aerosol spray, U.S. Patent No. 1800156 was grantedin 1931. The value generated to date from the aerosolspray does not depend on complementary technologies,whereas its function is applicable to products rangingfrom fire extinguishers to beverage and pharmaceuticaldispensers.

High Patent Complementarity/High PatentSpecializationA third configuration occurs when both patent comple-mentarity and patent specialization is high; a bundle ofcomplementary patents are needed to create a productwhere its use is restricted to a limited set of commercial-ization processes and product lines. Crop technologiesprovide relevant examples of this type of intellectualproperty. Bayer has developed LibertyLink® technol-ogy making crops resistant to pigweed (U.S. Patent No.RE36449), whereas Monsanto’s Intacta RR2 Pro™ tech-nology is intended for making crops insect resistant(U.S. Patent No. 6982367). The value of the technolo-gies depends on whether they are used together to makecrops resist both weeds and insects (PRNewswire 2013).Yet this combination has a limited and specific use ofcreating pest-resistant soybean and corn.

High Patent Complementarity/Low PatentSpecializationIn the final configuration, patent complementarity is sub-stantial whereas patent specialization is minimal; patent

Dow

nloa

ded

from

info

rms.

org

by [

205.

175.

118.

86]

on 2

9 A

pril

2016

, at 1

2:29

. Fo

r pe

rson

al u

se o

nly,

all

righ

ts r

eser

ved.


bundling is necessary to create value for commercializa-tion, and such bundles are applicable to a broad arrayof commercializing entities. For example, dynamic inte-grated circuits are used in a wide range of productmarkets. Field Programmable Gate Array (FPGA) tech-nology (U.S. Patent No. 5,687,325) allows an integratedcircuit to be configured by a customer or a designer aftermanufacturing. FPGA technology relies on a bundleof complementary patents. Most manufacturers requireenergy-efficient programmable logic devices, thus thevalue of FPGA’s is contingent on complementary powercontrol technologies for integrated circuits (e.g., U.S.Patent No. 5,675,808, U.S. Patent No. 6,993,669). Fur-thermore, FPGA technology has low specialization as itis embedded in products related to the automotive, tele-com, wireless, defense, and broadcast industries.

Alternatives for Allocating IntellectualProperty RightsOur analysis considers the organization of value chainactivities for patents that have been granted—that is,their deployment, enforcement, and commercialization.Patent ownership, and thus discretion over its deploy-ment and enforcement, may be held by the inven-tor, an independent distributor, or a commercializingentity. In bilateral licensing arrangements, deploymentand enforcement activities conferred through owner-ship reside with the inventing entity; market contractsmediate the direct relationship between the inventorand the commercializing entity. With firm integration,patent and commercialization assets are owned by oneentity responsible for a patent’s deployment, enforce-ment, and commercialization. Patent pools entail shareddeployment and enforcement amongst participating poolmembers. PAE intermediation occurs when PAEs securepatent ownership and specialize in deploying and enforc-ing them, licensing their rights to commercializingentities.

When deploying and enforcing patents entail fewmeasurement challenges, bilateral licensing agreementsbetween inventors and commercializing entities wouldyield the most efficient outcome. Market forces pro-vide clear price signals and transaction costs are heldin check. However, when there are substantial measure-ment challenges associated with patent deployment andenforcement, greater efficiency may be derived fromgovernance alternatives that economize these costs. Asdiscussed below, each alternative brings its own set ofcosts and benefits.

Firm IntegrationOne means of mitigating the transaction costs associatedwith bilateral licensing agreements is to integrate patentdeployment, enforcement, and commercialization withinone entity that purchases outright the patents it needs

for its commercialization activity. In such cases, when-ever commercializing units within the firm uncover newapplications for the patents owned by the firm, there isno need to renegotiate contractual restrictions as wouldbe the case with bilateral licensing arrangements. Thosewho deploy patent rights and those who commercializeproducts based on these patents are on the same team.Instead of relying on market contracts to mediate therelationship between patent owners and commercializingentities, internal fiat is used (Masten 1988). Measuringinfringement by would-be licensees is a nonissue, andany associated transaction costs can be avoided.

However, there are countervailing costs from integrat-ing patent deployment and commercialization that lim-its the prevalence of this type of integration; in suchcases, high-powered market forces and fine-tuned incen-tives are attenuated. Similar to the difficulties associatedwith integrating invention and commercialization activi-ties (Argyres and Liebeskind 2002), efficiently integrat-ing patent deployment and commercialization within onefirm can be hampered by competing internal incentives.Managers responsible for product commercialization aretypically incentivized to maximize profits from theirrespective product lines; thus, they are motivated to sup-press outside competition by restricting the licensing outof owned patents to others, even if such licensing wouldenhance profits overall (Fosfuri 2006). Because patent-owning commercializing entities primarily focus on gen-erating revenues from products developed internally,they tend not to vigorously pursue licensing revenues,leaving money on the table by not fully appropriating thevalue of their patents (Arora et al. 2004, Chien 2010).All else being equal, it is more challenging to motivatethose responsible for deploying patents to fully appropri-ate the value from these patents if they are employed byan integrated firm (those that deploy, enforce, and com-mercialize) than a specialized firm that depends solelyon licensing revenues for survival. Integrated firms eitherincur the costs of implementing managerial control sys-tems to measure and incentivize patent deployment pro-fessionals or accept the lost revenues from not fullyappropriating patent value. The net difference in costsbetween firm integration and bilateral licensing agree-ments is the transaction cost savings realized from inte-grating deployment and commercialization minus anycosts from compensating for the loss of market incen-tives by instituting managerial control systems to moni-tor and incentivize deployment professionals, or the lostrevenue from not fully appropriating patent value.

Although integrating patent deployment, enforcement,and commercialization would eliminate the need to mea-sure infringement by those firms that would otherwisehave been licensees, it does little to economize the costsof measuring infringement more broadly by those firmswho might never have been licensees. Whether patentsare owned by an entity specializing in invention or an

Dow

nloa

ded

from

info

rms.

org

by [

205.

175.

118.

86]

on 2

9 A

pril

2016

, at 1

2:29

. Fo

r pe

rson

al u

se o

nly,

all

righ

ts r

eser

ved.


integrated firm, the costs of contending with infringe-ment by likely nonlicensees remain.

Patent PoolsPatent pools are collaborative arrangements formed bypartner firms that contribute their relevant patents to ajointly owned organization that bundle them together andthen license the bundle back to members for commer-cialization. Thus, responsibility for patent deploymentand enforcement is shared among members of the pool.Patent pools benefit member firms by eliminating theneed to negotiate multiple contracts or renegotiate themwhen new commercialization opportunities arise. Mea-suring the contribution of each patent relative to the bun-dle is a one-time setup cost for the pool. By offeringa standard priced single-bundle license contract (Lerneret al. 2007), patent pools avoid the costs from measur-ing the value of the bundle relative to each members’commercialization processes or customizing the bundleto individual members’ needs. However, similar to thecase of integrated firms, the motivation to fully exploitthe licensing revenue potential from intellectual propertycontrolled within a patent pool is muted. The primarymission of a patent pool is to provide its members accessto the intellectual property within its domain with a sin-gle price license to the entire bundle; maximizing returnsfrom licensing arrangements with those outside of thepool is secondary.

As compared to firm integration which eliminatesthe costs of coordinating and negotiating with multi-ple owners by fully integrating patent deployment andcommercialization within one entity, some coordinationchallenges remain for patent pool members. To obtainhigher revenues from their patents, some pool membersmay conceal their complete holdings of complementarypatents until pools have been formed. To protect the col-lective membership from opportunistic members, elab-orate contractual safeguards need to be established andenforced (Joshi and Nerkar 2011). Thus, although trans-action costs associated with patent pools are lower thanthey would be if each member contracted with eachother member bilaterally for patent rights, they are morethan what would occur if all patents were owned by oneentity.

Patent pools are a hybrid form of governance wheretransaction costs are at an intermediate level between thecosts that occur through bilateral licensing contracts andthose through firm ownership of intellectual property.However, similar to firm integration and bilateral licens-ing arrangements, patent pools do little to limit the costsof monitoring infringement by firms outside the pool.They typically lack specialized capabilities to efficientlydetect and litigate infringement.

PAE IntermediationPAEs are a breed apart from inventors and commercial-izing entities. They generally operate in a manner similar

to venture capital and private equity funds by raisingmoney from both corporate and institutional investors tobuy assets from those who would have difficulty mone-tizing them on their own. However, rather than purchas-ing equity in early stage companies, PAEs buy patentsfrom individual inventors, corporations, universities, andgovernment research labs, and license them to com-mercializing entities. Table 2 lists some of the leadingPAEs in terms of patent holdings. Notably, none of themexisted prior to the year 2000.

PAEs are fluidly structured, mixing and matching per-sonnel to create temporary teams composed of tech-nology and industry specialists, and other functionalgeneralists.2 Their activities include patent procure-ment, deployment (licensing), and enforcement (infringe-ment detection, litigation). Procurement professionalsare technology specialists who have in-depth expertisein a narrow technological domain, and are responsiblefor negotiating with inventors and purchasing patents.Licensing professionals have general expertise in craft-ing and negotiating licensing agreements with commer-cializing entities for a broad range of technologies. Todetect any infringement of their patent holdings, PAEsemploy industry specialists who reverse engineer com-mercial products within their industry of specializationand map patent claims onto these products. Becausepatented technologies may be used across multiple indus-tries, those responsible for detecting infringement withintheir designated industry work closely with in-house tech-nology specialists. A team of lawyers is responsible forpursuing suspected infringers in hopes of a settlement, oras a last resort, in a court of law. Although salaries andbonuses for these professionals are based on how wellthey work together and contribute to PAE revenue, onePAE executive described how his firm relies on a selectionprocess to ensure motivated employees; only true believ-ers in the mission of creating a new and socially beneficialmodel for patent markets are hired.

Opportunities to share best practices from operationsin one technological domain to another are prevalent forlicensing and litigation. For example, professionals cre-ating licensing contracts or litigating infringement fortelecommunication technologies can apply much of whatis learned from licensing and litigating software tech-nologies. Such opportunities are less prevalent in pro-curement and infringement detection, where expertise isspecific to a technology or industry.

PAEs function as distributors that intermediate rela-tionships between inventors and commercializing enti-ties. In general, distributors enhance efficiency bymatching the supply of varietal goods from multipleproviders to the demand for this variety from individ-ual consumers. Without them, providers and consumerswould have to contract directly with each other, resultingin numerous transactions. For example, five consumerscontracting individually with five single-good providers

Dow

nloa

ded

from

info

rms.

org

by [

205.

175.

118.

86]

on 2

9 A

pril

2016

, at 1

2:29

. Fo

r pe

rson

al u

se o

nly,

all

righ

ts r

eser

ved.


Table 2 Leading Patent Assertion Entities

Entity U.S. patent holdingsa Focusb Year of foundingb

Intellectual Ventures 251000 Networking, semiconductors, biomedical, energy, software 2000Round Rock Research 31500 Semiconductors 2010Rockstar Consortium 21400 Networking, Internet 2011Conversant Intellectual

Property Management21200 Semiconductor memory, networking, automotive 2007

Acacia Technologies 11700 Automotive, medical, networking, computers 2000Unwired Planet 11100 Mobile Internet 2000IPG Healthcare 11050 Software, electronics 2008Global OLED Technology 800 Organic light emitting diodes 2009Scenera Research 350 Software, location based services, Internet, computers 2002

aPatentfreedom.com.bCompany websites.

would generate a total of twenty-five transactions; a sin-gle distributor performing the bundling function wouldreduce this total to ten—five for the providers and five forthe consumers.

Similarly, PAEs can reduce the number of transactionsfor patent deployment which would otherwise occurthrough bilateral licensing agreements between inven-tors and commercializing entities. Within a PAE, indus-try and technology specialists work closely to identifyindustry trends, patent clusters, and potential licensingopportunities that drive patent procurement. Licensingprofessionals can then mix these patents into bundlesthat meet the needs of licensees. Commercializing enti-ties benefit from one-stop shopping for intellectual prop-erty, reducing the costs of locating relevant intellectualproperty and its owners; inventors benefit from not hav-ing to market their patented intellectual property directlyto potentially a vast array of commercializing entities(Coughlan 1987).

The bundling function provided by distributors is par-ticularly beneficial when it also economizes transactioncosts from having to measure and price hard-to-valuegoods (Kenney and Klein 1983). De Beers, the pri-mary distributor of raw diamonds, exemplifies how a dis-tributor can diminish such measurement issues throughbundling. Although each diamond in a given bundle maybe above or below the bundle average for the variousattributes (e.g., shape, quality, color), bundle price isbased on average attributes. De Beers prevents buyersfrom disaggregating a bundle into individual diamondsand haggling for lower prices. Although buyers overpayfor diamonds of quality below the bundle average, andunderpay for those above the bundle average, the pur-chasing process is simplified.

Measuring patent value can be even more difficultthan it is for diamonds because of the potential for com-plementarity between them. PAEs provide simplifyingpricing services similar to De Beers. Following the ini-tial identification of possible distribution opportunitiesby industry and technology specialists, licensing profes-sionals work closely with reverse engineering teams to

spot potential infringers and additional sources of rev-enue for a prospective patent bundle. By developing rel-atively precise estimates of the value of these patentbundles, revenue models are generated. With the aid ofthese models, and working alongside industry experts,procurement professionals assess the marginal contribu-tion of the various patents to their respective bundles,and negotiate sale prices with inventors. Through PAEintermediation, inventors no longer have to coordinatewith other inventors of complementary patents or mea-sure the value of their patents for various applications.Likewise, commercializing entities are spared from hav-ing to assess the value of each patent within a bundle ofdesired patents.

Moreover, PAEs economize the costs of patent deploy-ment and enforcement through their scale of operationsand any learning that occurs through repeated deploy-ment and enforcement. Because PAEs specialize inpatent deployment and enforcement, their efficient scaleof operations is determined by these activities; profes-sionals conducting them can be employed at an efficientscale. In contrast, when deployment and enforcement isintegrated within firms that also commercialize, the effi-cient scale of operations is more likely determined bythe activity of commercialization, the dominant activityin terms of capital intensity. For example, assume thatthe efficient scale for the commercialization efforts of acommercializing entity is ten products that are based ona relatively limited number of patents owned by the com-mercializing entity. In such cases, employing deploy-ment and enforcement professionals at an efficient scalemay not be viable with so few patents. The same is likelytrue for entities specializing in discovery; their efficientscale in terms of producing patentable inventions is notlikely to match that which is needed to achieve efficiencyin patent deployment and enforcement. Unimpeded bydiscovery and commercialization pursuits, PAEs havethe flexibility to optimally scale their operations.

Although large integrated firms (e.g., IBM) may havethe necessary scale of patent holdings to potentiallyachieve patent deployment and enforcement efficiencies

Dow

nloa

ded

from

info

rms.

org

by [

205.

175.

118.

86]

on 2

9 A

pril

2016

, at 1

2:29

. Fo

r pe

rson

al u

se o

nly,

all

righ

ts r

eser

ved.


similar to PAEs, their lack of specialization hamperstheir ability to do so for two reasons. First, the incentiveto fully appropriate value from a patent will be some-what diluted for integrated firms because their survivalis buffered by their commercialization pursuits. In con-trast, PAE survival rests solely on monetizing patentsthrough licensing revenues by means of effective patentdeployment and enforcement; market incentives remainintact. Second, the commercialization activities of inte-grated firms further weakens their ability to enforce thepatents they own because their products render thempotential targets of countervailing infringement claims.By not pursuing commercialization, PAEs avoid suchdynamics and have greater bargaining power vis-à-visinfringing parties to secure licensing revenues.

The source of PAE efficiencies lies not only intheir specialization in patent deployment and enforce-ment, but also the tight coordination of these appro-priately scaled activities. Deployment and enforcementcan become mutually fortifying in a virtuous circle.With each deployment of a patent or patent bundlethrough licensing, licensing professionals broaden theirunderstanding of its commercial breadth and possi-bly identify ancillary commercial applications. Suchinformation aids patent enforcement professionals indetecting prospective infringers and pursuing additionallicensing opportunities, providing further insight intoadditional commercial applications.

Overall, PAEs convert the costs of patent deploymentand enforcement from what would otherwise be viewedas transaction costs associated with licensing agree-ments between inventors and commercializing entities,into an efficient production function. They enable themost granularity in terms of production specialization;inventors focus on invention, PAEs are responsible forpatent deployment and enforcement, and commercializ-ing entities concentrate on commercializing the patentsthey license.

However, PAE intermediation adds another point ofcostly bureaucracy between inventors and commercial-izing entities. PAEs incur management costs to mea-sure employee performance and incentivize them to fullymonetize their portfolio of patents—costs that detractfrom the value of combining intellectual property andcomplementary inputs for their commercialization. Suchcosts may not be justified in conditions where PAE inter-mediation does not generate substantial efficiency gains.

An Analysis of Governance AlternativesWhen might each governance alternative dominate theothers in terms of efficiently allocating intellectual prop-erty rights? Comparative economic organization eval-uates the mix of conditions (in our case, intellectualproperty configurations) and governance alternativesthat are most efficient in terms of minimizing total

costs including transaction, management, and production(Williamson 1991). Costs associated with patent deploy-ment, enforcement, and commercialization will dependon the intellectual property configuration and the gover-nance alternative used to allocate patent rights.

Low Complementarity/Low SpecializationWhen patents do not need to be combined with oth-ers to create value, and are broadly applicable acrossa variety of products and industries, bilateral licensingarrangements between inventors and commercializingentities are likely the most efficient means for allocatingpatented intellectual property. Because patent comple-mentarity is low, commercializing entities do not facethe prospect of having to haggle with multiple own-ers of complementary patents. Moreover, measuring themarginal contribution of each patent to the value of com-mercialized products is comparatively trivial when thereare few complementary patents to account for. Becausepatents in this configuration are applicable to a broadarray of industries and potential licensees, high-poweredmarket forces maintained through bilateral licensingagreements are particularly critical for efficient patentdeployment. Integrating patent ownership and commer-cialization within a firm is likely to lead to money beingleft on the table because integrated firms emphasize rev-enues from their own commercialization activities at theexpense of pursuing supplementary licensing revenue.

Nonetheless, bilateral licensing agreements for patentsin this intellectual property configuration incur trans-action costs from inventors having to monitor their li-censees for infringement, and renegotiate contracts aslicensees uncover unanticipated opportunities to use theintellectual property. Ownership of these patents by com-mercializing entities tightly integrates their deploymentand commercialization, reducing such transaction costs.However, renegotiating with licensees is reasonablystraightforward because of a one-to-one correspondencebetween patent and product (i.e., low complementar-ity), and the ease of measuring the contribution of thesepatents to the value of final products. Avoiding thesenominal transactions costs by integrating patent own-ership and commercialization is not likely worth fore-going powerful market incentives that are maintainedthrough bilateral licensing arrangements, and encouragerelatively efficient patent deployment. Unburdened fromhaving to deploy and enforce patents, licensees can focussolely on their commercialization pursuits and scale theiroperations accordingly.

Because patent bundling is unnecessary, the manage-ment costs associated with either patent pools or PAEintermediation are not justified. When patented intellec-tual property is applicable to a wide range of productsand industries, there is an increased risk of infringementby entities other than licensees; PAE intermediation

Dow

nloa

ded

from

info

rms.

org

by [

205.

175.

118.

86]

on 2

9 A

pril

2016

, at 1

2:29

. Fo

r pe

rson

al u

se o

nly,

all

righ

ts r

eser

ved.


could potentially economize on the costs of enforce-ment through scale efficiencies. However, because thereis a one-to-one correspondence between patent and prod-uct, infringement and its damages will be relativelyeasy to measure. Potential infringers will be deterredbecause proving their infringement would be relativelystraightforward.

Overall, the costs to deploy and enforce patents inthis configuration are a small percentage of total costsand can be borne by inventors, even if conducted some-what inefficiently (relative to PAE intermediation). Insum, total costs comprising management, transaction andproduction costs across the three types of entities (i.e.,inventor, distributor, commercializing entity) are mini-mized when patents in a low complementarity/low spe-cialization configuration are owned by inventors andlicensed directly to commercializing entities. Recentempirical evidence supports the notion that general pur-pose technologies with relatively few interdependencies(e.g., software security algorithms) are exploited primar-ily through bilateral licensing agreements (Gambardellaand Giarratana 2013).

Proposition 1. When both patent complementarityand patent specialization are low, bilateral licensingagreements between inventors and commercializing enti-ties are the most efficient means to allocate intellectualproperty rights, relative to the alternatives.

Low Complementarity/High SpecializationSimilar to when patents have low complementarity withother patents and have extensive commercial applica-tions (i.e., low specialization), when they have low com-plementarity and have limited commercial applications(i.e., high specialization), patent pools offer little ben-efit; avoiding onerous transaction costs associated withmeasuring the marginal value of complementary patentsand negotiating with multiple owners is immaterial. Byspecializing in enforcement at an efficient operationalscale, PAE intermediation could potentially economizeon monitoring for their infringement and associated lit-igation. However, because patent complementarity islow and there is a one-to-one correspondence betweenpatent and product, detecting infringement and deter-mining damages are relatively easy. Moreover, the num-ber of potential infringers is limited due to the patent’snarrow applicability. In such cases, the use of PAEswould involve costly intermediation and bureaucracy inexchange for little benefit.

Whether bilateral licensing or firm integration is moreefficient for low complementarity/high specializationpatents depends on whether there would be mutualdependence between the owners of such patents andthe owners of commercializing assets. If the value of apatent asset depends on whether it is combined with aspecific set of commercialization assets, and the value of

such commercialization assets also depends on whetherit is combined with the specific patent asset, noneof these assets could easily be reassigned for alterna-tive uses (Teece 1986). Such asset specificity and themutual dependency it creates between transacting par-ties exacerbate transaction costs by forcing both partiesinto a small numbers bargaining situation. If contractualchanges become necessary, opportunistic partners cantake advantage of those that have few alternative usesfor their customized assets, and are locked into the rela-tionship. Any losses suffered by each contracting partyfrom haggling diminish the value created from combin-ing their assets. To keep such opportunistic behaviorin check, elaborate ex ante contractual safeguards andcontingency clauses may be instituted and expensive expost monitoring conducted to ensure that one’s exchangepartner is not being exploitive. At some point, transac-tion costs become so burdensome that firm integrationwhere disputes are resolved by fiat may be more efficientthan market contracts.

Transaction costs resulting from specificity betweenpatents and commercialization assets are a function ofmutual dependency. If either the patent or commercial-ization assets can be easily reassigned for alternativeuses, the incentive to integrate the ownership of patentand commercialization assets is absent for at least onetransacting party. When a patent is applicable to a broadarray of product lines and industries as with the previousconfiguration, opportunities to reassign it to alternativeuses are plentiful; the value of the patent will not bedependent on specific commercialization assets and thereis little risk of mutual dependency. However this is notthe case when a patent has narrow commercial applica-bility; mutual dependency between patent and commer-cialization asset owners is a real possibility. Integratingpatent ownership with commercialization when there ismutual dependency will squelch the high-powered mar-ket incentives that would have existed through bilaterallicensing. Yet, because these patents hold little valueunless they are combined with a specific set of com-plementary assets, deployment opportunities beyond theintegrated firm’s commercialization pursuits are few; therisk of an integrated firm leaving money on the table byunderdeploying the patents they own is low, renderingcostly internal incentive structures for patent deploymentunnecessary. When the value of both low complementar-ity/high specialization patents and the assets to commer-cialize such patents are highly dependent on them beingcombined specifically with each other, integrating patentownership with commercialization activities is likely tobe more efficient than bilateral licensing. Total costsacross the three entities potentially involved (i.e., inven-tor, distributor, commercializing entity) are minimized.

As described above, drug molecules fit within thelow complementarity/high specialization configuration.

Dow

nloa

ded

from

info

rms.

org

by [

205.

175.

118.

86]

on 2

9 A

pril

2016

, at 1

2:29

. Fo

r pe

rson

al u

se o

nly,

all

righ

ts r

eser

ved.


To avoid small numbers bargaining situations, drug man-ufacturers that make molecule-specific investments intesting and branding often prefer outright ownership ofdrug molecule patents prior to commercializing them(Lerner and Merges 1998). Integrated ownership of apatented drug molecule and its dependent commercial-ization assets will likely occur if the value of thepatented molecule is also highly dependent on beingcombined with such assets. Because of the low cost ofdetecting infringement and determining damages, theseactivities can readily be conducted in-house, albeit inef-ficiently as compared to when done by PAEs.

Proposition 2A. When there is substantial speci-ficity between low complementarity/high specializationpatents and commercialization assets, integrating patentownership and commercialization is the most efficientmeans to allocate intellectual property rights, relative tothe alternatives.

When there is no mutual dependence between the own-ers of low complementarity/high specialization patentsand the owners of commercializing assets, bilaterallicensing agreements between inventors and commercial-ization entities are likely to be most efficient.

In such cases, the benefits of maintaining marketincentives for the deployment of patent assets and/ortheir complementary commercialization assets are likelyto outweigh any transaction cost savings from inte-grating patent ownership and commercialization. In theinstance where the value of these patents does notdepend on them being combined with a specific set ofcomplementary assets, inventors deploying them throughbilateral licensing have the flexibility to reassign themto alternative uses if necessary. The contractual chal-lenge of small numbers bargaining is diminished andalternative licensees within the same narrow commercialapplication can readily be pursued. High-powered mar-ket incentives ensure that such patents are deployed effi-ciently and their full value is appropriated by inventors.

Likewise, if the value of commercialization assetsdoes not depend on them being linked with specific lowcomplementarity/high specialization patents, owners ofsuch assets can easily reassign them to alternative uses ifnecessary. By contracting for the requisite patent rightsas opposed to securing patent ownership, commercializ-ing entities are not saddled with patent deployment andenforcement responsibilities, and can reap the cost ben-efits from specializing in commercialization. The patentprotecting the method for DNA cloning (US4237224)is one example of a patent that has value independentof other patents, and is highly specialized for a nar-row application, yet deployed through bilateral licensing.This cloning method is used in conjunction with genericlaboratory assets that can easily be reassigned. As aresult, inventors and patent owners Stanley Cohen ofStanford and Herbert Boyer of University of California

established a licensing program that has led to over 450license agreements and generated over $255 million inrevenues since the patent was granted in 1980.

Proposition 2B. When there is little specificity be-tween low complementary/high specialization patentsand their complementary commercialization assets,bilateral licensing agreements between inventors andcommercializing entities are the most efficient meansto allocate intellectual property rights, relative to thealternatives.

High Complementarity/High SpecializationWhen patents have high complementarity with otherpatents and such bundles have limited commercial appli-cation, patent pools are particularly efficient in allocatingintellectual property rights compared with alternatives.In such a configuration, transaction costs associatedwith bilateral licensing would be extensive because ofthe need for bundling; each viable commercializingentity would have to identify and transact with indi-vidual patent owners. Because accurately measuring themarginal value of each patent to the value of the entirepatent bundle is difficult and fraught with error, eachpatent owner represented in a patent bundle may haggleover price with the other patent owners and commer-cializing entities in an attempt to extract greater licens-ing fees.

Patent pools mitigate such costs through bundling.They provide a coordinated pricing mechanism to reducelosses stemming from independent price setting by mul-tiple patent holders. For example, the sewing machinewars that occurred in the early 1850s threatened tostop all sewing machine production because of constantinfringement litigation between the major manufactur-ers (Lampe and Moser 2010). To resolve these disputes,the I.M. Singer Company and three other manufactur-ers created the first patent pool in 1856 by pooling theirnine complementary patents essential to manufacturinga sewing machine. Doing so lowered overall licensingfees and associated transaction costs. Licensing fees thatwere collected were used to protect this pool of patentsfrom being infringed by nonmembers.

Integrating deployment, enforcement, and commer-cialization of such patent bundles within a single firmwould completely alleviate the costs of coordinatingmultiple patent owners, but at the expense of relativelystronger market incentives within patent pools to effi-ciently deploy bundled intellectual property. Relying onthe independent distribution capabilities of PAEs offersanother alternative for patents that have high comple-mentarity and limited commercial application. AlthoughAEs can coordinate pricing across inventors and effi-ciently bundle patents, they do so at higher manage-ment costs than patent pools. The overhead from relyingon specialized deployment and enforcement capabilities

Dow

nloa

ded

from

info

rms.

org

by [

205.

175.

118.

86]

on 2

9 A

pril

2016

, at 1

2:29

. Fo

r pe

rson

al u

se o

nly,

all

righ

ts r

eser

ved.


found within PAEs to allocate intellectual property is notlikely justified when bundled patents have few commer-cial applications. The number of potential infringers willbe restricted, and the costs of monitoring infringementand any subsequent litigation will be a small percentageof total costs.

Total management, transaction and production costsacross the three entities possibly involved (i.e., inventor,distributor, commercializing entity) are minimized whenpatents associated with a high complementarity/highspecialization configuration are allocated through patentpools.

Proposition 3. When both patent complementarityand patent specialization are high, patent pools are themost efficient means to allocate intellectual propertyrights, relative to the alternatives.

High Complementarity/Low SpecializationScale efficiencies in patent deployment and enforcementprovided by PAEs will be most beneficial when thesecosts comprise a large percentage of total costs. A mis-match between the efficient scale for inventing and theefficient scale for patent deployment/enforcement, as islikely to occur through bilateral licensing agreementsbetween inventors and commercializing entities, will beparticularly costly when patents need to be bundled, andthese bundles are broadly applicable across a varietyof industries and product lines. Under such conditions,patent deployment costs in terms of identifying comple-mentary patents, measuring marginal patent values, andnegotiating with multiple parties are likely to be sub-stantial, along with any savings from scale efficiencies.Through unified ownership and patent bundling, PAEsreduce the total number of transactions that would other-wise occur through bilateral licensing; haggling betweenindividual inventors, and between inventors and com-mercializing entities is inhibited; pricing challenges thatplague patent bundles diminish. Employing a cadre ofprofessionals to efficiently create and price customizedbundles of patents is viable when their cost can be spreadover a large array of patent bundles, commercial appli-cations, and licensees.

Likewise, scale efficiencies in patent enforcement areparticularly valuable for patent bundles that have broadcommercial applicability. Enforcement costs will be asignificant component of total costs when the potentialpool of infringers is expansive, and measuring infringe-ment damages for each patent is difficult due to it beingone component of a larger bundle. By specializing inenforcement, PAEs can more efficiently enforce broadlyapplicable patent rights than can individual inventors.It is the combination of high patent complementarity(generating high deployment costs) and low patent spe-cialization (generating high enforcement costs) wherethe scale efficiencies of PAEs provide the greatest pay-off in cost reduction vis-à-vis bilateral licensing. Bymaintaining separation between patent deployment and

commercialization, PAE intermediation preserves themarket incentives associated with bilateral licensing tofully deploy patent rights, while shedding the scale inef-ficiencies associated with bilateral licensing where inven-tion and patent deployment/enforcement activities areintegrated.

Integrating patent ownership and commercializationis also likely to be relatively inefficient vis-à-visPAE intermediation when patents need to be bundled,and these bundles are broadly applicable. Similar tothe mismatch in efficient production scales betweenthe activities of invention and patent deployment/enforcement, any mismatch in efficient production scalesbetween commercialization and deployment/enforcementwill be relatively costly for patents in this configura-tion. A commercializing entity could acquire additionalpatents beyond what is needed for their commercial-ization in an attempt to establish scale efficiencies inpatent deployment/enforcement. Moreover, large-scalecommercializing entities with extensive patent portfolioswould appear to have the requisite scale to efficientlydeploy and enforce any patents they may wish to own.However, combining deployment and commercializationdilutes high powered market incentives that are particu-larly critical to fully appropriate value from complemen-tary patents that are broadly applicable to a vast arrayof potential licensees. Furthermore, because such patentscould be infringed on by many, the threat of countervail-ing infringement claims would be excessive, and severelyimpede commercializing entities from effectively enforc-ing them.

Although patent pools can quell the inefficient hag-gling among multiple patent holders that often occurswhen there is substantial patent complementarity, theylack the deployment and enforcement efficiencies, rela-tive to PAEs, to adequately customize bundles of patentsfor a wide array of uses, monitor the diversity of poten-tial infringers, or appropriately price the bundle acrossvarying applications.

In sum, efficiencies in patent deployment and enforce-ment are particularly beneficial for high complementar-ity/low specialization patents because patent deploymentand enforcement costs are likely to be substantial. Unen-cumbered by inventing and commercializing activities,PAEs are able to provide efficiencies from specializing indeploying and enforcing such patents, and reduce costsotherwise borne by inventors and commercializing enti-ties; total management, transaction and production costsacross the three entities possibly involved (i.e., inventor,distributor, commercializing entity) are minimized.3

Proposition 4. When patent complementarity ishigh, and patent specialization is low, PAE intermedia-tion is the most efficient means to allocate intellectualproperty rights, relative to the alternatives.

Table 3 succinctly summarizes the results of our com-parative analysis.

Dow

nloa

ded

from

info

rms.

org

by [

205.

175.

118.

86]

on 2

9 A

pril

2016

, at 1

2:29

. Fo

r pe

rson

al u

se o

nly,

all

righ

ts r

eser

ved.


Table 3 Efficient Governance Alternatives

Patent specialization

Low High

Patentcomplementarity

Low Licensing

Low asset specificity

Licensing Integration

High asset specificity

HighPAE

intermediationPatent pools

Patents as Incomplete Contracts BetweenInventors and SocietyOur analysis suggests that, when there is substantialcomplementarity between patents, and these patent bun-dles are applicable across a broad array of product lines,PAE intermediation is more efficient than other gover-nance alternatives. What then underlies the emergenceof these organizations? The growing imperfections inthe patenting process in conjunction with rapid technol-ogy evolution are creating conditions that promote PAEintermediation in patent markets.

Patent protection is granted only to those claimsthat are deemed novel. Patents are negotiated con-tracts between self-interested inventors and United StatesPatent and Trademark Office (USPTO) patent examinersacting on behalf of society, and responsible for assess-ing the novelty of inventors’ claims. To determine thenovelty of a claim, a patent examiner reviews relatedprior art found in previously granted patents or pub-lications. The USPTO patent examiner then comparesthis prior art to claims in a new patent application toassess whether the invention is novel (Cotropia 2009).The examiner is ultimately responsible for determiningwhat intellectual property is included within the patentand what is excluded. In contrast to well-defined prop-erty rights specified in boilerplate title deeds for physicalassets such as land and equipment, the boundaries ofintellectual property are often ambiguous.

An inadequate assessment of prior art and future tech-nological trajectories by boundedly rational patent exam-iners can generate incomplete patent contracts betweeninventors and society. Patents whose intellectual prop-erty overlap that of other patents are often granted whenpatent examiners have inadequate information on priorart (Mack 2006). For example, Apple’s voice recognitionpatent was rejected nine times before being accepted byan examiner as a valid patent, despite substantial priorart overlapping the same intellectual property claimedin Apple’s patent application (Duhigg and Lohr 2012).Such outcomes create patent complementarity; commer-cializing entities require access to a host of relatedand overlapping patents to legitimately develop productswithout infringing on the intellectual property of others.These overlapping patents may be generated by a num-ber of different inventing entities, making it difficult for

commercializing entities to search, value, and contractfor all the relevant patents. In many technology sec-tors, overlapping patents have become the norm becauseof an increase in patent applications, limited USPTOresources for adequately reviewing these applications,and conscious patenting strategies by firms designed toblock competitors’ product development or artificiallygain competitive parity (Somaya 2012).

Uncertainty over how technology is likely to evolvealso impedes patent examiners from anticipating exclu-sions and specifying precise terms when issuing a patent.Because of the inability to anticipate relevant exclusions,patents and patent bundles often become applicable toa broader range of product lines over time. For exam-ple, at the time that a patent for physically cutting andpasting photographs of people’s faces into cartoons wasgranted, neither the inventor nor the patent examinercould have envisioned such tasks being conducted elec-tronically through advanced graphics software. Conse-quently, the claims specified in the patent did not excludethis technological possibility. When a firm eventuallydeveloped such software, it was sued for infringement bythe owner of the original patent, which could be used toclaim infringement because technological advances hadessentially broadened its commercial application (Fed-eral Trade Commission and Department of Justice 2012).Rapid technological expansion can broaden the commer-cial applicability of a patent or patent bundle, and createconditions where PAE intermediation may be more effi-cient than its alternatives.

To the extent that the technology within various sec-tors is evolving at an increasing pace, leading to greaterpatent complementarity and commercial breadth, PAEintermediation within these sectors will also increase.

Proposition 5. PAE intermediation will be moreprevalent within technology sectors that are rapidlyevolving.

DiscussionOur analysis provides insight into when and howPAEs may function in the allocation of intellectualproperty rights. They are particularly efficient in deploy-ing and enforcing patented intellectual property relativeto other governance alternatives when complementaritywith other patents is substantial and specialization interms of commercial application is minimal. Such con-ditions are especially acute in technology sectors thatare rapidly evolving. By taking into consideration thesedistinctions in patent complementarity and specializa-tion, we address the unique characteristics of intellectualproperty assets to apply transaction cost considerationsmore precisely to patent markets. Accounting for theoverall management, transaction, and production costsacross inventors, PAEs, and commercializing entities

Dow

nloa

ded

from

info

rms.

org

by [

205.

175.

118.

86]

on 2

9 A

pril

2016

, at 1

2:29

. Fo

r pe

rson

al u

se o

nly,

all

righ

ts r

eser

ved.


provides a more complete perspective on when variousgovernance alternatives are likely to play a role.

In addition to bilateral licensing, patent pooling, anddirectly commercializing, PAEs offer another way forfirms to appropriate value from the patented intellectualproperty they develop. Our work provides guidance forfirms on when to sell portions of their existing patentholdings to PAEs and when to pursue alternatives. Formany firms, their patent holdings are likely to be dis-persed across a number of intellectual property config-urations (e.g., low complementarity/low specialization)that we have conceptualized. Categorizing patent hold-ings based on these configurations may provide a start-ing point for managers responsible for them.

Barring government regulations, PAE intermediationis likely to continue to grow and alter the landscape ofhow intellectual property rights are allocated. Prior toPAEs coming of age, patent pools were developedto allocate rights to complementary patents applicableto a broad array of product lines. For example, the3G-mobile communications patent pool was formed in2001, whereas the digital data transfer interface poolwas formed in 1999 (Lerner et al. 2007). Going forward,PAEs provide a viable alternative to patent pool forma-tion (Bekkers et al. 2006). Similarly, PAE intermediationmay ultimately reduce the occurrence of technology-driven acquisitions; in some cases, PAEs may provide amore efficient means for divesting patents than integrat-ing patent deployment, enforcement and commercializa-tion through corporate takeovers.

PAE intermediation may also shift the locus of inven-tion and reduce the need for commercializing entitiesto rapidly expand their own patent portfolios throughinternal development. To freely commercialize products,firms often need to navigate thickets of complementaryintellectual property rights controlled by competitors(Hall and Ziedonis 2001). These competitors can usetheir patents to allege infringement, impeding the sale ofrival products (Shapiro 2001). Targets of such behaviorcan invent around their competitors’ patents by generat-ing their own patents having claims overlapping those oftheir competitors’ patents, and establishing some formof competitive détente through cross-licensing or the for-mation of patent pools. By reducing transaction costs,PAE intermediation may incentivize the production ofsuch patents by entities specializing in invention. Insteadof expanding their own patent portfolios, commercializ-ing entities may be able to license bundled complemen-tary patents from PAEs, and do so at a lower cost thandeveloping them internally.

It is important to note that although inefficient gov-ernance alternatives are likely to be rejected by poten-tial participants, in the short run, efficient governanceis not necessarily optimal for all parties. PAE inter-mediation may increase costs for commercializing enti-ties as they incur licensing fees that they would not

otherwise. Indeed, this is the side of the story thatone is likely to read in the popular press (Blumbergand Sydell 2011). Although rampant infringement maydampen technological progress and hamper commercial-ization activity in the long run, commercializing entitiesenjoy lower commercialization costs because of rela-tively inefficient patent enforcement by individual inven-tors vis-à-vis PAEs. Nonetheless, these lower costs comeat the expense of inventors.

Furthermore, our analysis does not suggest that PAEintermediation is necessarily beneficial for economicgrowth or social welfare, nor addresses its effect onconsumers. Economizing transaction costs can generatesecond-order counterproductive outcomes at the societallevel depending on the legal context and other institu-tional forces (North 1990).4 Whether PAEs exploit ormitigate the effects of an imperfect patenting processis open for debate. One possible outcome from PAEefficiencies in patent deployment and enforcement isthat PAEs extract licensing fees from “frivolous” patentsand hamper the commercialization of socially benefi-cial products. Furthermore, by deploying and enforcingsuch patents, PAEs may incentivize inventors to wasteresources on developing frivolous patents. Commercial-izing entities would eventually pass on the costs of thesefrivolous patents to consumers, harming social wel-fare. Moreover, there are structural differences betweenwhen PAEs assert their patents against commercializingentities and when commercializing entities assert theirpatents against other commercializing entities. In thecase of two commercializing entities, each can threatento block the other’s product sales through infringe-ment litigation; this mutual destruction scenario encour-ages some form of cross-licensing settlement. Suchsettlements have proliferated due to the abundance ofcomplementary patents being granted to commercializ-ing entities (Hall and Ziedonis 2001). However, becausePAEs do not compete in product markets, there is nomutually assured destruction scenario. As some suggest,this structural difference enables PAEs to legally extortcommercializing entities through excessive royaltieswithout fear of retribution (Blumberg and Sydell 2011).

Whether a patent or lawsuit is frivolous, or whethercertain PAE behavior is coercive is in the eye of thebeholder, and ultimately determined by the courts. Ouranalysis assumes that PAEs operate within the rules ofthe game. If policy makers wish to quell what theybelieve to be counterproductive PAE intermediation, theycan do so by minimizing conditions where PAEs deployand enforce intellectual property rights more efficientlythan other governance alternatives. Reducing both patentoverlap and the granting of overly broad patents wouldbe one means of doing so. Recent changes in how patentapplications are assessed may play a role. The AmericaInvents Act of 2012 makes it easier for firms to anony-mously submit any prior art they believe is relevant to

Dow

nloa

ded

from

info

rms.

org

by [

205.

175.

118.

86]

on 2

9 A

pril

2016

, at 1

2:29

. Fo

r pe

rson

al u

se o

nly,

all

righ

ts r

eser

ved.


their competitors’ patent applications in hopes of pre-venting competitors from obtaining overly broad patents.Other changes to the law enable firms to lodge post-grantopposition to newly granted patents.

An alternative conclusion from our analysis may bethat PAE intermediation mitigates the effects of inher-ently imperfect patent contracts between inventors andsociety by efficiently bundling and enforcing broadlyapplicable patents with overlapping boundaries. With-out the intermediation provided by PAEs, inventors andcommercializing entities would be burdened with exces-sive transaction or management costs, particularly insectors with rapidly evolving technologies where over-lapping and broadly applicable patents are likely toappear. Such cost burdens may discourage innovation.To the extent that the granting of patents naturally entailssome imperfection and uncertainty, PAE intermediationmay be socially beneficial by efficiently resolving theerrors made by the USPTO. Although PAEs have beendisparaged in the popular press for being highly litigious(Blumberg and Sydell 2011), this may be a functionof the type of patents they are likely to manage; PAEintermediation is more efficient than alternatives in theallocation of intellectual property rights for patent bun-dles that are broadly applicable. However, the threat ofinfringement is also substantial for this configuration.The litigiousness of PAEs is likely a natural outcome ofconditions where they are the most efficient alternativefor allocating intellectual property rights.

AcknowledgmentsThe authors thank the Arthur W. Buerk Center ofEntrepreneurship at the University of Washington for its finan-cial support. Senior Editor Nicholas Argyres and two anony-mous reviewers provided excellent guidance throughout thereview process.

Endnotes1Although responsibilities for patent enforcement can be fur-ther disaggregated from their deployment and outsourced, cre-ating efficient market contracts between patent owners andenforcement professionals is often difficult, because of prob-lems in measuring the effort of individuals contracted todetect infringement and enforce patent rights. Thus, it is oftenmost efficient to combine patent deployment and enforcementwithin a single firm; assuming these activities are integratedsimplifies our analysis without unduly limiting inferences.2Information regarding the internal structure of PAE is basedon interviews conducted by the authors. We initially identifiedcurrent and former employees through LinkedIn and alumnidatabases. Although we were repeatedly told of confidential-ity agreements and restricted communications with journalistsand academics due to past derision of PAEs in the popularpress, we were able to secure interviews spanning 45–90 min-utes with five current and former employees from two separatePAEs. Their titles ranged from portfolio architect to seniorvice presidents.

3The limited empirical evidence available is consistent withthe notion that PAEs tend to acquire patents that are broadlyapplicable (Fischer and Henkel 2012).4For example, in some legal contexts, bribery or other nefar-ious activities can be used to minimize transaction costs orenforce contracts. However, such activities may have harmfulsocial costs.

References

Argyres NS, Liebeskind JP (2002) Governance inseparability andthe evolution of US biotechnology industry. J. Econom. Behav.Organ. 47(2):197–219.

Arora A, Ceccagnoli M (2006) Patent protection, complementaryassets, and firms’ incentives for technology licensing. Manage-ment Sci. 52(2):293–308.

Arora A, Ceccagnoli M, Cohen WM (2008) R&D and the patentpremium. Internat. J. Indust. Organ. 26(5):1153–1179.

Arora A, Fosfuri A, Gambardella A (2004) Markets for Technol-ogy: The Economics of Innovation and Corporate Strategy (MITPress, Cambridge, MA).

Barzel Y (1997) Economic Analysis of Property Rights (CambridgeUniversity Press, New York).

Bekkers R, Iversen E, Blind K (2006) Patent pools and non-assertion agreements: Coordination mechanisms for multi-partyIPR holders in standardization. Paper for the EASST 2006 Conf.,Lausanne, Switzerland, August.

Bessen J, Meurer M (2012) The direct costs from NPE disputes. Lawand Economics Research Paper 12–34, Boston University Schoolof Law, Boston.

Blumberg A, Sydell L (2011) When patents attack! This AmericanLife (July 22), http://www.thisamericanlife.org/radio-archives/episode/441/when-patents-attack.

Chien CV (2010) From arms race to marketplace: The new com-plex patent ecosystem and its implications for the patent system.Hastings Law J. 62(December):297–356.

Cohen WM, Nelson RR, Walsh JP (2000) Protecting their intellec-tual assets: Appropriability conditions and why U.S. manufactur-ing firms patent (or not). NBER Working Paper 7552, NationalBureau of Economic Research, Cambridge, MA.

Cotropia CA (2009) Modernizing patent law’s inequitable conductdoctrine. Berkeley Tech. Law J. 24(2):723–783.

Coughlan AT (1987) Distribution channel choice in a market withcomplementary goods. Internat. J. Res. Marketing 4(2):85–97.

Duhigg C, Lohr S (2012) The patent, used as a sword. New York Times(October 7), http://www.nytimes.com/2012/10/08/technology/patent-wars-among-tech-giants-can-stifle-competition.html?_r=0.

Eisenberg RS (2011) Patent costs and unlicensed use of patentedinventions. University of Chicago Law Rev. 78(1):53–69.

Federal Trade Commission, Department of Justice (2012) TranscriptPatent Assertion Entities Activities Workshop, Washington, DC.

Fischer T, Henkel J (2012) Patent trolls on markets for technology—An empirical analysis of NPEs’ patent acquisitions. Res. Policy41(9):1519–1533.

Fosfuri A (2006) The licensing dilemma: Understanding the deter-minants of the rate of technology licensing. Strategic Manage-ment J. 27(12):1141–1158.

Galasso A, Schankerman M, Serrano CJ (2013) Trading and enforcingpatent rights. RAND J. Econom. 44(2):275–312.

Gambardella A, Giarratana MS (2013) General technological capabil-ities, product market fragmentation, and markets for technology.Res. Policy 42(2):315–325.

Dow

nloa

ded

from

info

rms.

org

by [

205.

175.

118.

86]

on 2

9 A

pril

2016

, at 1

2:29

. Fo

r pe

rson

al u

se o

nly,

all

righ

ts r

eser

ved.

http://www.thisamericanlife.org/radio-archives/episode/441/when-patents-attack

http://www.thisamericanlife.org/radio-archives/episode/441/when-patents-attack

http://www.nytimes.com/2012/10/08/technology/patent-wars-among-tech-giants-can-stifle-competition.html?_r=0

http://www.nytimes.com/2012/10/08/technology/patent-wars-among-tech-giants-can-stifle-competition.html?_r=0


Gans JS, Hsu DH, Stern S (2002) When does start-up innovationspur the gale of creative destruction? RAND J. Econom. 33(4):571–586.

Gilbert RJ (2004) Antitrust for patent pools: A century of policy evo-lution. Stanford Tech. Law Rev., No. 3, http://stlr.stanford.edu/STLR/Articles/04_STLR_3.

Hagiu A, Yoffie DB (2013) The new patent intermediaries: Platforms,defensive aggregators, and super-aggregators. J. Econom. Per-spectives 27(1):45–65.

Hall BH, Ziedonis RH (2001) The patent paradox revisited: Anempirical study of patenting in the US semiconductor industry,1979–1995. RAND J. Econom. 32(1):101–128.

Jacobides MG (2005) Industry change through vertical disintegra-tion: How and why markets emerged in mortgage banking. Acad.Management J. 48(3):465–498.

Joshi AM, Nerkar A (2011) When do strategic alliances inhibit inno-vation by firms? Evidence from patent pools in the global opticaldisc industry. Strategic Management J. 32(11):1139–1160.

Kenney RW, Klein B (1983) The economics of block booking. J. LawEconom. 26(3):497–540.

Lampe R, Moser P (2010) Do patent pools encourage innovation?Evidence from the nineteenth-century sewing machine industry.J. Econom. History 70(4):898–920.

Lerner J, Merges RP (1998) The control of technology alliances:An empirical analysis of the biotechnology industry. J. Indust.Econom. 46(2):125–156.

Lerner J, Strojwas M, Tirole J (2007) The design of patent pools:The determinants of licensing rules. RAND J. Econom. 38(3):610–625.

Ludlow T, Thumm M, Girones A (2008) Backing up theclaim: Reverse engineering enables successful patent licensing.Nouvelles—J. Licensing Executives Soc. 43(2):102–105.

Mack K (2006) Reforming inequitable conduct to improve patentquality: Cleansing unclean hands. Berkeley Tech. Law J. 21(1):147–175.

Masten SE (1988) Legal basis for the firm, A. J. Law, Econom.,Organ. 4(1):181–198.

McDonough JF (2006) The myth of the patent troll: An alternativeview of the function of patent dealers in an idea economy. EmoryLaw J. 56:189–228.

Merges RP (1994) Of property rules, coase, and intellectual property.Columbia Law Rev. 94(8):2655–2673.

Milgrom P, Roberts J (1990) Bargaining costs, influence costs, and theorganization of economic activity. Alt JE, Shepsle KA, eds. Per-spectives on Positive Political Economy (Cambridge UniversityPress, New York), 57–89.

North DC (1990) Institutions, Institutional Change and EconomicPerformance (Cambridge University Press, New York).

PRNewswire (2013) Bayer CropScience and Monsanto enter intocross-licensing agreements for next-generation and enablingtechnologies. Accessed May 15, 2013, http://www.prnewswire.com/news-releases/bayer-cropscience-and-monsanto-enter-into-cross-licensing-agreements-for-next-generation-and-enabling-technologies-203185651.html.

Reitzig M, Henkel J, Heath C (2007) On sharks, trolls, and theirpatent prey–unrealistic damage awards and firms’ strategies of“being infringed.” Res. Policy 36(1):134–154.

Romer PM (1990) Endogenous technological change. J. PoliticalEconom. 98(5):S71–S102.

Rosenkopf L, Nerkar A (2001) Beyond local search: Boundary-spanning, exploration, and impact in the optical disk industry.Strategic Management J. 22(4):287–306.

Shapiro C (2001) Navigating the patent thicket: Cross licenses, patentpools, and standard setting. Jaffe A, Lerner J, Stern S, eds. Inno-vation Policy and the Economy (MIT Press, Cambridge, MA),119–150.

Somaya D (2012) Patent strategy and management an integrativereview and research agenda. J. Management 38(4):1084–1114.

Teece DJ (1986) Profiting from technological innovation: Implicationsfor integration, collaboration, licensing and public policy. Res.Policy 15(6):285–305.

Williamson OE (1979) Transaction-cost economics: The governanceof contractual relations. J. Law Econom. 22(2):233–261.

Williamson OE (1991) Comparative economic organization: The anal-ysis of discrete structural alternatives. Admin. Sci. Quart. 36(2):269–296.

H. Kevin Steensma is professor of management and amember of the faculty at the Foster School of Business, Uni-versity of Washington-Seattle. He received his Ph.D. fromIndiana University in Bloomington. His research interestsinclude strategic alliances, knowledge spillover, innovation,and intellectual property rights.

Mukund Chari is a Ph.D. candidate at the Foster Schoolof Business, University of Washington, Seattle. His researchinterests include innovation and intellectual property rights.

Ralph Heidl is assistant professor of management and amember of the faculty at the Lundquist College of Business,University of Oregon. He received his Ph.D. from Universityof Washington. His research interests include multilateral col-laboration, innovation, and intellectual property rights.

Dow

nloa

ded

from

info

rms.

org

by [

205.

175.

118.

86]

on 2

9 A

pril

2016

, at 1

2:29

. Fo

r pe

rson

al u

se o

nly,

all

righ

ts r

eser

ved.

http://stlr.stanford.edu/STLR/Articles/04_STLR_3

http://stlr.stanford.edu/STLR/Articles/04_STLR_3

http://www.prnewswire.com/news-releases/bayer-cropscience-and-monsanto-enter-into-cross-licensing-agreements-for-next-generation-and-enabling-technologies-203185651.html




Documents

A Comparative Analysis of Patent Assertion Entities in Markets for Intellectual Property Rights