‘Software Ownership is Killing Innovation’—Controversial Author Calls for a Reboot

Creative destruction, once a hallmark of progressive capitalism, is no longer working. Powerful companies that control software and silo valuable data are impeding innovation and threatening society.

That is the conclusion of a recently published book, The New Goliaths: How Corporations Use Software to Dominate Industries, Kill Innovation, and Undermine Regulation. The author, James Bessen, vilifies the impact of software ownership and the ability of patents and other rights to effectively promote innovation and encourage competition.

Bessen is Executive Director of the Technology and Policy Research Center at Boston University. He also teaches at Boston University’s School of Law. By his own admission, he is less a trained economist than a former entrepreneur interested in innovation policy. He believes that software and data have become too valuable to be owned like other inventions. Proprietary ownership slows what he calls diffusion, getting innovation into the hands of businesses and people who can use and improve it.

He believes the most dominant tech players – Amazon, Microsoft, Google and Apple – have already started down the road to diffusion by opening their cloud-related code and using open-source platforms. They are not the first to do so, but among the most important. Linux is perhaps the best-known kernel to be open and royalty free. He believes that by “unbundling” their offerings, especially in relation to cloud services, big tech and others can facilitate diffusion. Bessen concedes that sharing code for them is less about altruism than smart business.

If not the big five, who are today’s proprietary software proponents? Bessen does not name names. He is convinced, however, that keeping technology developments out of the public domain, i.e., subjecting them to IP protection, has a negative effect on innovation. Bessen seems to be saying that if the largest tech companies with the most to lose are willing to freely share important developments, why shouldn’t others?

If invention rights are unclear, as he says, often leading to disputes and litigation, and especially those that relate to software and data, why not simply improve patent certainty? That would create diffusion through commerce, which would be good for jobs and society. It would also provide leverage to small and medium enterprises (SMEs) and independent inventors and create competition, potentially threatening the leadership of some businesses.

“Unfortunately, there is a fundamental difference between patenting a tangible invention and patenting a concept: it is often much harder to identify the boundaries of the property right for software patents. When property boundaries are poorly defined, parties are more likely to sue. (p. 251).”

My understanding is that software is not a “concept” but a series of instructions, some drawing on existing directions, that when properly deployed can permit significant efficiencies. While the law may not always consider this an invention, creators, businesses and investors often do.

‘Superstar Capitalism’

The New Goliaths examines at length, and with clarity, the history of business software and the evolution from an enhancement bundled without much thought and no direct cost with hardware sales to a more transparent and collaborative stand-alone product that can be used by customers and competitors alike. When building a standard, this can make sense. IBM is exhibit A for software’s evolution, suggests Bessen.

The company went from including software with hardware sales in the 1960s, to selling or leasing it as a separate product, which often required expensive tailoring to meet clients’ needs, to a less proprietary status when it embarked on a cloud initiative after the acquisition of Red Hat in 2019. The software shared still needs adapting to client systems, the data interpreted, and the server storage paid for.

One way to look at it – although I am uncertain that Bessen would agree – is that some types of software have become the proverbial razor and products like storage and consulting the required blades.

Bessen cautions government to guard against the effects of what he calls “superstar capitalism.” He says that policy challenges are necessary to find ways “to promote greater diffusion.” He does not believe that stronger, more certain rights will achieve wider use by making licensing easier and more profitable for a range of businesses and reliability more consistent (p. 227).

In superstar markets, he says, “where technology is used to differentiate firms,” IP owners have insufficient incentives to license or unbundle in other ways. This is the reason why policy is needed to spur unbundling.

“Proprietary software has fostered the rising and persistent dominance of large firms; it has slowed diffusion of new technology, reducing productivity growth; it has restricted the opportunities for workers to gain new technology-related skills, contributing to growing economic inequality; and it has increased the power of large firms both relative to consumers and relative to government regulators.”

Is private ownership the culprit or is it the lack of patent certainty? The reluctance of large technology businesses to out-license their software inventions, and refusal to in-license others’ and pay for what they need, has not helped. From a shareholder perspective it may be smarter for these firms to game the system by efficiently infringing, putting the blame on patent licensors; from a broader economic and societal view, probably not.

The implied framing is dubious: While Big Tech is stepping up to share, operating companies, like non-practicing trolls, are reaching down to take.

Accelerating Diffusion?

“Open platforms,” explains Bessen, “may be poised to change the economy. Where proprietary software systems have slowed the diffusion of new technology, unbundling may accelerate diffusion, bringing new industry dynamism and innovation, revering the negative effects of proprietary software systems on productivity growth, income inequality, and geographical divergence (p. 245).”

He gives credit to Amazon, Google, Microsoft, IBM, and even Apple, for providing open access to cloud-related code. They have recognized the value of sharing certain software and how by doing so those inventions can be improved and used to everyone’s benefit. It is not only in their best interest, he says, but in the interest of innovation and society. That many software engineers abhor the restrictions imposed by patents and copyrights encourages a vocal cheering section.

Killing Creative Destruction

“There is a broad swath of the economy – way beyond a handful of big tech companies in Silicon Valley,” Bessen told The New York Times in a recent Sunday Business feature. “There is an advantage to software that economists haven’t really reckoned with yet. Software isn’t accelerating creative destruction today. Software is suppressing it.”

In New Goliaths he challenges what he says is the ‘destruction myth’ that has its roots with Joseph Schumpeter’s theory of creative destruction, widely accepted as a catalyst for growth in capitalist economies. For free market proponents, it is healthy to challenge incumbents. But The New Goliaths suggests that it is more than software that should be shared. The book appears to provide a rationale for the continued growth of technology giants as long as they make available some vital pieces of their pie. Sharing proprietary inventions in this manner is an easier call for tech players who have an established business model to build on. For less well-capitalized businesses, and independent creators, that strategy is less of an option.

Since the 1990s, he says, the chances of unseating a dominant firm, one of the top three or four in an industry, has declined by half. Bessen does not attribute this to less reliable patents and IP-weakening court decisions, or to the establishment of the Patent Trial and Appeal Board (PTAB). Leveling the playing field has never been more difficult. It is doubtful whether eliminating or limiting proprietary inventions like software can help SMEs to remain competitive.

Dubious Roots

Bessen, readers may recall, is author with James Meurer of Patent Failure: How Judges, Bureaucrats, and Lawyers Put Innovators at Risk, the controversial 2008 book which claimed that patent litigation was running amok and patent trolls were destroying innovation.

The book is noted for data such as:

“We find that NPE lawsuits are associated with half a trillion dollars [my italics] of lost wealth to defendants from 1990 through 2010. During the last four years, the lost wealth has averaged over $80 billion per year [2008-2012].”

These figures have been repeatedly debunked. Adam Mossoff, professor of law at George Mason University, Senior Fellow at Hudson Institute and a Center for Intellectual Property Understanding board member, took Bessen and his partner in research crime, Michael Meurer, to task in a stinging article, “Repetition of Junk Science and Epithets Does not Make them True.”

A strong argument can be made that efficient infringement has cost businesses and invention owners, as well as investors and society, far more than the amount falsely cited by Bessen and Meurer. The kind of fear-mongering that that ‘Patent Failure’ promotes led in part to the adoption of the America Invents Act and the onerous PTAB, whose impact on innovation and the economy has yet to be fully measured.

The final two chapters, “The Next Unbundling” and “The New Information Economy,” are worth a detailed look.

Bessen’s premise is that proprietary software and data collection are by nature injurious to the health of innovation and society, and that large companies hold the most sophisticated software and extensive data. He is not wrong that software is difficult to frame and potentially dangerous to concentrate in the hands of a few. That is why, he says, we must create policy and licensing models that encourage fairer, more robust systems that will provide leverage to more businesses and balance. There is room for multiple strategies around IP ownership, as illustrated in a summary report from the Center for Intellectual Property Understanding (CIPU). That is not how Bessen sees it.

It is true that software collaboration can be a facilitator, but proprietary ownership be parsed and properly credited in a digital society – if there is motivation to do so. The technology is available to determine use of copyrighted content and trademark. A similar level of transparency can likely be achieved with patents, too.

The New Goliaths is provocative if not perplexing. The book is strongest when is makes the argument for sharing inventions that are relevant to other businesses, and weakest when it suggests that all companies that own software and other inventions should play by its new and better rules.

The term “patent troll” is used only once in the book and not until p. 251. It is not even mentioned in the index. However, it is the elephant in the room. Bessen cites his own work, providing a distorted and inaccurate description that damns non-practicing entities and indicts licensors. “The result of software patenting has been an explosion in patent litigation, much of it coming from patent trolls, parties who purchase or develop patents, not to introduce a new technology, but to extort payments through the threat of litigation.” He sees no technical, economic or societal benefit to independent licensing and fails to contextualize the real source of patent litigation. Owners – primary or secondary – often sue because they are forced to.

New Boogey Meme

The boogey meme for Bessen is no longer trolls. It is innovation policy and companies that want to maintain (in this case) software-related inventions. Remember, the shelf life of most software is a fraction of other inventions, so patents may not be that relevant to begin with.

Bessen neglects to mention that IBM remains the leading recipient of U.S. patent grants and a leading licensor. This is despite its increasing cloud participation and ownership of open-source leader Red Hat. My understanding is that a significant number of their new rights relate to software, and many expire before the first renewal. Whether diffusion is increased, he says, “will depend very much on government policy, and unfortunately, current policy has been moving in the wrong direction, partly as a result of the undue political influence of dominant firms.”

It is not clear whose undue influence he is referring to. Pharmaceutical companies? It has been well-established that undue political influence has emanated from some the largest technology businesses, frequently by way of lobbying and often directed at the White House. Their messages do not favor patent certainty, despite ownership of vast IP portfolios by many technology businesses. This paradox often perplexes astute lawmakers and investors. It should not.

With almost two-thirds of 2021 U.S. patent grants deemed software-related, how can invention occur without proprietary rights? How can those outside of the biggest and strongest companies compete? The difference with large, successful companies is that they have going for them what SMEs and inventors to not: brand recognition, capital, distribution, and an established customer base. They can afford to make available what other businesses might wish to remain private, at least for a few years.

Strong technology brands often obtain significant numbers of patents, but they do not rely on them the way many smaller competitors do (or used to do). They don’t have to. Patents were the anti-monopoly monopoly; the great leveler of the playing field that even the most powerful players had to respect. Today, patents are less of a defensive mechanism, and, therefore, less of a monetization tool.

“…the long arc of history has shifted this balance in favor of private incentives at the expense of diffusion, and it has shifted the ownership of those incentive rights from individual inventors to firms. Now, at a time when large firms are slowing the diffusion of technology, the tension between incentives and diffusion appears distinctly out of balance.”

Intelligent Innovation

Too many incentives? Too little diffusion? Surely intelligent innovation and IP policy can arrive at a balance.

“…it appears that courts, regulators, and legislators have tended to adopt the corporate view on these issues. That is, there appears to be cultural or ideological capture whereby policy makers have tended to acquire the outlook and concerns of the most-affected corporate interests. Policy makers and courts tend to privilege IP owners over the broader needs of society.”

“…changes in antitrust, intellectual property, and employment policy can go a long way toward nudging firms to unbundle and to promote diffusion of technology… When firms profit by using proprietary technology to differentiate their products from rivals’ offerings, their interests are not necessarily aligned with the social good.”

Bessen is not wrong in saying that access to data and systems to collect and manage them more efficiently provide a distinct advantage and sharing them would improve the future of many businesses. But software is the future of invention. Preventing or even discouraging software ownership would be more dangerous than compulsory licensing. It would serve as a kind of eminent domain and would likely spread to other inventions, and possibly content and creative expression.

If they are smart, leading technology businesses will try to remain a few steps ahead of potential competition and share what makes sense for them, not play politics with an already misaligned IP system. The next big thing will likely come out of nowhere, as it typically does. Fear on the part of incumbent businesses stems from the fact that, however well-prepared, they are likely to be blindsided by a development they are unable to own or control. Voluntarily offering royalty-free licenses on selective inventions can benefit other businesses, innovation and society. Limiting proprietary rights will not.