Issue #7 May 2005

After the gold rush: Patents, speculators, and innovators

The first American patent boom

"The country is flooded with patent monopolies... a great number of law suits arise, which are daily increasing in an alarming degree, onerous to the courts, ruinous to the parties, and injurious to society... It opens the door to frauds, which have already become extensive and serious... speculation in patent rights has become a regular business, and several hundred thousand dollars, it is estimated, are paid annually for void patents, many of which are thus fraudulently obtained."

When Senator John Ruggles of Maine first made this observation in 1836, the United States patent office was still in its infancy. First established by the Patent Act of 1790, the patent board consisted originally of three men: Secretary of State Thomas Jefferson, Secretary of War Henry Knox, and Attorney General Edmund Randolph. Their charter, in accordance with the Constitution, was to grant patents to "promote the progress of science and useful arts." It was decided that a patent would be granted to any invention declared sufficiently novel by two of these three patent officials.

It didn't take long for Jefferson and company to realize that the granting of patents was a difficult and time-consuming business. As early as 1791, the backlog of patent requests was long, but the spare time of the appointed "examiners" was short; Jefferson, a very busy man, was soon forced to admit that it was beyond his abilities to perform the painstaking research required to determine an invention's novelty. The pressures to approve new patents became so great that Jefferson, with the Patent Act of 1793, essentially transformed the patent board into a rubber-stamp entity.

For the next 43 years, patents were granted to whomever had the wherewithal to pay the application fees. Questions of novelty or validity were left entirely to the courts to decide, with completely predictable results. Patents multiplied; whereas 55 patents were issued from 1790 through 1793, nearly ten thousand patents were issued between 1793 and 1836. Patent speculation became rampant, lawsuits proliferated, and no coherent patent policy ever emerged from the judiciary. Jefferson's total abdication of any oversight responsibilities of the patent office turned out to be one of his sorest regrets, and he never lived to see the Patent Act of 1836 that would ultimately correct his mistake.

This story should seem familiar to observers of today's software patent industry; nearly 180 years later, Senator Ruggles' lament still rings eerily true. Patents work by granting potentially lucrative legal monopolies, to provide inventors with the "necessary" incentive to invent—and when standards for patentability are liberalized, for whatever reason, patent speculation is the obvious and natural result.

State Street and the software patent boom

In 1998, a new boom in patent speculation was triggered in the software industry by another dramatic liberalization of the standards of patentability: the State Street decision.

There's plenty of insightful analysis about the State Street ruling. The effect of this ruling, boiled down to its essence: software algorithms, which were previously held to be patentable only in very specific circumstances, and business processes, which were previously not patentable at all, both became broadly patentable. Specifically, software algorithms became patentable in any case in which the algorithm could be applied in any novel way to produce "a useful, concrete, and tangible result," even if that "tangible result" were merely a new business process.

There are already examples of companies in the post-State Street world that now pursue software patent speculation as their primary business model. Acacia Research, once a technology incubator, is now a patent purchaser; they've been extremely successful recently in licensing extremely broad streaming media patents to companies who would rather settle than fight. Intellectual Ventures was founded in 2000 by former Microsoft CTO Nathan Myhrvold, and according to Newsweek, the company is staffed almost exclusively with patent attorneys, pursuing their business model of hoarding patents. Myhrvold goes so far as to say that "intellectual property is the next software" and is currently studying to pass the patent bar exam. Another company that Myhrvold co-founded in 2001, ThinkFire, is managed entirely by patent licensing experts. ThinkFire's business methodology is neatly illustrated for the curious.

It's clear that there's plenty of money to be made in today's software patent gold rush. If Senator Ruggles is to be believed, there was plenty of money being made by the patent speculators of the 1830s, too. Making money, however, is not the Constitutionally defined purpose of our patent system—and it's not at all clear how our current software patent policy "promotes the progress of science and useful arts."

Innovating versus "inventing"

Brad Smith and Susan Mann of Microsoft wrote a paper for the Winter 2004 issue of the University of Chicago Law Review entitled "Innovation and Intellectual Property Protection in the Software Industry: An Emerging Role for Patents." In this paper, they suggest that the development of the software industry in the United States was a direct result of the development of a strong intellectual property policy. In response, Mark Webbink, Senior Vice President and Deputy General Counsel of Red Hat, wrote a paper for the Duke Law and Technology Review, in which he argues that innovation in the software industry has actually taken place despite the American patent system.

It can be difficult to quantify the correlations between patents and innovation in a given industry, but there are some indicators that seem to offer insight; Webbink looks closely at a couple of these in his paper.

One potential indicator is the number of patents required to protect a given invention. Pfizer makes billions of dollars with Viagra®; Merck makes billions of dollars with Zocor®. These two drugs are protected by a single patent each. On the other hand, Microsoft has fourteen issued patents and two patents pending on the positioning and movement of a cursor. "Rather than producing broad innovations to advance the software industry," Webbink notes, "information is being sliced and diced to the point that every trivial combination or extension of prior software technology is being accorded the same protection as a groundbreaking drug."

Another potential indicator is the relationship between a company's spending in research and development and its volume of patent filings. "In the summer of 2004," Webbink observes, "when Bill Gates announced that Microsoft would be increasing its annual patent filings from 2,000 to 3,000 per year, it was notable that there was no corresponding 50% increase in Microsoft spending for research and development."

Two researchers, James Bessen of the Boston University School of Law and Robert Hunt of the Federal Reserve Bank of Philadelphia, performed a more detailed analysis of the effects of software patent policy in their two papers, An Empirical Look at Software Patents and The Software Patent Experiment. Their research suggests that since the 1990s, in cases where a company's software patents have gone up, their R&D spending has gone proportionally down. "All else equal," they contend, "increases in software patent share were associated with decreases in research intensity. This suggests that in the 1990s, software patents substituted for R&D."

So despite the claims of patent attorneys to the contrary, more software patents do not necessarily mean more innovation; more patents just mean more patents. As Webbink puts it, "the block of cheese is the same size; the slices are simply thinner."

Filing patents, right or wrong

The notion that software patents are significantly flawed is fairly widely held, even among some companies that pursue software patents aggressively.

Once upon a time, the Oracle Corporation was a particularly outspoken opponent of software patents. Their one-time patent policy stated: "Patent law provides to inventors an exclusive right to new technology in return for publication of the technology. This is not appropriate for industries such as software development in which innovations occur rapidly, can be made without a substantial capital investment, and tend to be creative combinations of previously-known techniques." It's unclear whether Oracle still holds this position.

Given Oracle's erstwhile disapproval of patenting "creative combinations of previously-known techniques," though, it seemed particularly ironic when Oracle received precisely this type of patent in 2004 for their "self service system for website publishing." This patent was, of course, reviled for its breadth, its obviousness, and its utter lack of novelty.

Even Microsoft could not have achieved its dominance in software had software patents been prevalent prior to 1994. In a strategy memo Bill Gates sent to Microsoft employees in 1991, he freely admitted as much. "If people had understood how patents would be granted when most of today's ideas were invented, and had taken out patents, the industry would be at a complete standstill today," Gates said. Note, however, that this did not prevent Gates from using the flawed software patent system to his advantage as soon as he understood its implications. "The solution to this," he claimed in the memo, "is patent exchanges with large companies and patenting as much as we can."

So if no one really likes software patents much, and if so many people recognize the potential for harm in them, then why do so many companies pursue them? The answer, sadly, is that without significant changes to the current software patent regime, many companies feel that they simply have no choice.

Who loves software patents?

Clearly, our current software patent policy wouldn't exist if it didn't somehow benefit somebody. If software patents were regarded as a universal evil, they would already have been legislated out of existence. So who benefits from software patents?

Some of the biggest winners, according to Besson and Hunt, could be large companies with waning R&D productivity. "By the mid 1990s, software patents became a relatively inexpensive way to expand patent portfolios. This may have increased the attractiveness of a strategy that emphasizes patent rights over a strategy based on R&D. Such a change in strategy would be particularly attractive to mature firms if their R&D labs are not as productive as they once were," they hypothesize.

Other clear winners are patent attorneys, for whom the State Street ruling was a clear windfall. Patent attorneys make money when people sue each other over patents, or even when people merely threaten to sue one another over patents. It's also worth noting that the Patent Public Advisory Committee of the USPTO, a body "chosen to represent the interests of the diverse users of the USPTO," is comprised mainly of experienced patent attorneys, with no representation from the software industry at all. In fact, one of its members, Gerald Mossinghoff, is an advisory board member for ThinkFire, one of the aforementioned patent speculation companies.

Patents and copyright: Two models of protection

Here's what the Constitution says, specifically:

"The Congress shall have power... to promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries."

These Constitutional distinctions between "authors and inventors" and "writings and discoveries" are crucial. Because of this distinction, there are two completely distinct mechanisms for promoting the progress of science and useful arts: patents for "inventors" and "discoveries," and copyright for "authors" and "writers."

For years, software had been considered expression, and therefore adequately protected by the laws of copyright. In 1994, the USPTO held public hearings to explore "Use of the Patent System to Protect Software Related Inventions." Many leaders of the software industry were present.

Jim Warren of Autodesk said, "the system was not broken when there were no software patents."

Douglas Brotz of Adobe said, "the proper form of protection for software is copyright."

William Neukom of Microsoft admitted that "copyright has been and is an important and effective tool for the software industry," although he went on to add "that does not mean that there is no role for patent protection."

There are many proposed remedies to the current software patent dilemma, agreed upon by industry observers on all sides of the issue. In fact, Mark Webbink of Red Hat and Brad Smith of Microsoft agree on a number of measures that could be taken immediately to mitigate some of the brokenness of the software patent system: post-grant oppositions, third party participation in the patent examination process, and so forth.

Ultimately, though, patching broken parts of a software patent process is solving the wrong problem. As Doug Brotz from Adobe put it in those 1994 hearings, "software should not be patented, not because it is difficult to do so, but because it is wrong to do so."

After the gold rush

The Patent Act of 1836, which reintroduced the examination process and established the foundation for the modern American patent system, was passed on July 4, 1836. On July 13, 1836, US Patent Number One was issued, for traction wheels, to John Ruggles of Maine—the senator who wrote the Patent Act of 1836.

It's not likely that we will soon see an American politician so personally motivated to fix a broken patent system, as Senator Ruggles was. There aren't a lot of politicians writing code, and there aren't a lot of coders running for office.

There are clearly fortunes to be made from software patents—even if these fortunes wind up exclusively in the hands of large corporations, patent speculators, and intellectual property attorneys. Regardless of who gets rich from the software patent gold rush, though, America will eventually be forced to decide whether America's software patent policy really accomplishes its true purpose—innovation. Moreover, we will be forced to answer this question in a much more competitive and innovative world, in which one person's software innovation can be extended and improved upon by any person, anywhere, immediately.

The world will continue to innovate at an ever-accelerating rate—and if American patent policies pose barriers to innovation, then a handful of Americans will get rich, Americans will have less freedom to innovate, and the world will choose increasingly to innovate around us.

About the author

Greg DeKoenigsberg is the Community Relations Manager for Red Hat.