Paying for Patents
Why we should tax patents the same way we tax real estate.
IN 1987, IN THE EARLY YEARS OF THE INTERNET, CompuServe created a picture format called GIF. Simple to use and easy to create, GIF files exploded in popularity along with the web in the early 1990s. (If you're reading this online, the Legal Affairs logo you see in the upper left is a GIF.) But the technology turned out to conceal a barbed hook, one revealed only after the web came to rely on the GIF format.
One reason GIFs are so useful is that they're tiny, thanks to a file compression algorithm called LZW, which, unbeknownst to CompuServe, had been patented by another company in 1985. Software was rarely patented in those days and, given how similar LZW was to other compression technologies, CompuServe apparently assumed it hadn't been patented.
Six years after CompuServe rolled out the GIFs, Unisys, the Pennsylvania computing company that controlled the LZW patent, came knocking. CompuServe and the other graphics software publishers that relied on the program, including Adobe and Microsoft, were startledbut had no choice but to pay. From 1994 until the patent's expiration in 2003, Unisys imposed on the publishers a fee of up to $10 for each copy of GIF-creating software, which was indispensable to anyone who wanted his web pages to look better than a mimeographed copy. Any graphic artist depending on Photoshop, any small online businessthey all paid more for their software as a result of LZW's licensing fees. Because the patent had stayed hidden so long, the web had come to depend on a proprietary technology without realizing it.
Creators of GIFs joined others who found themselves ambushed by an unexpected patent claim. In the 1990s, Ford, General Motors, Chrysler, and other manufacturers had to pay compensatory damages of $1.5 billion for infringing on a bar code scanner patent, years after they had built scanners covered by it into their assembly lines. Because the patent system encourages such "submarine patents" by rewarding companies that fail to publicize their inventions, many more firms seem likely to face the same liability as the software publishers and the automakers.
The problem is serious, but there is a solution. Just as Bill Gates pays abnormally high property taxes on his $100 million Lake Washington estate, patents should be taxed according to what they're worth. Besides making future patent surprises less likely, such taxes would raise money that could help ensure the future of American innovation.
AS IT STANDS NOW, THE COSTS OF MAINTAINING A PATENT differ from the costs of maintaining most other physical things. A Hummer generally costs more to maintain than a Prius, and a Park Avenue penthouse requires more upkeep than a Texarkana trailer. But regardless of how much a patent is worth, maintaining it requires three periodic payments , totaling $7,000, over the patent's 20-year lifes pan. Whether the invention is a catfish lure or a piece of software at the heart of the web, the inventor has to pay the same fee. With government subsidies encouraging research that yields patents, it's not surprising that companies acquire plenty of patents and hold onto them, even if they're not utilized.
Consequently, it's easy to unknowingly and unintentionally infringe on someone else's patent. Although CompuServe relied on a presumably competent legal team, the company had no idea it was infringing on Unisys's patent for the first eight years following CompuServe's introdution of GIFs. Suppose a submarine required no fuel or upkeep. A reasonable strategy would be to keep it submerged at sea indefinitely, waiting for an enemy ship to pass overhead. Unisys employed a similar strategy: Given the low cost of holding on to the patent, Unisys had limited incentive to seek all profitable uses for it right away. But suppose that there were taxes on patents. This would encourage patent holders to think of their patents less as unlimited claims on an idea and more like any finite tool, subject to wear and tear. Every time that tax bill came due, the owners of the invention would rethink whether they were doing what they could to exploit it, if they should sell it to someone who would, or whether they should transfer it into the public domain.
The absence of a patent tax creates other unfortunate outcomes. With the current system, it's common for companies to enter into patent "arms races," spending time and effort to nail down patents on inventions they don't need and won't use. This occurs most often in specific technology industries, such as those creating semiconductors or software, where patents are so intertwined that companies must often engage in cross-licensingI'll give you access to some of my patents and pledge not to sue you if you give me access to some of yours and agree not to sue me, eitherin order to avoid legal entanglements.
To participate in this game, you need to control a portfolio of patents. The software maker Oracle has opposed applications for patents by other software firms and its lawyers have said that it "has been forced to protect itself by selectively applying for patents which will present the best opportunities for cross-licensing between Oracle and other companies who may allege patent infringement." By making it expensive to hold on to large patent portfolios, a tax would discourage companies from acquiring unneeded ones and would calm some of the frenzy of these patent races.
MOST PROPOSED SOLUTIONS FOR THE PROBLEM of submarine patents focus on legality, not economics: shortening the amount of time that patents last, hiring more and better-trained patent examiners, even banning patents wholesale in certain fields like software.
But none of these solutions gets to the root of the problem. Intentionally or not, it can be very profitable to hide, mothball, or neglect patents instead of putting them to use. The other proposals don't provide one of the benefits that a patent tax would: putting money into the system and, in an equitable way, helping fund the research that often lies behind patent discovery.
Despite the appeal of an Edisonian idyll where an inventor is struck by genius while toiling alone in his workshop, most valuable inventions today are the products of extensive and expensive teamwork. Take the pharmaceutical industry. The National Institutes of Health's $28 billion budget helps build, equip, and staff the corporate laboratories of big pharmaceutical companies that concentrate resources on producing patents. The research team at Abbott Laboratories that developed Ritonavir, a patented drug that dramatically boosts the effectiveness of HIV treatments, was initially organized in 1988 under a $3.5 million NIH grant. According to a company official, "Abbott believed at the time that the market . . . was too small to justify a major R&D expenditure." In other words, no government funds, no Ritonavir.
These are worthwhile investments that government should make. Since creating innovative products and putting them to use leads to higher tax receipts, they probably return more money to government coffers than they cost. Under a new patent tax system, the companies that benefit the most would return the most to the national treasury.
ONE CHALLENGE TO PUTTING THIS IDEA INTO PRACTICE involves assessment: How do we decide what each patent is worth? But there's a compelling response. Every patent holder should value his own possession.
Here is how the system might work. Those with the most information about a patent will be asked to gauge its worth, eliminating the cost of government assessors. If you believe someone has infringed upon your rights, you can sue or use the threat of a lawsuit to challenge and, if necessary, change his behavior. You will, however, be held to your self-assessment. If you claim a modest value to reduce your tax bill, you limit the amount for which you can sue, and you invite others to infringe. If you choose a large value to maximize your leverage in lawsuits and scare off trespassers, you pay for it with increased taxes. Self-assessment works in other instances, like in setting the coverage of an insurance policy, and for similar reasons. If you avoid high premiums by insuring your home for less than its value, you'll be in real trouble if it burns down.
Suppose Goliath acquires a patent on slingshots without intending to put it to use. If his self-assessed value for the patent is low because he has no plans to market slingshots, he'll be discouraged from claiming high damages if David infringes on it. If his valuation is steep solely as a barrier to potential infringers like David, Goliath will pay the consequences on April 15. Such binding valuations will also make it easier for a judge who knows nothing about slingshots to rule a lawsuit frivolous when that's warranted. If Goliath spends millions to pursue a claim that the giant himself has valued at $10,000, the judge will have reason to be skeptical.
Would this system destroy the hopes of the next Edison, working away by lamplight, sending in patent applications that he can't afford to pay taxes on if they're granted? Not likely. If his patent isn't valuable, high taxes won't stretch him. If it is valuable, he's likely to reap the benefits in the marketplace, or even through a loan, and he could leave behind working by lamplight. Property owners may pay property taxes while holding their noses, but they realize that these taxes help to pay for fire departments, sewers, and other services that protect their investments. Patent owners should start thinking of their intellectual property that way, too.