By Andy Reinhardt A proposed law to create Europewide rules for patenting software was soundly defeated July 6 in Strasbourg, France, after five long years of wrangling among the European Commission, the European Parliament, and aggressive lobbyists. In the wake of "no" votes in France and the Netherlands on the proposed European Constitution, and a bitter rift over the European Union budget in Brussels, the patent fiasco fuels a deepening sense of crisis in Europe.
In the end, nearly everyone voted against it (648-14) -- even the proposal's original backers, because no fewer than 170 amendments had so confused matters that initial supporters preferred no law to a bad law. Industry groups that had pushed for uniform software patents in Europe, such as the Microsoft-backed Computing Technology Industry Assn., the Business Software Alliance, and the European Information & Communication Technology Assn., put on a brave face, noting that the defeat prevented "harmful amendments" from being enacted.
SPECIAL INTERESTS' SABOTAGE. The sad truth is that everybody lost. Europe still has no clear solution to handling one of the most strategic and vexing issues of the Information Age: In what circumstances and on what terms should software innovations be patentable?
Many Europeans had hoped legislators would find an approach that avoided the worst excess of the U.S. system while affording appropriate protection to European innovators. Instead, the process was hijacked by interest groups whose intransigence and ideology scuppered intelligent compromise.
The biggest hypocrisy is that everyone claimed to be acting in the interest of innovation and small businesses. Now, with no new rules to go by, businesses large and small don't know what will happen to software patents they've already been granted by national authorities or the European Patent Office -- and which, in some cases, they have already used as the basis for global cross-license deals.
GOOD INTENTIONS. Moreover, they're unsure whether future Europe-based R&D will be subject to appropriation by other companies in the region. And especially for resource-strapped small companies that want to protect inventions, obtaining patents in Europe and elsewhere will be as difficult and expensive as ever. This is no way to encourage innovation.
The effort to craft an alternative in Europe started out well. From the beginning, most parties agreed that it should be possible to patent an invention -- a new antilock-braking system, say, or a blast-furnace control module -- that contained embedded software. But consensus broke down when people started discussing whether a piece of software that runs on a PC, or even an algorithm by itself, could be granted patent protection.
In one camp were patent attorneys, some policymakers, and large multinationals such as Microsoft (MSFT), SAP (SAP), and Nokia (NOK). They generally support software patents as a means to protect innovation by letting companies reap rewards from their R&D investments.
PHILOSOPHICAL OBJECTIONS. Of course, many large companies have already obtained hundreds or thousands of patents in the U.S., Asia, or individual European countries. They also have the financial means to defend their patents -- and use them to keep competition at bay.
On the other side were many smaller outfits struggling to innovate without running afoul of patents. These were joined by some economists who question the value of patents in spurring innovation. A 2004 study by Boston University visiting professor James Bessen and Robert M. Hunt of the Federal Reserve Bank, for instance, found that the existence of software patents can actually reduce R&D intensity.
But the most vociferous opponents were backers of open-source software, for whom patents are anathema to their philosophy and business model. They issued fiery manifestos proclaiming software patents "a danger to democracy" and a threat to Europe's position as a hotbed of open-source software development.
VAGUE AND CONTRADICTORY. The proposed legislation on the issue bounced back and forth for years through the halls of European government.
In September, 2003, the EU Parliament passed an amended version of the law that upended the Commission's original intent and drew such a tight limit on software patents that the Commission refused to accept it (see BW, 12/23/03, "Crafting a Better Patent Law").
After a retooling, it went back to the Parliament again -- and this time, the amendments were so numerous, vague, or contradictory that almost every thumb was pointed down. Each side pushed too far, and both lost.
TOO BROAD A BRUSH. Europe is back to square one, only worse. While the rival camps have bickered, the European Patent Office and its national counterparts have issued tens of thousands of software patents.
With little legislative guidance, and assuming that some sort of Europewide rules eventually would come into effect, they granted patent protection to ideas as diverse and far-reaching as a technique for audio compression that includes the MP3 format, a technology for video-on-demand transmission of live events, and a scheme for adding tabs to Web pages. Clearly, it's in the interest of society and the technology sector that such broad, and likely indefensible, patents not be enforced.
What's frustrating is that Europe was really onto something in questioning the value of software patents. As the evidence from the U.S. and other countries with liberal patent regimes suggests, patents can be a Pandora's Box -- one whose contents choke software innovation by miring programmers in endless legal battles.
CONFUSION TRIUMPHS. Given that perspective, the failure of European legislators (and lobbyists) to craft a compromise verges on the tragic. They could have started by agreeing that embedded software deserves protection, which would have helped thousands of small manufacturers whose innovations could now be subject to more skepticism if the key breakthrough is achieved through programming.
Likewise, a clear procedure for peer review, rapid challenges, and quick publication of patents would have created a model system the U.S. might have emulated. Last but not least, Europe could have gone on record saying that some kinds of software, including general-purpose programs and algorithms, are not patentable.
Instead, we've got: a European Patent Office whose mandate is confused and whose 30,000-plus software patents may be slowly overturned by European courts, a patchwork of national regulations that protect only applicants filing within the jurisdiction, and the possibility that some of Europe's largest and most successful tech companies will see the market value of their patent portfolios collapse.
NO GAIN WHATSOEVER. What of the small European businesses and innovators that opponents wanted to protect? They must now hire lawyers and pay fees to file patent claims in 25 different countries to gain protection just within the EU. Or they can try for a patent from the EPO, but that still requires costly translations and country-by-country enforcement. They're also no better protected against counterclaims by rival U.S. or Asian patent-holders in Europe than they were before.
The outcome in Europe entrenches a system that, like many national laws on the Continent, is increasingly out of step with global realities. The people who killed smart patent rules may have had noble intentions. But instead of scoring a hit for the little guy, they've actually put their compatriots at a disadvantage on the global stage. That hardly looks like progress. Reinhardt is a BusinessWeek correspondent in Paris