Why has the software industry seen an explosion of litigation lately? The courts played a big role, by making it easier to get patents in the 1980s and 1990s. But as this chart shows, the pendulum started to swing back in the opposite direction a few years ago:

Source: Searches of legal databases by Dennis Crouch
Source: Searches of legal databases by Dennis Crouch

The chart comes courtesy of Dennis Crouch, a law professor at the University of Missouri. He originally posted it in October. Above is an updated version that accounts for all of 2012.

It shows the number of times courts have cited Parker v. Flook, the 1978 case in which the Supreme Court took its strongest stance against patents on software. The case focused on using a computer to update the "alarm limit" for a chemical process. The court held that the patent effectively claimed a mathematical algorithm, which is too abstract to be eligible for patent protection.

The ruling has never been overturned (though its holding was narrowed by a 1981 ruling). And since all software consists of mathematical algorithms, some software patent opponents cite it as holding that software cannot be patented.

But the late 1970s proved to be the high-water mark for judicially-enforced limits on what could be patented. During the 1980s and 1990s, both the United States Patent and Trademark Office and the Federal Circuit Appeals Court (which hears appeals in patent cases) became more permissive. The result was a gold rush, as software, biotech, financial services, and other industries rushed to patent technologies that had previously been regarded as off-limits.

But then these firms began suing each other with their newly minted patents. According to one estimate, patent litigation now costs the economy tens of billions of dollars a year.

Perhaps spurred by bad media coverage over patent litigation, the courts began scrutinizing patents more closely. Christina Mulligan, a legal scholar at Yale Law School, points to two 2007 decisions by the Federal Circuit that invalidated patents for being outside the bounds of what could be patented, something it had not done in more than a decade.

The Supreme Court got into the act too, invalidating one patent in 2010 and another in 2012.

According to Mulligan, the courts had been so hostile to arguments that a patent was out of bounds that defendants didn't even bother raising them before 2007. But once the courts signaled that they were open to those arguments again, defendants began to raise them.

The results can be seen in the chart above. Not every citation of Flook means that a patent was invalidated. But it's at least a sign that the courts are wrestling with the limits the high court articulated more than three decades ago. And in the long run, stricter rules about what can be patented may reduce the amount of litigation that currently afflicts the economy.