US Supreme Court rules against software patents

It all started in the 1950s. The 1950s was a cyclical trough in the scope of patent eligibility. You can look at graphs of both the strictness of Supreme Court decisions regarding patents and patent rejections by the USPTO, and they peak mid century.

The 1952 Patent Act was principally written by [future] Judge Giles Rich. In 1956 he was nominated for a seat on the Court of Customs and Patent Appeals. He then spent four decades on the bench protecting and expounding upon his invention, expanding the scope of eligibility. Having been the author of the bill, his judicial opinions (even when in dissent) were extremely influential. You can trace the SCOTUS decisions in cases like Chakrabarty, Diehr, and State Street Bank to his opinions and arguments.

So the scope of patent eligibility was driven by both an overall change in business attitudes, as well as by the legal academy, with Rich and a few others in the vanguard. But of course it took awhile for their labor to bear fruit.

It was an ideologically driven movement; an ideology regarding the interplay of capital, property rights, and technological development. (And arguably a cousin to the law and economics movement.) All the concepts law students learn in patent law classes can be traced back to Rich and his ilk, because the truth of the matter is that before that era the conceptual underpinnings to patent law were never as sophisticated--or contrived--as they are today.

Like most ideologically driven movements, they never cared much about empirical evidence, as they were convinced about the validity of their concepts. Hopefully we're finally waking up not only to the obvious costs, but as well as to the flaws in modern arguments for patent monopolies.