Since the mid-20th century, conventional wisdom and prevailing economic theory have held that innovation is one of the main engines of GDP growth. In theory, if firms invest in research—and develop new products that increase revenue—then the growth rate for GDP will improve. And indeed, for many years the R&D-GDP link tracked in tandem across economic data charts. But in the early 1980s, the two parted ways: R&D spending recovered from a decline, but the GDP growth rate failed to follow the rebound.

video

What happened? The leading theory to explain the disconnect holds that R&D has gotten harder. Another theory suggests that firms have gotten worse at doing research and development. Neither theory suggests a solution to repair the R&D-GDP link. Anne-Marie Knott’s research offers a third explanation that provides unique insight into this broken link.

“The principal way that economists and managers have traditionally measured R&D output is by the number of patents a company holds,” explains Knott. “But only 10 percent of patents comprise 85 percent of the total value of all patents. And 50 percent of firms don’t even patent their R&D.” In other words, patents are an irrelevant measure of productivity.

In her research into innovation, firm strategy, and how they generate economic growth, Knott has developed a measure called the Research Quotient (RQ), which determines the value of productive R&D as it relates to a firm’s growth, profit, and market value.

Knott’s RQ measure is essentially a ratio of a firm’s revenues to its R&D investment. As a measure of innovation, it is universal: it can be calculated for any firm in any industry that conducts R&D. And her RQ measure is uniform because it relies on firms’ financial data rather than units such as patents.

In examining data for hundreds of firms going back to 1957 and calculating annual RQ measures for each firm, Knott made several interesting discoveries:

• There is no evidence to support the theory that R&D has gotten harder.
• Between 1957 and 2007, overall R&D productivity in the United States declined 65 percent.
• During the same period, R&D outsourcing rose dramatically.

Knott zeroed in on outsourcing as the culprit that broke the R&D-GDP link. She found that in the mid-1980s, firms slowed expansion of internal R&D departments in favor of shifting research to outside firms. In so doing, Knott argues, the funding firms lost valuable knowledge that is typically gained through the trial and error of the R&D process. That kind of institutional learning is often redeployed and applied elsewhere in a firm, leading to unexpected, but productive, ends. Knott concludes that firms are relinquishing potential assets when they outsource R&D and reaping zero returns.

Note: In Knott’s paper, “outsourcing” is limited to other US firms and does not include R&D services provided by universities or foreign firms.

“One of the things we know about R&D is that a lot of it ends in failure, much like entrepreneurial ventures,” explains Knott. “The Industrial Research Institute estimates that it takes 3,000 raw ideas before you have one commercial success, but the benefit is you are learning something along the way.”

Knott cites several examples in her paper, including a recent study of the banking industry’s adoption of Internet banking. The firms that outsourced the initial IT integration are less able to develop new applications and, accordingly, have lower revenues from their Internet operations.

In short, it’s not about spending more on R&D. Instead, it’s about keeping R&D in house to be more efficient and thus boost productivity and profits. “Reversing the R&D outsourcing trend,” Knott advises, “could repair the broken R&D-GDP link and eventually lead to a permanent increase in economic growth.”

CNBC created a new ranking of the most innovative US companies based on Anne-Marie Knott’s RQ metric and research. Read the CNBC RQ 50.