Are Innovation and R&D Yielding Decreasing Returns?
Given the pace of technological change, we tend to think of our age as the most innovative ever. But over the past several years, a number of economists have argued that increasing R&D efforts are yielding decreasing returns.
Has the ideas machine broken down?, asked The Economist in a January, 2012 article that examined the growing concerns that we may be in a long-term period of slow innovation despite our rapidly advancing technologies. “With the pace of technological change making heads spin, we tend to think of our age as the most innovative ever,” said The Economist. “We have smartphones and supercomputers, big data and nanotechnologies, gene therapy and stem-cell transplants. Governments, universities and firms together spend around $1.4 trillion a year on R&D, more than ever before.” But, perhaps these don’t quite compare with modern sanitation, electricity, cars, planes, the telephone, and antibiotics. These innovations, first developed in the late 19th and early 20th century, have long been transforming the lives of billions.
In a September, 2012 paper, Northwestern University economist Robert Gordon questioned the generally accepted assumption that economic growth is a continuous process that will persist forever. He wrote that the slow growth we’ve been experiencing in the US and other advanced economies isn’t cyclical, but rather evidence that long-term economic growth may be grinding to a halt.
The rapid growth and rising per-capita incomes we experienced at the height of the Industrial Revolution, - between 1870 and 1970, - may have been a unique episode in human history. Since the 1970s, US productivity and income growth have dipped sharply except for an Internet-driven productivity boost between1996 and 2004. Innovation may be hitting a wall of diminishing returns. There was little growth before 1800, and there might conceivable be little growth in the future.
A similar pessimistic view was expressed by George Mason University economist Tyler Cowen in his 2011 book The Great Stagnation. According to Cowen, over the past two centuries the US economy has enjoyed lots of low-hanging fruit, including a vast, resource rich land, waves of immigrant labor, access to education and the technological advances of the Industrial Revolution. But, Cowen believes that we are at a technological plateau, and wonders whether long term growth is still possible because the supply of low-hanging economic fruit is nearly exhausted.
Are Ideas Getting Harder to Find?, - a recent paper by economists Nicholas Bloom, Charles Jones and Michael Webb from Stanford and John Van Reenen from MIT, - shows that, across a wide range of industries, research efforts are rising substantially while research productivity is declining sharply.
“Economic growth arises from people creating ideas,” they write in the paper’s introduction. “As a matter of accounting, we can decompose the long-run growth rate into the product of two terms: the effective number of researchers and their research productivity. We present a wide range of empirical evidence showing that in many contexts and at various levels of disaggregation, research effort is rising substantially, while research productivity is declining sharply. Steady growth, when it occurs, results from the offsetting of these two trends.”
Moore’s Law, - the empirical observation that the number of transistors in a computer chip doubles approximately every two years, - illustrates these trends. The paper points out that the number of researchers required to double chip density today is 18 times larger than those required in the early 1970s. In the case of Moore’s Law, research productivity has been declining at a rate of about 6.8% per year.
The authors conducted a similar in-depth analysis in the agricultural and pharmaceutical industries. For agricultural yields, research effort went up by a factor of two between 1970 and 2007, while research productivity declined by a factor of 4 over the same period, at an annual rate of 3.7 %. For pharmaceuticals, research efforts went up by a factor of 9 between 1970 and 2014 while research productivity declined by a factor of 5, - an annual rate of 3.5%.
“Our robust finding is that research productivity is falling sharply everywhere we look. Taking the U.S. aggregate number as representative, research productivity falls in half every 13 years - ideas are getting harder and harder to find. Put differently, just to sustain constant growth in GDP per person, the U.S. must double the amount of research effort searching for new ideas every 13 years to offset the increased difficulty of finding new ideas.”
Why is research productivity declining, despite our increasingly advanced technologies? Let me briefly discuss two intriguing, potential answers to this important question.
In a 2009 paper, The Burden of Knowledge and the ‘Death of the Renaissance Man’: Is Innovation Getting Harder?, Northwestern economist Benjamin Jones argues that “If innovation increases the stock of knowledge, then the educational burden on successive cohorts of innovators may increase.” His theory is based on two simple observations.
First, “innovators are not born at the frontier of knowledge.” They must undertake considerable education to reach the frontiers of knowledge where the majority of innovation takes place. Individuals can only absorb knowledge at a limited rate, so their education occupies considerable time and a significant portion of their lives.
Second, the stock of knowledge has been rapidly expanding across most disciplines, over the past 150 years, and particularly over the past several decades. If reaching the frontiers of knowledge requires standing on the shoulders of giants, “one must first climb up their backs, and the greater the body of knowledge, the harder this climb becomes.”
Innovators can compensate for this increasing knowledge burden in two key ways. They can choose to learn more, thus continuing to lengthen their education. Or they can become more specialized, narrowing their area of expertise and forcing them to work in teams of innovators with complementary specialized expertise.
Jones presents evidence that both, longer educational periods and greater specialization, are actually happening. PhD’s have been taking longer in most fields, and additional postdoctoral training is often required for leading-edge academic and research positions. Analysis of a rich patent data set shows that the age of first patent has been increasing over time at a substantial rate. A similar analysis also shows that more and more research is being conducted by teams, and the size of the teams has been going up over the years. He further shows that teamwork and specialization are greater in fields with deeper knowledge.
Such a knowledge burden mechanism helps explain why productivity rates have not grown over the past several decades despite the large expansion of the overall research effort, - with potentially negative implications for long-run economic growth. It further suggests that the very nature of innovation is changing.
A second explanation for the decline in productivity growth is offered by MIT’s Erik Brynjolfsson and Andy McAfee in their 2014 article The Innovation Dilemma: Is America Stagnating?. The innovations in our digital age are fundamentally different from the physical industrial age innovations of the past two hundred years. As a result, the digital revolution is far from complete even decades after its onset, and its transformative impact on productivity, the economy and society is still decades into the future.
Digital innovation is recombinant in nature, based on building blocks and platforms. In this view, “the true work of innovation is not coming up with something big and new, but instead recombining things that already exist. And the more closely we look at how major steps forward in our knowledge and ability to accomplish things have actually occurred, the more this recombinant view makes sense.
As examples of recombinant innovation, e-mail and the Web were enabled by the Internet. Then came Web 2.0 which led to social networks and major social media applications. Cloud computing, the mobile Internet, and the Internet of Things have followed, leading to major new initiatives including smart cities and digital money. Data Science and AI have been emerging to take advantage of the big data being generated by all these digital devices, systems and applications, promising to usher an information-based revolution in a number of disciplines and human endeavo
“This progression drives home the point that digital innovation is recombinant innovation in its purest form,” write Brynjolfsson and McAfee. “Each development becomes a building block for future innovations. Progress doesn’t run out; it accumulates. And the digital world doesn’t respect any boundaries. It extends into the physical one, leading to cars and planes that drive themselves, printers that make parts, and so on.”
It takes considerable time, - often many decades, - to translate the benefits of technological innovations into productivity and economic growth. Doing so requires complementary innovations and investments, the reinvention of our economic and societal institutions, as well as new industry infrastructures, government policies, and educational opportunities. This was the case with the Industrial Revolution, where the biggest productivity benefits took place between roughly 1870 and 1970.
We’re living in a time of transformative technology advances, - the Internet, smartphones, cloud, IoT, AI - whose deployment and impact on economic and productivity growth are still lagging. Moreover, the more transformative the technologies, the longer it takes for them to be embraced by companies and industries across the economy. We’re still in a long in-between period, where multiple, overlapping technologies are continuing to emerge from R&D labs into the marketplace, but because of their profound transformative nature, their full deployment is still ahead of us.