When Will Productivity Gains Kick In?
The ongoing paradox is why productivity growth is slow despite breakneck tech advances
By Irving Wladawsky-Berger
“One of the most enduring macroeconomic puzzles of the last several decades is the pervasive slowdown in productivity growth across industrialized nations, despite breakneck advances in information and communication technologies (ICT) and electronics,” wrote economists Daron Acemoglu, David Autor, and Christina Patterson in their recent paper Bottlenecks: Sectoral Imbalance and the US Productivity Paradox.
Over the past three decades, IT and electronics patents and productivity growth have been moving in starkly different directions. The total number of patents granted between 1990 and 2010 grew from 99,000 to 208,000, with IT and electronics patents accounting for 80% of the increase. And, while total factor productivity (TFP) — generally considered the primary contributor to GDP growth — grew by 1.7% in the US between 1997 and 2005, its growth has been minimal, just around 0.4% since 2005.
Abundance or Exploitation?
“How can these facts be reconciled?” ask the authors. Are we in a “new age of abundance” driven by our increasingly smart machines? Or have we entered “an age of slower growth because the most-impactful technologies have already been developed and exploited?”
Their answer is to wait a while longer.
In a 2018 paper, The Productivity J-Curve, Erik Brynjolfsson, Daniel Rock, and Chad Syverson explained that historically transformative technologies — e.g., the steam engine, electricity, semiconductors, the internet, AI — are the defining technologies of their times. But, historical patterns tell us that there’s generally been a significant time lag between the marketplace acceptance of a potentially transformative technology and its broader impact on industries, economies, and societies.
While these technologies have great potential from the outset, realizing that potential requires major complementary investments including business process redesign; innovative new products, applications and business models; the re-skilling of the workforce; and a fundamental rethinking of the very nature of production.
In their new paper, Acemoglu, Autor, and Patterson offered a conceptual framework that explains the conflicting trends between technology advances and productivity growth: technological advances and innovation in any one industry rely on complementary innovations in the inputs and ideas from the supplier industries in their sector. [AI and Chat GPT may be accelerating the trend even faster than in the past.]
When innovation is unbalanced in a sector, it holds back aggregate productivity growth by creating innovation bottlenecks.
“Our conceptual framework thus emphasizes that a balanced distribution of technological advances across sectors is important for the viability of further innovations. … Our perspective also emphasizes how a more balanced distribution of technological progress (and research and development) can improve aggregate productivity performance.”
Bottlenecks to Growth
The authors illustrate how bottlenecks emerge, and how their resolution accelerates innovation and growth, with concrete examples from three relatively recent technologies: rechargeable batteries, transistors, and GPS.
Rechargeable batteries. High-energy rechargeable batteries have played a major role in powering laptops, personal mobile devices, electric cars, and other important 21st century innovations. Rechargeable batteries were a bottleneck technology until the 1970s because, for over one hundred years, the best available such batteries, lead-acid batteries, had low energy density, a slow charging rate, a short life cycle, and the potential for releasing explosive hydrogen gas while recharging. The rechargeable battery bottleneck was finally overcome by lithium-ion batteries, invented in 1973 and refined through the 1980s. In 1991, Sony began producing and selling the world’s first rechargeable lithium-ion batteries.
Transistors. Though the 1950s, electronic devices, telephone lines, radios, transmitters, audio amplifiers, and early computers, were bulky, fragile and slow because the electromechanical switches and vacuum tubes these products used were clear bottleneck technologies. The transistor, invented in ATT’s Bell Labs in 1947, was first commercialized in the 1950s with simple products like transistor radios. Transistors then enabled the rapid growth of computers starting in the 1960s. Moore’s Law succinctly captures the impressive advances of transistors and integrated circuits over the following decades. “The transistor is estimated to be the most-manufactured device in history, at 13 sextillion (1021) units to date, with billions more produced each day.”
Global Positioning System. “Traditional navigation was supplemented with radio-positioning systems in the 1970s, but these tools suffered from either poor accuracy or limited geographic coverage and hence did not penetrate beyond military and commercial shipping applications.” GPS was developed by the US military to overcome these shortcomings. GPS satellites were first launched in 1978 and opened to worldwide public use in 1983. Over the years, GPS has enabled a number of important innovations, including “precision agriculture, mining, and oil exploration; atomic-precision time information for synchronization of power transmission systems; remote surveying for geology and weather prediction; and innumerable consumer-facing services such as ride-hailing, targeted advertising, and object trackers.”
The authors developed a mathematical model to formalize their conceptual framework, namely that slow growth stems in part from an unbalanced distribution of innovation.
They then empirically validated their hypothesis using data from a variety of sources, including citation linkages, patents, and productivity growth data from 462 manufacturing industries and 42 non-manufacturing industries. Let me summarize their key findings.
· “An industry’s productivity growth is augmented by the mean productivity growth of its suppliers (measured by TFP or innovation) and, crucially, it is hampered by the variance of their productivity growth.”
· There is a powerful linkage between the innovativeness of a sector or firm and the imbalances it faces across its upstream (idea-supplier) sectors. The greater the dispersion, that is, the variance or spread of productivity growth among an industry’s suppliers, the more of a negative influence it has on that industry’s growth opportunities. Doubling the variance of the suppliers’ TFP growth in a sector is associated with a roughly 0.9 percentage point slower productivity growth for that sector.
· “The dispersion of TFP growth among key industries has increased significantly over the last several decades.” This higher dispersion essentially explains all of the aggregate productivity slowdown in manufacturing between the 1970s and 2007. For example, if the variance in TFP growth in manufacturing had remained at its 1977–1987 level for the subsequent two decades, US manufacturing productivity would have grown twice as rapidly in 1997–2007 as it did.
· “The variance of supplier TFP in manufacturing increased over this period both because lagging industries failed to grow and because leading industries pulled away from the rest.”
The fastest-growing industries, defined as those that had the largest impact on supplier TFP variance between 1997 and 2007 include electronic computers, computer storage devices, and semiconductors. Slow-growing industries that became the biggest bottlenecks over the same time period include petroleum refineries, pharmaceutical preparation, printed circuit assembly, and turbine generators.
And the most bottlenecked industries, defined as those that were held back by the uneven innovation across their suppliers, include surgical and medical instruments, gas engines, and industrial valves.
These empirical results hold for the entire economy. Within the manufacturing sector (where TFP is better measured) they are present in different subperiods and with alternative measures of productivity dispersion. “We also verify that these patterns are not driven by outliers, nor are they exclusively due to the rapid advances in computers and electronics sectors (though these sectors do play a central role in our results).”
Domino Effects
In addition, “we document analogous patterns using international data and establish that dispersion in productivity among key domestic and international supplier industries has also been a major impediment to productivity growth for several leading OECD economies.”
“We view our results as suggestive of a potentially important linkage between (endogenous) productivity bottlenecks and productivity growth. … Our analysis further suggests that, following major breakthroughs in sectors acting as bottlenecks, there should be an acceleration of both industry and aggregate productivity growth.”
This blog first appeared October 10 here.