“…challenges misconceptions that distort our economic debates.” - Arthur Brooks, President of the American Enterprise Institute
“…reminds us that inequality sends a signal of what society lacks most, in America’s case, entrepreneurship and risk taking.” - Lawrence Lindsey, CEO, The Lindsey Group, former Director of the National Economic Council
“…a fresh argument for the productive value of inequality.” - David Autor, Professor of Economics, Massachusetts Institute of Technology
“…a comprehensive explanation of the modern economy.” - Julian Robertson, Founder, Tiger Management
“…a must-read for serious students of economic policy.” - Glenn Hubbard, Dean, Columbia Business School, and former Chairman of the Council of Economic Advisers
“…a comprehensive explanation of the modern economy.” - Julian Robertson, Founder, Tiger Management
“Unintended Consequences is full of substance, it is one of the must-read books of the year, and once I finish it I will be giving it a second read through right away.” - Tyler Cowen, Professor, George Mason University
“Unintended Consequences provides a provocative interpretation of the causes of the global financial crisis and the policies needed to return to rapid growth. Whether you agree or not, this analysis is well worth reading.” - Nouriel Roubini, New York University; Chairman, Roubini Global Economics
“…a very valuable contribution.” - Larry Summers, former Secretary of the Treasury and director of the National Economic Council, president emeritus, Harvard University
“There are an amazing number of good ideas and interesting points made in Unintended Consequences. The thinking underlying it, and the obvious depth of understanding of the author, are very impressive.” - Steven Levitt, coauthor of Freakonomics; 2004 John Bates Clark Medal
“…serious thinking for serious thinkers. …a thought-provoking blueprint for growing middle- and working-class incomes.” - Mitt Romney, former Governor of Massachusetts
“…challenges misconceptions that distort our economic debates.” - Arthur Brooks, President of the American Enterprise Institute
AI Summary.An analysis of tens of thousands of responses on policy preferences and sociopolitical beliefs finds that widely used AI chatbots nudge users away from extreme positions toward more moderate, expert-aligned stances. The author estimates Grok guides conversations toward the centre-right, while OpenAI's GPT, Google's Gemini, and China's DeepSeek nudge toward a centre-left worldview.
John Burn-Murdoch Financial Times
Date Posted:
Is Database:
Database
Is Important:
Important
John Burn-Murdoch finds that while social media amplifies fringe views, LLMs “nudge people away from the most extreme positions,” suggesting AI could be “technocratising,” in contrast to “social media’s radically democratising influence.”
Do AI chatbots nudge users toward moderate political views?
Core argument: Across tens of thousands of responses on policy preferences and sociopolitical beliefs, the author finds that the most widely used AI chatbots nudge users away from the most extreme positions toward more moderate, expert-aligned stances.
I used a dataset of tens of thousands of responses to questions on policy preferences and sociopolitical beliefs to test whether and how the most widely used AI chatbots shape conversations about politics and society. The results strongly support the theory of AI chatbots as depolarising and technocratising. While different AI platforms behave in subtly different ways, all of them nudge people away from the most extreme positions and towards more moderate and expert-aligned stances. On average, Grok guides conversations about policy and society towards the centre-right — a rightward push for most people but a moderating nudge towards the centre for those who start out as conservative hardliners. OpenAI’s GPT, Google’s Gemini, and the Chinese model DeepSeek all exert similarly sized nudges towards a centre-left worldview — a slight leftward nudge for most people but a moderating push away from fringe leftwing positions. These are only the results of one analysis, and it is also possible that usage patterns and the models themselves will evolve in ways that deviate from the patterns I found. But at the very least, there is cause for optimism that the next information revolution may take us in a less corrosive direction than the last.
Takeaways by Macro Roundup® AI
Across tens of thousands of responses on policy preferences and sociopolitical beliefs, the author finds that the most widely used AI chatbots nudge users away from the most extreme positions toward more moderate, expert-aligned stances.
Grok on average guides policy and society conversations toward the centre-right, pushing most users rightward while moderating those who begin as conservative hardliners toward the centre.
OpenAI's GPT, Google's Gemini, and the Chinese model DeepSeek exert similarly sized nudges toward a centre-left worldview, slightly leftward for most users and away from fringe left-wing positions for others.
Related Articles:
The End Of The Gatekeepers— Social media has expanded who produces and broadcasts political information, and an analysis of the ideological positions of those sharing political content finds extreme views on both left and right are heavily over-represented relative to traditional media, where moderate positions were well-represented. The author argues this has eroded norms that kept radicals on the fringe.
China Punishes ‘Excessively Pessimistic’ Social Media Users— China's internet regulator has launched a campaign to remove online content that incites "excessively pessimistic sentiment," panic, or defeatist ideas such as "hard work is useless." Authorities have punished bloggers who advocated working less, an influencer who said forgoing marriage and children made financial sense, and a commentator who said China's quality of life lags Western countries.
Teens Used To Be Happier. What Changed?— The share of US high school seniors describing themselves as "very happy" peaked at 24.8% and remained around 24.4% before falling to 11.4% at its low, recovering only to 12.7%, well below the 17.9% recorded before the pandemic. The decline coincided with smartphones and social media becoming pervasive.
AI Summary.The annual compensation gap between top industry and academic artificial intelligence researchers widened from roughly $294,000 to $1.5m, coinciding with accelerating academic out-migration toward large incumbent technology firms. Researchers who move to industry cut paper production by about 63% while increasing patent filings nearly sixfold, consistent with a shift from open science toward proprietary innovation.
Ufuk Akcigit, Craig Chikis, Emin Dinlersoz, and Nathan GoldschlagUniversity of Chicago
Date Posted:
Is Database:
Database
Is Important:
Important
In 2001, 48% of AI researchers were in industry; by 2019—two years after the landmark transformer paper—68% were. Over this period, the compensation gap btw industry and academia for top 1% researchers widened markedly, from ~$294k to ~$1.5mm annually.
How does the pay gap shift AI research from publishing to patenting?
Core argument: The annual compensation gap between top industry and academic AI researchers widened from roughly $294,000 to $1.5 million, a fivefold increase since 2001, per a dataset linking publication records to administrative employer–employee data.
Using a novel dataset linking publication records to administrative employer–employee data, we document profound changes in AI talent allocation over two decades. The compensation gap between top industry and academic AI researchers widened from roughly $294,000 to $1.5 million annually—a fivefold increase since 2001. This divergence coincided with accelerating academic out-migration, particularly among early-career researchers, and an increasing concentration of talent flows toward large, incumbent technology firms rather than startups. These patterns reflect a structural reorganization of frontier AI research. Researchers who transition to industry reduce paper production by approximately 63% while increasing patent filings nearly sixfold, consistent with a shift from open science toward proprietary and commercially oriented innovation. Even among those who remain in academia, moonlighting has risen substantially. Frontier AI research has become increasingly capital intensive and concentrated within large firms. Universities may no longer be able to compete at the frontier of computational scale, but they continue to play a central role in preserving open science, training general human capital, and providing outside options in the market for research talent.
Takeaways by Macro Roundup® AI
The annual compensation gap between top industry and academic AI researchers widened from roughly $294,000 to $1.5 million, a fivefold increase since 2001, per a dataset linking publication records to administrative employer–employee data.
AI researchers who move to industry reduce paper production by approximately 63% while increasing patent filings nearly sixfold, consistent with a shift from open science toward proprietary, commercially oriented innovation.
Academic out-migration of AI researchers accelerated, particularly among early-career researchers, with talent flows concentrating toward large, incumbent technology firms rather than startups, while moonlighting rose among those remaining in academia.
Related Articles:
Ed Conard Debates Furman On “The Expected Value of Risk Taking”— A debate on U.S. fiscal policy argues that spending increases, rather than the tax increases needed to fund them, slow growth, and that higher taxes would reduce the expected value of America's innovative risk-taking. The argument holds that cutting debt below 100% of GDP is critical.
Never Enough: Dynamic Status Incentives in Organizations— Data on over 5,000 German fighter pilots show performance rose markedly as pilots approached the eligibility threshold for each medal in a tiered awards ladder and slumped immediately after receiving it, with the effect repeating at each successive rung. The authors estimate the ladder added 3,000–5,000 aerial victories, 5–10% of the total.
AI Summary.US labor productivity has grown at an average pace of 2.8% since the start of 2023, more than double the average rate for the decade through 2019. Economists caution that quarterly productivity figures are volatile and single readings carry limited information.
Enda CurranBloomberg
Date Posted:
Is Database:
Database
Is Important:
Important
American labor productivity has grown at an average pace of 2.8% since the start of 2023, twice the 2009–2019 rate.
Has labor productivity growth doubled its pre-pandemic pace?
Core argument: U.S. labor productivity rose 2.8% in the fourth quarter of 2025 and has grown at an average 2.8% pace since the start of 2023, more than double the average for the decade through 2019.
Labor productivity rose 2.8% [in Q4 2025]. Economists typically don’t read too much into one quarter, since the numbers tend to jump around. Still, the trend has been ticking up. After big swings in the pandemic period, productivity has grown at an average pace of 2.8% since the start of 2023. That’s more than double the average for the decade through 2019.
Takeaways by Macro Roundup® AI
U.S. labor productivity rose 2.8% in the fourth quarter of 2025 and has grown at an average 2.8% pace since the start of 2023, more than double the average for the decade through 2019.
Economists caution against reading too much into a single quarter of productivity data, since the figures tend to jump around, though the trend has been ticking up.
Related Articles:
Economic Growth Slowed in Fourth Quarter, Hurt by Government Shutdown— U.S. gross domestic product rose at a 1.4% inflation-adjusted annual rate in the fourth quarter, below the 2.5% economists surveyed expected, weighed down by a record-long government shutdown and slower consumer spending. Full-year growth was 2.2%, down from 2.4% the prior year and the weakest since 2022.
An AI Productivity Boom? Don’t Count Your (Productivity Data) Chickens— Measured productivity can rise from compositional effects when lower-wage, lower-productivity workers leave the labor force, as happened during Covid and 2009, and immigration policy changes could produce a similar effect. Much of the recent productivity boost appears to come from capital utilization tied to increased investment in AI rather than workers becoming more productive using AI.
Waiting for the AI J-Curve— Artificial intelligence has yet to appear in employment, productivity, or inflation data, or in profit margins and earnings expectations outside the largest technology firms. With fierce competition pushing large language model prices toward zero for users, value creation depends on sector-level implementation, and the author argues AI will likely enhance labor in some sectors rather than replace it across all.
AI Summary.The authors estimate that researcher efficiency, measured as the responsiveness of patenting to knowledge inputs, has been flat or rising over the last five decades. However, they find a negative time trend in firm growth conditional on patenting, indicating that growth, not ideas, is getting harder to find.
Teresa Fort, Nathan Goldschlag, Jack Liang, Peter Schott, et al.U.S. Census Bureau
Date Posted:
Is Database:
Database
Is Important:
Important
Efficiency in the production of “ideas,” as indicated by the elasticities of patent grants, external citations, and breakthroughs with respect to 6 measures of R&D input, is generally rising, but “growth, conditional on ideas, is getting harder to find.”
Are ideas harder to find, or growth from ideas?
Core argument: The authors estimate that researcher efficiency, measured as the elasticity of patenting to knowledge-input stocks, has been flat or rising over the last five decades.
[Our production estimation] demonstrates that researcher efficiency in terms of the elasticity of patenting to knowledge input stocks is flat or rising over the last five decades. It also shows that average patenting by semi-decade across firms, after controlling for knowledge-input use and firm fixed effects, declines for some measures of patenting but is flat or rising for others. Given these results, why has aggregate growth not kept pace with rising researcher efforts? To address this question, we examine the link between ideas and growth. We examine the link between patenting and firm growth, allowing the relationship to vary over time. [Equation 7, referenced in Figure 12, in which the left hand side variable is firm growth]. Figure 12 reports the semi-decade fixed effects fixed for these measures of US patenting activity across five samples of firms and for the same six knowledge inputs. These fixed effects capture average patenting in a particular sample and time period, after controlling for those firms’ average patenting and changes in knowledge inputs. We generally find a negative trend in these fixed effects over time, indicating that growth, conditional on ideas, is getting harder to find. The secular decline in these time effects might reflect an aggregate ‘crowding out’ of ideas as the pool of total ideas grow.
Takeaways by Macro Roundup® AI
The authors estimate that researcher efficiency, measured as the elasticity of patenting to knowledge-input stocks, has been flat or rising over the last five decades.
Average patenting by semi-decade across firms, after controlling for knowledge-input use and firm fixed effects, declines for some patenting measures but is flat or rising for others.
Semi-decade fixed effects linking patenting to firm growth generally trend negative across five firm samples and six knowledge inputs, indicating that growth conditional on ideas is getting harder to find, which the authors suggest might reflect an aggregate crowding out of ideas.
Related Articles:
Is It a Bubble— Transformational technologies have historically generated excessive enthusiasm and investment, producing more infrastructure than needed, asset prices that prove too high, and losses for those funding them, while also accelerating adoption. AI enthusiasm is argued to fit this pattern, with debt financing having the potential to magnify the effects.
Are Ideas Getting Harder to Find?— The authors estimate U.S. research productivity has fallen by a factor of 41 since the 1930s, averaging declines above 5% per year across sectors. Sustaining constant growth in output per person requires doubling research effort every 13 years, challenging growth models that assume constant research productivity.
Is The Great Stagnation Actually Just a ‘So-So’ Stagnation?— Long-running U.S. tests show population average reading scores moved from 255 to 256 out of 500 and math from 266 to 271, despite educational attainment rising more than three years. The author argues that if added schooling has not raised productivity, measured total factor productivity growth has been understated, shrinking the apparent "Great Stagnation" by about a third.
AI Summary.Data on over 5,000 German fighter pilots show performance rose markedly as pilots approached the eligibility threshold for each medal in a tiered awards ladder and slumped immediately after receiving it, with the effect repeating at each successive rung. The authors estimate the ladder added 3,000–5,000 aerial victories, 5–10% of the total.
Leonardo Bursztyn, Ewan Rawcliffe and Hans-Joachim VothNational Bureau of Economic Research
Date Posted:
Is Database:
Database
Is Important:
Important
Performance of Nazi fighter pilots rose as they neared eligibility for a medal and fell off upon receipt, prompting periodic offerings of new medals. This “hierarchical system of status-based incentives” raised net shootdowns by an estimated 5 to 10%.
Do tiered status awards repeatedly boost worker performance?
Core argument: The authors estimate that the German air force's ladder of successive medals added 3,000–5,000 aerial victories to German pilots' wartime tally, 5–10% of the total 54,835 victories in their data.
We provide direct evidence from WWII that one organization, the German air force, successfully created a ladder of highly coveted medals, providing status-based incentives for fighter pilots. The number of rungs on the ladder increased periodically as more workers earned the top awards. The successive medals potentially provided pilots with an incentive to keep exerting effort. Our study demonstrates that this system successfully increased performance. Of the total 54,835 aerial victories in our data, 39.5% were scored by the 10% most able pilots by our measure of ability. Moreover, 8.6% were scored the top 1% of pilots alone. We first employ a modified event-study design and data on the monthly performance of over 5,000 German pilots to demonstrate that pilot performance increases markedly as they approach the eligibility threshold for a medal. Success rates in aerial combat immediately slump after receiving the award, suggesting a sharp change in incentives and effort. We show that this incentive effect is weaker for pilots who already possess other markers of high social status, in line with predictions from a standard signaling game. As pilots approach the successive medal in the ladder, the same effort path repeats itself. Our empirical results show that a hierarchical system of status-based incentives can repeatedly elicit effort due to status concerns. The net effect of the medal ladder was substantial—we estimate that it added between 3,000 and 5,000 victories to German pilots’ tally during the war (5-10% of the total). [Only 19% of the sample survived the war.]
Takeaways by Macro Roundup® AI
The authors estimate that the German air force's ladder of successive medals added 3,000–5,000 aerial victories to German pilots' wartime tally, 5–10% of the total 54,835 victories in their data.
Using a modified event-study design on monthly performance of over 5,000 German pilots, the authors find performance rises markedly as pilots approach a medal's eligibility threshold and slumps immediately after the award, with the pattern repeating for each successive medal.
The top 10% of pilots by the authors' ability measure scored 39.5% of all aerial victories and the top 1% scored 8.6%, while the incentive effect was weaker for pilots already holding other markers of high social status.
Related Articles:
Killer Incentives: Status Competition and Pilot Performance during World War II— Analysis of over 5,000 German World War II fighter pilots finds that public peer recognition was associated with higher victory scores among skilled pilots without higher mortality, while average pilots gained little in victories but died at higher rates. The authors conclude status-based awards can have uneven effects in high-risk settings.
Ed Conard Debates Furman On “The Expected Value of Risk Taking”— A debate on U.S. fiscal policy argues that spending increases, rather than the tax increases needed to fund them, slow growth, and that higher taxes would reduce the expected value of America's innovative risk-taking. The argument holds that cutting debt below 100% of GDP is critical.
Long-Term Shareholder Returns: Evidence From 64,000 Global Stocks— A study of about 64,000 global stocks estimates net global stock market wealth creation of $75.7tn, with returns highly concentrated: the top 159 firms (0.25%) account for half of net wealth creation and the top 1,526 (2.39%) account for all of it. Fewer than half of stocks beat one-month US Treasury bills, 42.6% of non-US and 44.8% of US stocks.
AI Summary.The authors find that a minority of entrepreneurs are "transformative" — generating disproportionate productivity gains, scoring high on IQ tests, being well-educated, and hiring research and development workers. They argue that economies with more such entrepreneurs and R&D workers sustain higher long-run productivity growth.
Ufuk Akcigit, Harun Alp, Jeremy Pearce, and Marta PratoFederal Reserve Bank of New York
Date Posted:
Is Database:
Database
Is Important:
Important
Danish data show that the propensity to become an R&D worker or a “transformative entrepreneur” who pursues innovation is highly correlated with IQ. Productivity growth is driven by the “symbiotic relationship” between such entrepreneurs and R&D workers.
Which entrepreneurs drive disproportionate productivity gains?
Core argument: A distinct minority of entrepreneurs are "transformative" — those hiring at least one R&D worker — and the authors find these founders generate disproportionate productivity gains while tending to have high IQ scores and strong education.
We study who becomes an entrepreneur, along with their hiring and business decisions, and find that a distinct minority are “transformative.” These individuals, who generate disproportionate productivity gains, tend to have high IQ scores, be well-educated, and hire technical (R&D) workers. We identify entrepreneurs as individuals who register as primary founders of firms with at least one employee. Among this group, we distinguish transformative entrepreneurs, those who hire at least one R&D worker, from non-transformative entrepreneurs, who do not employ technical personnel. R&D workers are defined as individuals employed in occupations with high patenting intensity, reflecting their direct involvement in innovation-related activities. The shares of entrepreneurs and R&D workers vary across IQ deciles. The pattern is reversed across the two groups: individuals with higher IQs are more likely to be R&D workers, but less likely to be entrepreneurs. The data support the idea of productivity growth being driven by the symbiotic relationship between transformative entrepreneurs and R&D workers. For policymakers, the lesson is that when an economy has more R&D workers and transformative entrepreneurs, they sustain higher long-run productivity growth.
Takeaways by Macro Roundup® AI
A distinct minority of entrepreneurs are "transformative" — those hiring at least one R&D worker — and the authors find these founders generate disproportionate productivity gains while tending to have high IQ scores and strong education.
Across IQ deciles the pattern reverses between the two groups: individuals with higher IQs are more likely to be R&D workers but less likely to be entrepreneurs, with entrepreneurs identified as primary founders of firms with at least one employee.
The authors conclude that economies with more R&D workers and transformative entrepreneurs sustain higher long-run productivity growth, which they attribute to the symbiotic relationship between the two groups.
Related Articles:
Transformative and Subsistence Entrepreneurs: Origins and Impacts on Economic Growth— Only 17% of firms are "transformative" entrepreneurs, defined as those hiring a research-and-development worker at some point; higher IQ and educational attainment strongly predict transformative entrepreneurship, though the likelihood of becoming an entrepreneur overall declines with IQ. The authors argue education subsidies address financial constraints more effectively than uniform startup or research subsidies.
The Economics of Inequality in High-Wage Economies— Information technology created investment opportunities exceeding the supply of properly trained talent, while trade with low-wage economies, low-skilled immigration, trade deficits, and aging demographics eased constraints on low-skilled labor and risk-averse savings. The author argues this framework, rather than rising cronyism, explains high returns to success, slowing productivity growth, and low interest rates.
The Extreme Shortage of High IQ Workers— Semiconductor manufacturing requires workers of above-average intelligence, with the most demanding roles needing IQs in the top 1 in 1,000 — a pool of about 164,000 people in the US, of whom perhaps 100,000 can be placed in such professions. The author argues this scarcity explains reliance on high-IQ workers abroad and on smarter tools and non-human intelligence.
AI Summary.Since 1988, total factor productivity in the information technology sector has risen about 178%, versus about 12% across all non-IT sectors combined, while aggregate productivity rose about 40%. The IT sector accounts for roughly 45% of cumulative aggregate productivity growth, once sectoral linkages to downstream industries are weighted in.
Joel David, Gustavo de Souza, Adalyn SchommerFederal Reserve Bank of Chicago
Date Posted:
Is Database:
Database
Is Important:
Important
Btw 1988 and 2023, the US IT sector’s TFP rose by 178%, while TFP in all the non-IT sectors combined rose by only 12%. Taking account of sectoral linkages, the IT sector contributed ~17pp to TFP growth since 1988, and the non-IT sectors about 20pp.
How much of US productivity growth comes from the IT sector?
Core argument: The IT sector's total factor productivity rose 178% over the 36 years since the start of 1988, versus 12% for all non-IT sectors combined, while aggregate TFP increased 40%.
For a given % change in any particular sector’s TFP, the effect on aggregate TFP is equal to the sector-level change multiplied by a weight equal to the ratio of that sector’s nominal gross output to the total nominal value added of the entire private business sector. The total change in aggregate TFP is the sum of the weighted changes in sector-level TFPs. The “Domar weights,” [which] sum to more than one since the numerator includes sales of intermediate goods instead of only value added, incorporate the effects of sectoral linkages—when sector i becomes more productive, the sectors that use the output of sector i become more efficient and expand production, as do the sectors that use the output of those sectors as inputs (including possibly sector i), and so on. Fig.3, panel A shows that aggregate TFP has increased by a factor of 1.4, or 40%, in the 36 years since the start of 1988, while the IT sector’s TFP has increased by a factor of about 2.78, or 178%. In contrast, TFP in all the non-IT sectors combined has increased by a factor of only about 1.12, or 12%. Fig.3, panel B shows that the IT sector has contributed approximately 17 percentage points to cumulative TFP growth since 1988, and the non-IT sectors about 20 percentage points. Thus the IT sector accounts for roughly 45% of cumulative aggregate productivity growth.
Takeaways by Macro Roundup® AI
The IT sector's total factor productivity rose 178% over the 36 years since the start of 1988, versus 12% for all non-IT sectors combined, while aggregate TFP increased 40%.
The IT sector contributed roughly 17 percentage points of cumulative aggregate TFP growth since 1988 against about 20 percentage points from non-IT sectors, accounting for roughly 45% of cumulative aggregate productivity growth.
Aggregate TFP change equals the sum of sector-level TFP changes weighted by Domar weights—each sector's nominal gross output over total private business value added—which sum to more than one and incorporate sectoral linkages.
Related Articles:
Is The Great Stagnation Actually Just a ‘So-So’ Stagnation?— Long-running U.S. tests show population average reading scores moved from 255 to 256 out of 500 and math from 266 to 271, despite educational attainment rising more than three years. The author argues that if added schooling has not raised productivity, measured total factor productivity growth has been understated, shrinking the apparent "Great Stagnation" by about a third.
Assessing the Implications of a Productivity Miracle— If artificial intelligence can perform economically valuable cognitive tasks at very low marginal cost, consumers would buy far more of the cheapened goods and spend the savings in other categories, expanding activity and employment in sectors facing new demand. Where supply is inherently limited, such as real estate, those savings could bid up prices and produce inflation.
Productivity During and Since the Pandemic— Total factor productivity growth since early 2021 has run only slightly above its 2010–19 average, suggesting the 2020 productivity surge reflected cyclical dynamics layered on a continuation of the slow pre-pandemic trend. Productivity has since grown somewhat faster than the authors' analysis predicted, but remains well below the 1995–2004 pace.
AI Summary.Using US Census Bureau micro-data on manufacturing, the authors estimate that both the output elasticity of research and development and its marginal returns have risen, alongside rising technological rivalry and obsolescence. They conclude ideas are not getting harder to find, but innovations are more transient, so rising R&D need not raise aggregate productivity growth.
Yoshiki Ando, James Bessen and Xiupeng WangSocial Science Research Network
Date Posted:
Is Database:
Database
Is Important:
Important
Low productivity growth in the US manufacturing sector btw 1976 and 2018 was due to “rising technological rivalry” that “render[ed] rivals’ technologies obsolete and innovations more transient.”
Are ideas getting harder to find, or just more transient?
Core argument: Using US Census Bureau micro-data on manufacturing from 1976 to 2018, the authors find that both the elasticity of total factor productivity with respect to R&D and the marginal returns to R&D have risen sharply.
R&D investment has grown robustly, yet aggregate productivity growth has stagnated. Is this because “ideas are getting harder to find”? This paper uses micro-data from the US Census Bureau to explore the relationship between R&D and productivity in the manufacturing sector from 1976 to 2018. We find that both the elasticity of output (TFP) with respect to R&D and the marginal returns to R&D have risen sharply. Exploring factors affecting returns, we conclude that R&D obsolescence rates must have risen. Using a novel estimation approach, we find consistent evidence of sharply rising technological rivalry and obsolescence. These findings suggest that R&D has become more effective at finding productivity-enhancing ideas, but these ideas may also render rivals’ technologies obsolete, making innovations more transient. Because of obsolescence, rising R&D does not necessarily mean rising aggregate productivity growth. [Flagged by Tyler Cowen].
Takeaways by Macro Roundup® AI
Using US Census Bureau micro-data on manufacturing from 1976 to 2018, the authors find that both the elasticity of total factor productivity with respect to R&D and the marginal returns to R&D have risen sharply.
The authors conclude that R&D obsolescence rates must have risen, and using a novel estimation approach find consistent evidence of sharply rising technological rivalry and obsolescence.
The authors argue that R&D has become more effective at finding productivity-enhancing ideas, but that such ideas may render rivals' technologies obsolete, so rising R&D does not necessarily mean rising aggregate productivity growth.
Related Articles:
The R&D Puzzle in U.S. Manufacturing Productivity Growth— Research and development spending in U.S. manufacturing has been rising relative to output and hours worked while productivity has been flat, which suggests the ability of research and development to boost productivity growth has been declining.
Are Ideas Getting Harder to Find?— The authors estimate U.S. research productivity has fallen by a factor of 41 since the 1930s, averaging declines above 5% per year across sectors. Sustaining constant growth in output per person requires doubling research effort every 13 years, challenging growth models that assume constant research productivity.
Something Is Getting Harder But It’s Not Finding Ideas— The number of researchers in the US economy has grown to 23x its 1930 level while the economy's growth rate has been constant or declining, contradicting the assumption that constant research input yields a constant growth rate. Bloom et al. explain this divergence by assuming ideas get harder to find as more are discovered.
