CoChat
Sign Up
Cited by 715Labor market dynamics and wage inequalityGlobal trade and economicsHistorical Economic and Social Studies

Tasks, automation, and the rise in U.S. wage inequality

Open access
Acemoglu, Daron, Restrepo, Pascual·2022·Econometrica
DOI 10.3982/ecta19815
Open this paper in CoChat — free accountOpen PDF

Abstract

We document that between 50% and 70% of changes in the U.S. wage structure over the last four decades are accounted for by relative wage declines of worker groups specialized in routine tasks in industries experiencing rapid automation. We develop a conceptual framework where tasks across industries are allocated to different types of labor and capital. Automation technologies expand the set of tasks performed by capital, displacing certain worker groups from jobs for which they have comparative advantage. This framework yields a simple equation linking wage changes of a demographic group to the task displacement it experiences. We report robust evidence in favor of this relationship and show that regression models incorporating task displacement explain much of the changes in education wage differentials between 1980 and 2016. The negative relationship between wage changes and task displacement is unaffected when we control for changes in market power, deunionization, and other forms of capital deepening and technology unrelated to automation. We also propose a methodology for evaluating the full general equilibrium effects of automation, which incorporate induced changes in industry composition and ripple effects due to task reallocation across different groups. Our quantitative evaluation explains how major changes in wage inequality can go hand‐in‐hand with modest productivity gains.

Integrity signals

Registry

Crossref · 10.3982/ecta19815

Cites

50 of 50 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 7 cited sources checked · 0 supported · 0 flagged

  • Findingreasoning

    Automation reduced the relative, and in some cases real, wages of workers specializing in routine tasks in industries undergoing rapid automation.

    “In particular, automation reduced the relative, and in some cases real, wages of workers specializing in routine tasks in industries undergoing rapid automation”
    Quote in the paperSubstantial claim
  • Findingown data

    Results remain unchanged when controlling for other forms of capital deepening and TFP growth unrelated to automation.

    “Our results also remain unchanged when we control for other forms of capital deepening and TFP growth unrelated to automation.”
    Quote in the paperSubstantial claim
  • Findingown data

    The negative relationship between wage changes and task displacement is unaffected when controlling for changes in market power, deunionization, and other forms of capital deepening and technology unrelated to automation.

    “The negative relationship between wage changes and task displacement is unaffected when we control for changes in market power, deunionization, and other forms of capital deepening and technology unrelated to automation.”
    Quote in the paperSubstantial claim
  • Findingown data

    There is robust evidence in favor of the relationship between wage changes and task displacement.

    “We report robust evidence in favor of this relationship”
    • Also stated in: § 1. INTRODUCTION · boosted
      “Using both measures, we find a strong association between task displacement and wages.”
    Quote in the paperSubstantial claim
  • Findingown data

    Proxies of automation account for 45% of the observed changes in industry labor shares from 1987 to 2016.

    “These proxies of automation account for 45% of the observed changes in industry labor shares from 1987 to 2016.”
    Quote in the paperSubstantial claim
  • May combine two claims. 56
  • Classified as probably not a substantial claim. 17
  • The quote only partly states this claim. 1

  • May combine two claims.

    Findingown data

    The result that direct task displacement explains 50-70% holds regardless of controlling for standard forms of SBTC, which account for 10% of changes in wage structure.

    “This holds regardless of whether we control for standard forms of SBTC (e.g., allowing the productivity of workers to evolve as a function of their education levels and gender), which account for 10% of the changes in the wage structure.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Changes in industry concentration and markups, Chinese import competition, and deunionization do not appear to have a sizable effect on the U.S. wage structure.

    “and these factors themselves do not appear to have a sizable effect on the U.S. wage structure.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingnone given

    Since 1980, real wages of workers with a post-graduate degree rose, while real wages of low-education workers fell or remained stagnant.

    “while the real wages of workers with a post-graduate degree rose, the real wages of low-education workers fell or remained stagnant.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Offshoring-induced task displacement has similar effects on wages, though offshoring accounts for a smaller share of observed changes in task displacement and wage structure than automation.

    “Finally, we also show that offshoring-induced task displacement has similar effects on wages, though offshoring accounts for a smaller share of the observed changes in task displacement and the wage structure than automation.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Findingnone given

    The real earnings of men without a high-school degree are now 15% lower than they were in 1980.

    “The real earnings of men without a high-school degree are now 15% lower than they were in 1980.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Sizable distributional effects are accompanied by small increases in average wage level, GDP, and TFP.

    “These sizable distributional effects are accompanied by small increases in the average wage level, GDP, and TFP.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Once task displacement measures are included, most differences between workers with different education levels disappear, and there is no longer any differential wage growth for workers with a college degree relative to those with a high-school degree.

    “However, most of the differences between workers with different education levels disappear once our task displacement measures are included in columns 2 and 3”
    • Also stated in: § 4.3. Task Displacement versus SBTC · plain
      “Notably, there is no longer any differential wage growth for workers with a college degree relative to those with a highschool degree.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.
    • The authors state this as an assumption: The two direct task displacement measures are uncorrelated with other trends affecting wages except through automation-driven task displacement.

  • May combine two claims.

    Findingown data

    The 2SLS estimate for the effect of task displacement using all three proxies is -1.23 (s.e. = 0.19), and the implied share of variance explained is 50%.

    “The 2SLS estimate for the effect of task displacement, -1 23 (s = 0 19), and the implied share of variance of wage changes explained by task displacement, 50%, are similar to those in column 3 of Table I”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.
    • The authors state this as an assumption: The two direct task displacement measures are uncorrelated with other trends affecting wages except through automation-driven task displacement.

  • May combine two claims.

    Findingown data

    Some demographic groups experienced a 25% direct reduction in task shares due to automation between 1980 and 2016, while others saw no change.

    “some demographic groups experienced a 25% direct reduction due to automation between 1980 and 2016, while others saw no change in their task shares.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.

  • May combine two claims.

    Findingown data

    There is a strong correlation between task displacement and real wage changes from 1980 to 2016, with groups experiencing the highest levels of task displacement seeing their real wages fall or stagnate.

    “These plots reveal a strong correlation between task displacement and changes in real wages, with groups experiencing the highest levels of task displacement seeing their real wages fall or stagnate.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.
    • The authors state this as an assumption: The two direct task displacement measures are uncorrelated with other trends affecting wages except through automation-driven task displacement.

  • May combine two claims.

    Findingown data

    The relationship between task displacement and real wage changes is not driven by secular trends adversely affecting some groups and is not present between 1950 and 1980.

    “They demonstrate that the relationships depicted in the top two panels are not driven by secular trends adversely affecting some groups and are not present between 1950 and 1980-a period that predates major advances in automation.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Offshoring contributes to task displacement and depresses real wages of exposed groups, but it explains only 12% of the variation in wage changes.

    “As expected, offshoring also contributes to task displacement and depresses real wages of exposed groups, but it only explains 12% of the variation in wage changes.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.
    • The authors state this as an assumption: The two direct task displacement measures are uncorrelated with other trends affecting wages except through automation-driven task displacement.

  • May combine two claims.

    Findingown data

    Workers in the top quintile of the direct task displacement distribution saw their real wage decline by 12%, while workers in the least exposed groups enjoyed real wage growth of about 26%.

    “it shows that workers in the top quintile of the (direct) task displacement distribution saw their real wage decline by 12%, while workers in the least exposed groups enjoyed real wage growth of about 26%.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.

  • May combine two claims.

    Findingown data

    Conditional on task displacement, exposure to industry labor share declines and relative specialization in routine jobs do not explain real wage changes.

    “Column 4 shows that these two variables themselves do not explain real wage changes (conditional on task displacement), while task displacement remains very strongly correlated with wage changes.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.
    • The authors state this as an assumption: The two direct task displacement measures are uncorrelated with other trends affecting wages except through automation-driven task displacement.

  • May combine two claims.

    Findingown data

    Task displacement explains more than 50% of the overall changes in the wage structure, while education dummies explain less than 17%.

    “Task displacement also explains more than 50% of the overall changes in the wage structure, while the education dummies explain less than 17%.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.
    • The authors state this as an assumption: The two direct task displacement measures are uncorrelated with other trends affecting wages except through automation-driven task displacement.

  • May combine two claims.

    Findingown data

    In specifications allowing differential trends by baseline wage level, task displacement explains about 40% of observed wage changes, while differential trends by education and baseline wages explain 18% and 7% of the variation.

    “The results of these demanding specifications are similar to those in the first three columns, and our task displacement measure explains about 40% of the observed wage changes, while differential trends by education and baseline wages explain, respectively, 18% and 7% of the variation.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.
    • The authors state this as an assumption: The two direct task displacement measures are uncorrelated with other trends affecting wages except through automation-driven task displacement.

  • May combine two claims.

    Findingown data

    Using the preferred measure, a 10 percentage point higher task displacement is associated with a 4.4 percentage point decline in employment between 1980 and 2016 and a 3.5 percentage point increase in non-participation.

    “using the estimates from column 5 based on our preferred measure, we see that a 10 percentage point higher task displacement is associated with a 4.4 percentage point decline in employment between 1980 and 2016, and a similar 3.5 percentage point increase in non-participation.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.
    • The authors state this as an assumption: The two direct task displacement measures are uncorrelated with other trends affecting wages except through automation-driven task displacement.

  • May combine two claims.

    Findingown data

    Including controls for differential evolution of supply raises the explanatory power of task displacement; task displacement explains 63%-72% of changes in the U.S. wage structure, while education dummies explain 4%-18%.

    “Now, task displacement explains 63%-72% of the changes in the U.S. wage structure, while the education dummies continue to explain a small portion (4%-18%) of the variation.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.
    • The authors state this as an assumption: The two direct task displacement measures are uncorrelated with other trends affecting wages except through automation-driven task displacement.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Panels C and D for commuting zones yield results similar to Panels A and B.

    “Panels C and D repeat this exercise for 54 demographic groups defined by a coarser grouping of gender, education, age, and race, but now across 722 U.S. commuting zones (for a total of 20,768 observations).The results are similar to those in Panels A and B.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Controlling for exposure of workers to occupations in the bottom tercile of the overall wage distribution and its interaction with industry automation has no effect on task displacement results, and these variables themselves are not significant.

    “Table A-VI controls for the exposure of workers to occupations in the bottom tercile of the overall wage distribution and the interaction between this exposure and industry automation.This has no effect on our task displacement results and these variables themselves are not significant.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Including workers' exposure to industries with declining unionization rates and exposure to Chinese import competition, task displacement measures have similar coefficients to Table I and continue to explain about 50% of changes in the U.S. wage structure.

    “With both controls, our task displacement measures have similar coefficients to the ones we saw in Table I and continue to explain about 50% of the changes in the U.S. wage structure.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Controlling for workers' exposure to industries with rising sales concentration and to markup changes using three alternative estimates, task displacement effects are similar to baseline and exposure to concentration or markups has little explanatory power for wages.

    “In all specifications and with either measure of task displacement, our results are similar to the baseline estimates presented in Table I.For example, the effects of task displacement range between -1 31 and -1 42, while exposure to changes in concentration or markups have little explanatory power for wages.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Even with the markup correction, markup changes do not have a robust effect on wages and explain no more than 3% of the variation in wage changes.

    “Even with this correction, markup changes do not have a robust effect on wages and explain no more than 3% of the variation in wage changes.”
    Quote in the paperSubstantial claim

  • May combine two claims. The quote only partly states this claim.

    Findingown data

    With demographic group fixed effects absorbing national trends, task displacement has negative and significant but smaller effects on wages, especially in columns 2-3 and 5-6.

    “We find negative and significant but smaller effects of task displacement (especially in columns 2-3 and 5-6).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Controlling for exposure to changes in industry capital to value added ratios and TFP growth, exposure to these industry variables has no direct impact on wages and does not alter the relationship between task displacement and wages.

    “Columns 1-2 and 5-6 in Panel A of Table V further explore the role of capital intensity and industry productivity by controlling for exposure to changes in industry capital to value added ratios and TFP growth.”
    • Also stated in: § Confounding Trends: Capital, TFP, Deunionization, Imports, and Markups · plain
      “from non-automation technologies, exposure to these industry variables has no direct impact on wages and does not alter the relationship between task displacement and wages.25”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Using a measure of task displacement that partials out the component of industry labor share changes driven by markups does not affect the paper's conclusions, and point estimates for task displacement remain sizable and precise.

    “Panel B goes one step further and uses a measure of task displacement that partials out the component of industry labor share changes driven by markups.28 This correction does not affect our conclusions, and our point estimates for the effects of task displacement remain sizable and precise.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In stacked-differences models with two periods, 1980-2000 and 2000-2016, results are similar to but smaller in some specifications than those in Table I.

    “The results in both panels are similar to, but smaller in some specifications than, those in Table I.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    In general equilibrium, task displacement generates a 21% increase in the college premium (80% of the observed increase).

    “Second, in general equilibrium, task displacement generates a 21% increase in the college premium (80% of the observed increase) and a 22% increase in the post-graduate premium (55% of the observed increase).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingnone given

    Task displacement due to automation accounts for the observed decline in the labor share (by construction) and the observed increase in the capital-GDP ratio over this period.

    “Finally, task displacement due to automation accounts for the observed decline in the labor share (by construction) and the observed increase in the capital-GDP ratio over this period.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    These alternative economic trends do not appear to play a major role in the evolution of the U.S. wage structure between 1980 and 2016 and have negligible effects on our estimates.

    “These alternative economic trends do not appear to play a major role in the evolution of the U.S. wage structure between 1980 and 2016 and have negligible effects on our estimates.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingnone given

    Two thirds of the rise in average real wages is due to educational upgrading of the workforce, which is not present in the model.

    “In contrast to this small automation-induced increase in productivity, in the data TFP grew by 35% during this period, and average real wages rose by 29% (though two thirds of the latter is due to educational upgrading of the workforce, which is not present in our model).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Task displacement leads to sizable increases in wage inequality, but only small productivity gains.

    “Most notably, task displacement leads to sizable increases in wage inequality, but only small productivity gains-thus providing a possible resolution to a puzzling feature of the U.S. data.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Findings are confirmed for the 1980-2007 period, which avoids persistent effects of the Great Recession.

    “In Appendix Table B-8, we confirm our findings for the 1980-2007 period, which avoids any persistent effects of the Great Recession.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationnone given

    Technologies directly improving the productivity of skilled labor do not involve any displacement and always increase the wages of unskilled workers.

    “In contrast, technologies directly improving the productivity of skilled labor do not involve any displacement and always increase the wages of unskilled workers.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    Automation displaces workers from tasks where they had comparative advantage, reducing their relative wages and even possibly their real wages.

    “automation displaces workers from tasks where they had comparative advantage, reducing their relative wages and even possibly their real wages.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based). May combine two claims.

    Interpretationnone given

    Automation corresponds to increases in the productivity of capital at tasks previously assigned to labor and leads to the displacement of workers from these tasks.

    “Automation and offshoring: automation corresponds to increases in the productivity of capital (or reductions in the cost of producing this type of capital) at tasks previously assigned to labor and leads to the displacement of workers from these tasks.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductreasoning

    The reduced-form analysis misses three indirect channels via which automation affects wages in general equilibrium: common productivity effect on real wage level changes, industry composition changes, and ripple effects.

    “Although our reduced-form analysis documents a strong negative relationship between direct task displacement and relative wage changes across worker groups, it misses three indirect channels via which automation affects wages in general equilibrium.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    There is a unique equilibrium in the model; output, wages, and the capital share in GDP can be expressed as functions of task shares.

    “PROPOSITION 1-Equilibrium: There is a unique equilibrium.In this equilibrium, output, wages, and the capital share in GDP, s K , can be expressed as functions of task shares:”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    The direct task displacement measure represents the direct reduction in group g's task share due to automation.

    “This measure represents the direct reduction in group g's task share due to automation.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    Automation always increases the capital share and reduces the labor share of value added.

    “Equation ( 9) also shows that automation always increases the capital share and reduces the labor share of value added-an observation that will be at the core of our measurement approach, in Section 2.4.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    The elasticity of substitution between factors is not equal to λ, but σ ≥ λ.

    “Second, despite appearances, the elasticity of substitution between factors is not equal to λ, but σ ≥ λ.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    Task displacement played both an unequalizing and a polarizing role in the wage distribution.

    “displacement has been particularly high during this period for groups in the middle of the wage distribution-thus playing both an unequalizing and a polarizing role.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.

  • May combine two claims.

    Interpretationown data

    The baseline results document a strong association between task displacement and real wage changes and support the presumption that this association captures the direct causal effect of automation-driven task displacement.

    “Overall, our baseline results document a strong association between task displacement and real wage changes, and support our presumption that this association is capturing the (direct) causal effect of automation-driven task displacement experienced by a demographic group.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.
    • The authors state this as an assumption: The two direct task displacement measures are uncorrelated with other trends affecting wages except through automation-driven task displacement.

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    The automation-driven labor share decline quantifies the direct losses of routine tasks experienced by workers in an industry.

    “The automation-driven labor share decline quantifies the direct losses of routine tasks experienced by workers in an industry.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    The second task displacement strategy exploits actual measures of automation, such as adoption of industrial robots, dedicated machinery, and specialized software.

    “Our preference for the second strategy is rooted in the fact that it exploits actual measures of automation, such as adoption of industrial robots, dedicated machinery, and specialized software.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductnone given

    The 49 industries can be consistently tracked in Census and BLS data and cover the entire nongovernment sector.

    “These 49 industries can be consistently tracked in Census and BLS data, and cover the entire nongovernment sector.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    When λ = 1 and Assumptions 1 and 2 hold, direct task displacement can be measured as equation (14).

    “Appendix A-3 of the Supplemental Material shows that when λ = 1 and Assumptions 1 and 2 hold, (direct) task displacement can be measured as”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Appendix A-3

      Not checked in the free run.

    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.

  • May combine two claims.

    Interpretationown data

    Much of the change in the U.S. wage structure between 1980 and 2016 is due to task displacement, with a minor role for standard (factor-augmenting) SBTC.

    “These results are the basis for our claim that much of the change in the U.S. wage structure between 1980 and 2016 is due to task displacement, with a minor role for standard (factor-augmenting) SBTC.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.
    • The authors state this as an assumption: The two direct task displacement measures are uncorrelated with other trends affecting wages except through automation-driven task displacement.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    More conservative standard errors confirm the main findings.

    “More conservative standard errors (for our baseline specification in column 3) are presented in Table B-3 in the Supplemental Material and confirm our main findings.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Adao, Kolesár, and Morales (2019)

      Not checked in the free run.

    • cites Borusyak, Hull, and Jaravel (2022)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    The first task displacement strategy is valid when λ = 1, there are no changes in markups, and there are no other influences on industry wages such as changes in worker rents.

    “This strategy is valid when λ = 1 (so that factor prices do not affect the labor share), there are no changes in markups, and there are no other influences on industry wages (such as changes in worker rents).”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: The observed decline in an industry's labor share can be entirely attributed to automation for the first task displacement measure.

  • May combine two claims.

    Interpretationown data

    Task displacement has been at the root of the changes in the wage structure from 1980 to today, while other forms of SBTC had limited explanatory power.

    “In summary, our results show that task displacement has been at the root of the changes in the wage structure from 1980 to today, while other forms of SBTC had limited explanatory power.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.
    • The authors state this as an assumption: The two direct task displacement measures are uncorrelated with other trends affecting wages except through automation-driven task displacement.

  • May combine two claims.

    Study design and conductown data

    The paper's regressions control for observed changes in industry composition but do not separate industry shifts induced by automation, missing one component of the total impact of automation.

    “Second, although our regressions control for observed changes in industry composition, they do not separate industry shifts induced by automation, missing one component of the total impact of automation.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    Task displacement-and not so much rising market power-has played a defining role in the surge in U.S. wage inequality over the last four decades.

    “These results imply that task displacement-and not so much rising market power-has played a defining role in the surge in U.S. wage inequality over the last four decades.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    Task displacement at the regional level matters and has a precisely-estimated negative impact on wages, in line with the paper's theory.

    “These estimates imply that task displacement at the regional level matters and has a precisely-estimated negative impact on wages, which is in line with our theory.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductreasoning

    The paper's regression estimates focus on direct effects of automation via task displacement and do not account for ripple effects, which also impact the wage structure.

    “Third and most importantly, our regression estimates focus on the direct effects of automation via task displacement and do not account for the resulting ripple effects, which also impact the wage structure.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    The propagation matrix is well defined, θ_gg ≥ 0 captures competition for tasks between groups, and its row sum ε_g is always between 0 and 1.

    “In Appendix A-2, we prove that is well defined and has several properties.First, θ gg ≥ 0 captures the extent of competition for tasks between groups g and g.Second, we show that the row sum of , which we denote by ε g , is always between 0 and 1.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductreasoning

    The reduced-form analysis did not allow for ripple effects, which were thus partially captured by the task displacement measures.

    “Our reduced-form analysis did not allow for ripple effects, which were thus partially captured by our task displacement measures.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    Direct task displacement generated by automation is the exogenous impulse that leads to a change in task shares, while ripple effects encoded in the propagation matrix and changes in industry composition determine the full general equilibrium implications of this task displacement.

    “The formulas in this proposition clarify that the direct task displacement generated by automation is the (exogenous) impulse that leads to a change in task shares, while ripple effects encoded in the propagation matrix and changes in industry composition determine the full general equilibrium implications of this task displacement.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    Automation is distinct from trade in final goods and from other technologies that do not generate task displacement and impact labor demand mainly by changing industry composition.

    “This reiterates that automation is distinct from trade in final goods and from other technologies that do not generate task displacement and impact labor demand mainly by changing industry composition.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    Because η < 1, automation induces a shift toward sectors, such as services, that have less automation and increases the demand for workers specialized in these sectors.

    “Because η < 1, automation induces a shift toward sectors, such as services, that have less automation and increases the demand for workers specialized in these sectors.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    Automation can generate meaningful real wage declines, which contrasts with the canonical SBTC model, where technological progress is predicted to increase the real wages of all groups.

    “This result highlights how automation can generate meaningful real wage declines, which contrasts with the canonical SBTC model, where technological progress is predicted to increase the real wages of all groups.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Interpretationreasoning

    The formulation imposes homotheticity, which can be relaxed by allowing expenditure shares to additionally depend on the level of consumption.

    “This formulation imposes homotheticity, which can be relaxed by allowing expenditure shares to additionally depend on the level of consumption.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: The demand system formulation imposes homotheticity.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationnone given

    The second part of the assumption ensures that an equilibrium satisfying the tie-breaking rule always exists.

    “The second part of the assumption is to ensure that an equilibrium satisfying this tie-breaking rule always exists (an equilibrium without this rule always exists).”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    The idea that routine tasks are easier to automate is the main premise of Autor, Levy, and Murnane (2003) and is in line with several studies that document a decline in routine jobs following automation.

    “The idea that routine tasks are easier to automate is the main premise ofAutor, Levy, and Murnane (2003) and is in line with several studies that document a decline in routine jobs following automation, including Ace-”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Autor, Levy, and Murnane (2003) source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationreasoning

    In the limit case where tasks can be traded across regions with no transaction or transport costs (or labor is perfectly mobile), task displacement reduces the wages of all workers in a given group by the same amount across all regions.

    “For example, in the limit case where tasks can be traded across regions with no transaction or transport costs (or labor is perfectly mobile), one would expect task displacement to reduce the wages of all workers in a given group by the same amount across all regions.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    The importance weight depends on the revenue share of the task in total costs, and hence the productivity of the factor performing the task has an exponent equal to the elasticity of substitution minus 1.

    “In particular, this importance weight depends on the revenue share of the task in total costs, and hence the productivity of the factor performing the task has an exponent equal to the elasticity of substitution minus 1.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationnone given

    If H is CES with elasticity η, then sY i (p) = α i • p 1-η i.

    “For example, if H is CES with elasticity η, then sY i (p) = α i • p 1-η i.”
    Quote in the paperNot a substantial claim
67 of 67

  • The authors state this as an assumption.

    Workers can only produce non-overlapping sets of tasks.

    “ASSUMPTION 1: 1.Workers can only produce non-overlapping sets of tasks (i.e., ψ g (x) > 0 only if ψ g (x) = 0 for all g = g).”

  • The authors state this as an assumption.

    The Inada condition ensuring finite output is satisfied.

    “In Appendix A-2 of the Supplemental Material, we derive an Inada condition that ensures finite output (in the one-sector case, this condition implies A λ-1 k • k < 1), and we assume throughout that it is satisfied.”

  • The authors state this as an assumption.

    When a task can be produced at the exact same unit cost by different factors, it is allocated to capital or to the type of labor with the higher index.

    “Throughout, when a task can be produced at the exact same unit cost by different factors, we assume it is allocated to capital or to the type of labor with the higher index”

  • The authors state this as an assumption.

    Each factor has a strict comparative advantage for some tasks.

    “and we also assume that each factor has a strict comparative advantage for some tasks.”

  • The authors state this as an assumption.

    Labor is supplied inelastically.

    “Labor is supplied inelastically, and we denote the total supply of labor of type g by ℓ g .”

  • The authors state this as an assumption.

    Industry outputs are combined into a single final good using a constant returns to scale aggregator.

    “We assume that industry outputs are combined into a single final good (aggregate output) using a constant returns to scale aggregator, H(y 1 y I ).”

  • The authors state this as an assumption.

    Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.

    “ASSUMPTION 2: Only routine tasks can be automated and, within an industry, different groups of workers are displaced from their routine tasks at a common rate.”

  • The authors state this as an assumption.

    Factor-augmenting and productivity-deepening technologies augment skills associated with education and gender.

    “we assume that these technologies augment skills associated with education and gender”

  • The authors state this as an assumption.

    Workers in the government sector experienced no automation.

    “we assume that workers in the government sector experienced no automation.”

  • The authors state this as an assumption.

    The two direct task displacement measures are uncorrelated with other trends affecting wages except through automation-driven task displacement.

    “Our identifying assumption is that the two (direct) task displacement measures are uncorrelated with other trends affecting wages-except through automation-driven task displacement.”

  • The authors state this as an assumption.

    The non-automation component of specialized software and dedicated machinery is orthogonal to other factors affecting industry labor shares.

    “Our exclusion restriction does not impose that all software and dedicated machinery are automation technologies, but it requires that their non-automation component is orthogonal to other factors affecting industry labor shares.”

  • The authors state this as an assumption.

    The observed decline in an industry's labor share can be entirely attributed to automation for the first task displacement measure.

    “Our first and simpler strategy assumes that the observed decline in the labor share of an industry, -d ln s L i , can be entirely attributed to automation.”

  • The authors state this as an assumption.

    Task allocation is unaffected by wages (no ripple effects).

    “This Jacobian summarizes the impact of a change in wages on task allocation (which was equal to zero under Assumption 1).”

  • The authors state this as an assumption.

    The common diagonal term of the propagation matrix is the same for all groups (θ_gg = θ for all g).

    “The assumption of common diagonal term is consistent with our reduced-form analysis, which did not find evidence of significant heterogeneities in the effects of task displacement across groups.”

  • The authors state this as an assumption.

    Each group's automation-driven task displacement is orthogonal to the unobserved wage effects (exclusion restriction).

    “To estimate the parameters of this system, we impose the exclusion restriction that each group's (automation-driven) task displacement, Task displacement direct g , is orthogonal to u, or equivalently to υ, which leads to the moment conditions”

  • The authors state this as an assumption.

    The demand system formulation imposes homotheticity.

    “This formulation imposes homotheticity, which can be relaxed by allowing expenditure shares to additionally depend on the level of consumption.”
16 of 16

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Acemoglu and Restrepo (2022) attribute 50 to 70 percent of observed changes in the US wage structure between 1980 and 2016 to the task displacement of worker groups specialized in routine work in industries experiencing rapid automation.

    “Acemoglu and Restrepo (2022) attribute 50 to 70 percent of observed changes in the US wage structure between 1980 and 2016 to the task displacement of worker groups specialized in routine work in industries experiencing rapid automation.”

    From Labor Market Consequences of Generative AI: Early Evidence from Norway (2026) · cites this paper as Acemoglu and Restrepo, 2022

  • Other technologies, like industrial robots, have been used to automate repetitive or systematic tasks that are often performed by less-educated workers.

    “while others, like industrial robots, have been used to automate repetitive or systematic tasks that are often performed by less-educated workers (65, 66).”

    From The impact of generative artificial intelligence on socioeconomic inequalities and policy making (2024) · cites this paper as [65]

  • The displacement of labor by "so-so technologies" (e.g. self-checkout kiosks or automated phone systems) that offer little or no productivity gain, along with diminished worker voice due to intensified monitoring and surveillance, can be harmful to long-run productivity and other social goals like job satisfaction.

    “While some amount of this is unavoidable, the displacement of labor by “so-so technologies” (e.g. self-checkout kiosks or automated phone systems) that offer little or no productivity gain, along with diminished worker voice due to intensified monitoring and surveillance, can be harmful to long-run productivity and other social goals like job satisfaction (65, 68).”

    From The impact of generative artificial intelligence on socioeconomic inequalities and policy making (2024) · cites this paper as [65]

  • Workers who in earlier decades would have qualified for mass expertise jobs in clerical, administrative and production occupations were instead shunted into non-routine, hands-on service occupations, placing downward pressure on already low wages for this work.

    “Workers who in earlier decades would have qualified for mass expertise jobs in clerical, ad-ministrative and production occupations were instead shunted into non-routine, hands-on service occupations.This placed downward pressure on already low wages for this work”

    From Applying AI to Rebuild Middle Class Jobs (2024) · cites this paper as Acemoglu and Restrepo, 2022

  • There is no economic law that says that the forces of automation and new work creation are equal and offsetting; recent evidence indicates that the former is outpacing the latter.

    “There is no economic law that says that the forces of automation and new work creation are equal and offsetting; in fact, recent evidence indicates that the former is outpacing the latter (Acemoglu and Restrepo, 2022;Autor et al., forthcoming).”

    From Applying AI to Rebuild Middle Class Jobs (2024) · cites this paper as Acemoglu and Restrepo, 2022

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Reported real-wage effects for high-school dropouts disagree with Table VIII

    Major

    The introduction reports automation-driven real-wage declines of 8.8% for high-school dropout men and 2.3% for women. Table VIII's preferred automation-driven estimates instead report declines of 8.41% and 3.40%, respectively; neither difference is explained by rounding.

  • Table V's unionization and import-competition columns are reversed in the discussion

    Minor

    The discussion assigns declining unionization to columns 3 and 7 and Chinese import competition to columns 4 and 8. Table V's headings assign Chinese import competition to columns 3 and 7 and de-unionization to columns 4 and 8.

2 of 2

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Acemoglu, D., & Restrepo, P.
Affiliations
  • Department of Economics, MIT
  • Boston University (US)
Venue
Econometrica
Publisher
The Econometric Society
Published
2022
Volume
90
Issue
5
Pages
1973-2016
ISSN
0012-9682 (ISSN-L)
DOI
10.3982/ecta19815
OpenAlex
W3174318537
Cited by
715
References
50
Citations by year
2026: 273 · 2025: 232 · 2024: 137 · 2023: 49 · 2022: 15 · 2021: 4 · 2020: 1 · 2019: 2 · 2013: 1
FWCI
87.00
Type
journal article
Language
en
Open access
Licence
Locations
Econometrica (publishedVersion, cc-by-nc-nd)
Topics
  • Social Sciences › Economics, Econometrics and Finance › Economics and Econometrics › Labor market dynamics and wage inequality
  • Social Sciences › Economics, Econometrics and Finance › General Economics, Econometrics and Finance › Global trade and economics
  • Social Sciences › Economics, Econometrics and Finance › Economics and Econometrics › Historical Economic and Social Studies
Keywords
wage inequality, automation, routine tasks, skill-biased technological change, job displacement, industry structure, productivity improvement, market power, union decline, capital deepening

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-09.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-09.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review