The Race between Machine and Man: Implications of Technology for Growth, Factor Shares and Employment
Summary (4 min read)
1 Introduction
- The accelerated automation of tasks performed by labor raises concerns that new technologies will make labor redundant (e.g., Brynjolfsson and McAfee, 2012, Akst, 2014, Autor, 2015).
- Automation allows firms to substitute capital for tasks previously performed by labor, while the creation of new tasks enables the replacement of old tasks by new variants in which labor has a higher productivity.
- The results in this case are similar, but the conditions for uniqueness and stability of the balanced growth path are more demanding.
- 4 equilibrium incorporating capital accumulation and directed technological change, but also because tasks are combined with a general elasticity of substitution, and because the equilibrium allocation of tasks critically depends both on factor prices and the state of technology.
2.1 Environment
- All tasks and the final good are produced competitively.
- A new (more complex) task replaces or upgrades the lowest-index task.
- Each task is produced by combining labor or capital with a task-specific intermediate q(i), which embodies the technology used either for automation or for production with labor.
- In Section 4 the authors relax this assumption and allow intermediate producers to make profits so as generate endogenous incentives for innovation.
2.2 Equilibrium in the Static Model
- These two special cases ensure that the demand for labor and capital is homothetic.
- The unit cost of production for tasks i ≤ I, on the other hand, depends on min { R, Wγ(i) } reflecting the fact that capital and labor are perfect substitutes in the production of automated tasks.
- In the static model, this will be the case when the capital stock is not too large, which is imposed in the next assumption.
- An increase in I∗—which corresponds to greater equilibrium automation—increases the share of capital and reduces the share of labor in this aggregate production function, while the creation of new tasks does the opposite.
- 14The increasing labor supply relationship, (11), ensures that the labor share sL = WL RK+WL is increasing in ω.
3 Dynamics and Balanced Growth
- The authors then investigate the conditions under which the economy admits a balanced growth path (BGP), where aggregate output, the capital stock and wages grow at a constant rate.
- The authors conclude by discussing the long-run effects of automation on wages, the labor share and employment.
3.2 Long-Run Comparative Statics
- The authors next study the log-run implications of an unanticipated and permanent decline in n(t), which corresponds to automation running ahead of the creation of new tasks.
- Because in the short run capital is fixed, the short-run implications of this change in technology are the same as in their static analysis in the previous section.
- Moreover, the asymptotic values for employment and the labor share are increasing in n.
- The dotted line depicts the case where wI(n) is large relative to R, so that there are significant productivity gains from automation.
- In contrast to the concerns that highly productive automation technologies will reduce the wage and employment, their model thus shows that it is precisely when automation fails to raise productivity significantly that it has a more detrimental impact on wages and employment.
4.1 Endogenous and Directed Technological Change
- This ensures that the unique equilibrium price for all types of intermediates is a limit price of ψ, and yields a per unit profit of (1−µ)ψ > 0 for technology monopolists.
- These profits generate 21 incentives for creating new tasks and automation technologies.
- The authors also assume that this compensation takes place with the new inventors making a take-it-or-leave-it offer to the holder of the existing patent.
- Developing new intermediates that embody technology requires scientists.
- For notational convenience, the authors also adopt the normalization G(0) = κNκI+κN .
4.2 Equilibrium with Endogenous Technological Change
- The authors first compute the present discounted value accruing to monopolists from automation and the creation of new tasks.
- To simplify the exposition, let us assume that in this equilibrium n(t) > max{n̄(ρ), ñ(ρ)}(ρ), so that I∗(t) = I(t) and newly-automated tasks start being produced with capital immediately.
- The condition σ̂ > ζ guarantees that the former, positive effect dominates, so that prospective technology monopolists have an incentive to introduce technologies that allow firms to produce tasks more cheaply.
- Proposition 6 also shows that, for κIκN > κ, the unique interior BGP is globally stable provided that the intertemporal elasticity of substitution is infinite (i.e., θ = 0), and locally stable otherwise (i.e., when θ > 0).
- In summary, Proposition 6 characterizes the varieties of BGPs, and together with Corollary 2, it delineates the types of changes in technology that trigger self-correcting dynamics.
5 Extensions
- In this section the authors discuss three extensions.
- First the authors introduce heterogeneous skills, which allow us to analyze the impact of technological changes on inequality.
- Second, the authors study a different structure of intellectual property rights that introduces the creative destruction of profits.
- Finally, the authors discuss the welfare implications of their model.
5.1 Automation, New Tasks and Inequality
- To study how automation and the creation of new tasks impact inequality, the authors now introduce heterogeneous skills.
- This extension is motivated by the observation that both automation and new tasks could increase inequality: new tasks favor high-skill workers who tend to have a comparative advantage in complex tasks, while automation substitutes capital for labor in lower-indexed tasks where low-skill workers have their comparative advantage.
- When ξ < 1, as the time during which a task has existed tends to infinity, the productivity of low-skill workers relative to high-skill workers converges to γL(i, t)/γH(i) = γH(i) ξ−1, and limits to zero as more and more advanced tasks are introduced.
- If ξ = 1, in the unique BGPWH(t) and WL(t) grow at the same rate as the economy, the wage gap, WH(t)/WL(t), remains constant, and capital, low-skill and high-skill workers perform constant shares of tasks.
5.2 Creative Destruction of Profits
- The authors modify their baseline assumption on intellectual property rights and revert to the classical setup in the literature in which new technologies do not infringe the patents of the products that they replace (Aghion and Howitt, 1992, and Grossman and Helpman, 1991).
- This assumption introduces the creative destruction effects—the destruction of profits of previous inventors by new innovators.
- Here πI(t, i) and πN (t, i) denote the flow profits from automating and creating new tasks, respectively, which are given by the formulas in equations (23) and (24).
- The next proposition focuses on interior BGPs and shows that, because of creative destruction, the authors must impose additional assumptions on the function ι(n) to guarantee stability.
- In their baseline model, the key force ensuring stability is that incentives to automate are shaped by the cost difference between producing a task with capital or with labor—by lowering the effective wage at the next tasks to be automated, current automation reduces the incremental value of additional automation.
5.3 Welfare
- The authors study welfare from two complementary perspectives.
- In particular, suppose that there exists an upward-sloping quasi–labor supply schedule, Lqs(ω), which constrains the level of employment, so that L ≤ Lqs(ω).
- Crucially, the reduction in employment resulting from automation now has a negative impact on welfare, and this negative effect can exceed the positive impact following from the productivity gains, turning automation, on net, into a negative for welfare.
- Interestingly, new tasks increase welfare even more than before, because they not only raise productivity but also expand employment, and by the same logic, the increase in labor supply has a welfare benefit for the workers (since they were previously constrained in their employment).
6 Conclusion
- As automation, robotics and AI technologies are advancing rapidly, concerns that new technologies will render labor redundant have intensified.
- In their model, this takes the form of the introduction of new, more complex versions of existing tasks, and it is assume that labor has a comparative advantage in these new tasks.
- When the long-run rental rate of capital is not so low relative to labor, their framework generates a BGP in which both types of innovation go handin-hand.
- The authors consider their paper to be a first step towards a systematic investigation of different types of technological changes that impact capital and labor differentially.
- Third, in this paper the authors have focused on the creation of new labor-intensive tasks as the type of technological change that complements labor and plays a countervailing role against automation.
Did you find this useful? Give us your feedback
Citations
407 citations
271 citations
219 citations
217 citations
References
6,398 citations
[...]
4,598 citations
3,234 citations
2,935 citations
"The Race between Machine and Man: I..." refers background in this paper
..., Brynjolfsson and McAfee, 2012, Akst, 2014, Autor, 2015). The recent declines in the share of labor in national income and the employment to population ratio in the US economy, shown in Figure 1,1 are often interpreted to support the claims that as digital technologies, robotics and artificial intelligence penetrate the economy more deeply, workers will find it increasingly difficult to compete against machines, and their compensation will experience a relative or even absolute decline. Yet, a comprehensive framework incorporating such effects, as well as countervailing forces, remains to be developed. The need for such a framework stems not only from the importance of understanding how and when automation will have these transformative effects on the labor market, but also from the fact that similar claims have been made, but have not always come true, about previous waves of new technologies. Keynes (1930), for example, famously foresaw the steady increase in per capita income in the 20th century from the introduction of new technologies, but incorrectly predicted that this would create widespread technological unemployment as machines replaced men. Economic historian Robert Heilbroner confidently stated in 1965 that “as machines continue to invade society, duplicating greater and greater numbers of social tasks, it is human labor itself — at least, as we now think of ‘labor’ — that is gradually rendered redundant” (quoted in Akst, 2014), while another observer of mid-century automation, economist Ben Seligman, similarly predicted a future of work without men (Seligman, 1966). Wassily Leontief was equally pessimistic about the implications of new machines. He drew an analogy with the technologies of the early 20th century that made horses redundant and speculated “Labor will become less and less important. . . More and more workers will be replaced by machines. I do not see that new industries can employ everybody who wants a job” (Leontief, 1952). This paper is a first step in developing a conceptual framework which both shows how machines replace human labor and why this may or may not lead to the disappearance of work and stagnant wages. Our main conceptual innovation is to introduce into a unified framework both automation replacing tasks previously performed by labor and the creation of new complex tasks where labor has a comparative advantage.2 The role of these new tasks is well illustrated by the technological and organizational changes during the Second Industrial Revolution, which not only involved the replacement of the stagecoach by the railroad, sailboats by steamboats, and of manual dock workers (1)Figure 1 presents the estimate trends in the employment to population ratio for potential workers aged 25-64, nonfarm business sector labor share and productivity. The trends are computed using the Hodrick-Prescott filter with parameter 6.25. See Karabarbounis and Neiman (2014), Piketty and Zucman (2014), and Oberfield and Raval (2014) for more detailed evidence on the decline of the share of labor in national income....
[...]
...6 Automation then tends to increase inequality by taking jobs from unskilled labor. The creation of new complex tasks also increases inequality at first, since skilled workers have comparative advantage in such tasks, but reduces it over longer periods as new tasks are standardized and can employ unskilled labor more productively. This extension formalizes claims in the literature suggesting that both automation and new, more complex tasks, increase inequality, but also pointing out that short-run dynamics following such technological changes might be quite different — especially from their medium-term implications in the case of new labor-intensive tasks. Our second extension establishes that under different assumptions on patents and the resulting creative destruction effects, there are similar qualitative forces, but the model might generate multiple and/or unstable steady-state equilibria. Our paper relates to several literatures. It can be viewed as a combination of task-based models of the labor market with directed technological change models.7 Task-based models have been developed both in the economic growth and labor literatures, dating back at least to Roy’s seminal work (1955). The first important recent contribution is Zeira (1998), which proposed a model of economic growth based on capital-labor substitution and constitutes a special case of our model when technology (both automation and the set of tasks) are held fixed....
[...]
...6 Automation then tends to increase inequality by taking jobs from unskilled labor. The creation of new complex tasks also increases inequality at first, since skilled workers have comparative advantage in such tasks, but reduces it over longer periods as new tasks are standardized and can employ unskilled labor more productively. This extension formalizes claims in the literature suggesting that both automation and new, more complex tasks, increase inequality, but also pointing out that short-run dynamics following such technological changes might be quite different — especially from their medium-term implications in the case of new labor-intensive tasks. Our second extension establishes that under different assumptions on patents and the resulting creative destruction effects, there are similar qualitative forces, but the model might generate multiple and/or unstable steady-state equilibria. Our paper relates to several literatures. It can be viewed as a combination of task-based models of the labor market with directed technological change models.7 Task-based models have been developed both in the economic growth and labor literatures, dating back at least to Roy’s seminal work (1955). The first important recent contribution is Zeira (1998), which proposed a model of economic growth based on capital-labor substitution and constitutes a special case of our model when technology (both automation and the set of tasks) are held fixed. Acemoglu and Zilibotti (2000) developed a simple task-based model with endogenous technology and applied it to the study of productivity differences across countries, illustrating the potential mismatch between new technologies and the skills of developing economies (see also Zeira, 2006, Acemoglu, 2010). Autor, Levy and Murnane (2003) suggested that the increase in inequality in the U....
[...]
..., Brynjolfsson and McAfee, 2012, Akst, 2014, Autor, 2015). The recent declines in the share of labor in national income and the employment to population ratio in the US economy, shown in Figure 1,1 are often interpreted to support the claims that as digital technologies, robotics and artificial intelligence penetrate the economy more deeply, workers will find it increasingly difficult to compete against machines, and their compensation will experience a relative or even absolute decline. Yet, a comprehensive framework incorporating such effects, as well as countervailing forces, remains to be developed. The need for such a framework stems not only from the importance of understanding how and when automation will have these transformative effects on the labor market, but also from the fact that similar claims have been made, but have not always come true, about previous waves of new technologies. Keynes (1930), for example, famously foresaw the steady increase in per capita income in the 20th century from the introduction of new technologies, but incorrectly predicted that this would create widespread technological unemployment as machines replaced men....
[...]
...This assumption builds on Schultz (1965) (see also Greenwood and Yorukoglu, 1997, Caselli, 1999, Galor and Moav, 2000, Acemoglu, Gancia and Zilibotti, 2010, and Beaudry, Green and Sand, 2013). (7)On directed technological change and related models, see Acemoglu (1998, 2002, 2003a,b, 2007), Kiley (1999), Caselli and Coleman (2006), Gancia (2003), Thoenig and Verdier (2003) and Gancia and Zilibotti (2010)....
[...]
2,627 citations
Related Papers (5)
Frequently Asked Questions (15)
Q2. What are the contributions in "Nber working paper series the race between machine and man: implications of technology for growth, factor shares and employment" ?
The authors examine the concerns that new technologies will render labor redundant in a framework in which tasks previously performed by labor can be automated and new versions of existing tasks, in which labor has a comparative advantage, can be created. In an extension with heterogeneous skills, the authors show that inequality increases during transitions driven both by faster automation and introduction of new tasks, and characterize the conditions under which inequality is increasing or stable in the long run. Stability is a consequence of the fact that automation reduces the cost of producing using labor, and thus discourages further automation and encourages the creation of new tasks.
Q3. What are the future works in "Nber working paper series the race between machine and man: implications of technology for growth, factor shares and employment" ?
Incorporating the possibility of such “ middling tasks ” being automated is an important generalization, though ensuring a pattern of productivity growth consistent with balanced growth is more challenging. Second, there may be technological barriers to the automation of certain tasks and the creation of new tasks across industries ( e. g., Polanyi, 1966, Autor, Levy and Murnane, 2003 ).
Q4. What is the function that determines whether this effect is complete or incomplete?
since the function Γ limits to 1 over time, the parameter ξ determines whether this standardization effect is complete or incomplete.
Q5. What is the source of non-homotheticity in the general model?
The source of non-homotheticity in the general model is the substitution between factors (capital or labor) and intermediates (the q(i)’s).
Q6. What is the first-order negative effect of automation on welfare?
But now, because workers are constrained in their labor supply choices, the lower employment that results from automation has a first-order negative effect on their welfare.
Q7. What are the new tasks that are being replaced by labor?
while industrial robots, digital technologies and computer-controlled machines replace labor, the authors are again witnessing the emergence of new tasks ranging from engineering and programming functions to those performed by audio-visual specialists, executive assistants, data administrators and analysts, meeting planners and computer support specialists.
Q8. What is the effect of rents on the labor supply relationship?
If the elastic labor supply relationship results from rents (so that there is a wedge between the wage and the opportunity cost of labor), there is an important new distortion: because firms make automation decisions according to the wage rate, not the lower opportunity cost of labor, there will be a natural bias towards excessive automation.
Q9. What is the function of the assumption that labor has a competitive advantage in tasks with a?
the authors impose the following assumption:Assumption 1 γ(i) is strictly increasingAssumption 1 implies that labor has strict comparative advantage in tasks with a higher index, and will guarantee that, in equilibrium, lower-indexed tasks will be automated, while higher-indexed ones will be produced with labor.
Q10. how much is the value of a limit price for all types of intermediates?
This ensures that the unique equilibrium price for all types of intermediates is a limit price of ψ, and yields a per unit profit of (1−µ)ψ > 0 for technology monopolists.
Q11. What is the structure of equilibrium in a task-based framework?
The authors characterize the structure of equilibrium in such a model, showing how, given factor35prices, the allocation of tasks between capital and labor is determined both by available technology and the endogenous choices of firms between producing with capital or labor.
Q12. What are the reasons why the factor prices are not compatible with the modern office?
Just to cite a few motivating examples for this assumption: power looms of the 18th and 19th century are not compatible with modern textile technology; assembly lines based on the dedicated machinery are not compatible with numerically controlled machines and robots; first-generation calculators are not compatible with computers; and bookkeeping methods from the 19th and 20th centuries are not compatible with the modern, computerized office.
Q13. How do the authors explain the production of tasks i?
Although tasks i ≤ The authorare technologically automated, whether they will be produced with capital or not depends on relative factor prices as the authors describe below.
Q14. What is the effect of automation on wage inequality?
As a result, low-skill workers are progressively squeezed into a smaller and smaller set of tasks, and wage inequality grows without bound.
Q15. What is the need for additional empirical evidence on how automation impacts employment and wages?
and perhaps most importantly, their model highlights the need for additional empirical evidence on how automation impacts employment and wages (which the authors investigate in Acemoglu and Restrepo, 2017) and how the incentives for automation and the creation of new tasks respond to policies, factor prices and supplies.