The available dataset also allows Van Reenen and Machin, as well as Berman, Bound and Machin (1998) to make considerations on the within/between industries change in skill proportion, both with respect to employment and wage-bill. INSERIRE FIGURA? Pag 1222 van reenen. The aggregate change in nonproduction wage bill share, nonproduction employment share and high education employment share are explained for the most part by within-industry changes in all the seven OECD countries considered. Berman et al. go further and argue how SBTC has a pervasive effect, backing their assertion presenting empirical evidence that the greater skill upgrading occurred in the same industries across different countries. Going back to the data collected by Van Reenen and Machin, the sectors with the biggest changes are computers and nonelectrical machinery, professional goods, paper, printing and publishing. The authors also look at cross-country correlations in industrial R&D intensities for fifteen manufacturing industries, which are all positive, large and significant at a 0.05 p-value. Again, this reinforces the conclusion that SBTC has a pervasive character.
These analyses inform us of a temporal link between the rise in skilled labor wage-bill share (as well as employment share) and R&D expenditure in several countries, although it is yet not clear what it is that connects one’s skill with the technology in use.
At this point the concept of technology-skill complementarity must be introduced. According to this notion, it is the greater value resulting from the parallel use of technologies and skills that causes technical change to become skill-biased. Different theories suggest different sources for such complementarity.
Krusell et al. (2000) developed a framework to better understand the importance of skilled biased technological change in which the elasticity of substitution between capital equipment and skilled labor is lower than that between capital equipment and unskilled labor. This means that equipment capital is more likely to substitute less educated employees than more educated ones. According to this view, the driving force of skilled biased technological change would be the cheapening of equipment capital (especially after 1975) and the consequent rise in its use (again notably after 1975). Given such rise in the stock of equipment capital and capital-skill complementarity, the consequence is an increase in the relative demand for skills and a larger skill premium.
Another formulation of the technology-skill complementarity derives from the view of human capital of Nelson and Phelps (1966), who base their argument on the shared assumptions that education enhances one’s ability to receive, decode and understand information, and that the processing and interpretation of data enables one to perform many jobs. They rank occupations based on the degree to which adaptation to change and learning during performance are required. At the bottom are functions that are highly routinized for which judgment effort required is nearly constant over time. On the other end of the scale are jobs that require the ability to exercise judgment and to keep up with the current technology and innovations. They believe that better educated people behave like innovators, by fostering the adoption and the diffusion of new technologies and they validate this idea by bringing the example of the educated and uneducated farmers. Being the educated farmer aware of the expected payoffs and risks that new products and processes might bring, and being keener in discriminating between worthy and unworthy opportunities, he will be quicker to adopt new technologies. The relatively less educated farmer, on the other hand, does not possess either the knowledge or the willingness to adopt a new technology, and he waits until he has concrete evidence of its profitability. In this framework, technology is relatively more complementary to high-skilled labor, which is less negatively affected by the disturbance of innovations and for which it is less costly to adapt to them. Two interesting considerations about this second formulation of the technology-skill complementarity are worth mentioning. First, this view is consistent with the TFP slowdown that occurred in most developed economies during the 1980s. As maintained by Nelson and Phelps, learning is a fundamental feature of new technology adoption, and learning takes time. Once innovations reach firms, before they can be included in the production processes a period of adjustment must be underwent by even the most educated employees. The second consideration regards the rise of the skill premium and its alleged provisional nature: it seems like it is only a matter of time until less educated workers catch up with more skilled ones. With time, the wage differential should disappear until the next episode of technical change. It is interesting to note the difference with the first formulation of the technology-skill complementarity, for which the effect of the falling price of equipment capital is permanent.
The third view on technology-skill complementarity considers the effect of technical change on the organizational structure of firms. Milgrom and Roberts (1990) claim that the increased use of technology reduces costs of communication, coordination and data storage, as well as costs of monitoring and supervision of employees. This allows firms to operate with a flatter organizational structure. In this new, less tall arrangement there is no need for workers to concentrate on routinized, specialized tasks: employees can now form teams and perform a variety of more integrated functions. In a way that is similar to the second formulation, more educated, skilled or able workers are likely to find it easier to perform multi-tasking activities and adapt better to this transformation.
Autor et al. (2006) develop a model to explain how a decline in the price of computer capital, also referred to as computerization, may lead to job polarization. They classify occupations in Manual, Routine and Abstract task-intensive jobs, based on their task content. Manual task-intensive occupations are non-routine, low-skilled occupations both in terms of earnings and education, e.g. those performed by truck drivers, janitors, waiters. Routine task-intensive jobs are intermediate in terms of skills required and correspond to occupations such as that of bookkeepers, cashiers, telephone operators. Finally, abstract task-intensive jobs require non-routine cognitive capabilities such as problem solving and coordination and are usually characterized by high earnings and education.
The model builds on three premises. First, computer capital is a close substitute for human labor in Routine tasks (both cognitive and manual). This complementarity is given precisely by the “routineness” (Autor, Levy and Murnane ) of these tasks: they can be expressed with rules and codified procedures and therefore can be carried out by computers, which only have to follow precise commands. On the other hand, there are some tasks that, although they result trivial to humans, are impossible to perform for machines, due to the difficulty experts face in turning them into algorithms the computers can follow. This brings us to the second assumption: Manual tasks, due to their non-routine character, are neither substitutable nor complemented by technology. The third assumption is that
Routine task inputs (performed by either human labor or computers) are complements to Abstract tasks.
As explained by Autor et al. (2003), this complementarity – which refers to the relationship between Abstract tasks and technology, unlike the one discussed in the previous section, which explains in a more general way the link between skills and technological change – takes three forms. First, by taking care of all the Routine tasks, computers allow the supply of workers endowed with Abstract skills to expand, given that skilled workers now have to concentrate on those kinds of tasks only. Second, a larger supply of Routine inputs, given by the higher speed at which computers operate with respect to human labor, increases the marginal productivity of workers performing Abstract activities that rely on those inputs. Third, because workplaces are now “computerized”, the comparative advantage of workers in such an environment becomes the ability to perform Abstract tasks.
Going back to the model developed by Autor, Katz and Kearney, workers are divided into two categories based on their educational attainment: College graduates, who are able to perform Abstract tasks and High School graduates, who can freely decide to supply Routine or Manual task inputs. As already pointed out, what aliments their model is the exogenous decline in computer capital price over the last decades of the 20th century.
Output is produced using the following Cobb-Douglas production function: with α,β,γ ∈ (0,1),α+β+γ=1. A, R and M represent respectively Abstract, Routine and Manual tasks inputs described above. While Abstract and Manual tasks are provided by labor inputs, LA and LM, Routine inputs can come from either computer capital, denoted by K and measured in efficiency units, or by suppliers of LR. Being computer capital a perfect substitute for Routine labor inputs, it is supplied in a perfectly elastic way to Routine tasks and its price is ρ, which we know is falling at an exogenous rate.
The amount of education that each worker possesses is exogenous. The fraction θ∈(0,1) represents the share of workers in the economy who are High School graduates, while the remaining 1-θ the share of College Graduates. The latter are endowed with one efficiency unit of Abstract skills, which are inelastically allocated to the supply of Abstract tasks; the former possess one efficiency unit of Manual skills, ηi efficiency units of Routine skills and 0 efficiency units of Abstract skills. Whether High School graduates will decide to devote their endowments to Routine or Manual tasks depends on the wage they would get in the two scenarios. Given w¬¬¬m and wr, wages paid to Manual and Routine tasks respectively, each HS graduate will devote their efficiency unit to Manual tasks as long as η_i
This essay has been submitted by a student. This is not an example of the work written by our professional essay writers. You can order our professional work here.