Introduction
In this chapter we discuss the recent three decades of data on technology, productivity, and labor market outcomes. In particular, we explore the hypothesis that technological change has affected the labor market in various ways.
We argue that (i) there is ample evidence indicating significant capital-embodied and/or skill-biased technological change and that (ii) this kind of technological change would plausibly lead to many of the transformations in the labor markets that we have observed. On the one hand, we are interested in possible implications of non-neutral technological change - of the kind we think we have experienced - on variables like wage inequality, unemployment, labor share, and unionization. On the other hand, we explore the possibility that the labor market can be used as an additional source of evidence of non-neutral technological change, a testing ground of sorts.The past 30 years are particularly informative because they have contained rather important trend changes in several variables. We have seen a productivity slowdown common to all industrialized countries and common to almost all industries, together with continuing structural change away from manufacturing and toward services. An exception to this widespread productivity slowdown was the fast and accelerating productivity growth of industries producing investment goods, in particular those producing equipment. Only very recently has there been a more widespread acceleration of productivity growth. Of course, in this context we are arguably in the midst of an “Information Technology Revolution”. We also discuss evidence of changes in the workplace - in how production within firms is organized - possibly reflecting underlying changes in technology.
In the labor market, we have seen a sharp increase in wage inequality in the United States contrasting a roughly flat development in Europe, whereas we have witnessed a strong increase in European unemployment and no trend in U.S.
unemployment.[164] The organization of labor markets seems to have changed too: for example, unions have lost prevalence during this period, and to the extent there have been unions, centralized bargaining has been replaced by decentralized bargaining in many sectors. Are all these developments consistent with basic economic theory and a short list of underlying technological driving forces? We argue that they are. To make our argument more convincing, we also put the past three decades in a historical perspective, going back as far as the early 20th century with data on technological change and the skill premium.One distinctive feature of this literature is that the many different ideas have been presented in a wide variety of theoretical frameworks ranging from the neoclassical Ramsey-Cass-Koopmans growth model to the Schumpeterian endogenous growth model; from the traditional McCall search model to the Lucas-Prescott island economy; from the Mortensen-Pissarides matching model to the competitive directed search framework; and from the Bewley-Aiyagari incomplete-markets model to Arrow- Debreu economies with limited enforcement. We think two main reasons exist for the lack of a unified framework of analysis. First, this field of research is still relatively young; second, departing from the competitive model in studying labor markets is fairly natural, and many alternative frameworks exist that incorporate frictions. The main drawback of the lack of a unifying framework - we will repeat it often in the chapter - is that making structurally based quantitative comparisons between different mechanisms is difficult.
To us, these heterogeneous approaches pose a formidable challenge in the exposition. Our solution has been to give priority to presenting a range of ideas, using a variety of theoretical setups, rather than to discuss in great detail a few more specific frameworks. This approach has necessitated a summarization of some rather rich models in a few key equations, which misses some of the elegance and richness of the original frameworks.
We hope, however, that our spanning a wide spectrum of ideas and macroeconomic effects of technological change helps paint a picture that is broader and that, at least in an impressionistic way, suggests that the main underlying hypothesis we are proposing is quite reasonable.The presentation of the ideas in this chapter is organized into four parts. In the first part, Section 2, we review the main trends in the data on technological change and labor market inequalities. We then cover two kinds of theories that could account for the data.
In the second part of the chapter we cover “neoclassical” theory (i.e., models where wages directly reflect marginal productivity). We view the firm as hiring labor of different skill levels in a competitive and frictionless labor market. Wages, thus, will be influenced by technology in a very direct way. Similarly, the returns to education, ability, and experience, which we discuss in detail in Sections 3 and 4, respectively, will be directly tied to changes in technology. Therefore, within these kinds of theories, the shape of the production function of the firm is crucial. We then move beyond the production function of the firm or, rather, we attempt to go inside it. In particular, Section 5 explores the possibility that the organization of the workforce also has changed within firms. These transformations, of which there is some documentation, are arguably also a result of the kind of technological change we look at in this chapter. We point, in particular, to a recent literature that explores how firms are organized and how the IT revolution, by inducing organizational changes in the firm, had a substantial impact on wage inequality.
The second class of theories we cover, in the third part of the chapter, relies more on frictions in the labor market and deals more directly with how this market is organized. Here, technological change can still directly influence wages but there are new channels. For one, wages may not only reflect marginal productivity.
Moreover, now unemployment is more in focus and is a function of technology, and since unemployment - through workers’ outside option - may also feed back into wages, the picture becomes yet more complex. In the context of how wages are set, we furthermore argue in Section 6 that the importance of unions and their modus operandi are influenced by technology and, more generally, that labor income as a share of total income may respond to technological change in the presence of unions. An important point that we make in Section 7 is that “luck” can be a key part of wage outcomes for individuals active in a labor market with frictions, such as the search/matching frameworks, and that the “return to luck” can be greatly affected by technology as well. Finally, government participation in labor markets - labor-market “institutions”, in the form of unemployment benefits, firing costs, and so on - likely interacts with technology in determining outcomes, and Section 8 completes the third part of this chapter by analyzing the interaction between technological shocks and labor market institutions in the context of the comparison between the United States and Europe.The fourth and final part of the chapter asks the “So what?” question: given the significant transformations observed, is a government policy change called for? Our discussion here is very brief. It mainly points out that a basic element underlying any decisions on policy, namely, what the welfare outcomes of the changes in wages, unemployment, and so on, are for different groups in society, is studied only partially in the literature so far. Studies of changes in expected lifetime income of different groups exist, but it is reasonable to assume that risk matters too, especially with trend changes as large as those observed (at least to the extent they are hard to foresee and insure). In Section 9 we therefore cover some examples of more full-fledged attempts to look at the distribution of consumption and welfare outcomes of the changes in technology/labor market outcomes. Finally, Section 10 concludes the chapter.
2.