As summarized earlier, the causes of economic change can be divided between shocks and policies. Let us start with the shocks.
1.1.1 Shocks: industrial revolutions (and the pandemic)
It is certainly true that economic policies, and in particular the modification of these policies which started in the 1980s, played a significant role in determining the current evolution of the world economy, but these policies would have been impossible, or would have resulted in different outcomes, without the rapid spread of technological innovation throughout the world.
In spite of the unanimous recognition of the crucial role played by technological progress, the period from the last two decades of the twentieth century to the first two decades of the twenty-first century often lacks adequate interpretation since scholars tend to confuse two rather different waves of technological progress: the third and the fourth industrial revolutions. This confusion results from a considerable degree of overlap between the two phenomena, since both occurred close in time and used similar technologies. But, albeit apparently similar, the last two industrial revolutions are in fact crucially different from one another and impacted economic reality in rather different ways.
It is no simple task to propose comprehensive labels and time spans for complex processes like industrial revolutions. In any case, the beginning of the first industrial revolution is conventionally placed in 1770 and was based on the development of the steam engine, whereas the beginning of the second industrial revolution is placed in 1870, and was based on the widespread development of electric power. During these revolutions, technological progress basically relocated workers from one sector (at first from agriculture, later from manufacturing) to another (in the beginning, to manufacturing, later to services) with a temporary increase in short-term unemployment but no notable impact on long-term unemployment. Furthermore, generally speaking, this displacement had a positive impact, at least in the longer term, on the income of the relocated workers.
The beginning of the third industrial revolution is usually placed in the 1970s and was characterized by the diffusion of the digitalization of information. The digitalization of information had a considerable impact on consumption behaviors and product innovation, with the invention of personal computers, cellular phones, digital cameras, etc. But it had the greatest impact on the industrial structure: productive processes became more flexible and required more skilled and fewer unskilled workers (and/or fewer workers performing routine tasks).
The consequences of the latter process can be, and indeed have been, interpreted in two different ways. Some scholars 1 refer to it as skill-biased technical change, 2 or skill-biased technological change, a process that caused a reduction in the demand for unskilled workers, which could easily be replaced by machines, and an increase in the demand for skilled workers, i.e., workers capable of using the new machines. As a consequence, the wages of unskilled workers fell while the wages of skilled workers rose, so that wage polarization and inequality also increased. 3 Other scholars 4 maintained that such an approach was unable to grasp some important empirical evidence, and focused instead on a slightly different scenario, which they called task-biased technical change or routine-replacing technical change. According to this latter interpretation, technological progress mainly impacted negatively on workers performing routine tasks by bringing their wages well below those of workers performing nonroutine tasks. 5
It is worth noting that skill-biased technical change and routine-biased technical change are not simply different ways of interpreting the same phenomenon, but rather different phenomena that impact in different ways on the labor market. This is so since skill-biased technical change impacts positively on wages and income for higher-skilled workers, and so ameliorates the situation for countries with a higher share of skilled workers over the total workforce; symmetrically, it impacts negatively on wages and incomes for lower-skilled workers, and so worsens the situation for countries with a smaller share of skilled workers over the total workforce. Thus, this approach sees an increase not only in inequality between low- and high-skilled workers but also in the inequality between skilled countries and unskilled countries. On the contrary, task-biased technical change impacts on the single tasks of the production process, and in particular on routine tasks, that can more easily be carried out by machines or computers (software programs) independently of whether the same tasks had previously been carried out by low- or high-skilled workers. In this case, inequality will increase between workers performing different tasks and not between skilled and unskilled workers or countries: technological progress hits workers performing a single task within a given category of workers, e.g., bank tellers, but not workers performing a different task within the same category of workers, e.g., bank clerks. Furthermore, it can no longer be said that skilled workers and countries can avoid the negative impact of technological progress while unskilled workers and countries cannot: acquiring skills by investing in human capital might therefore not be a solution.
For the cases of both skill-biased and task-biased technical change, the general idea, supported by many empirical studies, 6 is that technological progress spreading from the third industrial revolution may have raised unemployment and lowered wages for unskilled workers and/or workers performing particular (generally routine) tasks, but has not increased long-term unemployment in general, since redundant workers have easily been reabsorbed in other jobs or tasks (albeit often with a lower wage or worse working conditions).
So, the third industrial revolution may have contributed, together with other causes (in particular, globalization and markets liberalization), to wage polarization and to increase in inequality (between workers and countries), but not to increase in unemployment. We will describe the consequences of technological progress for the labor market in greater detail in Section 1.2.
Apart from the impact on the labor market, the third industrial revolution also had a significant impact on other economic phenomena, most of which linked to the globalization process. In particular, thanks to ICT technological progress it became possible to move financial capital from one country to another almost instantaneously, to implement just-in-time production, to develop new financial instruments, and to coordinate and facilitate the management of multinational companies operating over different countries (thus opening the way beyond Coase’s idea that there is a bureaucratic limit to the size of a firm). All these are characteristics of globalization, which we discuss in Section 1.1.3.
However, the ICT revolution also made room for another, more recent, industrial revolution, the fourth industrial revolution (Schwab 2016). At the center of this new wave of technological progress is, again, the development of ICT (in this case in the form of artificial intelligence) and the entry of robots endowed with artificial intelligence into the productive process. Robots and artificial intelligence change everything, and in particular change the impact of the new technologies on wages, employment, and the labor market.
A number of different definitions of artificial intelligence can be found in the theoretical literature (see, e.g., Wang 2008). According to Minsky (1968, p. v), artificial intelligence is “the science of making machines do things that would require intelligence if done by men. It requires high-level mental processes such as: perceptual learning, memory and critical thinking.” For the limited purposes of this book, we will define artificial intelligence as the capacity of computer systems to follow a principle of rationality and reasoning similar to that followed by the human mind. Whereas, according to Freeman (2015, p. 2):
[t]he term “robots” refers broadly to any sort of machinery, from computer to artificial intelligence programs, that provides a good substitute for work currently performed by humans (…) it does not matter whether a robot/machinery has a humanoid appearance, as long as it can perform human functions.
In what follows we will refer mainly to robots endowed with artificial intelligence as the key feature of the new process, ...