One reason why grand theories of capitalist development are unlikely to work for every place at every time has to do with the bargaining power of labor. As noted above, labor unions gained in political power after the New Deal. And as the economist Henry Farber and colleagues have shown, higher union density is associated with lower levels of wage inequality.71 Another reason is that technology works in mysterious ways. Technological progress has at times replaced workers, putting downward pressure on wages and thereby serving to increase the share of national income that goes to capital. And at other times, it has favored labor. The relationship between technological progress and the income distribution is not monotonic: some technologies may increase inequality, while others reduce it. It depends on the degree to which technological changes are replacing or enabling. And it depends on whether the supply of workers with the right skills is able to keep pace with demand.
In the end, for most people, the main source of their income is not physical or financial capital, but human capital. The wealth of workers is in their skills. It is from their human capital that they make their living. How human capital can affect the income distribution is not hard to see: empirical studies have shown that 77 percent of the variation in workers’ earnings stem from individual characteristics.72 As we shall see in chapter 10, the lack of education and training can even exclude entire social groups from the growth engine.
Most available evidence suggests that technological change was primarily skill-biased in the period 1870–1970, serving to drive up the wages of skilled workers relative to those of the unskilled. But if technology increased the relative demand for skilled workers, why did wage inequality not grow as well? The leading explanation for the great levelling comes from pioneering work by Jan Tinbergen that conceptualized patterns of inequality as a race between technology and education.73 Empirical work by Claudia Goldin and Lawrence Katz, two Harvard University economists, has shown that this view does a good job of explaining patterns of American wage inequality up until the 1970s.74 Indeed, even if technological progress favors skilled workers, growing wage inequality does not have to be the result. The return to human capital depends on demand as well as supply. As long as the supply of skilled workers keeps pace with the demand for them, the wage gap between skilled and unskilled workers will not widen. While a number of short-run events and government interventions contributed to the great leveling, the most pervasive force—and certainly the best documented one—behind its long-run egalitarian impact was the upskilling of the American workforce, which depressed the skill premium.75 Enabling technological change and the expansion of education provided the principal forces for convergence. From 1915 to 1960, the relative skill supply grew about 1 percent ahead of demand on an annual basis, causing wage differentials to become compressed. This pattern stands in contrast to the post-1980s period, when the demand for skills outpaced their supply.76
While the demand for skills was in part met by market forces, its supply also depended in large part on educational policies and new institutions that served to increase access to education and training. One institutional invention in particular was essential to making the gains from technological inventions widely shared: public schooling. The period 1910–40 is commonly referred to as the high school movement in American educational history. In 1910, only 9 percent of American youths obtained a high school diploma. In 1935, 40 percent did. In the period 1900–1970, over 70 percent of the increase in years of schooling was due to secondary school attendance. Some places adopted public schooling faster than others. The early adopters were characterized by greater community stability, a higher degree of ethnic and religious homogeneity, income and wealth equality, and higher levels of wealth. Thus, in short, social and financial capital helped foster the formation of human capital.77
The obvious reason why the high school movement happened was demand. Education was the best investment people could make for their children in the age of the Second Industrial Revolution. In the period 1890–1920, white-collar occupations, which typically required a high school diploma, paid about double the amount of occupations that didn’t. High schools soon became a training camp for work—and for life more generally. As late as 1870, most Americans still worked in agriculture, and those employed in manufacturing worked in industries of the Industrial Revolution such as cotton, silk, and woolen textiles. All of these jobs required little formal training: only 10 percent of working Americans were employed in jobs that required education beyond elementary school. Child labor remained a persistent opportunity cost to education. By the turn of the century, boys as well as adult women typically earned half the hourly wage of adult men. Though school attendance was already compulsory in most states, some were reluctant to enforce it, as industry depended on access to cheap labor. Consequently, especially in the South, where child labor persisted longer, inequality was transmitted between generations as only relatively well-off families could afford to send their children to school.
A virtue of the Second Industrial Revolution was that it helped reduce the opportunity cost to education by raising skill requirements. In 1900, only four women and six boys were employed in the automotive industry across the entire country. For Americans who wanted their children to escape the drudgery of farming, a high school degree was a ticket to better-paying work and gradually became essential for the majority of jobs. Office workers, including bookkeepers, clerks, and managers, found themselves handsomely paid for their education. High school education was less common among blue-collar workers, but those who did stay on in school were overwhelmingly found in jobs associated with the Second Industrial Revolution. They were electricians, auto mechanics, electrical engineers, machinists, and so on. In 1902, a manager at the Decree Tractor Company made clear it that he would not “take boys in the office unless they are at least high school graduated.… In the factory we like the boys to have a high school education if possible.”78 By 1920, a full quarter of the labor force was working in occupations in which at least a high school degree was expected. Some leading technology firms, like General Electric, even required some years of high school for their apprentices. And the demand for skills only continued to expand thereafter. In the petroleum industry, for example, educational standards were continuously rising in the postwar period, for production and supervisory workers alike. At one refinery, management made a high school education a job requirement across the board in 1948. And in 1953, a preemployment test was introduced for applicants for production jobs, aimed at determining “an individual’s ability to memorize, concentrate, observe, and follow instructions”; it also tested mathematical knowledge, such as algebra and geometry, at the level of a high school sophomore.79
Another survey, looking at the introduction of automatic reservation systems in the airline industry, similarly shows that more sophisticated skills were demanded as technology progressed. With airline traffic increasing to fifty million passengers in America in 1957—up by more than 300 percent over just a decade—manual methods constituted a key bottleneck to the capacity of airline companies to handle flight reservations. The new system significantly changed the content of jobs and the need for training:
Classes for training instructors began while the equipment was being installed.… The airline lengthened this indoctrination training—which had covered from 5 to 7 days—to 8 to 10 days. After a week of subsequent on-the-job training, under his supervisor, the employee receives an additional 26–33 hours of advanced classroom instruction.… Seven new technician jobs were set up in connection with maintaining the new system. The technicians, who were previously employed as repairmen in the airline’s radio shop, had worked directly and constantly on equipment. In contrast, the technician now works alone in an air-conditioned, noiseless control room. He works in his street clothes, and the only time he has direct contact with the automatic equipment is during preventive maintenance tests or on occasions when the equipment is out of order. The technicians were given specialize
d training by the manufacturer of the system, and attended classes 1 day a week for about 6 months.… A group of professional jobs concerned with electronic data-processing research was also created, following the advent of the new reservation system. This group is comprised of five systems engineers. These professionally trained persons perform duties which involve planning systems development and extending electronic methods to all clerical activities of the company. Their annual salaries start at $7,000. The qualifications for systems engineers include education at college level and cover a variety of airline experience. It is interesting to note that 4 of the 5 men in the group have college degrees in business administration and the social sciences. All have had considerable and varied work experience with the company.80
In contrast to grand theories of capitalism, the race between education and technology is a simple and robust empirical observation. It does not by any means exclude the possibility that other forces shape the trajectory of American inequality. Macroeconomic shocks, unions, tax policy, and financial-sector regulation, just to name a few, have all played a role in shaping American inequality. Even Kuznets and Piketty in their earlier work pointed to factors beyond their later theories. Before Kuznets advanced his exceedingly optimistic theory asserting that capitalist development will cause inequality to automatically fall in the long run, he explicitly mentioned the role that economic shocks might play. And in an earlier work with Emmanuel Saez, Piketty suggested that “one could indeed argue that what has been happening since the 1970s is just a remake of the previous inverse-U curve: a new industrial revolution has taken place, thereby leading to increasing inequality, and inequality will decline again at some point, as more and more workers benefit from the innovations.… Explanations pointing out that periods of technological revolutions such as the last part of the nineteenth century (industrial revolutions) or the end of the twentieth century (computer revolution) are more favourable to the making of fortunes than other periods might also be relevant.”81
This is also the favored interpretation of the economist Branko Milanovic, who has recently put forward the idea of Kuznets waves accompanying every new technological revolution. His work does indeed show that the trajectory of inequality in Britain during the Industrial Revolution looks astoundingly similar to that of the computer revolution in America.82 However, such an interpretation immediately raises the question of why American inequality during the Second Industrial Revolution seemingly followed a different pattern. The reason is that different economic models apply to different technological revolutions. As we have seen, the race between technology and education does a good job of explaining trends in the labor market over the first three-quarters of the twentieth century. But such models only apply when technological progress is of the enabling sort. This stands in stark contrast to the first seven decades of the Industrial Revolution in Britain, when the technological progress was primarily replacing and many people adjusted poorly (chapters 4 and 5). In this regard, as we shall see in chapter 9, the computer revolution more closely resembles the experience of the Industrial Revolution.
Conclusion
The Industrial Revolution didn’t create the middle class, but it surely facilitated its growth. The spread of the factory system prompted the rise of industrial capitalism, and with it the expansion of the commercial and industrial bourgeoisie. Yet the history of the Industrial Revolution was not just the triumph of capital. The fact that the term “white-collar” first entered common usage during the first half of the nineteenth century suggests that labor markets were undergoing rapid change as industrialization picked up in pace. By the mid-nineteenth century, white-collar jobs supported a relatively prosperous group of families that we could also term middle class. The ascent of mechanized industry was accompanied by wage polarization as the earnings of white-collar workers pulled away from those of production workers. As discussed above, mechanization during the classic years of industrialization replaced relatively skilled artisan craftsmen with machines operated by the less skilled. Middle-income craftsmen saw their jobs disappear as mechanized factory production took over. Artisans of all kinds—cabinetmakers, watchmakers, shoemakers, and so on—closed their shops as the factories produced ever-growing quantities of goods. But as the establishments became larger and required more professional administrators, white-collar jobs expanded from 1850 onward. A more nuanced picture shows a hollowing out of the labor market, to the detriment of artisan craftsmen but to the benefit of the white-collar middle class. In America, newly collected data shows that the wages of white-collar workers were already growing steadily before the Civil War.83
Thus, in the nineteenth century, the ranks of the middle class were swelled by a growing number of managers and other skilled professionals, who took on an array of increasingly complex administrative and managerial tasks in the ever-growing factories. The best evidence of their societal prominence and relative affluence may be archaeological. The conditions in which the middle class lived in the late nineteenth century are most visible when one walks the streets of the old manufacturing towns in the eastern United States, from Cambridge, Massachusetts, to Hartford, Connecticut. Members of the middle class “were the first residents of the iconic nineteenth-century brownstones of New York City and the substantial Italianate and Queen Anne houses built in cities and towns throughout the northern tier of states.”84 Yet as a percentage of the American population, the middle class was still a small group. Occupational statistics provide some perspective on the percentage of middle-class households. In 1870, at the dawn of the Second Industrial Revolution, 8 percent of American workers were classified as managers, professionals, or proprietors (see table 1).
While the wages of production workers were lagging behind, the machines that replaced artisans could not run on their own. Working-class people, who gathered in the emerging factory towns, were the ones who ran the machines. The American urban working class first appeared in the early 1800s, when the economy began to industrialize, and expanded enormously over the next century as manufacturing jobs attracted millions of migrants from Europe and the American hinterland. The factories required operatives, who were “less skilled than the artisans they displaced in the sense that an artisan could fashion a product from start to finish, while the operative could perform a smaller set of tasks aided by machinery.”85 This is not to suggest that operatives did not have any skills. Factory workers had to learn how to operate the machines on the job. Though early textile machines were designed to be tended by children, the adoption of steam power eventually increased the demand for skilled operatives. In chapter 5, we saw that Engels’s pause persisted as long as technological change served to replace skilled artisans with children. This was also the time when British workers regularly rioted against the mechanized factory. But things changed with the adoption of steam power, as adults regained their comparative advantage in production, where they found their skills augmented by increasingly sophisticated machines. Larger and more complex machinery also meant that a growing number of skilled engineers and mechanics were needed to design, install, and maintain the equipment. Thus, though the wages of the white-collar workforce rose significantly relative to those of the blue-collar workers below them, the wages of production workers—like machinists, furnace men, and textile weavers—increased during the second half of the nineteenth century. To be sure, the sons of manual workers rarely moved up into the ranks of the middle class. The worlds in which the white-collar middle class and production workers lived were still separated in the late nineteenth century.86 Working-class men and women could at best aspire to a middle-class lifestyle, as could their children. But the Second Industrial Revolution meant that they finally could achieve it.
The technologies that defined the twentieth century allowed the production workers to attain a lifestyle superior to that of the upper classes in the early nineteenth century. Luxuries like hot running water and central heating systems began to be installed in large homes of the wea
lthy in the 1880s, and soon trickled down to working-class homes in the early twentieth century.87 At the same time, an array of electrical inventions became available to American households during the early part of the twentieth century, relieving working-class housewives of some of their burdens. Mass production naturally targeted the mass market, and other key inventions, like the automobile, soon became available to the bulk of the population as well. As Gordon’s account of the automobile revolution illustrates, there was a car for everyone:
The Cadillac, Lincoln, and the Chrysler Imperial were for the ancien régime of inherited wealth and for the executive suite. The four-hole Buick Roadmaster connoted the vice president, while the three-hole Buick Century was for the rising midlevel executive, the owner of the local retail business or restaurant. Farther down the perceived chain of status were the Oldsmobile, the Pontiac, and the ubiquitous Chevrolet, America’s best-selling car year after year, eagerly bought by the new unionized working class that, in its transition to solid middle-class status, could afford to equip its suburban subdivision house with at least one car, and often two.88
Working Americans did not just gain from technological change in their capacity as consumers. Perhaps more importantly, twentieth-century mechanization was primarily augmenting, and the few who lost their jobs to machines mostly faced decent alternative job options, as is reflected in the unprecedented wages taken home by blue-collar Americans: “As the wages of industrial workers increased in the thirty years after World War II, husbands were increasingly able to support a family lifestyle that included a modest home, a car, ample food and clothing, and perhaps even a vacation trip using the camper sitting in the driveway; more and more working-class families had enough income and consumption to reach the lower rungs of the broad American middle class.”89 The baby boom was in part a reflection of the growing optimism of young families, which created additional demand for goods and services and prompted the continued expansion of manufacturing and the creation of new labor-intensive services. During this period, a young male high school graduate could expect to find a secure job at a decent wage. The American economy was able to generate sufficient opportunity for blue-collar workers to attain a middle-class lifestyle on the basis of nothing more than their wages. The middle class, at its peak, was a diverse blend of white- and blue-collar workers. The result is reflected in the compression of the income distribution, prompting President John F. Kennedy to note that “a rising tide lifts all the boats.”90 Members of Marx’s proletariat began to join the ranks of the middle class, which explains why worker resistance to mechanization became a distant memory.
The Technology Trap Page 25