The Technology Trap
Page 30
Why, then, are not firms moving to the hinterland, where labor and housing are cheap? The answer is that innovative companies want to stay close to other such companies and skilled people. Giles Duranton and Diego Puga are two economists who have aptly conceptualized cities as nurseries for innovation, where firms in their early stages need to experiment to grow.37 During this innovative phase, firms benefit from the exchange of ideas that is facilitated by the density of cities and their proximity there to related businesses, which increases the demand for the relatively skilled. These nursery cities serve as incubators for new job creation. When a prototype has been developed and operations have become more standardized, it makes economic sense to relocate to places where real estate is cheaper and the costs of production lower. New jobs, in other words, will eventually spread to other locations. As long as nursery cities do not churn out new jobs faster than they diffuse geographically, convergence in employment will follow. But new jobs spread only as they become standardized, and since the beginning of the computer age, jobs that have become standardized have not diffused across the country: they have either been automated away or sent abroad. The flourishing cities of America have become nursery cities for innovation. But the rest is done abroad or by machines.
The places where work has been replaced, rather than complemented, by machines, are the ones that are in decline. The uneven map of multipurpose robots depicted in figure 16 makes a simple point about technological progress. Automation—like most other economic trends—will not occur everywhere in the same way or at the same pace. More than half of America’s robots are in just ten states, most of which are in the eastern heartland—where male joblessness and life dissatisfaction is the highest.38 As a matter of fact, Michigan alone has almost as many robots as the entire American West. And like the presence of robots, joblessness has not risen uniformly across America. The share of middle-aged men who do not go to work is 51 percent in Flint, Michigan, but only 5 percent in Alexandria, Virginia. The economists Benjamin Austin, Edward Glaeser, and Lawrence Summers recently conducted a detailed study of the geography of joblessness among men in their prime. Some of the authors’ findings may not come as a surprise. One is that on average, men are more likely to go to work in places where educational attainment is high.39 Another is that men are less likely to go to work if durable manufacturing was historically a key source of employment in the area where they live. Indeed, as noted above, since the dawn of the computer revolution, new jobs have overwhelmingly appeared in cities with skilled populations. Meanwhile, in places specializing in routine manufacturing work, automation has had the opposite effect: it has replaced workers rather than creating new functions for them.
FIGURE 16: The U.S. Geography of Industrial Robots, 2016
Source: International Federation of Robotics (database), 2016, World Robotics: Industrial Robots, Frankfurt am Main, https://ifr.org/worldrobotics/; S. Ruggles et al., 2018, IPUMS USA, version 8.0 (dataset). https://usa.ipums.org/usa/.
Note: This figure shows the number of industrial robots per thousand workers across America in 2016. Darker shades correspond to more robots per thousand workers. County boundaries are based on maps from IPUMS NHGIS (www.nhgis.org).
Because jobs have come and gone so unevenly across America, the computer revolution has made the country less flat. Where new tech jobs have appeared, the local service economy has been given a boost. Moretti estimates that each new tech job creates sufficient demand to support another five jobs in a given city. Where blue-collar jobs disappeared, in contrast, the local economy has taken a hit. Each manufacturing job lost, Moretti finds, costs another 1.6 jobs in the local service sector.40 This process has meant that while skilled cities have prospered, smokestack cities have become obsolete. As late as 1970, average incomes were higher in Cleveland, Ohio, and Detroit, Michigan, than in skilled cities like Boston, Massachusetts, and Minneapolis, Minesota. But in recent decades, skilled cities have pulled away while manufacturing cities and industrial towns have faded. In the words of Glaeser, “If these patterns continue, we may come to see an ever more uneven America with rich, successful, skilled regions that compete well worldwide and poor, unskilled regions that are repositories of despair.”41
11
THE POLITICS OF POLARIZATION
What happens to liberal democracy when the social fabric fractures and the middle class starts to shrink? Societies with extreme gaps between rich and poor have historically been more prone to oligarchy and populist revolutions. A broad middle class, as many political scientists have pointed out, is an essential pillar of a stable democracy. Indeed, the long persistence of extreme inequalities helps explain why liberal democracy did not arrive earlier. In preindustrial societies, the landowning elites had little interest in extending the franchise, while avoiding starvation was the main preoccupation of the poor. Without a middle class with a different set of expectations, there was little demand for democracy. Barrington Moore’s classic Social Origins of Dictatorship and Democracy is perhaps best known for the blunt remark “no bourgeoisie, no democracy.”1 Though this assertion has been subject to much criticism, Moore’s point was not that the bourgeoisie must always produce democracy. What he argued was that the replacement of the landed elite was essential in bringing democracy about, as was the case in England—where industrialization set the stage for democratic transitions.
While social scientists have long noted a close correlation between economic development and democracy, it is not immediately clear what is driving that relationship. However, one compelling explanation is that industrialization spins off new social groups that begin to demand more political power as they get richer. In Political Order and Political Decay, Francis Fukuyama provides a vivid account of how the Industrial Revolution changed the underlying nature of society in ways that challenged the old authoritarian order.2 The rise of democracy surely had much to do with the spread of values favoring equality, but these ideas didn’t come out of a vacuum. The profound changes set in motion by the Industrial Revolution not only generated sustained economic growth but also dramatically changed the composition of societies by creating and mobilizing new groups of people—notably, the bourgeoisie and the factory working class. Thus, Fukuyama’s road to liberal democracy starts from Karl Marx’s theory of social class. According to Marx, the first new social class to emerge out of the old feudal order was the bourgeoisie. This class included merchant townsmen who had grown wealthy through trade and invested heavily in the factory system and thus the Industrial Revolution. Industrialization in turn mobilized Marx’s second new class, the proletariat, whose members left rural areas for the emerging factory towns. These groups were excluded from political participation in the feudal order, but as they grew richer and became more organized, they demanded more political power—which in turn created pressures for democracy.3 Fukuyama writes:
Increasing industrialization induced peasants to leave the countryside and enter the working class, and by the beginning of the twentieth century they were the largest social group. Under the impact of expanding trade, the number of middle-class individuals began to swell, first in Britain and the United States, then in France and Belgium, and by the late nineteenth century in Germany, Japan, and other “late developers.” This then set the stage for the major social and political confrontations of the early twentieth century.… The key insight is that democracy is desired most strongly by one specific social group in society: the middle class.4
Democracy and the Middle Class: A Very Brief History
Marx predicted that the proletariat’s struggle against the bourgeoisie would prompt a socialist revolution, but in practice, industrialization meant the incorporation of working people into a broader middle class. However, as the Luddite uprisings and other machinery riots suggest, the objectives of the bourgeoisie and the average worker were hardly aligned from the onset. One obvious reason was Engels’s pause, during which labor did not see the benefits of industrialization. In fact, Mar
x’s belief that capitalism was set for a crisis of overproduction, in which mechanized industry extracted surpluses from working people, which led to greater wealth in the hands of the few and the impoverishment of the proletariat, seemed quite plausible at the time. Cascading levels of inequality, Marx predicted, would eventually create a shortfall in aggregate demand and lead to the collapse of the capitalist system. Such a crisis, he argued, could be avoided only though a revolution in which the proletariat would take ownership of the means of production and redistribute the benefits of mechanization. Had Engels’s pause persisted longer, some version of this outcome might well have occurred. But mercifully it didn’t.
We saw in chapter 5 that things changed around the mid-nineteenth century. As Marx was writing, the modern growth pattern emerged. By the closing decades of the Industrial Revolution in Britain, productivity growth had accelerated with the adoption of steam, and real wages had begun to rise in tandem, as the nature of technological progress shifted and people acquired new skills. As noted above, augmenting technical change has a symbiotic relationship with the bargaining power of labor, while worker-replacing technical change works in the opposite direction. During Engels’s pause, early textile machinery replaced skilled craftsmen in production. As the domestic system gradually vanished, new jobs emerged in the factories, but the textile machines were designed to be tended by children.
However, the arrival of more complex machines, following the more widespread adoption of steam power, broke the deskilling pattern according to which skilled craftsmen were replaced by child labor. Heavy machinery required more skilled workers of greater physical strength in the factories, and technical change turned from worker replacing to augmenting. This served to increase the bargaining power of labor, as workers’ skills became more valuable over time. And consequently, workers began to see the benefits of mechanization in their pockets. While a causal link is hard to establish, it is even harder to explain the end of widespread resistance to machinery in isolation from the arrival of the modern growth pattern. As we have seen, workers and the trade unions representing them rarely questioned the desirability of mechanization thereafter.
The modern growth pattern came about in the absence of organized labor or any significant government intervention, but that should not be taken as downplaying the importance of the labor movement. Another reason why Marx’s socialist revolution did not happen is that throughout the industrializing West, workers joined forces in trade unions and began to push for democracy, higher wages, safer working conditions, and more redistribution of wealth and income. The political branch of the labor movement de facto accepted a laissez-faire regime with respect to mechanization, but they insisted on establishing a tax-financed welfare system. From the mid-nineteenth century onward, industrialization put labor on a virtuous cycle, where mechanization increased workers’ earning power and trade unions increased their bargaining power. These two processes were mutually reinforcing. The augmenting nature of technological change made workers’ skills more valuable, thus increasing labor’s bargaining power. The labor movement successfully took advantage of this by agitating for workers’ rights to organize formally and vote.
To be sure, the two components of liberal democracy—the rule of law and universal suffrage—are distinguishable political goals. And quite naturally, they tended to be favored by different groups.5 In Britain, the commercial and industrial bourgeoisie, which provided the base support for the British Liberal Party in the nineteenth century, was more interested in legal protection for private property and free trade policies, while the working class was more interested in democracy. The Great Reform Act of 1832 was the first of a series of reforms on Britain’s long road to universal suffrage. It is justly regarded as “a major turning point in English history,” as it helped trigger democratization and key economic reforms like the implementation of the personal income tax in 1842 and the repeal of the Corn Laws four years later.6 Because Parliamentary reform had become a party-political question, supported by Whigs and Radicals (who would go on to form the Liberal Party) and opposed by most Tories, a Whig majority in the House of Commons was necessary for the Great Reform Act to make it through Parliament. The chief motive of the Liberals was to strengthen their own position by increasing the power of Parliament and reducing that of the crown, though over time some Whigs came to support expanding the franchise for its own sake. What the coalition of Liberals had in common was that they were in favor of more personal freedom, checks on the powers of the crown and the church, and above all free trade. However, their electoral success depended on working-class agitation. The economists Toke Aidt and Raphael Franck have shown that “the reform-friendly Whigs would not have obtained a majority of seats in the House of Commons in the 1831 election had it not been for the violence of the Swing riots.”7 The social unrest brought about by the rioters’ fear that threshing machines would take their jobs set in motion a process of democratization. The perceived threat of social upheaval, as political scientists have noted, prompted the ruling classes to extend the franchise.8 Voters and patrons who didn’t support Whigs and Radicals in earlier elections, the authors show, voted for the candidates promoting Parliamentary reform, but only after experiencing the violence caused by machinery uprisings, giving the reformers a majority in Parliament.
The Great Reform Act of 1832 can hardly be described as a working-class victory: people who didn’t own any property remained politically disenfranchised. But workers did help set the wheels of democratization in motion. From then on, electoral competition between the Liberals and the Tories created pressures for the continued extension of the franchise, meaning that labor gradually came to represent a growing percentage of the electorate. This was spurred on by the struggles to expand trade unions, which prompted the rise of the Labor Party and eventually its displacement of the Liberals as the main opposition to the Tories. However, in the mid-nineteenth century, long before the rise of the Labor Party, the Tories had already shifted their position from representing wealthy landowners to gathering support from the new middle class. As the historian Gertrude Himmelfarb has argued, Benjamin Disraeli’s decision to pursue the Reform Act of 1867, which extended the franchise to about 40 percent of the male population, reflected the belief that the Tories were a national party, capable of appealing to a wider share of the populace.9
As more and more people won the franchise, workers in the industrial West used their newfound political power to vote for welfare state policies and social legislation. In Growing Public, Peter Lindert shows that between 1880 and 1930, heavily voting democracies taxed more and spent more on social transfers—such as aid to the poor, elderly, sick, and unemployed.10 One reason why welfare spending was so restricted before the twentieth century, Lindert convincingly demonstrates, is that workers lacked political voice. In the classic years of the Industrial Revolution, when Engels’s pause persisted, the British government even rolled back welfare spending. Before the Great Reform Act, Britain was the one place where poor relief had become much more generous. In the light of workers’ lack of political clout, this might seem puzzling. Why did landowners tax themselves 2 percent of gross domestic product (GDP) to help the poor? At the time, no government had ever done so.
Following in the footsteps of George Boyer, Lindert argues that rural landlords, who depended on seasonal hires, had an interest in keeping farmers in the area to secure access to cheap labor. Poor relief provided a way of keeping workers from emigrating to Britain’s emerging industrial centers. But things changed as the new merchant class gained political clout. The Great Reform Act extended the franchise to merchant manufacturers in urban centers. Industrialists saw little benefit in supporting a system that kept workers in stagnant rural areas. The rise and fall of the Poor Laws, in other words, were shaped by the self-interest of the groups that controlled the levers of political power.
* * *
While some people have always worried about the wealthy using their econ
omic power to bend democracy to their own ends, there has also been concern over voting majorities using their democratic power to tax away the fortunes of the rich. In Democracy in America, published in 1835, Alexis de Tocqueville noted that universal suffrage, which “gives the government of society to the poor,” will lead to the poor redistributing the wealth of the rich because in virtually every nation, the vast majority of citizens do not have any property.11 “Tocqueville’s observation has received modern expression, with an optimistic twist, in leading theories of income redistribution offered by political scientists,” in the words of Jacob Hacker and Paul Pierson.12 Median voter theories, so popular among political scientists and economists alike, build on the premise that in a majority-rule voting system, the decisive swing voter—or the median voter—ultimately drives politics and the supply of redistributive policies. Because the median voter almost invariably earns less the than the national average, he or she will seek more income redistribution through government. This theory then leads to the simple prediction that more redistribution will follow from greater inequality.13