REDISTRIBUTION IS indeed necessary. And though it remains uncertain what the best redistributive mechanisms are—whether some form of universal guaranteed income or state provision of welfare programs—the provision of goods must go beyond the means for basic sustenance. To lead a worthwhile life, people need to have access to the material and educational means for constructing meaningful lives for themselves. To help them in this endeavor, policymakers should massively expand access to the arts, to education and to other cultural programming.

At the same time, concerns (shared by many conservatives) about the dangers of too much power in a centralized authority are not without merit. That’s why much of this money should be allocated to cities, regions and nonstate actors like voluntary associations. Towns and individuals, states and NGOs, should have the freedom to experiment. And by pairing locally administered public programs with more participatory means of governance, people will be able to decide on spending priorities and have a greater stake in their communities.

All of this will, to be sure, require high levels of taxation on the owners of capital and equipment. Such transfer payments do, admittedly, create risks—the most obvious of which may be that a generous welfare state with high taxation could make the bulk of the population unwilling to work. This won’t be a big problem, however. Many people will still want to work because of the prestige and purpose associated with the scarce employment opportunities in an automated economy. Progressive labor policies could increase the amount of work sharing, giving more people the opportunity to do part-time and limited work that still leaves them with ample leisure time. And since the overall demand for human labor will rapidly shrink in an automated economy in any case, it would be a boon, not a problem, if large sections of the population decide to opt out of work entirely.

Obviously, there is a lot more work to be done in thinking through how forms of redistribution might preserve a broad-based middle class in a heavily automated economy. Nevertheless, the policy challenge is ultimately much less serious than the political challenge. The really daunting problem is not how to design tomorrow’s welfare state; it is to make political movements shaped in an industrial age, and only just starting to understand the political realities of a service economy, accept that a radically expanded form of welfare for the middle class will be needed in the automation age.

For now, most politicians on both the left and the right continue to believe that their constituents can preserve a broad middle class in the same way in which it was initially created: by means of a mass-labor economy that creates rapidly rising wages. But to help the middle class—and democracy—survive, politicians need to recognize that job creation can no longer be the principal tool, nor the foremost goal, of economic policy. Redistribution on a significant scale is indispensable.

But while redistribution is economically feasible, it is far from clear that it will be adopted in time—and the grounds for pessimism are especially strong in the United States.

The first and most significant reason why America may have special difficulty adapting has to do with campaign finance. A growing body of empirical research makes clear that wealthy donors have outsize power both in selecting their favorite candidates and in swaying actual policy outcomes. In America’s privately funded electoral system, which allows for unlimited contributions by the superrich, a small slice of very wealthy donors can thus play a powerful gatekeeping role, effectively disbarring any politician who threatens their interests. And since elites would be hard hit by any redistributive policy, they have strong material incentives to avoid redistribution—as well as strong psychological incentives to convince themselves (and the politicians they select) that people can adapt to the changing economic circumstances through small adjustments.

A second reason is that Americans are more reluctant than their European counterparts to turn to the state to solve their economic problems. This distrust in government has led to increased outsourcing and downsizing of government, which has effectively crippled the capacity of the state to accomplish the kind of necessary large-scale intervention that is needed to confront automation. Worse, it has limited the capacity of policymakers to build up their own in-house expertise, increasingly forcing them to rely on whatever policy resources are provided to them by the private sector. This may lead politicians to persist with proincumbent business policies that are no longer up to the task of dealing with the much more fundamental challenge facing us in the coming decades.

Finally, America has more veto powers than any other developed democracy, favoring continuity over change, and the worsening status quo over desperately needed experiments with new policies. All in all, the picture isn’t pretty. In a gridlocked two-party system, crippled by a massive lobbying community that has little incentive to embrace large-scale redistribution, the possibility for major action seems low indeed. The political challenges of the automation revolution are not only vast. Worryingly, they are of a kind that the American political system will find especially difficult to meet.

IF PESSIMISTIC predictions about the economic consequences of automation come to fruition, the future looks bleak. Even so, some hope remains. Democracies, after all, have proven surprisingly resilient in the past. In economic crises, leaders tend to dig in their heels, insisting on an increasingly pure form of their ideology. But when the crisis lasts long enough, their ideological rigidity can give way to surprising moments of flexibility, especially if public opinion demands it. In the age of the Great Depression, FDR challenged the conventional wisdom (even within his own party) that markets are self-correcting, and revolutionized the state’s role in the economy. In the stagflation of the 1970s, Nixon, to the horror of many Republicans, implemented wage and price controls.

Public-opinion trends suggest that the foundations for a similar moment of ideological flexibility are growing. With only a small number of Americans experiencing increases in their living standard, and a majority getting poorer, deep economic anxiety has spread throughout society. Under the right circumstances, that anxiety could be channeled into support for the kind of redistribution that the automation age will necessitate. Indeed, a much wider portion of the electorate already favors higher taxes on the rich and expanded welfare benefits for ordinary citizens than the manifestos of mainstream politicians would suggest. If the public were to vote directly, more redistribution would be adopted in a landslide.

Moreover, while the promise to create more jobs still dominates contemporary politics, citizens are undoubtedly open to a future in which they might be freed from the necessity of toil. The current crop of politicians may talk up job creation—but polls also show that most Americans don’t like their jobs. As one recent Gallup poll found, nearly 90 percent of workers described themselves as either “not engaged” or “actively disengaged” from their work. A politician who would advocate unleashing the robots, promising a combination of generous welfare benefits and a radically shortened workweek—or even a universal basic income—might quickly build a substantial following.

Another reason to hope that there may be more potential for ideological flexibility in American politics than meets the eye is that both the left and the right can draw on their own traditions in imagining new departures. The left’s attachment to labor, for example, is long and it goes deep. And yet, work has, at the most fundamental level, always been the condition rather than the goal of its core constituency. At its origins, in the early nineteenth century, the left either condemned industrial society or saw it as a necessary step in the progression towards a materially plentiful future. The Luddites attacked the machines, the Utopian socialists imagined communities where toil would no longer be required, and even the mainstream of the worker’s movement long preserved the promise that work would eventually cease. Karl Marx not only predicted that the state would eventually fade away but that any need for real work would, too. The world in which one could decide to hunt in the morning, fish in the afternoon and be a critical critic in the evening is, for all intents and purposes, the world of automation.

But just as the withering away of the state receded into an ever more distant future in real-world Communist regimes, so too did the withering away of work eventually drop out of the imagination of the global left. Since the voters of the left have, over centuries, spent much of their time at work, and since that work was often backbreaking, it is only natural that the left not only aimed to improve the conditions of that work—but also came to invest the activity itself with dignity and meaning. In this way, a movement that had been rooted in radical criticism of the degradation of the human spirit, involved in forcibly having to accept employment in a capitalist economy, slowly morphed into a movement devoted to expanding the number of people who would find employment in a capitalist economy.