“Amazon needs only a minute of human labor to ship your next package,” read a CNN headline last October. The company has revolutionized its warehouse operations using an army of 45,000 robots and other technologies. Previously workers known as “pickers” would walk among shelves to find goods. Now robots bring the shelves to them; pickers select goods, scan them, and put them into bins; after robots whisk the shelves away. A network of automated conveyer belts then sends the bins to “packers,” who spend just fifteen seconds on each, sealing boxes with tape that is automatically dispensed at the perfect length. “By the time you take an Amazon delivery off your stoop, walk into your home, find a pair of scissors and open the brown box,” the story intoned, “you’ve already spent nearly as much time handling the package as Amazon’s employees.”
The story is hardly exceptional. Each week, it seems, another magazine, book, or think tank sketches a dystopian near-future in which new technologies render most workers unnecessary, sparking widespread poverty and disorder. Delivery drivers, the thinking goes, will not be needed when there are drones or autonomous cars staffed by robots, and Starbucks baristas and fast food workers will be redundant when a tablet can take your order and a machine can prepare it. Some even envision more skilled jobs at stake: robots repairing our homes, caring for the elderly, or nursing patients back to health. As President Obama warned in his farewell address, “The next wave of economic dislocations . . . will come from the relentless pace of automation that makes a lot of good, middle-class jobs obsolete.”
An economic challenge of this magnitude requires ambitious solutions, and many in public debates have converged around a basic income. The idea is simple: the state would provide regular cash grants, ideally sufficient to meet basic needs, as a right of citizenship or lawful residency. Understood as a fundamental right, basic income would be unconditional, not means-tested and not contingent on previous or current employment. It would help sever the link between work and welfare, provide income security for all who are eligible, and perhaps mitigate growing inequality. It could also enable people to provide unpaid care work or community service, start new businesses, or get an education.
While this widespread attention to the problems of work and equality is welcome and overdue, and while a properly designed basic income would have many virtues, we need to be clear about the policy’s justifications, merits, and limits. As noted above, basic income proponents often pivot off the threat of widespread technological unemployment. But students of capitalism have been predicting labor’s demise ever since they identified and named “capitalism” itself. Is this time different? Consider what has happened at Amazon: warehouse robotics lowered prices and increased sales, and in early February the company announced plans to hire one hundred thousand more workers across the country.
Yet the news is still not good. Technology has transformed work in ways that have to do with political economy, not resource distribution. Amazon workers spend less than a minute per package because the company requires them to work at a furious pace, and it can afford to hire them by the thousands in part because it pays fairly low wages. Amazon also outsources many deliveries to third-party vendors whom it pays by the package, thus avoiding duties under wage per hour, workers’ compensation, and collective bargaining laws. Increasing the pace of work and outsourcing are not new, of course, but information technologies make such efforts easier and more profitable. With computer analysis of barcode scans, for example, Amazon can track the efficiency of pickers, packers, and drivers without necessarily setting eyes on them. Technology is not a substitute for menial labor in this story but rather one among many tools to keep labor costs down by exerting power over workers.
This account changes the case for a basic income. Many of today’s basic income proponents are libertarians and view the policy as a means of compensating losers, or as an excuse to repeal wage per hour or collective bargaining laws. Few are concerned about public goods, workers’ and capital owners’ entitlements within the firm, the power of various social groups, the ability of workers to organize collectively, and the question of what constitutes good work, not just jobs.
An alternative case for basic income draws from classic commitments to social democracy, or an economic system in which the state limits corporate power, ensures a decent standard of living for all, and encourages decent work. In the social democratic view, however, a basic income would be only part of the solution to economic and social inequalities—we also need a revamped public sector and a new and different collective bargaining system. Indeed, without such broader reforms, a basic income could do more harm than good.
This agenda will of course make zero progress during the Trump administration. But questions surrounding work and rising inequality are not going away. After all, Trump exploited fears of a jobless or insecure future in his campaign, signaling a return to our industrial heyday, with good-paying factory jobs implicitly promised to whites, men, and Christians. On the left, meanwhile, there is grassroots energy and momentum to think big and to address these issues head on, in all their complexity. But we still need a vision of good work and its place in our society, one that recognizes how our economy—and our working class—have changed dramatically in recent decades. I do not think for a moment that I have all the answers. But I do think an ambitious agenda around technology, work, and welfare can be a focal point and political resource for organizers, and perhaps even candidates, in the years to come.
Consider the life of a truck driver forty years ago versus today. In 1976 long-haul truck drivers had a powerful, if flawed, union in the Teamsters and enjoyed middle-class wages and excellent benefits. They also had a remarkable degree of autonomy, giving the job a cowboy or outlaw image. Drivers had to track their hours carefully, of course, and submit to weigh stations and other inspections of their trucks. But dispatchers could not reach them while they were on the road, since CB radios have limited range. Truckers would call in from pay phones, if they wanted.
No longer. Trucking companies today monitor drivers closely through “telematics” devices that gather and analyze data on their location, driving speed, and delivery efficiency. Some even note when a driver turns the truck on before fastening his seat belt, thereby wasting gas. As sociologist Karen Levy has shown, some long-haul trucking companies use telematics to push drivers to drive for all the hours permitted per day under federal law, at times waking them up or even overriding drivers’ own judgments about whether it is safe to drive. UPS has used the technologies to reduce its stock of drivers, and many have noted the stress that “metrics-based harassment” puts on workers.
While the specter of self-driving vehicles is out there, this is the current reality for many drivers and will be for the foreseeable future. We have seen stunning advances in autonomous vehicles in recent years, but there is a vast difference between driving on a highway or broad suburban streets in good weather conditions and navigating narrow and pothole-filled city streets, not to mention making the actual delivery to houses, apartments, and businesses. As labor economist David Autor and others have argued, we are nowhere close to fully automated production or distribution of goods, since so many jobs involve nonrepetitive tasks. In other words, the reports of the death of work have been greatly exaggerated.
Technological development is nevertheless altering the political economy of labor markets in profound ways. As we can see in the truck driver example, many firms are deploying information technologies to erode workers’ conditions and bargaining power without displacing them.
And of course truck drivers are not alone. Many other firms today use advanced information technologies to push for more efficiency, in the process reducing workers’ discretion, ultimately requiring them to work harder, faster, and for less. For example, where once taxi drivers’ folk knowledge of the optimal path from A to B in a crowded city was a valuable skill, now Uber and Lyft can calculate the best route through GPS technology and machine learning processes based on data gleaned from hundreds of thousands of trips.
Other technological innovations make it easier—which is to say more efficient—to purchase labor without entering formal employment relationships and accepting the attendant legal duties. In the past firms tended to employ workers rather than contractors, or to pay employees above-market wages, in scenarios where it was difficult to train or monitor them. Workers who felt valued in this way would work diligently and remain loyal toward firms, ultimately reducing overall labor costs.
Again Uber’s model helps illustrate. The company’s app reduces consumers’ and drivers’ search costs significantly. Rapid scalability reduces Uber’s costs of identifying and contracting with new drivers and riders; its GPS-based monitoring of its drivers enables it to know whether they are speeding or otherwise driving carelessly and whether they are accepting a sufficient number of fares; and its customer rating system enables it to manage an enormous workforce without managerial supervision. The net result is an economic organization of global scope based largely on contract where the firm disclaims any employment relationship toward its workers and therefore any employment duties toward them.
To be clear, there are powerful arguments that Uber drivers meet the legal test for employment, given the company’s pervasive control of their work and its economic power over them. But given the ambiguities of current law, Uber has few economic incentives to bring drivers inside the firm, making them employees, or to extend them generous wage and benefit packages. Similarly Amazon’s analytics help it to keep wages low: with barcode scanners tracking pickers’ and packers’ efficiency, the company does not have to pay workers as well to keep them motivated.
Finally, extensive data about market structures and consumer demand can enable firms to exert power over their suppliers or contractual partners, driving down costs—and therefore wages and conditions—through their supply chains. Walmart has long leveraged its unparalleled market data to estimate the lowest possible price suppliers will accept for goods, putting downward pressure on their profits and their workers’ wages. Amazon does the same today, and franchisors such as McDonald’s set prices and detailed product specifications for their franchisees.
Many firms today have substituted algorithmic scheduling for middle-managers’ local knowledge, using data on past sales, local events, and even weather forecasts to schedule work shifts. A Starbucks employee, for example, has little schedule predictability since she is at the mercy of the algorithm, and a McDonald’s worker can be sent home early if computers say sales are slow. This push to limit labor costs through finely tuned scheduling practices also alters workplace norms, since workers cannot appeal to a computer’s emotions in asking for more or less time, a raise, or a slower pace of work. The net effect of all of this is that power in our labor and product markets is increasingly concentrated in a few hands.
Crucially such management techniques and new production strategies are often more efficient than the status quo. Amazon has undeniably lowered prices for goods through its use of automation. Similarly a recent MIT study calculated that just three thousand multi-passenger cabs using a version of Uber’s algorithm could serve Manhattan’s need for taxis. The potential benefits here are staggering, especially if coupled with a modern mass transit system: shorter commutes, less car ownership, less pollution, and more urban space.
But the line between innovation and exploitation is far from clear. While some workers will thrive as their unique skills and talents are rewarded by new technologies, many others will have less autonomy, less generous wages, less time for social connection, and unpredictable schedules. And under current laws, we can expect such trends to accelerate. Our labor and employment laws still envision the economy of the 1930s, which was dominated by massive industrial firms with hundreds of thousands of direct employees. Those laws rarely touch modern “fissured” work relationships such as Uber’s relationship with its drivers, Walmart’s relationship with its suppliers’ workers, or McDonald’s relationship with its franchisees’ workers. Those laws also limit workers’ ability to unionize or bargain effectively since they encourage bargaining at the firm or even plant level whereas today’s modal workplace is growing ever smaller. Workers have fewer and fewer means to exert power on their own behalf.
How would a basic income impact workers and firms in this context? It would surely protect workers against the economic harms of unemployment and underemployment by giving them unconditional resources, and it would enable them to bargain for higher wages and to refuse terrible jobs. But a basic income would do little to reduce corporate power, which is a function not just of wealth but of the ability of firms to structure work relationships however they wish when countervailing institutions—such as a powerful regulatory state—are absent or ineffective. Yes, a basic income would make it easier for workers to organize and demand reforms—Andy Stern dubbed it “the ultimate permanent strike fund”—but the threat of termination or retaliation would still prevent many workers from protesting or striking in the first place.
And of course many have argued that a basic income would make minimum wage and collective bargaining laws less necessary, since workers’ material needs would be met by the state. But cash benefits and reasonable wages are not morally equivalent. In Robert Solow’s memorable phrase, the labor market is a “social institution” governed in part by norms of reciprocity and mutual respect. Workers often demand higher wages from larger and more profitable employers. And they work less diligently when they feel disrespected. So there is a major difference between a generous wage on the one hand and a meager wage supplemented with cash from the state on the other. Generous wages help make firms’ economic responsibility roughly commensurate with their economic power. Meager wages signal disrespect, and state transfers are impersonal.
In fact, without labor market regulations in place, the impact of a basic income on very low-wage work could be disastrous. If a basic income were not extended to green card holders, guest workers, or “irregular” immigrants (those who enter or stay without authorization), such workers would be far cheaper to employ in menial jobs, at which point they would be permanently enshrined as a laboring underclass. (Currently minimum wage laws apply regardless of work authorization, on the grounds that differential treatment would undermine standards for all.) A basic income could even be designed to serve white nationalist ends. In fact, far-right European parties have often embraced the welfare state as a means of defending citizens against a purported tide of immigrants.
A standalone basic income also will not ensure equal access to quality education, health care, mental health care, housing, and transportation. Liberal markets systematically fail to provide such goods to the poor and working class, for the simple reason that they often are not profitable when provided on equitable terms. Giving cash to individuals to purchase them will not suddenly change matters.
Popular debates have largely ignored these limits of a standalone basic income, an oversight that is not entirely accidental. As a tax-and-transfer program, basic income would be consistent with a wide variety of political-economic systems, including neoliberal capitalism, social democracy, and various forms of socialism, but much of the basic income literature has a libertarian streak. On the right, Milton Friedman proposed a negative income tax in large part because he hoped it would reduce government bureaucracy. On the left, Philippe van Parijs’s now-classic argument for the policy held it would maximize what he called “real freedom” better than standard welfare-state policies. And Silicon Valley’s take exemplifies an “everyday libertarianism,” which views market results, including pre-tax incomes, as presumptively fair.
Unlike many libertarians, basic income proponents accept the necessity and fairness of income and wealth taxation. But a basic income is still no cure for the moral ills of liberal markets. Since labor cannot be separated from workers, it will never be a classic commodity, and labor markets will never be stock exchanges for faceless buyers and sellers. Low wages carry a stigma that low bids for soybeans never will. In the long run, companies cannot treat workers or even consumers as line items on a balance sheet without risking a revolt. Uber is now paying a high price for ignoring this lesson. The company’s flagrant disregard for our moral economy and its open embrace of a cutthroat, winner-take-all libertarianism made it a pariah in many quarters well before it faced allegations of a workplace culture of sexual harassment.
In my view, the more compelling arguments for basic income are rooted in commitments to equality as well as freedom. Take Thomas Paine’s Agrarian Justice (1797), which prefigured the later case for social insurance. The earth, he argued, had been “the common property of the human race” in its natural state. Private property enabled the few to profit from the earth’s resources, but all are entitled to compensation through a “citizens dividend.” G. D. H. Cole updated Paine in the twentieth century, arguing for a “social dividend” on the grounds that all production is “a joint result of current effort and of the social heritage of inventiveness.” Auto manufacturers did not discover electricity, after all, and Silicon Valley did not invent the Internet. Rather than a means of maximizing freedom, a basic income would help meet our duties toward one another.
Workers cannot appeal to a computer’s emotions in asking for more or less time, a raise, or a slower pace of work.
A still more compelling case for basic income builds on Elizabeth Anderson’s and Philip Pettit’s (quite distinct) arguments that a just society must ensure that no group of citizens is subordinate to another. Extreme poverty causes subordination since it forces us to beg or to work at terrible jobs. Means tests are equally degrading since food stamps and similar programs tend to restrict what we can buy, again stigmatizing the poor. This argument is quite different from the libertarian case for a basic income since it does not view basic income as a replacement for the welfare state. Rather it asks basic income to serve a discrete and limited purpose: making sure nobody falls through the cracks.
And if one agrees that we need to root out subordination, we will need to do much more than pass a basic income. A just society would not just eliminate penury and then leave people to their fates. It would also strive for a fair distribution of power. This point strikes me as obvious, but we can miss it by focusing on unemployment and poverty. A society in which a few make decisions and the many take orders is an oligarchy, not a democracy.
This is what I mean by the “social democratic” case for a basic income: it would help build a post-industrial welfare state by alleviating dire poverty. But it is unlikely to pass and would do little good unless coupled with other efforts to ensure broadly dispersed power, including a substantially revamped public sector and new and stronger regulations around work.
The public sector agenda should begin with a massive investment in human services. This would include primary, secondary, and higher education; childcare and elder care; health care; and mental health services. All of these are critical to human flourishing and economic growth, especially in a technologically advanced economy. There is also the added benefit that these sectors are among the least vulnerable to automation since they require interpersonal communication and human judgment: turning them into basic rights of citizenship would create millions of jobs.
Extensive job training and placement programs for unemployed workers would reduce the devastations associated with job losses. This should be coupled with a guarantee that the government will stand as employer of last resort, as William Darity and Darrick Hamilton have advocated. Classic WPA-style public works would be one option, but we might do more good by thinking locally: retrofitting houses and other smaller buildings for energy efficiency, as john a. powell and others have suggested, or building and repairing local parks and schools, or training unemployed workers for jobs in childcare, education, elder care, or health care. Some subset of the projects could be identified through participatory budgeting or other deliberative processes and carried out in partnership with local governments or organizations.
Then we should consider unconditional cash benefits in some form, ideally a generous basic income, especially if technological unemployment becomes significant. But we need to design it correctly. If certain immigrants are denied a basic income, then we must establish a transparent pathway to citizenship and ensure that they enjoy generous wages. Currently many states restrict convicted felons from voting or collecting many public benefits. If that pattern holds with a basic income, that group could also become a pool of menial labor.
Paying for all of this would not be easy, of course, which is why it may make sense to take baby steps. Unconditional cash grants to parents, for instance, would go a long way toward alleviating poverty and would be less politically controversial than unconditional benefits for the able-bodied and childless. Or perhaps a participation income, in which citizens would be entitled to cash benefits on the basis of some qualified civic service each year, though this would be less desirable than a public jobs guarantee combined with other generous benefits.
As we wait for the politics to catch up with the policy, however, the “baby steps” approach can inch us closer to meeting social needs, developing a highly trained workforce and rebuilding faith in the public sector.
Then there are the near-intractable problems of improving private sector work and limiting corporate power. Of course what makes work “good” is hard to define, in part because it is so dependent on social context and individuals’ preferences. We look back fondly on the industrial era in part because manufacturing jobs delivered high wages and benefits, but those jobs only became “good” after workers organized and forced firms to raise wages and reduce hours. Manufacturing jobs were also physically and emotionally punishing, and factory workers’ pride was often based on a perverse ethic of masculine suffering.
It is easier to say what bad work is. There is certainly no shortage of it these days: many workers earn low pay, have long or unpredictable hours, and are vulnerable to arbitrary treatment. Raising the minimum wage and reducing the standard workweek to thirty-five or even thirty hours would help. So would the public investments in human services discussed above, since those new jobs would require significant judgment (ensuring a degree of worker autonomy) and current law extends many democratic norms to the public sector already. A public jobs guarantee would give workers the security of knowing that an alternative always exists if a private-sector job proves unendurable.
The centerpiece of reform efforts should be the encouragement of collective bargaining between workers, their employers, and whatever other firms enjoy economic power over them. That will require reforms to bring our labor laws into sync with the reality of what work now looks like. Rather than requiring workers to organize shop by shop, we could encourage bargaining at the corporate level, such that all McDonald’s workers, for example, would be in one bargaining unit, regardless of whether their restaurants are franchises or owned by the parent corporation. Better still, we could put all fast food workers nationwide in one bargaining unit empowered to negotiate with an association of fast food companies. Similarly workers could be granted bargaining rights with the firms at the top of their respective supply chains: Uber drivers would have bargaining rights vis-à-vis Uber, and Walmart and Target would have duties toward the workers who produce, process, and move goods to their shelves, including production workers, warehouse workers, farm workers, janitors, and many others.
Without labor market regulations, the impact of a basic income on low-wage work could be disastrous.
This would work a major change in our labor law system, but the elements are already being built on the ground. On the company side, growing market concentration makes such bargaining quite plausible technically, if not politically. A national association of retail workers need only drive a settlement with Walmart, Target, Macy’s, Gap, and a few others to raise standards. Where once taxi drivers would need to negotiate with hundreds if not thousands of medallion owners, now drivers can negotiate with Uber and Lyft directly.
Unions and other worker organizations have been dealing with these issues for decades and have developed workable models of organizing and bargaining that demonstrate proof of concept. For example, the Coalition of Immokalee Workers (CIW), an organization of farm workers in Florida, has pressed many major retailers and restaurants to join a “fair food program” under which they commit to paying more for produce and to rooting out slavery, corporal punishment, and other abuses in the fields. CIW succeeded despite having no collective bargaining rights at all and no legal employment relationship with the retailers and restaurants involved. They instead relied on high-visibility worker action, consumer boycotts, and creative media strategies.
Legal scholar Kate Andrias has argued that these efforts reflect an emerging model of unionism she dubs “social bargaining,” in which workers demand that firms accept responsibility through their supply chains to a degree that exceeds the letter of the law. This strategy relies on public protest as well as more conventional union strategies—a kind of bargaining in the public green—and often utilizes state legislative power to backstop organizing efforts.
National legislation could encourage robust social bargaining in a variety of ways. Some states allow their departments of labor to constitute “wage boards” empowered to set wages within particular industries upon consultation with labor, firms, and the state. The federal government could follow suit, expanding the mandate of such boards to include issues of employee status, hours of work, and other basic standards. The federal government could also make it easier for workers to obtain union representation without enduring a contentious organizing process. Drawing from the Seattle legislation that established collective bargaining rights for Uber and Lyft drivers, firms could be required to disclose lists of workers to any worker organizations that meet basic indicia of independence and capacity.
It is crucial to emphasize the utility of social bargaining in today’s economy. It would better reflect contemporary production relationships among firms, suppliers, and workers. It may even be relatively acceptable to large firms, insofar as it would seek to equalize wages across a sector and would not encourage detailed bargaining over worksite-specific minutiae, which employers have good reason to resist. And it could be carried out through organizations less formal than classic unions, which seem more appealing to contemporary workers. Of course, as with the public benefits outlined above, passing labor law reforms will not be easy. But how else can we realistically enhance workers’ bargaining power?
These three policy shifts—a basic income and other economic security guarantees, vastly expanded social programs, and new rules to encourage social bargaining—would supplement and reinforce one another. Better educational policies should help employers by ensuring a mobile, highly skilled workforce, and public health care and other social insurance would reduce the costs of employment. A basic income and a public jobs guarantee would enable workers to stand up for themselves more readily. Unions representing broad swaths of the precarious workforce would have incentives to push for a robust welfare state and even a basic income. Those unions could also reduce elites’ domination of our politics, which may otherwise prevent implementation of a basic income, limit its generosity, or set it up to fail.
As many will realize, this institutional arrangement looks a bit like the Scandinavian “flexicurity” model—a portmanteau of “flexibility” and “security”—which combines high wages; extensive welfare and job training programs so workers can move between jobs; and relatively flexible employment policies that enable firms to hire, fire, and reassign workers at will. Such economies are quite open to technological innovation, but these institutions help ensure that its benefits are shared more equitably than they are in the United States. Not coincidentally, Scandinavian welfare states seem to be evolving toward a basic income, as the policy would fit nicely within flexicurity. But collective bargaining plays a crucial role in that system: without powerful unions, it is not clear that flexicurity would have developed in the first place, much less endured.
A basic income is a simple and elegant way to redistribute resources. But there are no simple, elegant solutions to complex political and economic challenges. A decent future of work and welfare requires a basic income—and much more.