How to Modernize the Rules of Work to Fit the Times
by seth d. harris and alan b. krueger
seth d. harris, a former deputy secretary of labor, is a distinguished scholar at Cornell's School of Industrial & Labor Relations. alan b. krueger, a former chairman of President Obama's Council of Economic Advisers, is a professor of economics and public affairs at Princeton.
Illustrations by Michael Wertz
Published May 2, 2016.
Uber driver an employee or an independent contractor?
That isn't a question of semantics. Employees qualify for a host of legally mandated benefits and protections, including rights to organize and bargain collectively, civil rights safeguards, workers' compensation insurance, overtime pay and the minimum wage. Indeed, an extensive social compact has emerged in the relationship between employees and employers over the past century. The former cede control over their work lives (and, to some extent, their economic futures) to the latter, and thereby make themselves economically dependent. In return, employers provide the minimum level of economic security required by law, and often more.
Along with redressing the inherent imbalance in bargaining power between individual workers and employers, these laws improve the efficiency of the labor market by correcting market failures, like information asymmetries between worker and employer – as in, "do I need a mask to prevent exposure to chemical X?" – or discrimination that would otherwise deny qualified workers access to more-productive jobs.
Independent contractors, in contrast, are expected to use their bargaining power as independent businesses to negotiate their own compacts. The law doesn't guarantee them a minimum wage, nor do they have much protection against discrimination. But unlike workers formally designated as employees, they have the freedom to work when they want, where they want and how they want. So, whether workers are classified as employees or independent contractors matters a great deal to their ability to call on the law to protect them.
There is currently much uncertainty as to whether your Uber driver – or Lyft driver, or TaskRabbit handyman or Thumbtack personal trainer – will be ruled an employee or an independent contractor by regulators and the courts. On the one hand, these workers have some characteristics of independent contractors, including being able to choose when to work and whether to work at all. They typically use their own equipment. They may also work simultaneously with multiple "intermediaries" (that is, the companies with online apps that connect them with customers) or conduct personal tasks while they are working with an intermediary.
On the other hand, they have some of the characteristics of employees. The intermediaries retain some control over the way workers perform their work, such as setting their fees or determining the sort of equipment that must be used, and they may, in effect, fire workers by prohibiting them from using their services. Those workers are thus not independent businesses as that term is usually defined.
Labor, employment, and tax law currently only recognize those two traditional legal classifications. Yet increasingly, workers are falling into a gray area between the two, and that is inefficient and potentially unfair for all concerned.
Neither Fish Nor Fowl
Even if one sets aside the complications introduced by digitally enabled work, the relevant labor and employment law is hardly a model of clarity. In fact, there are multiple laws that apply different tests for determining whether a worker is an employee or an independent contractor. These tests typically involve factors that are weighed against one another, but legal precedents and regulatory interpretation don't offer uniform guidance regarding how to apportion the weights. No single factor is decisive, courts and administrative agencies warn, and the tests do not require the satisfaction of all factors.
It is thus up to judges, juries and administrators to decide. For example, the IRS's test for tax liability places weight on whether workers receive fringe benefits in determining whether they are employees, while the test mandated by the Fair Labor Standards Act does not.
Wait, it gets worse. The Fair Labor Standards Act's test puts weight on whether a worker exerts independent judgment, while the test used for the Employee Retirement Income Security Act does not. And the amount of weight to place on whether workers exert independent judgment relative to the amount of weight to place on whether they can set their own hours is unspecified. Similar inconsistencies and ambiguities arise in the National Labor Relations Act, the Civil Rights Act and state labor and employment laws.
Moreover, when workers simply cannot be classified unambiguously, the United States (unlike some other countries) does not have a default rule that automatically assigns them to employee status, until proven otherwise. Too often, judges, regulators and employers decide what result they want, and then make their analyses fit the desired outcome. That's not how the law is supposed to work, and it risks stifling innovation, undermining efficiency and denying workers and employers a fair deal on the aforementioned social compact.
Judges and administrative bodies asked to decide difficult cases are increasingly requesting new rules that would allow them to avoid conflicting decisions on the classification of workers in the gig economy. In a case involving the classification of Lyft drivers under the Fair Labor Standards Act and state law, for example, Judge Vince Chhabria of the United States District Court for the Northern District of California wrote in May 2015:
As should now be clear, the jury in this case will be handed a square peg and asked to choose between two round holes. … Some factors point in one direction, some point in the other, and some are ambiguous. Perhaps Lyft drivers who work more than a certain number of hours should be employees while the others should be independent contractors. Or perhaps Lyft drivers should be considered a new category of worker altogether, requiring a different set of protections. But absent legislative intervention, California's outmoded test for classifying workers will apply in cases like this. And because the test provides nothing remotely close to a clear answer, it will often be for juries to decide.
The parties in this case reached a settlement in January 2016 (pending Judge Chhabria's approval), with Lyft agreeing to pay the drivers $12.25 million and to make two concessions. The company agreed to bar drivers only for cause and to pay the costs of arbitration in the future. The drivers remained independent contractors.
Shannon Liss-Riordan, the Boston-based lawyer who represented the drivers, said she believed the settlement was fair "given the risks we faced in litigation." But the case plainly did not resolve the uncertainty surrounding the classification of on-demand workers. Indeed, the fact that both sides were willing to settle rather than take their chances with a jury indicates the risk and ambiguity that they perceive in the current system. A similar suit against Uber Technologies in California, brought by the same lawyer, is pending.
Different coasts, different legal interpretations. A ruling by the Department of Economic Opportunity in Florida in December 2015 found that the facts clearly supported the case that Uber drivers were independent contractors, and it took issue with rulings in two other states:
Uber is a technology platform that, for a fee, connects transportation providers with customers seeking transportation. The agreement between drivers and Uber specifies that the relationship is one of independent contractor, and the actual course of dealing confirms that characterization. Drivers have significant control over the details of their work. Drivers use their own vehicles and choose when, if ever, to provide services through Uber's software. Drivers decide where to work. Drivers decide which customers to serve. Drivers have control over many details of the customer experience. Drivers may provide services through, or work for, competing platforms or other companies when not using the Uber application. On these facts, it appears that Uber operates not as an employer, but as a middleman or broker for transportation services.
Two other jurisdictions – California and Oregon – have in recent months come to the opposite conclusion, finding that drivers using the Uber software are employees under the law of those states. These opinions are not persuasive, as they largely ignore the contract, and misconstrue the actual course of dealings between the parties. To a significant extent, these rulings appear to rest on the fact that Uber could not be in business without drivers. But the same is true of all middlemen. Uber is no more an employer to drivers than is an art gallery to artists.
Other cases like these are seemingly inevitable. No one could sensibly expect much clarity to emerge from this legal process, given the inconsistency and obsolescence of the underlying statutes. The likeliest outcome is conflicting decisions doled out over the course of many years, and the resulting legal morass will likely slow the development of innovative new sectors and services, create inefficiencies for all parties involved in the emerging internet-intermediated labor market and impose high transaction costs from extensive litigation.
Late last year, we proposed a new category for independent workers that gives legal recognition and protections to workers who fall in the gray area between employees and independent contractors (see "A Proposal for Modernizing Labor Laws for Twenty-First-Century Work: The 'Independent Worker,'" The Hamilton Project). Our goal was to modernize labor, employment and tax law to be more efficient and fair in preparation for the growing number of on-demand workers and perhaps to preempt the coming avalanche of litigation that is likely to leave all sides worse off.
Under our proposed definition, independent workers are distinguished from employees because the independent workers control whether they work, when they work, how long they work, where they work and (for the most part) the manner in which they provide a customer with service and the duration of that service. Intermediaries can have some control over the way independent workers perform their tasks, but less than traditional employers. We propose that these workers receive legal benefits and protections that are appropriate for workers who have this kind of substantial control over their work, but still establish important economic relationships with the intermediaries that differ from the arm's-length contracts typifying independent businesses.
We think that regardless of whether they work through an online or offline intermediary, independent workers should qualify for many of the benefits and protections that employees receive. These include the freedom to organize and bargain collectively, civil rights protections, mandatory tax withholding and worker-company shared contributions for the payroll taxes that fund Social Security and Medicare.
Economists and business analysts will no doubt squabble over who would ultimately bear the cost, as they do over benefits now provided to traditional employees. We think there is solid evidence that some (even most) of these costs likely will be borne by workers in the form of higher commissions and fees they pay to intermediaries in the long run, just as supply-and-demand pressures in the traditional labor market induce employers to reduce wages to make up for a sizable percentage of the employee benefits and taxes they must cover. If nothing else, requiring intermediaries to provide benefits – such as paying for half the Social Security payroll tax – will make it easier to compare the after-tax compensation of gig work with traditional employment. We also expect that it will significantly increase tax compliance, to the benefit both of the independent workers and the U.S. Treasury.
Since it is conceptually difficult to attribute some independent workers' hours on the job to any single intermediary – and it is often unclear whether they are on the job or not – they would not qualify for hours-based benefits, including overtime pay and the minimum wage. Further, independent workers would rarely, if ever, qualify for unemployment insurance because these benefits are premised on a decision by the employer to discharge or lay off a worker without cause rather than a decision by a worker with broad discretion to choose whether to work through an intermediary. As a result, independent workers would not be covered by unemployment insurance and intermediaries would not be required to fund that program. Intermediaries would, however, be free to pool independent workers for purposes of providing medical insurance and other benefits at group rates without the risk that their relationship would subsequently be treated as an employment relationship.
Our goal is to find a middle way to ensure that gig-economy workers share America's workplace social compact without jeopardizing the benefits that come from innovation and flexible work arrangements. We were guided by three principles:
- Legally required benefits and protections should be appropriate for workers who control when and whether to work.
- Independent-worker status should be neutral with respect to employment status.
- Creating a new classification for independent workers should make the labor market operate more efficiently. In particular, companies that use independent workers should survive and grow because they have a better service, not because they take advantage of regulatory arbitrage.
A natural concern here is that some employers may restructure their work relationship to convert employees to independent-worker status. That would be very unlikely, however, for two reasons.
First, employers would have to give up a tremendous amount of the control they currently exert in order to make the switch. How many employers would be willing to tell their employees that they could come to work whenever they wanted, and only if they wanted, without any advance notice? That would be a great deal of control to sacrifice in exchange for very little, since independent workers would be entitled to virtually all of the legally required benefits that employees receive, save overtime pay, the minimum wage (which is unlikely to be binding anyway) and unemployment insurance.
Second, employers can already organize work to classify (and, in too many cases, misclassify) their workers as independent contractors, who receive none of the legal benefits for which independent workers would qualify in the new category. That is, there are greater cost savings to be had in the current, badly flawed system by rejiggering work to meet the legal definition of independent contractor. Indeed, we suspect that many more workers would gain legal protections than would lose them under our proposed system.
The Gig Economy By The Numbers
A variety of estimates suggest that the number of gig workers, defined as those working through an online intermediary, is currently relatively small. Our own estimate, based on the number of Google searches for major online platforms, including Uber, Lyft, Grubhub and others, puts the figure at about 600,000, or less than a half-percent of total U.S. employment. Lawrence Katz and Alan Krueger estimated that, in the fall of 2015, 0.5 percent of the U.S. work force provided labor services to customers through an online intermediary, such as Uber or TaskRabbit. The McKinsey Global Institute estimated that fewer than 1 percent of all workers are providing services through the online economy. And an estimate by the JPMorgan Chase Institute, based on bank deposits from 30 work-related Internet platforms, such as Uber and TaskRabbit, into individuals' bank accounts, puts the figure around 1 percent of the U.S. population. This last analysis further estimates that a cumulative 4 percent of the population has worked through an online platform at some point.
Although the number of workers currently working through online intermediaries is relatively low, the rate of growth is quite high. For example, Uber, which is the largest online intermediary, has doubled the number of its U.S. workers each six months for the past few years. Technological advances are likely to lead to further disintermediation of the workplace that is expected to propel growth in freelancing and gig work in the future.
It should also be noted there are hundreds of thousands more workers, like many taxi drivers, Avon direct-sales workers and others who currently work with an offline intermediary. Katz and Krueger, for example, found that twice as many workers work through an offline intermediary as through an online one. Many of these workers would likely be classified as independent workers under our proposal. At current growth rates, the number of independent workers providing labor services through online intermediaries would likely exceed the number working through offline intermediaries within a few years.
A Grand Bargain
Our plan is meant to start a conversation. We recognize that those vested in the current system and those who firmly believe their side will prevail in the legal wrangling would have reason to oppose it. As Sen. Mark Warner of Virginia put it in discussing our proposal when the Brookings Institution rolled it out, "They've managed to piss off both sides with their project already, which is a pretty good starting place to be."
There was something for all sides to dislike (and like) about our proposal. Unions and some labor-connected think tanks and academics dislike the exclusion of the minimum wage, as well as the exclusion of some other benefits and protections. Intermediaries, for their part, object to allowing independent workers to organize and to bargain collectively. Of course, we did not offer our proposal to serve any particular interest. Rather, our objectives are to maximize the size of the economic pie – the joint surplus, as economists would put it – and to ensure a fair distribution of that surplus to those who contribute to creating it.
Given that litigation over employee and independent-contractor status for gig workers is still in its early stages, we suspect that both sides will have to lose some important cases before they will be ready to seriously consider a new legal classification for workers who don't fit comfortably in either status. That is, all sides may need to be shocked into the realization that the 20th-century (or 19th-century) legal structure is inappropriate for some newly emerging sectors of the contemporary labor market.
This circumstance is reminiscent of the situation that existed when the workers' compensation insurance system was created. Workers' comp laws were enacted in most states at the beginning of the 20th century, when the legal system for compensating workers and their families for workplace injuries proved inefficient, burdensome and too uncertain for all sides concerned.
The goal of our proposal is to suggest a detailed solution to the unnecessary tumult being created by our outdated, uncertain and inefficient legal regime before it frustrates or stifles intermediaries, independent workers and customers alike. Others, perhaps motivated by different principles, may suggest other solutions. We welcome a vigorous debate on these issues. As in any grand bargain, there is room for disagreement and space for compromise once both sides recognize that change can benefit all parties, compared with the status quo.