Abstract
I would like to begin by thanking Senator Mark Warner for his kind introduction. I am particularly appreciative that he took the time to come here this afternoon given all the other critical work he is doing. He has taken the leadership in Congress—and I would say in the Nation—on emerging workplace issues. He called me late last year, just after I read the news that he was appointed Vice Chairman of the Senate Intelligence Committee, and I congratulated him on his new position, and he responded by saying something like: “Thanks, but I’d much rather spend my time developing policies that are appropriate for independent workers in the emerging gig economy.” I also want to thank Ken Prewitt for his remarks and the American Academy of Political and Social Science for selecting me for this prestigious award and for giving me the opportunity to present the Moynihan Lecture today. 1
To be associated with Daniel Patrick Moynihan is a tremendous honor. I was inspired by his career and one connection I have always felt that I have with Senator Moynihan is that his career in the federal government began at the U.S. Department of Labor as Assistant Secretary for Policy, Planning and Research. That position was actually created for him. When Bob Reich asked me to be Chief Economist of the Labor Department in 1994 he told me that what he was hoping to achieve by creating this new position was what was done for Daniel Patrick Moynihan—to bring in scholars who would enrich the government with their work and then go back and enrich the Academy with what they had learned. Daniel Patrick Moynihan was the perfect role model for that process.
There are too many friends, family members, and distinguished colleagues in the audience to acknowledge all of them, but I want to acknowledge that my parents are here. My parents have inspired me throughout my career, but especially for the topic of my lecture today. My dad was a CPA. (I used to say that he is a CPA, but he recently informed me that he finally let his license expire.) He became self-employed in the early 1970s and I remember clearly that when he made that decision we had a family meeting in our living room and he informed my brother, sister, and me that he was going to go out on his own. “Don’t worry,” he said, “we will still be covered by mom’s health insurance.” My mom was a public school teacher, with good health insurance. I have thought about that story often in recent years because the Affordable Care Act (ACA) has helped to do for many entrepreneurs what my mom was able to do for my dad, namely, provide access to health insurance. I will have more to say about this later in my talk.
The motivation for my lecture today is that the U.S. job market has been changing for several decades, with steady growth in self-employment. IRS data on Schedule C filings for self-employment income clearly show that self-employment is rising as a share of total employment, while household survey data from the Current Population Survey (CPS) show a downward trend, which suggests to me that the job market has changed so quickly that many workers are fundamentally confused about the nature of their employment relationship (see Figure 1). 2 The distinction between being self-employed and being a traditional employee matters, in part, because self-employed workers are responsible for providing for their own safety net, as well as securing other benefits and work protections, and for assuring that they are paid for work that they complete. The employer-employee relationship and a legal apparatus developed over many years handles these tasks for traditional employees. I think we need to update our labor laws to keep up with the evolving nature of work relationships.
IRS vs. CPS Data on Self-Employment as a Share of CPS Total Employment
Several factors are fueling the growth in self-employed workers, and one important one is technology. Work is becoming more standardized and more of a commodity. Probably everyone in this room uses Microsoft Office. It is easier to reach out to the external job market and obtain inputs when necessary if work is standardized. More recently, the growth of the gig economy (sometimes called platform work or digitally intermediated work) has boosted the trend toward self-employment. Demographics are also part of the explanation, as older workers are more likely to be self-employed, and our workforce is aging. And, finally, part of the trend is a result of companies misclassifying workers, which Weil (2014) calls a fissuring of the workplace; this trend is unfortunate and needs to be closely monitored and policed.
I suspect the trend toward self-employment will continue. Our labor laws and safety net were developed for a bygone era, when almost all workers could expect continuing employment relationships with companies. I think the question that we face going forward is, How do we extend the safety net to the large and growing group of independent workers, particularly in an age when there seems to be tremendous opposition to universal mandates that compel participation? This is the case even though some of our great advances over the last century—Social Security being a prime example—were only possible because the federal government could pass a universal mandate. And Social Security has passed the test of time; it is the ultimate portable benefit. It applies to workers who are self-employed and to traditional employees who receive W-2 tax forms. It does not matter which employer you work for. I am sure that President Roosevelt and his advisors did not anticipate Uber or TaskRabbit, but Social Security works well even for that sector of the economy. The challenge that we face now is how do we create in our existing system, which has protections and benefits for the roughly 85 percent of the workforce in a traditional job, additional protections and benefits to the rest of the workforce? That is the topic I address today.
I learned when I worked in Washington that when you confront a difficult policy challenge, you basically have three approaches to choose from. First, you could enunciate a set of principles, and challenge others to create policies that satisfy those principles. I will do a little bit of that, I have to confess. Second, you can appoint a commission and hope the commission solves the problem, or at least takes the pressure off until a solution could be found. Or, third, you can make concrete proposals to actually solve the problem. I will follow the third approach, but because I am uncertain about the right direction to go, I will also do something else that I learned in Washington: I will report the pros and the cons—arguments in favor, arguments against—my proposals, in an attempt to advance the discussion of policy options concerning independent workers.
Princeton Self-Employment Survey
Figure 2, drawn from Katz and Krueger (2016), shows that older workers are more likely to be self-employed, and there was strong growth in self-employment among older workers over the last decade. 3 Almost one in five workers over the age of 54 is working in self-employment on their main job. Although the Great Recession surely accelerated this trend, the growth in self-employment preceded the recession according to IRS data on 1099 filings and Schedule C filings (see Katz and Krueger 2017).
Percent of Workers Who Are Independent Contractors on Their Main Job, by Age
To better understand the nature and challenges of self-employed workers, I designed the Princeton Self-Employment Survey (PSES). This survey was conducted April 24–April 27, 2017, a week after Tax Day, and I will publicly share the results for the first time in this lecture. This was an online survey of more than ten thousand individuals using Qualtrics software and the Qualtrics panel of respondents. Although the Qualtrics sample does not hold up to the high standards that the Bureau of Labor Statistics and Census Bureau employ, as best I can tell it represents the self-employed population reasonably well. We also developed sample weights so that on average the sample is representative of the American Community Survey (ACS). 4 One of the features that I built into this survey is an A/B experiment where I asked the same question two different ways to randomly divided groups. Even if one is worried about the representativeness of the overall sample, at least in this survey experiment I had some control and could test the effect of question wording on this group of respondents.
The survey worked as follows. The first question was based on the basic monthly CPS class of worker questions on the main job. The responses are tabulated in Table 1. A total of 10,368 individuals started the survey, and 1,149 were screened out after the first question because their only work in the previous week was filling out questionnaires or because they did not work at all, and another 640 were dropped from my analysis sample because they resided outside the United States, were under age 18, or provided nonsensical (usually typing gibberish to open questions) responses. The remaining 8,579 individuals participated in the A/B experiment. Specifically, half of the respondents were randomly selected and asked the BLS Contingent Worker Survey (CWS) question on self-employment: “Last week were you working or self-employed as an independent contractor, an independent consultant, or freelance worker? That is, someone who obtains customers on their own to provide a product or service.” That phrasing, which designates Question A, potentially leaves out much self-employment work. How should an Uber driver, who does not obtain customers on his or her own, respond? And suppose you own your own business and are self-employed, you would be hard pressed to know how to answer that question. So I tried it a second way for the other half of workers, where I first provided a preamble to the BLS question and then asked the CWS Question: “Many people work in self-employment, on either a part-time or full-time basis, doing things such as working on construction jobs, selling goods or services in their businesses, or working through a digital platform or intermediary, such as Uber, Upwork or Avon. Last week, were you working or self-employed as an independent contractor, an independent consultant, or freelance worker? That is someone who obtains customers on their own to provide a product or service?” This version is designated Question B.
Question 1: CPS Class of Workers
NOTE: Sample size is 8,579 for the percentages because individuals under age 18, nonresidents, and gibberish responders are excluded. Weights are designed to match ACS total.
Includes tax-exempt and charitable organizations.
Respondents were 4.5 percentage points more likely to be classified as self-employed in Question B than in Question A: 32.2 percent versus 27.7 percent (p < .001), which highlights the sensitivity of reporting self-employment status depending on question phrasing. 5 Note also that these percentages are high. Many of these workers are self-employed on secondary jobs, which may go unreported in the CPS. According to the weighted class of worker on the main job question in Table 1, about 8 percent of the workers respond as self-employed. Interestingly when I look at this subset who reported that they were self-employed on their main job, there is not a significant difference in responses to the CWS question asked the A way or the B way. In fact, it reverses order and the Question A phrasing had a slightly higher proportion (see Figure 3). About 85 percent in either question indicated that they were self-employed.
A/B Question Test for Subsample That Responded Self-Employed to Main Job Question (N = 1,321)
One of the challenges involved in defining self-employment is that many workers combine self-employment activities with traditional employment, while others earn almost all of their income from self-employment. IRS data reported in Jackson, Looney, and Ramnath (2017) show that in 2014 about 60 percent of the self-employed earned virtually all of their income from self-employment, while many others earned a small share of income from self-employment activities. In the rest of my talk I focus on the group that is mainly self-employed—that is, those earning most of their income from self-employment. Specifically, I restrict the PSES sample to the group of individuals who indicated self-employment as their main activity based on the CPS class of worker question in Table 1. Almost one fifth of these individuals have traditional W-2-type jobs in addition to being mainly self-employed. My reason for focusing on those who are mainly self-employed is that workers who are connected to a traditional employer have more access to the safety net and have more access to benefits, such as health insurance coverage, than those who are primarily self-employed. The results of the A/B question experiment also suggest that membership in this group is less sensitive to survey question wording.
Table 2 compares the industry and demographic characteristics of the self-employed in the CPS with the PSES, and with employees in the CPS. Self-employed workers tend to be older than traditional employees. They also tend to be better educated. The self-employed are also over represented in construction and in professional, technical, and scientific positions; they are underrepresented in manufacturing compared with traditional employees. Almost two-thirds of the self-employed are men. The self-employed have higher earnings, on average, than traditional employees, and they have greater dispersion in earnings, highlighting the heterogeneity of the self-employed. The PSES provides some additional details on the self-employed. Most of the self-employed are in unincorporated businesses, and most report that they file Schedule C. About half say they work from home or a home office.
Characteristics of Workers by Class of Job
NOTE: All statistics utilize weighted data. Princeton survey weights are based on self-employed in 2011–15 American Community Survey.
The PSES included a question designed to gauge participation in the gig economy: “Last week, did you find any paying work through a digital platform, such as Uber, TaskRabbit or Handy? This is often called gig work.” A total of 14 percent of the self-employed responded yes. To put this figure in context, if you multiply 14 percent by the percent of all workers who are self-employed according to the CPS (about 8.5 percent), you come up with an estimate that a little over 1 percent of the workforce overall is in the gig economy. Of course, there are others who perform gig work who are not mainly self-employed, so that figure understates the total share of workers finding work online. Nevertheless, it is consistent with studies that find that the digitally intermediate labor market is growing exponentially (e.g., Farrell and Greig 2016).
The self-employed as a group are mostly satisfied working in self-employment. When asked, “Would you prefer to work for someone else rather than being self-employed?” a total of 83 percent responded “no.” This figure is 63.1 percent for the subset of self-employed workers who simultaneously hold a traditional job, and 63.5 percent for those who find work through a digital platform. Thus, a smaller majority of these workers seem to prefer their employment status.
For those who simultaneously held both a traditional employment job and self-employment position, I asked, “Suppose it was possible for you to work 5 more hours each week. How would you prefer to work: in self-employment at your typical hourly rate
Another finding that is consistent with this interpretation is evidence on “hour constraints.” Self-employed workers are often underemployed. In my survey, 40 percent of the self-employed worked less than 35 hours per week in their self-employment job. A third of all self-employed workers said that they would like to have more hours per week. Consistent with this finding, in the CPS the self-employed are more likely than traditional employees to be classified as part time for economic reasons. These findings suggest that one aspect of traditional employment is that it often entails an implicit contract for full-time work. The self-employed, by contrast, take on more risk, and have more volatility in income. Nonetheless, they seem by and large satisfied with this arrangement.
Work-Related Benefits
I consider two types of policies: grand, big ideas, that will not be implemented tomorrow, and incremental policies that could be implemented tomorrow. Both are important to consider. Here is a really ambitious, big idea: Hanauer and Rolf (2015) have proposed the idea of “Shared Security Accounts,” in which all workers would be covered by a universal system that provides health insurance, retirement benefits, paid leave, and so on. Employers and online platforms like Uber would contribute 25 percent of their workers’ compensation into a fund to pay for those benefits. Workers could choose which benefits they want. Although in practice the plan would be more complicated, which I explain shortly, this is the gist of the idea. I should also note that Shared Security Accounts are not a total head-in-the-sky idea. Washington State and New Jersey have considered legislation along these lines for self-employed workers.
So I asked workers about this proposal in my survey as follows: “Policy-makers have been discussing the idea of creating a fund to help self-employed workers obtain work-related benefits, such as health insurance and retirement savings, that they would be able to receive regardless of where they worked, and they could take with them if they changed jobs. Do you think this is a good idea?” Fully 81 percent thought it was a good idea; only 7 percent thought it was not a good idea. But I did not tell them that they would have to pay for it (or that whoever it was that they were working with would have to pay for it on their behalf), so maybe such strong support is not surprising. Nevertheless, workers seem to like the idea of Shared Security Accounts.
One of the core issues that arises in these proposals is what benefits should be included. The Washington State proposal looks to me like a cafeteria plan, where workers can choose benefits until their account runs out. So I wanted to probe which benefits were most important to the self-employed. Specifically, I gave respondents a list of eight potential benefits to choose from: life insurance, paid sick leave, paid family leave, retirement savings, disability insurance, unemployment insurance, workers’ compensation insurance, and health insurance. I randomly changed the order of these benefits across respondents and asked them to select and drag the benefits listed on the screen to put them in order from the ones that are most important to them to least important.
What did I find? Simply put: health insurance. More than 50 percent listed health insurance first, and another 15 percent put health insurance second. Retirement savings was a distant second in terms of desired benefits (see Figure 4 for the full results).
Ranking of Various Benefits by Self-Employed Workers
If I look at those who do not have health insurance versus those who do have health insurance, the picture looks about the same. This not surprising because in earlier work with Ilyana Kuziemka, we found that a large share of those who lacked health insurance pre-ACA had been turned down for health insurance because of pre-existing conditions, and they said they wanted health insurance. 6 So what I take away from this exercise is that the consensus of the self-employed is that they want assistance obtaining health insurance.
In the PSES, 78.6 percent of the self-employed were covered by health insurance. That is close to estimates from the CPS and IRS. Using IRA tax data, Jackson, Looney, and Ramnath (2017) find that 75 percent of the self-employed had health insurance in 2014, while 90 percent of W-2 workers had health insurance. This gap takes on added significance, and suggests a market failure, if you bear in mind that median income is 25 percent less for W-2 workers than for the self-employed.
I explored how the self-employed obtained their health insurance in the PSES. Almost 20 percent of the self-employed with health insurance reported that they obtained their insurance from healthcare.gov or a state exchange. Since these exchanges were only up and running for a few years at this time, this is a remarkable success story. Another 24 percent said they obtained health insurance from “the government or state (e.g., Medicaid or Medicare),” which is probably related to the expansion of Medicaid under the ACA. These findings are consistent with other evidence suggesting that health insurance coverage has increased significantly more for the self-employed than for the employed since the ACA took effect.
Job Training and Taxes
Next consider job training. Just 36 percent of self-employed workers in my survey said they received training in the last year. Interestingly, this figure was much higher (55 percent) for the subset of workers who had a traditional job in addition to being self-employed. Higher educated workers were also more likely to report having received training in the last year.
Across all self-employed workers, 16 percent participated in continuing education and 19 percent had skills training. Those were the two most popular forms of training. Only 8 percent of self-employed workers said they had safety training in the last year. This is a broader point, but if you look at some of the major safety disasters we have had in the last decade, such as the BP Oil Spill or the Gold King Mine catastrophe in Colorado, contractors have played a central role in causing safety and environmental hazards. I think these incidents and my survey finding of relatively little safety training raise the question of whether self-employed contractors are receiving enough safety training.
Most (62 percent) self-employed workers paid for their training themselves or had a family member pay for it. Only in 6 percent of cases was training paid for by a hirer or a customer, which again highlights the risk that contractors are not provided adequate safety training.
Taxes
Because the survey was conducted the week after taxes were due I asked, “Is the amount that you have to pay in income taxes more than you planned for this year, about the same as you planned for this year or less than you planned for this year or hadn’t paid taxes yet?” The good news is that only 6 percent admitted that they had not paid taxes yet. Fifty-six percent said they paid about the same amount as they planned for. Now I thought if I gave them a chance to complain about their taxes, more would have said they paid more rather than less than expected, but the remaining cases were close to symmetric: 17 percent said they paid more than they planned and 14 percent said they paid less than planned. I was surprised that more were not surprised by their tax bill.
I also asked if they had any unreimbursed expenses. A majority (56 percent) had no expenses over the year. Among those with positive unreimbursed expenses, the median expense was $50 and the 90th percentile was $600.
Policy Options for Benefits, Training, and Taxes
As background for discussing policy options, I feel obliged to first report my objectives for policies for independent workers. These are listed below, not necessarily in order of priority:
Policy should be neutral with respect to self-employment vs. traditional W-2 employment (e.g., tax treatment of health insurance should be equivalent).
Policy should avoid creating incentives for employers to convert W-2 jobs into self-employed jobs.
Policy should be economically efficient, which implies that policy for the self-employed should take advantage of scale economies and risk pooling, and avoid adverse selection.
Self-employed workers should be covered by the same essential protections and benefits as W-2 employees (i.e., social compact applies equally).
Policy should minimize administrative burdens and overcome well-established behavioral biases (e.g., status quo bias and hyperbolic discounting).
These objectives stem from a desire for economic policy to be efficient and fair. You will notice, however, that the objectives can sometimes be in conflict with each other. I do not view that as a fatal flaw, however, because in economics we think about trade-offs, so we can think about trade-offs when it comes to these objectives.
The desire for policy to be economically efficient is particularly challenging in this area, because, when it comes to insurance, markets often break down because of adverse selection and moral hazard problems, as Joe Stiglitz, a previous Moynihan Winner, showed in his path-breaking work with Michael Rothschild. A standard solution to these problems in insurance markets is to have universal coverage enforced by a mandate. Given political obstacles, the way we often solve these problems in practice is to pool risks within groups, which is what employer-provided insurance does.
With these objectives in mind, let us consider Shared Security Accounts. First consider the pros. The goal of Shared Security Accounts is to provide independent workers with the same safety net as traditional employees. In the Washington State bill the accounts would only apply to self-employed workers who work through a broker, such as Uber, or some other type of intermediary that connects workers to their customers. But, as I noted, one can think of applying the policy more broadly. Shared Security Accounts would have the benefit of taking advantage of scale economies since they pool workers across eligible intermediaries.
Shared Accounts also prevent free riding, which can be a major problem with our existing laws. The ACA helps to minimize this problem with health insurance, but a company like Uber is arguably free riding on other employers who provide health insurance benefits to Uber drivers, either because they have another job in addition to driving on the Uber platform or because their spouse is employed and they obtain coverage through their spouse’s employer. Another major pro is that it offers a comprehensive solution. And it maintains flexibility in that workers can take their benefits wherever they work; they are portable like Social Security.
What about the cons? I think the workers are likely to ultimately pay for this benefit because the costs of providing benefits are often shifted from firms to workers in the form of lower wages, so the next time I do this survey I will ask, “If you have to pay for this do you think it is a good idea?” Another con is that there is a risk of adverse selection if workers can choose their benefits from a cafeteria plan. In the Washington State bill there is also a significant problem with hours proration, meaning that if you work 20 hours a week as opposed to 40 hours a week, you qualify for only half the benefits. But the concept of work hours worked is ill-defined for many self-employed workers. An Uber driver, for example, could be reading a book or taking care of personal tasks while waiting for a customer (and can turn down the customer), so it is difficult to ascertain how much he or she worked in a given time period. The concept of work hours is quite different for the self-employed, and proration would be difficult or impossible to administer. Moreover, the Washington State proposal, while very important and in a way path breaking, only addresses a small portion of the population that struggles to obtain benefits. As mentioned, only 14 percent of independent workers are connected to a digital intermediary.
Another conceptual issue is that part of the compensation an independent contractor receives is a return on capital. If you are driving a car for Uber or Lyft, part of the income you are making is a return on the investment in your car, yet we would be requiring a contribution on that investment for your benefits.
Next consider health insurance. I think the most important lesson from my survey is that we should take steps to ensure that the health insurance exchanges work if we want the self-employed to continue to have access to health insurance, particularly when they are foreclosed from other options. Given that the debate on the ACA is hopefully settled for now, I will skip to the idea of a safe harbor.
One policy proposal that has gained some traction is to have a carve out for intermediaries that permits them to provide benefits without risk that their contractors will be deemed employees. Indeed, my sense is that many of the new online intermediaries would like to provide some benefits to their workers, but they refrain from doing so because they are worried that they will be classified as an employer if they provided access to benefits, such as life insurance.
Another, less extreme, solution to this problem is to support greater use of third parties to provide benefits. Stride and the Freelancers Union are examples of such third parties. There is a risk of adverse selection when it comes to third parties: an enterprising third-party benefit provider might deploy strategies to help only the healthy people, for example. This is a con.
Nevertheless, especially when I think about a group such as the Freelancer’s Union, I think that having more representation and voice for workers are positive outcomes in their own right, and my personal view is that I would be willing to sacrifice some economic efficiency from adverse selection to provide workers with greater representation and voice at work. Senator Warner has proposed the sound idea of supporting the states to have pilot projects—to experiment—on alternative ways of providing benefits. In an area where we have much to learn before going to a national scale, this seems quite safe and sensible.
One of the main concerns policy-makers have about self-employment is their enormous tax gap. The IRS estimates that about 65 percent of the income taxes owed by self-employed proprietors are not paid. 7 One way of addressing the tax gap is to have intermediaries assist with filing tax information with the IRS. For example, Uber could populate their drivers’ tax forms. Another idea worth considering is to have a standard deduction for the self-employed. If most self-employed workers have very low expenses, you can have a standard deduction like 10 percent or 5 percent of people’s income and most would be better off. This could help to increase tax compliance. It would certainly reduce record keeping. Such a proposal has been implemented in France.
The tax treatment of health insurance for the self-employed is disadvantaged compared to that of traditional employees in that the cost of providing health insurance for employees is excluded from their income, so it is not covered by the income tax or by payroll taxes. For the self-employed, however, health insurance expenses are excluded from income taxes but not from payroll taxes. With payroll taxes of around 15 percent, this creates a significant additional tax on the self-employed. That could easily be rectified through tax policy.
As mentioned, the self-employed receive relatively little job training. The IRS is tough on the deductibility of training expenses for the self-employed. Particularly when it comes to safety-related training, it would make sense for the IRS to be more permissive in allowing training deductions as a business expense. Congress could also enact tax credits to encourage job training, particularly for safety training, for self-employed workers.
Discrimination and Contract Enforcement: Findings and Policy Proposals
Two of the most interesting findings came from my survey related to discrimination against minority self-employed contractors and contract enforcement. The survey asked: “Have you ever felt that you were discriminated against by a client or customer because of your race, sex, age, ethnicity or religion?” Respondents had the opportunity to mark yes or no for each category. The key findings are summarized in Figure 5 for selected subgroups. The most disturbing result is that 47 percent of self-employed African Americans indicated that they had felt discriminated against because of race. 8 The corresponding figures were 22 percent for Hispanics and 25 percent for Asians; only 11 percent of whites said they had felt they had been discriminated against because of race. Looking at the other groups, young people (40 years old and below) were slightly more likely to say that they felt discriminated against than older workers, and that may well be related to the reason why young people are less likely to be self-employed. Last, 28 percent of women said they had felt discriminated against based on sex, compared to just 11 percent of men.
Perceived Discrimination, Selected Groups
These findings lead to an obvious policy proposal: Extend coverage under Title VII of the Civil Rights Act of 1964 to independent contractors. The self-employed currently have few options if they face discrimination. The arguments in favor of this proposal are obvious: (1) it would extend protections that already exist for employees to the self-employed; (2) discrimination is plainly unfair and economically inefficient if it originates from personal animus or ignorant stereotypes; and (3) there is an administrative system in place to enforce the policy. On the con side, one could argue that self-reported survey responses do not prove that workers were actually discriminated against, even if they report feeling discrimination. Relatedly, one could also argue that extending protection under the Civil Rights Act to self-employed workers could lead to frivolous lawsuits.
If you go back to the objectives for policy that I laid out, I think one could make a strong case for extending Title VII protections against discrimination to the self-employed.
Contract enforcement
I will conclude by talking about contractual issues. By coincidence, New York City this very week began implementing a law called the Freelance Isn’t Free Act that the City Council enacted at the end of last year to require that contractors be paid on time. Consequently, I asked in the survey, “Have you had an incident in the last year where someone hired you to do a job or project and you were not paid on time?” A total of 36 percent of respondents indicated “yes” that they had not been paid on time.
In about two-thirds of these cases contractors were paid within two months of when they thought payment was due, and in more than 10 percent of cases it took more than six months for the contractors to receive payment. The survey also asked, “Have you had an incident in the last year where you were not paid the full amount that you were owed for a job or project that you completed?” Some 27 percent of all self-employed workers said “yes” that they had had an incident in the last year when they were not paid in full. And among those who said they had a late payment, more than half reported that they had an instance when they were not paid the full amount they were owed.
New York City’s new law follows common sense contracting practices. It requires a written contract for any contract involving more than $800 in goods or services. The contract must contain (1) a list of services or products to be provided, (2) the rate and mechanism of compensation, and (3) the date at which the hiring party must pay the contractor or a mechanism for determining the date. There is a mechanism to try to resolve contract disputes and there can be penalties—up to two times damages, injunctive relief, and legal fees—if there is evidence of a pattern or practice of violations.
Arguments in favor of the Freelance Isn’t Free Act are that it addresses an important market failure and imbalance in the contractor market, it enshrines good contract practice into law, and it protects freelancers from exploitation and against retaliation. On the other side, critics could argue that the law is burdensome for the parties to administer, that it interferes with private contracts, and that it can make it harder for the hiring party to seek redress for substandard work. New York’s experience under this new law will provide valuable evidence for research going forward that should be of interest to other towns, cities, and states around the United States.
Conclusion
One of Daniel Patrick Moynihan’s famous sayings was, “Everyone is entitled to his own opinion, but not to his own facts.” The research reported in this lecture was an attempt to develop a base of facts that could be used to devise and debate policy for independent workers. The challenge of modernizing our labor and employment laws to address independent workers will likely grow in the future with the advent of digitally intermediated labor markets and the aging of the workforce. As a result, we will continue to have a sizable and growing share of the workforce that is left to its own devices when it comes to securing their safety net, obtaining training, preventing discrimination, and enforcing contracts. A key challenge for policy-makers and researchers in coming years is to develop policies tailored to independent workers to address the imbalances between the self-employed and traditional employees, and make the labor market operate more fairly and more efficiently. I am not sure if the policies that I considered in my lecture today are the best path forward, but I am convinced in the spirit of Daniel Patrick Moynihan that we need to develop a base of facts to advance the quality and outcome of the policy debate.
Footnotes
Notes
Alan B. Krueger is the Bendheim Professor of Economics and Public Policy at Princeton University. He has published widely on the economics of education, unemployment, labor demand, income inequality, and social insurance. He served as Chairman of President Barack Obama’s Council of Economic Advisers and was a member of his cabinet from 2011 to 2013. He was also Assistant Secretary for Economic Policy and Chief Economist of the U.S. Department of the Treasury in 2009–10, and Chief Economist at the U.S. Department of Labor in 1994–95.