Q&A: Assistant Professor Dmitri Koustas Discusses the US Gig Economy

Dmitri Koustas is an Assistant Professor at the University of Chicago Harris School of Public Policy. An economist focusing on labor issues, he is interested in why households take part in alternative work arrangements like the gig economy, which includes popular online platform services like Lyft and UberEats.

With a tightening economy, return-to-office mandates, a shrinking federal workforce, and other forces causing the United States labor market to constantly evolve, we sat down with Professor Koustas to discuss his work.

How would you describe your overall research agenda?

Overall, my research aims to understand how individuals and families adapt to the changing economic and work environment. The goal is to provide insights that can inform policies and practices to improve financial stability and well-being, and to better prepare households for the future of work. I am particularly interested in gig work, both older types of gig work and new platform gig work, such as rideshare, delivery services, and other often temporary, flexible jobs through online platforms.

The second area of my research agenda delves into the financial behaviors and decisions made by households. I examine how households manage financial uncertainties such as fluctuations in income – and yes, sometimes their response is to enter the gig economy. Recently, I've looked at how households responded to student loan forgiveness.

What does the future of work look like?

There's an old joke: it's difficult to make predictions, especially about the future. That’s certainly true when it comes to the future of work.

When I first started to study gig work, there was an existential concern that good wage and salary jobs were going to become a thing of the past, with most of those roles becoming contingent jobs in the gig economy. There were some flashy newspaper headlines claiming that we would all be gig workers by 2050. That was based off survey data that really did not offer an accurate picture, many of the surveys were ad hoc and non-representative, and certainly not meant to be compared over time.

My research uses tax data, and is comparable over time. It arrives at a different conclusion. My work has found that people are just as likely as before to be wage and salary workers as before, at least in their main job. Gig work, although growing in popularity, is not often replacing a full-time job but is instead filling in some income gaps for people across the United States.  

What trends are you seeing that will continue to influence how we work?

Coming out of the COVID-19 pandemic, people seem to have been getting more flexibility in their regular wage and salary jobs – so many have more remote work and ability to set their own hours, enabling gig work opportunities on top of their other employment.

There's a question about how persistent this is going to be. In the government, particularly with recent steps at the federal level, there has been some pushback against remote work. I would like to see the economy maintain some of the flexibility – I think it is good for workers and their families.

On the horizon, I think another existential challenge is the role of artificial intelligence. But the jury's still out as to whether this is going to be labor-augmenting or labor-replacing technology, on the whole. For gig workers, AI could mean self-driving cars, which is a clear labor-replacing technology. For other workers, it could mean more freedom from menial tasks. I think we're going to have to wait and see how artificial intelligence shakes out.

Who's performing platform gig work? How has that changed over time?

Lots of people do gig work. In fact, today, over 10% of the workforce makes at least some of their money outside of employee work, mostly on a part-time basis.

In our tax data analysis going back to 2000, we found it surprising that the share of the workforce engaging in gig work remained remarkably stable over much of the 2000s and 2010s, even with a very broad definition of gig work.

In this broad definition, we included anyone making earnings outside of employee arrangements in a variety of industries. This could be consultants, or it could be cleaners. It could include owner-operators in the trucking industries or other types of driving tasks. In the agriculture field, many are paid as non-employee workers – gig workers.

Beginning in 2012, we did start to see some growth in gig work. When we examined the sources of the growth, this seemed to be entirely new demand, driven by changes in technology, specifically the matching technology that platforms like the rideshare companies were now offering. Since then, the share of the workforce with gig income grew by approximately two to three percentage points. Though there are nearly 100 different platforms in the data we work with, 90% of the growth is from platforms for transportation and delivery.

Is the growth mostly full-time gig employment?

For many gig workers, it’s not their main job. We don’t see any significant increase in the likelihood that individuals are doing gig work on a full-time basis, making their full-time living from gig work or other forms of self-employment. Based on their fairly small earnings, we suspect that it’s largely secondary or supplemental employment. It seems to be augmenting other traditional wage employment over the course of the year. It’s higher for platform gig work, where about 85% have traditional work during the year, than other forms of non-wage work (like consultants or agriculture workers), where about 60% have other, wage-based employment. All of this has been stable over time.

What role did the pandemic have on platform gig work?

Ex ante, the effect of the pandemic was ambiguous.

Think back to 2020. We’re seeing many people being laid off from their regular jobs, something that generally gets people to enter the gig economy. They need extra money and the barriers to entry are low, so many would naturally enter the gig economy. But the largest part of the platform gig economy up until that point – rideshare – dramatically decreased because of stay-at-home orders.

Though rideshare fell, there was a huge demand shock for delivery. Since people were not traveling to restaurants, stores, or bars, they wanted more deliveries. In our research examining the pandemic, we find that the platform gig economy grew dramatically as a result, by millions of new workers. Nearly all of them, we think, were engaged in delivery.

Given the nature of this work and the risks of the pandemic, the platform gig economy became younger and more female. Women who felt uncomfortable having a passenger in their car may have felt more comfortable making a delivery, especially contact-free. By 2022, there was close to gender parity in platform gig work. This wasn’t true in the beginning, and it certainly wasn't true in the incumbent taxi industry, for instance.

In the broader gig economy outside of platform gig work, however, there was a significant decline, by around 10%.

Have those changes sustained over time?

The size of the general gig economy is still down today per the latest data available, except for platform gig work. There are more people making deliveries, but there are fewer people being paid as gig workers to do anything else.

There are a variety of reasons for this, but I suspect a big part of it is that regular employment has become more appealing in many ways. Regular jobs have become more flexible, something that in the past was only true in gig work. I suspect that some of the decrease in gig work broadly defined is due to that channel.

This seems like a big change in how people receive their income throughout the year. How has the Internal Revenue Service managed changes in how people work? And what does it need to do in order to be prepared in the future?

The short answer is that IRS hasn’t made a large number of changes.

Part of the problem we face is that, with the rise of platform gig work, many people are doing gig work for the first time; they may have been wage and salary workers before, and they might not understand that they have to file as self-employed workers or even pay taxes on the income that they receive.

With gig work, taxes are not being withheld, but there is still a self-employment tax that people must ultimately pay. Some people actually may do gig work because they view not having taxes withheld as a potential benefit, perhaps because they need the money upfront or because they see tax preferences associated with being self-employed. Others, like I said, just don’t know, and so there is some underreporting of income.

One thing that the IRS could do a better job at is making sure that the online platforms report labor income in a fair and consistent way. Some of the platforms are currently exploiting a loophole where they only report the income for some individuals reaching particularly high reporting thresholds.

Some issues need to be resolved by Congress. There was a great deal of discussion a few years ago about a third employment category that could better represent the modern workforce, as opposed to today’s binary system where you are either an employee (think W-2s) or a gig worker (think 1099s). Employees have many labor protections that go back to the 1930s, with the Fair Labor Standards Act and the like – hard-fought gains for workers. Gig workers by and large have none of these protections.

Switching gears a bit, what happens when people no longer have to pay student loans as a result of student loan forgiveness?

Beginning in 2021, the Biden administration changed the way that certain student loan forgiveness programs were administered, leading to over $130 billion in student loans being canceled — about 8% of total outstanding student loan debt, concentrated mostly among young people. In some cases, there were one-time adjustments, and in other cases, the forgiveness was simply sped up.

Our recent study of the effects of these programs revealed that once these student loans were forgiven, we see increases in homeownership, in auto ownership, and credit card debt. In the six months after forgiveness, the average American spent about nine cents for every dollar they have forgiven.

But that’s not all. We also find interesting changes in employment and earnings. People's propensity to work and earn falls by about 2%. We also see people change what kind of work they do.

There are a couple of ways to interpret this. Perhaps people are reducing the hours that they’re working, because they need less money to pay their student loan bill. Perhaps people have been locked into certain careers in order to pay off their student loans, and now with those constraints relaxed, they can choose other ways to make a living t. Or perhaps they are interested in becoming self-employed or starting a business.  All of this is worth further exploration.

The evidence is fairly clear that people were engaging in public service work because of the Public Service Loan Forgiveness program, which in some sense is a measure of the program’s success in encouraging people to enter public service who otherwise would not be. Once those loans are forgiven, they don’t have to be in those types of jobs anymore.

Is the American Dream changing?

I think the American Dream should constantly evolve — the economy's evolving, and people's dreams will evolve with it. The dreams of my grandparents, all immigrants from Greece to the United States, with little formal education, differed from my dreams. In many ways, I have a classic American Dream success story. My grandparents had grade school educations, and now their grandson is teaching at the University of Chicago.

At the same time, there are components of the American Dream that are constant over time. We all want a safer and prosperous country for ourselves and our children, one where hard work in school and in the workplace allows you to get ahead.

I am an optimist, and I do believe that the American Dream is alive and well. It is under threat in certain ways: it always is and always has been under threat. Good policy can help secure this dream for more people and for many generations to come, and I truly believe that high-quality research is crucial to help inform good policy.