The pandemic exposed the vulnerability of hourly service workers, who Harvard Kennedy School Professor Danny Schneider says are still being treated as disposable even after they were deemed “essential.”

Featuring Daniel Schneider
October 5, 2022
39 minutes and 18 seconds

Eight years ago, Harvard Kennedy School Professor Danny Schneider co-founded The Shift Project, which has built an unprecedented repository of data on scheduling and working conditions for hourly service workers. Analyzing the effects of the pandemic, Schneider says research shows that even as they were being lauded as heroes, working conditions for hourly service workers were deteriorating. But if there’s a silver lining, it was that the pandemic also shone a spotlight on the plight of workers who had previously been largely invisible as they dealt with low pay, ruthless algorithmic scheduling, and health problems related to stress and overwork. And as evidenced by recent successful efforts to unionize at places like Starbucks and Amazon, Schneider says hourly workers may now have a stronger voice in their own working environments. The question, he says, is, "Are corporate executives and policymakers actually listening?"

Episode Notes:

Daniel Schneider is the Malcolm Wiener Professor of Social Policy at the Harvard Kennedy School and Professor of Sociology in the Harvard Faculty of Arts and Sciences. Professor Schneider completed his B.A. in Public Policy at Brown University in 2003 and earned his PhD in Sociology and Social Policy from Princeton University in 2012. Prior to joining Harvard, he was a faculty member in the Department of Sociology at UC Berkeley and a Robert Wood Johnson Foundation Postdoctoral Scholar in Health Policy Research at Berkeley/UCSF. Professor Schneider’s research interests are focused on social demography, inequality, and the family. He has written on class inequality in parenting, the role of economic resources in marriage, divorce, and fertility, the effects of the Great Recession, and the scope of household financial fragility. As co-director of The Shift Project, his current research focuses on how precarious and unpredictable work schedules affects household economic security and worker and family health and wellbeing.

Ralph Ranalli of the HKS Office of Public Affairs and Communications is the host, producer, and editor of HKS PolicyCast. A former journalist, public television producer, and entrepreneur, he holds an A.B. in Political Science from UCLA and an M.S. in Journalism from Columbia University.

The co-producer of PolicyCast is Susan Hughes. Design and graphics support is provided by Lydia Rosenberg, Delane Meadows, and the OCPA Design Team. Social media promotion and support is provided by Natalie Montaner and the OCPA Digital Team.

Daniel Schneider (Intro): But what we also see is that these costs are intergenerational. And so far from teenagers being the ones who are working hourly service sector jobs, we find that large shares of workers are middle aged folks or older. In fact, 20% are over the age of 50 who are working in these jobs. And about a third of workers are parents to children. And those kids really suffer the consequences when their parents are exposed to unstable and unpredictable schedules.

Ralph Ranalli (Intro): Hello and welcome to the Harvard Kennedy School PolicyCast. I’m your host, Ralph Ranalli. During the COVID-19 pandemic, the lawn signs sprang up like mushrooms after a long rain. “Thank you essential workers!” they said. “Essential workers are our heroes!” But lawn signs don’t look after your kids when your boss schedules you to work on short notice. And thank you-s don’t make you any less contagious when you have to go to work with a fever because the rent is due and your employer doesn’t give sick days. Eight years ago, Harvard Kennedy School Professor Danny Schneider co-founded The Shift Project, which has built a massive repository of data on scheduling and working conditions for hourly service workers. Analyzing the effects of the pandemic, Schneider says research shows that even as they were being lauded as heroes, working conditions for hourly service workers were deteriorating. But if there’s a silver lining, it was that the pandemic also shone a spotlight on the plight of workers who had previously been largely invisible as they dealt with low pay, ruthless algorithmic scheduling, and health problems related to stress and overwork. And as evidenced by recent successful efforts to unionize at places like Starbucks and Amazon, Schneider says hourly workers may now have a stronger voice in their own working environments. The question, he says, is, "Are corporate executives and policymakers actually listening?" Danny Schneider is here to help us sort it out.

Ralph Ranalli: Danny, welcome to PolicyCast.

Daniel Schneider: Thanks so much for having me on. I'm looking forward to talking with you.

Ralph Ranalli: So you are best known these days for your work on The Shift Project, which is now the largest source of data on work scheduling for hourly service workers. You get reports from 84,000 workers in the retail and fast food sectors from across the country, including worker schedules, economic security, and the health and wellbeing of those workers and their families. Can you just go back a bit and walk me through how your work in this area got its start and sort of where and why along that evolution did you realize that it was important work?

Daniel Schneider: Yeah, thanks. I thought about the origin story, because in some sense that we ended up here is surprising. I'm trained as a sociologist, as a demographer, as a family demographer studying marriage and divorce, parenting and child wellbeing. So how did I come to be studying unstable schedules and paid sick leave? And I think that the easy answer is, they are actually pretty tightly related, but the longer story is a fortuitous set of conversations. Back in 2010, we were in a different economic crisis. We were in the great recession and there was a long tradition of understanding how crises like that affect families from the Great Depression to the Iowa farm crisis and then to the great first recession. And my colleague, Kristen Harnett, who's a professor at UC San Francisco, and I embarked on that kind of work, trying to understand how these massive shocks to employment, how foreclosures and delinquencies on mortgages, were affecting families.

And what became pretty clear in that work is that losing your job was a bad thing. And having your home foreclosed was definitely a bad thing, but so was the worry and the fear and the uncertainty that those events were imminent or at least highly possible for many workers. And really, that notion that there is this sort of short of the catastrophe, the uncertainty can really shape our demographic behavior, our health and wellbeing, our family life really grabbed our interest. And then a short time later, I moved to the Bay Area. I was connecting with some old friends, one of whom was a labor regulator at the Office of Labor Standards in San Francisco. And we were talking about this. Interestingly for how the project ended up, I think we were talking about this while I was taking a walk with my kid, exploiting my very flexible schedule over which I had a lot of control and able to balance parenting and work very well from within the academy.

And she told me about this interesting new policy area that was developing in San Francisco around secure scheduling legislation where that a law to govern the scheduling practices of many retail and food service firms. And as she described the problem to me, it became really clear. This was the instance of everyday uncertainty, not that extraordinary uncertainty that comes from worrying about losing your job in a recession, but just a relentless unpredictability. A predictability that is always there, predictably unpredictable, but you don't know exactly what form it takes. And that was really the spark that began us on chasing down this question. But what we quickly realized was that we were not going to be able to use our standard tried and true demographic methods of taking some data off the shelf and just sort of running with what was already available and easy to use.

Ralph Ranalli: So fast forward to today. Everybody's talking about all these buzz phrases, like the great resignation, quiet quitting, nobody wants to work anymore. How did we end up in this moment? And what do you think the opportunities are for this moment in terms of actually doing something about that unpredictable scheduling and its effect that it has on the people who do those kinds of jobs where they're vulnerable to it.

Daniel Schneider: I think the pandemic, in some sense, put a spotlight on an old problem, but maybe a spotlight on a problem that got worse during the pandemic. And what I mean by that is that service sector workers and hourly workers in sectors beyond, grocery and retail and fast food, have long contended with bad jobs, to use the technical term, right, jobs with low wages, with few fringe benefits, with little paid time off and with highly unstable and unpredictable work schedule, schedules that vary from day to day and week to week with little advance notice. But I think most of us didn't think about it much. And I think that it was in some sense, a "take it for granted" workforce. And I think there were various sorts of cultural schemas by which we made sense of what this was. It's the first rung on the ladder. It's just teenagers working a job after school. 

And what the pandemic showed us is that things were worse than many people had realized. They got worse in the pandemic. And we realized that wasn't, those schemas were not true. So the pandemic found this group of already precarious workers on the front lines having to go to work in person while professionals worked from home.

Ralph Ranalli: We were deeming them essential workers.

Daniel Schneider: Essential workers.

Ralph Ranalli: When they had never been deemed that before.

Daniel Schneider: And at the same time, those jobs got worse. Now, there was not just low wages and bad schedules, but sort of existential threats. It's a customer service job, and the American public was not very friendly customers necessarily during the pandemic. It was the time of real polarization and fear and anger playing out in restaurants and Walmarts around the country. And so these jobs got harder. And at the same time, they were recognized, as you say, as essential and newly visible. So that was an interesting moment. And I think on the one hand, it was a moment of real possibility for businesses to do something different, to recognize the humanity and the difficulty of their workers and the difficulties they face, for government to seize the moment and pass labor standards that might increase job quality, and for maybe the workers themselves to have a sort of moment of real consciousness about the conditions they were facing relative to what they realized were these white-collar professionals who could work from home, take sick days whenever they wanted, and otherwise be quite protected from the pandemic.

Ralph Ranalli: So it seems like we haven't grasped that moment though, that we're still in the process of trying to figure out what to do about making this situation better. Just for example, in my neighborhood, people don't talk about climate change necessarily, but they do talk a lot about the fact that two of the local Starbucks have to close at 3:00 PM because they don't have enough people working there. Why aren't we taking this more seriously? Or is it more of a gradual process, do you think, that is just kind of taking its time to unfold?

Daniel Schneider: I think in some sense there were these multiple possibilities for change that business might seize, take the high road, that government might raise the floor or that workers might find a new voice. And I think two of those things really have not happened to the extent they could have, but one might be happening. So the first is, why didn’t businesses just take the high road. In a global pandemic, the need for, for instance, paid sick leave has never been more clear. Perhaps businesses would see, even if not just an altruistic attitude, a sort of business case for voluntarily providing paid sick leave. We know from lots of research that when frontline workers show up to work sick, that spreads infection through communities, to customers, to other workers. And what we saw in this moment, it was a really good test case. That's not what happened. We did not see widespread voluntary expansion of paid sick leave in any durable way.

Ralph Ranalli: And I think, to me, it was actually eye opening reading your research on the amount of workers who showed up when they were sick. And it seems like it's actually constant, that the odds are there's probably—at any given time—someone showing up to work sick for whichever fast food establishment or retail store you're visiting. And if you really think about it, it's kind of disgusting, frankly, for lack of a better term. And you would think there would be more concern about that, but yet we look at something like the recent contract negotiations with the railroad workers, they got basically one sick day a year, and this was hailed as a major victory.

Daniel Schneider: I think sick days are a great example of the kind of impossible choices that we ask frontline workers to make. Frontline workers don't want to go to work sick. They don't want to be there and they don't want to get their coworkers and customers sick. But it’s not just that people may lack paid sick days, but that there is a constellation of labor practices that put people in an impossible situation. Not being paid really enough to live on means that you can't just take a day off. Having a regime of unstable scheduling, where you want just as many people in your store at the moment that you need them for customers and not a single soul more so that you can minimize labor cost is a recipe for people getting in trouble for calling out sick. When managers have a huge amount of discretion over who gets the good schedules and sufficient hours, they reward those who come into work no matter what, and always make their shifts and don't call out sick. And so it is this, in some sense, taking the high road means, yeah, let's give people paid sick days, but it's more than that. It is really rethinking a whole low road paradigm of staffing and compensation that is just prevalent at many of the largest firms in this sector.

Ralph Ranalli: What are some of the fallouts in terms of mental and physical health for the people who work these unpredictable schedule jobs that you've documented and that should be of concern, not to just the people who are in that situation, but to other people because it affects them as well, sort of the general public?

Daniel Schneider: I think what we see here is that work is really a fundamental determinant of health. It's an important social determinant of health. Who is healthy and who is not certainly hinges on medical care and access to insurance, but it is also, we know, fundamentally about somebody's social and economic context and work is an incredibly important part of that. And so when we talk about precarious jobs being bad for your health, that's not just because they're not paid very well and introduce economic stress, but it's also because they are incredibly volatile and uncertain. And living with that uncertainty, not knowing how many hours you're going to get or when they're going to occur or if you'll get sent home or called in, is just a stressful way to live. It's living on high alert in a sense. And what we have seen in our research is that workers have more instability and unpredictability, and I want to be clear, this is not flexibility that many white collar workers have or desire, it's instability imposed by the companies.

Ralph Ranalli: These workers open and close. It's closing then opening shifts, isn't there a name for it?

Daniel Schneider: Clopening shifts.

Ralph Ranalli: Clopening shifts. Where you close the store the night before, and then you're back to open in the morning.

Daniel Schneider: But if your store is open till midnight and you open at five, it's not a very long period of time.

Ralph Ranalli: And it's also being scheduled and rescheduled with very little notice and having to rearrange things like childcare.

Daniel Schneider: Absolutely. And so for workers themselves, we see that those workers sleep less well at night. That's a really robust finding across our studies and studies done by others in very, very different settings and different ways of estimating that. We see that these workers are just less happy. They're more depressed. They have much higher levels of psychological distress on standard scales. But it's not just the workers who are negatively affected, I think in a really ruthless view of the economy that I certainly don't ascribe to, you would say, well, listen, these are not good jobs. That's right. And they're going to have these costs and that's why workers should go to school and that's why they should train up and try to get better jobs. I don't think any job should cause people this much distress. 

But what we also see is that these costs are intergenerational. And so far from teenagers being the ones who are working hourly service sector jobs, we find that large shares of workers are middle aged folks or older. In fact, 20% are over the age of 50 who are working in these jobs. And about a third of workers are parents to children. And those kids really suffer the consequences when their parents are exposed to unstable and unpredictable schedules. Those children are more likely to have low quality sleep themselves, inconsistent bedtimes. Those kids have much more inconsistent childcare arrangements and more informal childcare arrangements. In a way, if you step back, of course, that's true. The instability parents experience will directly manifest in their kids' lives, and kids thrive on stability and routine. And that's exactly what this kinds of scheduling robs them of and it has consequences. These kids have more behavior problems, they're more internalizing and externalizing, sort of sad and mad behaviors for kids who are exposed to this kind of instability. And what it means is that these kids are not getting an equal chance out of the starting gate, that the difficult conditions their parents face at work constrain that equality of opportunity that we really see as central to our sort of American [inaudible 00:14:55] .

Ralph Ranalli: So you've studied, obviously, hourly service work, but you've also studied gig work, so-called gig work. Is there a functional difference between those two things? What is it, if so, and is one worse than the other?

Daniel Schneider: Yeah. In a way, when we turn our minds to the future of work, the thing that leaps to mind is the sort of platform or gig economy, because it feels in some sense, quite disruptive of what has been the status quo of work, where people were directly employed as a W2 employee of a company. These gig workers find themselves sort of independent contractors out on their own.

Ralph Ranalli: Like an Uber driver or a Lyft driver.

Daniel Schneider: A task rabbit, a tasker at one time. Now, I'd say in some ways what we see, even for W2 workers, is the future of workers here. They are dealing with algorithmic schedule control and surveillance on the job. So what's distinct for these gig workers? And I think the first thing that many gig workers lack is standard employment protections. Now, workers could use some more standard employment protections, W2 workers, but gig workers don't even get those. So a gig worker in California would not have the benefit of California's higher minimum wage. They would not have the benefits of the paid sick leave policies that accrue to them, of workers' compensation. And these are serious problems. And I think they become more serious when we realize that likely many of these workers are misclassified as independent contractors, that they are not true independent contractors. They are rather really more properly characterized as direct employees and companies are, in some sense, skirting their obligations under the law by treating their employees in this way.

What platform economies promise though, is a sort of notion of an independent entrepreneur, that you are your own boss. You are driving your taxi business around, and you can prosper, and that you have control. Unlike retail and food service workers, or W2 drivers who deal, as we've shown in our work, with really unstable and unpredictable schedules and little control, perhaps platform and gig workers can really choose when and how much they want to work. And that's an appealing promise, particularly in how I think Americans think of themselves. What a lot of research shows is, that's not really true and that for many gig workers who are driving and doing delivery, they're just getting by. They are not really successful entrepreneurs. We've shown this in some of our work where we can track the earnings even before taking out deductible expenses for drivers and for similar W2 retail and food service workers, and these drivers are making less money. The other one is this notion of control, and there's been some wonderful ethnographic and qualitative work done by sociologists that really takes this on, that the actual lived experience of working to the app is not one of freedom and control, but rather of a kind of algorithmic task master who needs you to be working when it needs you to be working and keeps you engaged on that app, even when you would like to stop working.

Ralph Ranalli: Yeah, algorithms seem to be a recurring theme on this show when we talk about problems. We've had Latanya Sweeney come in here and talk about algorithmic bias in the criminal justice realm. We've had Joan Donovan come in and talk about problems with algorithms when it comes to misinformation and social media bubbles. And here we are talking about algorithms again. When did this sort of algorithmic scheduling start really having a grip on these industries and started causing the problems that it's caused?

Daniel Schneider: Yeah, it's a really good question. Is it the case that this is all really new, that schedule instability and unpredictability is a new problem faced by service sector workers and those in related sectors, or is just how it's always been and it's just come more to the fore? And it ends up being a surprisingly hard question to answer, because we don't have much data on what workers were experiencing in the past. That's partly why we set up the Shift Project and started collecting all this data. But unfortunately, we didn't get around to it until 2016 so we're not able to look back in time that well. What we do know from some data sources that we have is it seems like there's a real uptick in work hour volatility around the great recession, actually, that that was a moment when for those who stayed on the job, everyone's hours got more volatile.

But what was interesting is that it was only for workers with less education, earning less money, that it was really sticky. In some sense, large employers didn't let a crisis go to waste and they took advantage of this moment of sort of unsettled economic times to institute a new set of practices that really enshrine this work hour volatility. And that kind of volatility can be done without an algorithm, a scheduling manager can just do this. But I think what human resource management software has let people do is come up with somewhat imperfect forecasts of how much staffing they will need based on a basic model, and then unfortunately, often staff people with little regard for them as humans and really much more is treating them as inputs to be slotted into spaces.

Ralph Ranalli: Right. It's sort of the commodification of workers as inputs and not as human beings, something which we've seen since the industrial revolution. But I think with the algorithms, you get this ruthless efficiency that maybe you don't get with a human schedule. And it's not just that, because I think you found that minority workers in food service and retail industries were something like 10 to 20% more likely to report canceled shifts, on-call shifts, clopenings, and things like that, so we've seen racial bias in other algorithms. Is that true here?

Daniel Schneider: I think what's going on in the scheduling process illustrates another part of the reality of algorithms, which is that they still need human help or tweaking. And that's certainly what we see in the scheduling process, is that often the algorithm or the human resource management software is used to run a first schedule. And unfortunately, that's often done without regard for when people are available to work or without regard to giving people any consistency in when they work. In this way, as you said, it's very akin to the sort of just in time inventory management, except now the problem is the inventory are people. But then things change, there's some wonderful work on The Gap by Susan Lambert and Julie Henley and colleagues where it turns out the algorithm doesn't really know when sales, big sales, are going to be announced by corporate or when trucks are going to show up to unload and need stocking in the stores And so managers have to adapt on the fly and they do this under constraint because it is common for a manager's compensation to be tied to hitting very tight labor budgets and not going over that labor budget. And that is an output that central or corporate or district managers monitor pretty closely. So how do you achieve that? How do you respond to unforeseen events, a rainstorm nobody expected and the restaurant is empty, a playoff win that nobody saw coming and the bar is full within these constraints of not really trying to minimize labor cost? 

And I think that is where we see, first, an additional element of instability creeping in, but also the room for bias to creep in. Because managers, like many Americans, have implicit or explicit racial and gendered biases really encoded into how that they're sort of cognitive schemes of the world. And so when somebody needs to get sent home, when somebody needs to get called in, that can often be the other. And that's what we find in our work is that a large part of why workers of color, particularly women of color have more unstable schedules is because they're much more likely to find themselves in situations where they are the other to their manager, their manager is not like them. And that becomes a cause of these inequalities in scheduling.

Ralph Ranalli: So how do you fix the algorithm problem? It's interesting because I used to work at the Harvard Innovation Labs, and one of the ventures there was actually a software company that was working on software that would allow people who worked these jobs where they had very flexible and changeable schedules to sort of trade shifts and do things more efficiently. But looking back on that, I'm thinking: Is that really the answer? To layer good technology on not-so-good technology? Or is it something else? Is it a policy fix?

Daniel Schneider: So before I came to Harvard, I spent eight years in the Bay Area and that really could have gone either way, but I think in the end it left me pretty skeptical of the tech fix. But here's the vision of it, and it's not unlike what this company was working to develop. I think they had some competitors. It turns out because there are a lot of these products out there in the world, the ability to try to tweak the tech to solve some of these problems and what those tweaks, or maybe really fundamental changes, we'll see look like or are a few different things. One is that workers might be able to update their availability for when they could really work and the algorithm might respect it when scheduling shifts. That might make a big difference. That's not quite control, but it's something.

Ralph Ranalli: It sort of gave them a voice. I think that's what this company was trying to do was kind of give workers a voice in inputs to the algorithm instead of just sort of being passive receptors.

Daniel Schneider: Absolutely. The other thing that these app apps can do, and they are apps that workers have on their phone, is give people the ability to swap shifts. And that in some sense is a little bit different, that's not the control over what's set or the predictability of your schedule, but it is the ability for the worker to drive changes to the schedule rather than just managers to drive changes onto the worker's schedule. And then the other innovation that I think we've seen less of in the world, but is coming is self-scheduling. The notion that workers might just pick their own shifts, that the algorithm wouldn't assign shifts at all, managers wouldn't assign shifts. Workers would have the ability to go in and select the shifts that they wanted and that worked for them. Now, things need to be worked out for this to be real because there's some shifts that nobody really wants to work and yet somebody's got to staff the store.

It's unclear who would get to pick shifts first and who gets to pick shifts second. Seniority systems, for instance, are really common in health care. Is that what would work in retail and food service settings? But a number of companies are forging ahead here and trying to implement these things. Walmart, most prominently, has rolled out a set of performs around their scheduling software to try to do this. We're working with Ikea right now on a field experiment to test some of these same changes to Ikea's scheduling systems to give workers more control, more voice. But we need to see, this is why we need to do research because we don't actually know how these things unfold and there are often unintended consequences.

Do we see, for instance, disparities in whose shift requests for swaps get accepted? Do we see inequalities in who gets to pick the most desirable shifts with self-scheduling? So we'll see, but I do think that the tech fix is not going to work on its own, at least, because even if effective, there will be real variance in who adopts these technologies. And so labor standards that lay out a basic floor of what workers are due, what we think as a society they are owed are essential. And we have seen in our work that these labor standards can make a difference for workers. Seattle was the second jurisdiction in the country to pass a secure scheduling ordinance requiring large food service and retail employers to provide at least two weeks notice of work schedules to hourly workers and then to compensate workers when they made changes, when employers made changes at the last minute.

Compensation didn't prevent these changes, it just said, workers' time has value. If you send someone home unexpectedly, you should pay them a little for that. If you ask them to stay later than expected, there should be a small bonus. And what we find is that the law did change workers' experiences of instability. The share of workers getting at least two weeks notice went up significantly. The share of workers with uncompensated cancellations or last minute timing changes went down, but not universally. We saw a significant improvement, but we know that compliance with these laws is imperfect. It can be difficult or employers often can just get away with not doing it.

Ralph Ranalli: How do you make a broad, generalized case that goes beyond just these people are people, they need to be treated like people, for a societal good, for passing these secure scheduling laws is a societal good that benefits everybody.

Daniel Schneider: I do think we can start from a sort of normative space where this is something that should be a common denominator for our working lives as Americans. And I think in some sense, those are calls rooted in a request for dignity and respect. They're not unlike what workers asked for before the passage of the Fair Labor Standards Act a hundred years ago or so. Eight hours for work, eight hours for rest, eight hours for what we will, this was a moral call and it resulted in still one of our most important labor standards. And to give that value is to recognize that it is not a natural law that firms must maximize shareholder value at all costs. That is not a property of physics. It is in some sense, just the water in which we happen to swim right now. It is our current paradigm, but it can be disrupted.

I think the second argument, why this is a good and just policy, but also a productive one is that these are really harms on workers and their children and that they cost workers substantially in terms of negative effects on their health, on their ability to achieve mobility, on their children's ability to have an equal chance out of the starting gate and that it likely imposes costs on society more generally. That poor health is an expensive proposition for workers who are not paid sufficient wages and have volatile hours turned to, if not their employer to provide a basis of living, then to social supports and social welfare organizations. And in this sense, employers are externalizing a cost that is never charged back to them, but instead is charged to all of us who subsidize this kind of low road behavior.

And then I think the third argument that I really don't think should be necessary to make is that there's a double bottom line here, that actually businesses are often quite shortsighted in adopting these precarious labor practices that, yes, you can see on the quarterly earnings that you've trimmed your labor costs and so performed well, but in the longer run, this is a much higher turnover workforce. These are less satisfied folks. They are providing less confident and warm customer service, for instance.

Ralph Ranalli: Yeah. I think you mentioned costs. I think you've mentioned some examples of companies that have done this and shown empirically that it's not necessarily a drag on the bottom line. Is there an example you can give of some companies that have done that?

Daniel Schneider: I think we see it really vividly just without even seeing a change in practices. Part of what the Shift Project gives us is the ability to look inside the sort of worker experience at dozens and dozens of firms. And there's one vision that, if you looked inside 20 fast food firms, you'd see the same thing, that it is an inevitable race to the bottom in terms of job quality. The firms have no choice. It's a competitive, ruthless market. That's not what you see. What you see is real variation across these firms in their labor practices. Scheduling at an In-and-Out Burger is really different, much more stable, much more predictable than scheduling at McDonald's or Wendy's. There is another way here, another business approach. But we also see it particularly in what's called The Gap study, which I mentioned earlier, which was a randomized store level intervention at Gap in the U.S. that gave workers in some stores access to a set of tools that return more schedule stability and predictability. What happened? Workers slept better at night. I mentioned that before. It's a really consistent finding across lots of different scheduling studies, but it wasn't just workers who benefited. What the study found was that year over year same store sales increased in part because of greater retention of more experienced workers who are more productive and more effective sales agents on the floor. So in that sense, it's not just conjecture. There's real evidence that these kinds of more sustainable and humane labor practices can return value to firms as well as workers.

Ralph Ranalli: And retraining new workers when your workers leave costs money.

Daniel Schneider: It costs money. And so our own work has shown real associations between exposure to instability and worker turnover, and that's costly. There are also a set of business practices that firms can adopt that make it more possible to give people consistent and predictable hours. One that comes up often is cross training. So if workers only know how to do one thing and that one thing doesn't need doing, then it feels really expensive to keep them at work and you want to send them home. But if workers know how to do a few things, and you trust your workers a little bit, then you open the possibility of really maintaining more stability and predictability in people's work hours, for instance.

Ralph Ranalli: One thing I found that was eye-opening when I was looking into some of your research, was your research on the effects on elderly workers of scheduling instability. They seem to have less of it, but it does seem to still affect them and affect them in a way that's sort of unique to who they are. Can you talk a little bit about what you found when you looked at elderly workers?

Daniel Schneider: Again, we have this characterization of service sector jobs as often just kids.

Ralph Ranalli: A 16 year old flipping burgers at McDonald's.

Daniel Schneider: After school. We know from Walmart greeters, that's not true, but it's a more general demographic proposition in the sector. About 20% of workers are over the age 50, for instance, at these large firms. And in some ways, what we see here is emblematic of a broader shift in the American workforce, where we are asking, expecting workers to work longer and to work to later ages. And it's not clear that jobs are designed particularly well to allow that. I think scheduling is one place where we see that. Now, what we do see is that older workers have somewhat less exposure to really unstable and unpredictable schedules than their younger counterparts. That's partly about tenure and seniority. It's probably partly also about the ability to work productively with managers and get assigned to better shifts.

Now, the other question though, is that there's still high levels of schedule instability and unpredictability for these workers. Does it matter in the same way for them, maybe. Older workers are sort of free of child raising responsibilities. They're not trying to juggle school and work. And we just don't see that, I think what we see instead is that these schedules have really dire consequences for their health and wellbeing, for their happiness, for their sleep quality and for their intention to keep working. In some sense, these jobs, which are so difficult to sustain, make it very difficult for older workers who are asking to keep working to do that. We're trying to follow up on some of that work now by actually tracking workers over time. Re-interviewing the same workers, 3, 6, 9 months later to see, are older workers not just unhappy with their jobs, but are they pushed out of these jobs? Are they pushed out of the labor force altogether, perhaps before they're ready, given age of social security benefits eligibility and the capacity of workers to save for retirement.

Ralph Ranalli: When you look at the possibilities of what can be done in this space, both from a policy standpoint, but also changing public perceptions, what are your hopes for what can be achieved, say five years from now in terms of making this a better space for these kind of workers, a healthier space and a healthier space just for us as a society?

Daniel Schneider: When we began our conversation, we sort of imagined three ways that the world could change for these workers and for their jobs. One is that companies might sort of take the high road. The other is that we might pass stronger and more progressive labor standards. The third is that workers might find a new voice and a new worker power. And I'm somewhat hopeful on some of these fronts. We've talked a little bit about, I think, a narrow but realistic path for employers that they might voluntarily adopt, that there might be a tech fix that some firms will embrace that could benefit workers. But I am not optimistic that firms will broadly embrace voluntary, paid sick leave provision, or much more stable and predictable schedules without that technological intervention. 

Labor standards really have the possibility to help workers, not just when their firms agree out of beneficence to do these things, or because they've seen a profit motive, but more broadly in the population. And we have seen over the last 15 years an explosion of state and local labor standards that have really, because they're passed in large states, and small states covered more and more workers in the American economy. And that work continues. There was a little bit of a pause during the pandemic period, but now again, we see states from Connecticut to Colorado considering secure scheduling laws and passing paid sick leave laws. And this does make a difference for workers, even short of a federal standard. At the federal level, we certainly know what an optimistic position would look like. It would look like passing a federal guarantee of paid sick leave and paid family and medical leave for all workers or even Elizabeth Warren and Rosa DeLauro's federal Schedules That Work Act. Congress has been effective recently in passing legislation like the IRA that matters for households and workers. And so I think we can try to stay optimistic.

And then the last one is that workers do seem to be finding a new voice and do seem to be organizing effectively. Unions have been on a long, not so slow decline in the United States for decades. And what we see at Amazon to some extent at Starbucks, and now, Trader Joe's and Chipotle, is a really new energy that seems to be taking off to be, as this research in areas are kind of catching fire. We've been tracking the attitudes of service sector workers towards unionization over the last six months or so. And what we see is some of that process, that when Starbucks workers—because that's where most of this action has been—when Starbucks workers vote to unionize near where you work, at a McDonald's or at Wendy's, those workers' attitudes towards utilization grow more favorable. And so in this way, there is the possibility that this movement continues to sort of self-power and to feed that fire of organizing. Now, to get from that to bargaining successfully with companies and winning what workers want is several more steps, but it is a change unlike any we've seen in some time.

Ralph Ranalli: Well, this is certainly a really interesting moment in history to be talking about these issues. And I just wanted to say thank you very much for coming in and walking through it with me. And it's been a great conversation. Thank you.

Daniel Schneider: Hey, thanks a lot. I really enjoyed it. I hope your listeners do too.

Ralph Ranalli (Outro): Thanks for listening. Please join us for our next episode, when we’ll talk to Harvard Kennedy School professors Erica Chenoweth and Zoe Marks about the recent protests in Iran and their important research on the connection between authoritarianism and violence against women. If you have a comment or a question about PolicyCast, please email us at policycast dot H-K-S dot harvard dot edu. And until next time, remember to speak bravely, and listen generously.