Spotlight Exclusives

Shift Scheduling Laws Can Boost Worker Well Being

Daniel Schneider and Kristen Harknett Daniel Schneider and Kristen Harknett, posted on

The working conditions of frontline workers have received increased attention since the onset of the COVID-19 pandemic last year—but difficult and even dangerous work environments have been a problem in this sector for decades. In 2016, Daniel Schneider, professor of Public Policy at Harvard’s Kennedy School, and Kristin Harknett, associate professor, University of California, San Francisco School of Nursing, Social & Behavioral Sciences, created The Shift Project as an effort to collect data on work scheduling for hourly service workers. Schneider and Harknett spoke with Spotlight about their recent analysis of the Seattle Secure Scheduling Ordinance, which went into effect in 2017 as one of the nation’s first laws to regulate workers’ schedule predictability. The conversation has been lightly edited for length and clarity.

Why don’t we start with explaining what the Shift Project does?

Schneider: The Shift Project really grew out of Kristen’s and my interest in understanding how day-to-day uncertainty for many low-wage workers shaped everyday life—how it shaped people’s sleep and stress and how it shaped their economic security and how they interacted with their kids and how their kids fared. We spend a lot of time thinking about poverty in terms of low income and wages, but there’s a lot more going on in these jobs than just folks not getting paid a lot. These jobs are precarious along multiple dimensions. And around 2015, we began to understand that unstable and unpredictable work schedules were a really important part of that picture. Now, we’re both trained as demographers and sociologists, so we turned to our standard social science data sets as we often do when we have an interesting question. And we came up empty.  This was an aspect of the contemporary economic experience, of the experience of work, that wasn’t well captured in the census or the CPS (Current Population Survey) or the PSID (Panel Study of Income Dynamics).

And so, we decided we needed to collect that data. And we began to do that by doing in-depth interviews with retail and food service workers to understand their experiences. And then we brought that to a larger scale, and we developed a method for surveying frontline retail, food service, pharmacy, grocery, hardware and big box workers, and others, to collect really detailed information about what they were experiencing on the job. Their wages, yes, but also how much notice they got of their work schedules, whether their schedules changed at the last minute, whether they had to work on call, how much control they had over when they could work, if they had any schedule flexibility at all. And what we discovered through this is that the days of a nine-to-five schedule were long gone, but so were the days of a regular night shift or a weekend shift. Instead, people were working schedules that were highly variable, that were unstable and unpredictable, where they had as little as a week or even a couple days’ notice of when they were going to work. And then once that schedule was published, it was often changed at the last minute, as employers sought to really carefully align staffing with customer demand. They never wanted somebody on the floor that wasn’t absolutely needed. And they never wanted to have any fewer workers than they absolutely needed. And so, people ended up being deployed as kind of a just-in-time resource, as employers shifted payroll risks onto individual workers. And that’s been the principal concern of the Shift Project, that we’ve tried to collect this new data on and really understand how it affects workers.

And do you collect data or an annual basis or how does that work?

We collect it on an ongoing basis. We do basically two survey waves a year. It’s a repeated, cross-sectional survey, and we field one wave in the fall and one wave in the spring. And in each wave, we target workers who are employed at one of about 150 of the largest retail and food service companies in the country. We have also embedded some longitudinal panel data collection in this design where we go back to people, having interviewed them in our baseline survey, and try to re-interview them to understand how things are changing.  And as Kristen will talk about, we also oversample workers in particular places that have passed some innovative legislation designed to regulate work schedules.

Which brings us to Seattle, right? Let’s start by just talking a little bit about the law and then you can get into the study looking at what the impact has been.

Harknett: Sure. Seattle passed the Secure Scheduling Ordinance, and it went into effect in the summer of 2017. And what this ordinance did was regulate the kind of routine unpredictability that Danny was just describing in work schedules for workers employed by large service sector firms. It had a few different components to it. The simplest was you need to give workers two weeks advance notice of their schedules. So, the first thing is schedules just need to be made more predictable. Workers need some lead time so that they can line up child care, so if they’re a student, they can plan their class schedule—generally so they can plan their lives around work. The other provision was that if you schedule workers with less than two weeks’ notice, or you change their schedule with less than two weeks’ notice, you need to compensate them. You give them what’s called predictability pay, and it’s sort of conceptually similar to overtime pay where you can ask workers to work extra, but you need to pay them extra. So, in the case of scheduling, you can ask workers to work with little notice, but you need to compensate them for that inconvenience, for dropping everything and coming in.

And if a worker has a shift scheduled and they come in and you say, oh, actually we don’t need, you need to pay them for half of that shift. The principle behind it is that we had really seen a pendulum shift where workers were bearing all of the risks of the uncertainty of the ups and downs of business demand. But now these laws, like Seattle’s, try to restore more balance where we recognize businesses will sometimes have to make changes at the last minute, but then workers deserve a little bit of extra compensation for that.

And this is one of a number of laws like this, right?

Yes. There are a handful of laws like this. San Francisco was the very first city to pass legislation like this, and then Seattle and a small city named Emeryville in California followed suit a couple of years later. And now Chicago and Philadelphia and New York City also have fair work week laws as well as the state of Oregon.

And what did you find was the impact of Seattle’s law?

Danny described how the Shift Project can target and collect survey information from service sector workers about their experiences with unpredictable schedules. And so, we did just that, in Seattle and a number of other cities that we could compare Seattle to. And we did that before the secure scheduling ordinance went into effect, so that we’d have a baseline of what things looked like before this law regulated work schedules. And then we went back and surveyed workers one and two years later to see how their conditions had changed.

And what did you find?

We found that the Seattle ordinance worked as intended in a number of respects. We saw that workers in Seattle were more likely to get that two-weeks’ notice that the law required. The increase in two-weeks’ notice was about a 20% increase compared to what we saw at the baseline. Now, not every worker got the two-weeks’ notice they were supposed to get, but we saw a sizable increase in the share of workers who got two-weeks’ notice in Seattle. We also saw a pretty big decrease in those last-minute schedule changes and in last-minute schedule changes without getting that extra compensation.

And one of the exciting things that we saw is that we looked beyond the effects on workers’ work schedules, because as Danny described, in our other research in the Shift Project, we knew that there was really strong connection between schedules and bottom-line measures of wellbeing. So, in Seattle, we had an opportunity to see if a law can rein in unpredictable schedules and at the same time, to see if there were downstream effects on things like sleep and material hardship and happiness. And in Seattle, we do see improvements in all those areas. After this law went into effect, we saw improvements in workers’ sleep quality. We saw reductions in their experiences of material hardship. And we saw increases in their reports of happiness—all the things that we had expected would be associated with these better schedules. And so, Seattle is really a proof of concept that a law like this can change schedules in the way it’s supposed to, and that when we do so, we see that workers experience improvements in their wellbeing as well.

And you were collecting this data pre-pandemic, right? How has that changed this conversation?

It’s a great question. All of the Seattle results that I described were pre-pandemic. And when the pandemic struck, the focus on conditions of service sector work really shifted dramatically. There had been a lot of momentum behind this fair work week conversation, but it all kind of ground to a halt in the pandemic and people were much more concerned with health and safety and availability of paid sick leave. And so unfortunately, that conversation about schedules receded in importance. I hope that schedules will come back and hopefully this article will do a part in that, because during a pandemic, it’s no less important to have some notice of your schedule and some predictability of your schedule, and maybe it becomes even more important in the context of the, all these extra stressors and all this extra uncertainty going on with daycares closing and school’s closing and working from home. So, I think schedules remain very important, but were underappreciated during the pandemic.

I have to say, I’m surprised by that. I would’ve thought that there would’ve been increased interest, given everything that’s going on.

Schneider: I think one thing we see here is about the dynamics of power. During the pandemic, it’s been a little bit of a rollercoaster when it comes to who holds power. In the early days of the pandemic, we saw millions of service workers losing their jobs and that was not a time when any elected officials were really embracing implementing new mandates on work schedules in the face of pretty consistent business arguments against it that say this will increase labor costs and reduce jobs. Now, that pendulum has swung back and we’re at a moment where workers seem to have in their hands a lot of power, a lot of labor market power as by some measures labor markets are quite tight and at least some serve sector employers are having trouble staffing up.

And so, that kind of power could now change things in two ways. One is that it could lead to policymakers feeling like they have the leeway to pass new laws and to set new standards. This is a time when jobs are hard to come by, where you can raise the floor. I don’t think we’ve seen a lot of action that is consistent with that getting done, at least at the federal level. The Build Back Better bill may or may not include any paid family and medical leave; initial efforts to rate the minimum wage did not go very far. That’s been somewhat of a surprise given this dynamic.

The other possible change is that if businesses really are having so much trouble finding workers, well, then maybe if they made their jobs more humane, they would have less trouble. And we’ve certainly seen stories in the press about businesses raising their wages—although we tend to hear about the businesses that raise their wages and not about the businesses that don’t. So, might we see some progress on schedules driven by that market logic?  I don’t think we’ve seen a lot of that sort of progress yet, at least in the data we’re tracking. But we might see some change. Walmart has said that they’re going to institute more schedule flexibility, letting workers swap shifts with each other or update their availability more often. We’re doing a project with IKEA testing out some of those same kind of changes to the underlying algorithm to give workers more control over their work schedules. But these things take time. Not that raising wages is easy, but changing scheduling practices is really a change to the fundamental way that these firms do business. Even if there was a business appetite, the ship is hard to turn.

Does the rise of telework really impact this? Or are these positions that tend to be not as affected by that?

Harknett: I would say the latter. Most of the sector workers don’t have the option to work from home. They still have to go in. One implication of that is that during the pandemic, when many people were told, “you have to go work from home,” service sector workers were alongside first responders, healthcare workers, who had to keep showing up for work in person and subjecting themselves to health risks. And so, it really did spark a conversation about service sector work as essential work, as work that is essential for meeting our basic, daily needs. I think there was some appetite to talk about things like hazard pay, but I think because the conditions were so uncertain for businesses, it was a hard moment to imagine requiring more predictability in work schedules. I think maybe now that things are feeling a little bit calmer, and businesses are open and staying open, it may be a good time for people to start re-appreciating how schedules are a really fundamental aspect of job quality conditions and quality of life for workers.

Has this ever been introduced or even discussed at the federal level?

Schneider: The federal Schedules That Work Act was last introduced in 2019. Its sponsor in the Senate was (Sen.) Elizabeth Warren (D-Mass.) and Rep. (Rosa) DeLauro (D-Conn.) in the House. There’s an element to that bill that’s about schedule stability and an element that’s about sufficient hours as well. But the federal vote is hard, and it may be that the more immediate prospects for legislative change will be at the state level.  States like Connecticut have several times introduced these bills and they’ve gotten close to passage and there was a similar effort in New Jersey that came close.



Kristin Harknett, associate professor, University of California, San Francisco School of Nursing, Social & Behavioral Sciences


Daniel Schneider, professor of Public Policy at Harvard’s Kennedy School

« Back to Spotlight Exclusives