Maria Fernandes worked three jobs at three different Dunkin’ Donuts stores in New Jersey. She napped when she could, often in her car between shifts. She tended to sleep with the engine running, and kept a container full of fuel in the back in case she ran out. On August 25, 2014, the container accidentally overturned and the car filled with fumes while she slept. Fernandes, 32, died there, still wearing her uniform.
Her death has been held up as a sign of the stresses facing America's low-wage workers, and the stresses that our jobs can put on our sleep. A new study by public health researchers at the University of Pennsylvania adds more evidence, suggesting that the main activity to which Americans sacrifice sleep is paid work.
And this sacrifice is not evenly distributed. Adults like Fernandes who work multiple jobs—estimated to be around 15% of the American workforce—are 61% more likely than others to report sleeping six hours or less on weekdays. Short sleepers also tend to have longer commutes, work irregular or shift hours, start work earlier, and get home later in the evening than normal sleepers, according to the research. The lower your family income and the lower the level of education you've received, the higher your odds are of being a short sleeper.
In other words, sleep has its own caste system.
"This study presents another opportunity to raise concerns about sleep patterns as both an unmet public health and a social justice problem," Lauren Hale, an associate professor of preventive medicine, writes in an essay accompanying the paper, which was published in the journal Sleep in December.
More than ever, working Americans are starved of sleep: Up to 30% of employed adults report routinely sleeping less than six hours a night, representing approximately 40.6 million workers. (The American Academy of Sleep Medicine recommends that adults get about seven to nine hours of nightly sleep for optimal health, productivity, and alertness.) Short sleepers are also potentially sacrificing their health and safety: Short sleep duration has been linked to obesity, diabetes, hypertension, stroke, coronary artery disease, and higher levels of mortality in general.
"My research has shown that individuals with lower socioeconomic status (i.e., less education, unemployed) are more likely to have inappropriate sleep duration and poorer sleep quality," says Hale, who is an editor of the journal Sleep Health.
There are many possible reasons for this, Hale says: Hours spent sleeping are lost amid higher levels of anxiety, more financial insecurity, poorer health, and less free time, with more time spent working low-income jobs and commuting to work. Additionally, "some of this may be related to the physical environment in which people sleep. Maintaining a regular, quiet, cool, and dark sleeping space may be a luxury that not all can afford."
Another study shows that African-Americans were over three times as likely as whites to report very short sleep—less than five hours—while Asians and non-Mexican Hispanics were two to three times as likely. Racism may also take a toll on these groups: Discrimination based on race is associated with shorter sleep and more sleep difficulties.
For Hale, who has been sounding an alarm about the importance of sleep for life trajectories and health outcomes, sleep poverty is another form of social inequality. And short sleeping times may eventually compound the effects of other forms of scarcity—of money, of time, of health, of opportunity—to further limit social mobility.
"When taken as a whole, the patterns show that those with lower levels of social status are more likely to sleep either too little or too much, categories which are associated with higher risks of mortality among a host of other adverse outcomes," Hale wrote in her commentary on the study. The research examined a wealth of publicly available data about how Americans spend their time, from the U.S. Census's American Time Use Survey, in the years between 2003 and 2011.
Having less than you need of any resource has been shown to limit long-term planning, increase anxiety, and sap both brainpower and willpower. Poor sleep, for example, increases cravings for high-calorie foods. Shorter sleep has also been associated with poor psychological health.
One group at the nexus of race, low income, and social status is domestic workers. Nannies, cleaners, and elder-care workers are predominantly black or Hispanic, female, and poor. And most domestic workers are excluded from basic labor protections, making them vulnerable to exploitation. A survey conducted by the National Domestic Workers Alliance showed that 25% of live-in domestic workers had responsibilities that prevented them from getting at least five hours of uninterrupted sleep during the week prior to being interviewed. Another report cataloged the typical day of a live-out nanny, which began at 5 a.m. and ended at 11 p.m.
"Stories abound about domestic workers who work from before dawn to well into the night; who are awakened in the middle of the night to perform minor tasks; whose sleeping quarters are in basements, in closets, on porches," says Linda Burnham, research director of the National Domestic Workers Alliance and coauthor of the report.
In 2013, California's Domestic Worker Bill of Rights included a provision requiring employers of 24-hour or live-in caregivers to provide eight hours of uninterrupted sleep in adequate sleeping conditions. California and New York are currently the only states that afford some labor protections to domestic workers.
"Deprived sleep for domestic workers can spiral into other health risks, including risk of hypertension, diabetes, heart disease, obesity, depression, anxiety, and premature mortality," says Jennifer Dillon, the communications director of the National Domestic Workers Alliance. "Domestic workers providing care without sufficient sleep may be more prone to mistakes and accidents, affecting not only their own health and safety, but of those under their care as well."
Domestic workers are also subject to what Dillon calls "time theft": that's time consumed by irregular shift hours, being on call for shifts that never materialize, and long unpaid commutes. And less time means less sleep.
In a recent blistering exposé of the human cost of the rise of workforce monitoring software for Harper's, Esther Kaplan highlights the same issue for freelancers who find work through marketplaces like oDesk and for workers at companies like FedEx and Century 21, citing a 2012 study of retail workers that found only 17% of surveyed workers—and just 10% of those who were part-time—had a set schedule. Only 30% received their schedule more than a week in advance.
In 2014, Starbucks announced that it would curb "clopenings"—shift schedules where the same worker is required to close a branch and open it again in the morning—after the New York Times profiled a barista and single mother who worked two clopenings in a single weekend, closing at 11 p.m. and reopening at 4 a.m.
Low-paid retail workers are highly likely to have a second or third job just to make ends meet. "Having multiple jobs and/or irregular hours affects one’s internal clock, making it difficult to get sufficient sleep at night," says Hale.
On the other end of the sleep spectrum are workers like pilots who are required by law to get a minimum amount of sleep. In 2014, the Federal Aviation Administration introduced new rules that require pilots to have 10 hours of rest, including eight hours of uninterrupted sleep, between each eight- or nine-hour shift. The new rules were introduced partly in response to the 2009 crash of Colgan Air Flight 3407, which killed 50 people, and in which pilot fatigue was determined to have played a part.
In other areas of transportation, like commuter and freight rail, where hours can be quite irregular and regulations are fewer, exhaustion can also take a heavy toll. In a five-year study conducted by the Federal Railroad Administration of main-track train collisions, 29 percent were found to be possibly fatigue-related. Still, despite a five-year-old federal law directing the FRA to issue regulations on sleep disorders and fatigue, such rules remain nonexistent.
One surprising finding in the sleep data is that first-generation immigrants have fewer sleep problems, and tend to sleep more than their U.S.-born peers.
"Some studies show that the more ‘American’ an immigrant becomes, the less sleep and more sleep complaints the immigrant gets," says Hale. "This negative acculturation is likely related to an entire lifestyle transition as immigrants take on American behaviors in terms of diet, media, and attitudes." In an international poll, Americans reported sleeping about 40 minutes less on workdays than respondents in the other countries surveyed.
So sleep may have a cultural aspect too, with higher levels of sleep poverty reflecting the priorities of a society. In a country where 9.8% of those surveyed about their sleep habits had multiple jobs; a third of the workforce is now engaged in contract, temporary, or freelance work; and social inequality is deeply entrenched, how can we tackle the sleep crisis?
Hale thinks that the sleep gap cannot be bridged simply by treating individuals, but will require structural and societal changes. "We should propose policies, programs, and interventions that give individuals more control over their sleep health, including more control over their schedules," says Hale. "This may include addressing public transportation challenges, encouraging flexible work hours among employees, and, for teenagers, delaying high school start time."
More research is needed, too. Hale underscores that the data doesn't offer evidence of causality. For instance, some people manage to get by on shorter sleep than others—true short sleepers, they're called—and thus intentionally select jobs with earlier schedules (and vice versa). Among the study's other limitations, she notes that people tend to overreport physiological sleep time, and the various data do not include various other factors that can influence sleep, like alcohol and caffeine consumption, sleep disorders, medication use, psychological well-being, room temperature, and light exposure.
Still, says Hale, the study offers valuable results and interesting details about how we've been sleeping. For instance, in 2008 and 2009, the prime years of the economic recession, Americans' average sleep duration was longer.
Another interesting finding: Freelancing might actually boost sleep. Self-employed workers achieve significantly more sleep than private-sector employees, and are 17% less likely to be short sleepers. Another recent study also reported that an increase in perceived control over work time was associated with more sleep and lower levels of fatigue and depression symptoms. Later start times help too. Workers who clock on before or at 6 a.m. get an average of six hours' sleep. Those who start work between 9 a.m. and 10 a.m. log around 7.29 hours.
But, says Hale, shift workers and those with multiple jobs are unlikely to be offered such flexibility. The broader solution, she insists, will have to come from a mix of personal behavior and changes to social policy. "At an even broader level, policies that improve educational and employment options, as well as increased quality health care for all Americans, will move us in a direction to reduce disparities in sleep."