Take a look at the above photo of a New York City subway platform and guess what day and time it was taken. If your snap glance absorbed only the crowd, you probably guessed a weekday rush hour. But look more closely. You don’t see grey-haired men in flannel suits with solemn faces, you see All The Young Dudes in jeans just kind of slouching there, dude-like. You don’t see businesswomen striding for the stairs, you see ponytails and a lime green T-shirt that wouldn’t fly even on the most casual of Fridays.

This is not the picture of a platform at morning or evening rush on a weekday in Manhattan. It’s the picture of a platform at half past one. In the morning. On a weekend. In Brooklyn. It’s also a sign of things to come.

The growth of midday, evening, and weekend transit use is not unique to this particular stop on the New York City subway. More critically, the rise of off-peak ridership is not unique to New York City or to subway systems, either. Metropolitan areas across the United States — whether their primary mass transit system is a metro rail or a commuter train or a bus network — are recognizing that city residents can’t get by on great rush-hour service alone. They need frequent, reliable transit all hours of the day and long into the night.

“The growth in transit ridership is happening in the off-peak hours,” says transportation planner David King of Columbia University. “It’s strange. You get on a train at five o’clock in morning and it’s jammed.”

Take the New York City subway in a broader sense. Since 2007, ridership on the weekends has grown at a much greater rate than ridership on the weekdays. During the period from 2007 to 2012, weekday ridership grew at just under 7 percent. During that same stretch, weekend ridership grew at just over 10 percent. A planning director at the Metropolitan Transportation Authority told the New York Times in 2011 that to find a similar explosion in weekend subway use you’d have to go back to a time when people worked six days a week.

“The New York City subway has seen tremendous growth on the weekends over the years,” says MTA spokesman Aaron Donovan. “Weekend growth has outpaced regular growth.”

Now head to the Midwest and take the bus system in Minneapolis-St. Paul. There, too, off-peak service demand has outpaced rush-hour growth along some bus corridors. In response, the Metro Transit agency in the Twin Cities expanded evening and weekend service last summer. Some off-peak frequencies have tripled — down to a bus every 20 minutes instead of one every hour. That puts service ahead of where it was even before the Great Recession. In other words, this isn’t just the economy recovering, it’s ridership surging.

“There’s many routes where the off-peak ridership is growing faster than the peak ridership,” says John Levin, director of service development at Metro Transit. “We’re always going and finding where we can free up resources and where we need to add resources, and it tended to be that we’ve seen the most need during the off-peak, in terms of the overall scale.”

And go to Los Angeles, where even commuter rail — the transport mode created specifically for rush-hour riders — has seen an off-peak and weekend bump in some metro areas. Metrolink spokesman Jeff Lustgarten says weekend ridership in May 2013 hit 21,315, a jump of nearly 30 percent on the year before. He says that while weekday ridership is steady, weekend growth has been in the double digits. In response to this off-peak demand, Metrolink began promoting weekend rides and recently doubled some Sunday service.

“Certainly commuter-based travel is always going to be a core component of overall ridership, but people who have recreational trips … they’re taking advantage of the system on the weekends,” says Lustgarten. “Generally speaking, people are looking for alternative means of getting around town.”

Looking for it on a weekend. In spring and summer. In Los Angeles.

•       •       •       •       •Transit experts have been making the case for off-peak service expansion for years. It’s oftencost-efficient. (Many drivers needed for rush hour get paid to sit around during the midday hours.) It’s always great for society. (Lower-income people use off-peak transit at much higher rates than wealthy people; a 2003 study found that 60 percent of off-peak riders made under $40,000 a year.) And there’s enormous growth potential. (Two-thirds of transit trips are notwork commutes, as the Commuting in America, 2013 chart below shows, making them strong candidates to occur outside rush hour.)

“There’s long been a recognition here that frequency improvements — especially off-peak frequency improvement — more than pay for themselves in terms of ridership,” says Metro Transit’s Levin. “When we doubled the frequency on one of our core routes a few years ago, we more than doubled the ridership.”

Best of all, the benefits of full-day service create a cycle that perpetuates more transit use across the board. That’s the main takeaway of a recent off-peak service analysis made on the Pascack Valley line of New Jersey Transit commuter rail. The agency introduced non-rush hour trains on that line in October 2007 — seven inbound and six outbound where there’d been no off-peak service before. In June 2010, Devajyoti Deka of the Alan M. Voorhees Transportation Center began conducting surveys and on-board focus groups with off-peak and peak riders alike, to see how the service change had influenced their behavior.

Without question, the addition of off-peak service on the Pascack Valley line took cars off the road. In a recent issue of Transportation, Deka and coauthor Thomas Marchwinski of NJT report savings of at least 12 million vehicle miles a year. More fascinating was the way off-peak trains affected rush-hour ridership. Roughly 5 percent of surveyed riders started using more peak trains once the off-peak service was introduced. And of all the passengers who said they’d go back to driving if off-peak service were cancelled, three in five were peak riders.

Deka believes that there’s a psychological element to off-peak service that transit agencies fail to appreciate. If people know a train can take you back anytime you need, they’re more willing to take the train in during rush hour in the morning. “They have this thing in the back of their mind that if they have to come back early they can come back early, or if they have to stay late they can stay late,” says Deka. “So there is this indirect benefit which you will not notice in ridership data.”

(As for that ridership data, Pascack Valley weekend ridership was up more than 20 percent in the first quarter of 2013 over the year before, while weekday was up 8 percent [PDF]. That trend held true across the whole NJT system: weekends up 12 percent, weekdays 3 percent.)

Considering the rationale for off-peak service has been around for years, the big question is why transit agencies are only now seeing enough fresh demand to do something about it. Some agencies point to changing travel habits among Millennials. Some experts see a broader but related shift in American auto dependency, with an increasing number of urban households living car-free. That’s true even in places without great transit systems — Detroit experienced a 5 percent increase in car-free households from 2007 to 2012 — suggesting economic roots.

Immigration might play a role in off-peak demand, too. Last year, Governing reported that immigration had surpassed domestic population growth in 135 U.S. metro areas, according to Census data. Such demographic shifts could have a big influence on the nation’s transport network, because low-income immigrants are much more likely to commute off-peak than their American-born counterparts (see evening rates below), says planning professor Michael Smart of Rutgers, who studies immigrant transportation patterns. They’re also more likely to use transit for the types of non-work trips that often occur off-peak; for instance, says Smart, they’re five times more likely to take transit to get groceries.

“It’s definitely true that immigrants are more likely to be using transit to get to work in odd hours,” he says. “But even more than that, they’re much more likely than the U.S. born — particularly low-income or low-skilled foreign-born people — to use transit for things that are not about a job.”

Then there are changing work patterns themselves. The rise of telecommuting means people traveling at non-traditional times for both labor and leisure. Such shifts, in turn, mean service workers must travel at off times to get to their jobs. The result, says David King, the Columbia planner, is a bifurcation of the labor market in which neither high-skill nor low-skill workers are tethered to a 9-to-5 workday — or a 9-to-5 transit system — as strongly as they used to be.

“That will dramatically change how we travel,” says King. “What that means for future investment priorities is also important.”

•       •       •       •       •Bay Area Rapid Transit is already weighing what off-peak demand might mean for tomorrow’s transit investments. BART has long been considered a hybrid commuter rail and metro core system: serving downtown San Francisco but also the suburban Bay area. The plans for 2025 and beyond, dubbed “Metro Vision,” call for tipping this balance toward the core end [PDF]. That means trains running every 15 minutes or better middays, late nights, and weekends — true “show up and go” service.

“That gets us less out of the commuter rail mindset and more to the metro mindset of frequent service for 18 hours a day, rather than just frequent service during the peak,” says Tom Radulovich, head of the BART board of directors. “Metro Vision, just the name of the project implies that at least the BART planners think we’re more of a metro than commuter rail. And this is what metros do — run frequent off-peak service.”

The ridership trends certainly point in that direction. Off-peak ridership on BART has grown steadily since mid-2011, often outpacing rush-hour rates. In October 2012, for instance, peak ridership grew 10 percent on the year before while weekday off-peak grew 14 percent, Saturday grew 21 percent, and Sunday grew 13 percent. The agency made off-peak expansions several years ago only to cut them during the recession, but it’s started making them again on what Radulovich calls the “shoulders of the peak.” Those first few trains after rush-hour service ended were just too crowded.

Radulovich sees a number of reasons for the rise in off-peak demand. Tech companies keeping unusual hours. Service workers returning to the job market on swing shifts. A declining rate of car-ownership among riders. Perhaps above all, a rise in residential and business development in and around BART stations — and not just those located downtown. Altogether it amounts to a culture of residents less reliant on the automobile for whatever trip purpose, at whatever trip time.

“I think those folks are going to want BART to run more frequently and be more convenient at more hours of the day,” he says. “They’re going to be interested in off-peak trips, they’re going to be interested in Saturday and Sunday frequency, they’re going to be interested in evening frequency, they’re going to be interested in late-night service, in a way that our traditional park and ride suburban constituency is not.”

Of course, if it were easy to build a full-scale all-day transit system, more cities would have done it. The challenges generally break down into money and politics (what doesn’t?). On the economic side, there’s a reluctance to shift resources away from rush-hour because that’s where ridership, and thus revenue, is more certain. Off-peak service means new operating costs, in the form of drivers and maintenance, and perhaps even new capital expenses. Since most fleet maintenance is done on weekends and nights — in a word, off-peak — some systems will need more vehicles to expand service into those periods.

At the cultural end, the low-income riders who stand to benefit most from increased off-peak service often have the weakest political voice. Some politicians carry a vehicle bias: they will see empty midday buses and trains and blast off-peak expansion as wasteful, even as they endorse highway lanes full of single-occupancy cars. Others have a rush-hour mindset: they come to work at that time, so everyone else must, too. These counterarguments aren’t always off-base. Most people do drive most places, and the biggest commute shares do occur at the peaks [PDF].

“The peak tendency has been amazingly consistent,” says Steven Polzin of the University of South Florida, co-author of the Commuting in America, 2013 series on commute trends. “One of the intriguing things is there’s been a decline of the ‘peak of the peak’ commuting, but not a lot.”

What that means is that the early adopters of tomorrow’s all-day transit systems are likely to be big agencies in major cities. That’s not to say smaller areas lack the popular demand or the institutional desire to go off-peak. Just recently Jacksonville, North Carolina, population 70,000,expanded bus service to the shoulders of the peak so more commuters could get to and from work. It’s more to say that “somebody has to change the tradition,” as Deka puts it, “and the big agencies are in a better position, I think, to change the tradition.”

Top image: Nolan Levenson courtesy Rudin Center / NYU.

This article is part of ‘The Future of Transportation,’ an Atlantic Cities series made possible with support from The Rockefeller Foundation.