Home New York City Transit On the effect and cause of the MTA’s ridership decline

On the effect and cause of the MTA’s ridership decline

by Benjamin Kabak

A Monday meltdown reinforces why transit ridership is down across NYC. (Source: MTA)

Monday morning dawned in New York City yesterday as it so often does these days: with a total MTA meltdown. The problem this time came about because of a year-long project to fix a stretch of tunnel under 4th Ave. about which the MTA forgot to tell riders and somehow messed up the GO. Thanks to a typo, the D, N and R were all running local and wooden plywood formed a surprising barrier on the express tracks. As complaints on social media piled up and no one knew what was happening, riders flocked to any other possible route — ferries, Ubers, whatever they could find.

Later in the day, the MTA determined that a typo caused the meltdown, and Aaron Gordon did a deep dive into the mess that was. As Gordon discussed, because the work order identified the wrong signal as the end point of the construction zone, D trains were routed onto the local tracks, and the entire 4th Ave. line south of Atlantic Ave. was thrown into disarray. Furthermore, since Transit had scheduled this work as a long-term change, the internal communications staff didn’t realize it required special word to passengers in the form of station posters and announcements. Thus, the powers-that-be at Transit simply didn’t know they needed to make an effort to bring word of these service changes to commuters during rush hour on a Monday morning.

The agency struck an apologetic tone. Sarah Meyer, Transit’s new-ish Chief Customer Officer, issued a lengthy statement about the issue (which notes how the blue wall actually improves service and eliminates the need for flagging). “We deeply apologize for our significant errors today and know that we need to do better,” Meyer (a long-time classmate of mine in elementary and high school) said. “We are working through our policies and procedures to ensure this does not happen again.”

But happen again it does time and time again. As Dave Colon detailed at Curbed (while linking to this post of mine on this very topic from 2011), the MTA constantly fails at communicating basic information about subway delays. As Colon noted, Transit’s twitter account had no update on the issue until 9:17 a.m., at least 40 minutes to an hour after initial reports started trickling in. No one anywhere knew what was happening, and this lag in getting information to the public is a near-daily problem these days.

Now, Monday’s delay was a particularly bad one, but this a long-winding way of getting to another point: Monday’s issue isn’t that rare and is illustrative of declining reliability of subway service and a main driver behind the alarming multi-year dip in ridership I charted last week. Following the release of the raw numbers, during the MTA Board meeting last week, agency executives presented a deep dive on ridership trends. What they unveiled wasn’t surprising: Subway ridership is on target for another two-percent dip this year, and the biggest declines in weekend ridership are from the hours of 9 p.m. to 5 a.m., not coincidentally when service gets less frequent and more likely to be plagued by track work-related changes.

Other than from 5 a.m. to 7 a.m., weekday subway ridership is down across the board with the biggest dips in evening and late night travel. (Source: MTA)

It’s hard to find good news in the MTA’s report. The only time ridership has steadily increased has been in the hours between 5 a.m. – 7 a.m., and while the morning rush is still above 2014 levels, by the end of this year, ridership in the 22-hour period from 7 a.m. – 5 a.m. will be at its lowest in six years. Overall, the biggest declines are in the Bronx and Queens, with these boroughs seeing subway ridership fall by 6 percent and 5 percent respectively this year.

Not coincidentally, again, the MTA notes that growth in the usage of for-hire vehicles — the Ubers, Lyfts and Vias of the world — increased markedly in 2017. Growth in FHV usage was 13 percent last year as 63 million more riders used for-hire vehicles in 2017 than in 2016. That’s a larger increase than from 2012-2016 combined, and when combined with the 199 million bike and ferry rides (and the corresponding 6 million rider increase last year), this increase is a one-for-one match with the 69 million rider decline in subway service.

The increase in for-hire vehicle and taxi usage corresponds closely with the dip in subway ridership. (Source: MTA)

So is Uber responsible for the MTA’s ridership dip? That is what MTA Chairman Joe Lhota seems to want the city to believe. As he said last week in response to a question to Dan Rivoli of The Daily News, “Some days, I drink Coke, and some days, I drink Pepsi.” It’s hard to read this as anything other than an attempt to blame personal choice and forces outside of the MTA for the decline in ridership. To me, though, this is both a gross misreading of the situation and an inverting of the cause and effect. The cause of the MTA’s ridership declines isn’t an increase in Uber use; rather, the effect of the MTA’s declining service reliability is an increase in the use of for-hire vehicles.

Notably, the MTA has charted steep declines in outer-borough routes (that is, travel between the boroughs rather than to Manhattan), and that can be explained by long and arduous trips that either require multiple transfers or multiple modes. With VC money still subsidizing for-hire vehicle fares, those New Yorkers who need to take these outer borough trips and can afford a taxi ride do so, particularly when subway service (and communication about that subway service) is a crap shoot. As one Uber rep said to The Wall Street Journal, “The best way to boost subway ridership is to improve service.”

These ridership declines are worrying trends. The spiral will continue to send those with means out of the subway system while New Yorkers who can’t afford alternate travel are left with an unreliable system that costs them time and money in delays and lost wages. The MTA should recognize that the cause is not Uber but rather the service, and Monday’s communications and travel meltdown was just another sign of the depths of the transit crisis.

You may also like


JJJ July 31, 2018 - 1:53 am

I’ve seen a lot of talk about service (ie, time), but very little about the other side of the equation: price.

Aside from becoming more ubiquitous, Uber, Lyft and other services (Via etc) have continuously hit lower price points. Remember, when Uber started in NYC, it was at the black-car level only. Then X. Then Pool. Now Express Pool.

And in that same period, MTA fares have gone up.

MTA plans to continue to raise prices, but that’s not how the economy works! A business cannot charge based on what they want to make, they need to charge based on what people are willing to pay, and that’s influenced by the competition.

Aside from MTA service being less reliable at night, Uber fares are cheaper than during commuting hours. Additionally, if youre with friends, splitting an Uber will be cheaper than the sum total of MTA swipes.

Does MTA publish how many monthly pass customers they have? Anecdotally, I know a few people who ditch the monthly in summer because they’re more willing to walk, use bikeshare, or even take the ferry. Once the monthly is gone, every swipe counts.

J Adlai July 31, 2018 - 8:50 am

This is exactly what I was thinking.

MTA needs to engage in an aggressive cost-cutting exercise. If they can’t do this, then as Pool and Pool express expand, ridership will continue to crater. There’s no reason why they can’t get the fare down lower than $2. Instead they’re talking about raising fares.

SEAN July 31, 2018 - 9:26 am

How would you propose doing such a cost cutting program.

J Adlai August 1, 2018 - 11:55 am

Adopting best practices with train staffing levels would seem to be the low hanging fruit. Decreasing the amount of staff needed for maintenance tasks, requiring higher workforce productivity, would be the next. And streamlining management structure/office staff seems like it could produce substantial savings, especially with what they reported the BSC initiative accomplished.

Phillip Roncoroni August 2, 2018 - 1:26 pm

If you want to call a prolonged labor strike “low hanging fruit,” sure.

Your comment also ignores the hundreds of millions of recurring cost savings achieved over the years in management/back office functions.

J Adlai August 3, 2018 - 2:12 pm

Thanks to the Taylor law, there won’t be a prolonged labor strike.

And the very last line of my comment addressed the previous savings achieved in management/back office functions.

Alon Levy August 3, 2018 - 2:17 pm

Labor can’t strike and management can’t unilaterally change the contract to reduce staffing.

Phillip Roncoroni August 2, 2018 - 1:25 pm

Lyft, Uber, etc. are all currently being subsidized by massive amounts of VC money, and the fares are artificially low. Uber lost $4.5 billion last year.

Alon Levy July 31, 2018 - 11:01 am

Carpooling is a very small share of urban travel, though. In the US the average occupancy per car in commute trips is 1.2. Intercity trips are more likely to be done in a group (often a family), but urban trips for the most part are individual, and there, the Pool just means you’re meandering through side streets to pick up and drop off other people.

The subway ridership decline isn’t about Uber. It’s about bad service and trips not taken. Uber is just a convenient villain that’s easy to overregulate; it serves the same purpose here that (((big banks))) do as a villain anywhere else.

JJJ July 31, 2018 - 12:03 pm

Alon, you are citing commuting statistics, not general travel statistics. Ridership is down during times of day where most people are not commuting. Those trips are more likely to be done with additional family members or friends. Additionally, when people report their travel, I would think they list taxi/ride service as their method even when inside a pool service because it is paid. At least to me, carpool implies free ride with the coworkers.

Alon Levy August 1, 2018 - 12:42 pm

You’re right that this isn’t about commutes, but I’m just skeptical that there’s a lot of carpooling even for non-commute trips within urban areas. If it’s between cities, or even between separate parts of the same region (e.g. Waltham -> Boston), then sure, but that seems somehow unlikely in New York. If you live in Morningside Heights then you could socialize elsewhere but then it’s unlikely you’d be going with other people also from Morningside Heights.

John August 3, 2018 - 2:55 pm

Carpooling can also refer to the ride sharing offered by Via, Lyft and Uber (to name a few). I know (anecdotally) of many people using those ride-sharing apps to not use the subway for commuting and non-commuting purposes.

I now when I go out at night will often check the apps for pricing to my destination. If the subway isn’t close to where I am or want to be and the ride-sharing price isn’t more than $5-$7, I will often use it. Before I would just take the subway.

Bob91 August 2, 2018 - 3:29 pm

Also a factor in Uber/Lyft pricing is the number of people. If you’re with a few friends, the total price of an Uber/Lyft is the same regardless of having multiple people, while the price of the subway is obviously per person on the subway. Oh yea, and on a weekend night its usually exponentially faster and more convenient.

John A. Noble July 31, 2018 - 9:20 am

Thank you for explaining what happened with the D/N/R. Yesterday I found out that N trains are going to run local from 36 to 59 for a year and none of us had any idea why; no signs had been posted, no outreach to customers in Bay Ridge whatsoever. It’s only going to add about three minutes to my commute, but that’s not the point. MTA completely dropped the ball here and I’m glad they acknowledged their incompetence.

Larry Littlefield July 31, 2018 - 9:57 am

Transit history buffs have seen this movie before, and recognized when it started in the mid-1990s.


“It would seem, looking back, that the New York City Transit Authority didn’t have a chance in the 1970s. Only 4 days into 1970, the fare was raised from 20 cents to 30 cents, with riders to Rockaway paying a double fare. The fare increase was supposed to plug up large deficits in operations, and whatever surplus was available could go to infrastructure repairs. With this fare hike, ridership declined, and this vicious cycle of fare hike / reduced ridership would repeat itself several times in the 1970s. Yet the required maintenance was never done.”

“The rapid transit infrastructure of NYC in the 1970s was suffering from the effects of ‘deferred maintenance’ initiatives started in the 1960s. The fiscal crisis of 1975 didn’t help matters. Conditions were so deplorable that it was amazing that trains even ran. If they did run, they were dark, or completely covered by graffiti. Track conditions were horrible, too – there were hundreds of “red flag” zones where subway trains had to slow to 10mph or less. Although the R-44 and R-46 subway cars were introduced during this time, the introduction was extremely problematic.”

What no one says is what actually saved the city, which has run up huge debts and pension funding shortfalls due to retroactive pension increases — which have been repeated. From 1970 to 1980 the CPI doubled, meaning the real cost (and value) of those Lindsay/Rockefeller debts fell by half. It was only period that public sector compensation actually fell relative to private sector compensation here.

“Even labor / management relationships suffered, with strikes threatened every two years. In 1980, there was a 12-day subway strike, the first since 1966.”

They had rejected something like a 10 percent raise per year. But ever since the, they’ve gotten richer and richer — mostly in years of not working – while almost everyone else outside the executive/financial class got poorer and poorer.

Larry Littlefield July 31, 2018 - 10:23 am

“So is Uber responsible for the MTA’s ridership dip? That is what MTA Chairman Joe Lhota seems to want the city to believe. As he said last week in response to a question to Dan Rivoli of The Daily News, “Some days, I drink Coke, and some days, I drink Pepsi.”

Will someone please demand some answers from Lhota on the impact of city non-funding in the 1990s, when he as Giuliani’s budget director, on today’s transit riders, and why that’s fair?

Klaus K. July 31, 2018 - 11:48 am

Given the state of the MTA’s finances, it seems like a strange priority for City Council Speaker Corey Johnson and Mayor de Blasio to institute a Fair Fares program (as trumpeted in all the local media on early June) when the MTA badly needs additional resources, is underfunded as it is, and is hiking fares again. Plus, I don’t see what the positive impact is on overall MTA service or the overall impression that service makes to add to the overcrowding by adding more riders who are simultaneously paying less (as in it just adds to the MTA’s falling revenue problem). The feedback loops for the MTA all seem negative. Why can’t anyone go in there and reduce cost by reducing head count.

Benjamin Kabak July 31, 2018 - 11:49 am

I’ll defend Fair Fares here: It’s a mobility and social justice program, and it’s a city subsidy so that the MTA doesn’t lose fare revenue. It’s inline with how a rich city should be promoting low-cost mobility for low-income residents who can’t afford other means of travel.

Klaus K. July 31, 2018 - 11:57 am

While you are correct on the surface, DeBlasio and Cuomo are arguing about the city contributing to the MTA budget to fix the subways, so in my view, Fair Fares is indeed taking money away from that. That money is immediately unavailable to the MTA for anything else and allows the mayor to claim he is investing in the subway, so the feud and the resulting dysfunction continues. Fair Fares will not make the subways better.

Rudie July 31, 2018 - 8:01 pm

Would it make more sense for the city to ask the MTA to reintroduce off-peak fares? Off-peak fares should be better from a social justice perspective due to the rider mix and take off some of the political resistance to fare increases. I just don’t get how the city politicians think further subsidizing peak ridership is responsible given all the system problems that are not being fixed in short order.

JJJ July 31, 2018 - 11:22 pm

I agree with Rudie. Lower off peak fares address both competition and equity. Thats when the MTA is at its worse, and roads are at their best (least congestion). Obviously, thats where ridership is suffering. Additional, lower income employees are more likely to commute off peak.

SEAN August 1, 2018 - 10:24 am

That’s what Metro Transit does in the twin cities, plus everyone including seniors & the disabled pay the peak fare – no discounts unless you have a pass. It’s similar with DC’s Metro subway & their fare structure.

RBC July 31, 2018 - 9:12 pm

Its also worth noting that declines in ridership also have to do with increased work-from-home arrangements. More companies are opting out of expensive commercial leases & giving employees remote access.

Not Uber August 1, 2018 - 8:54 am

We all know he doesnt really believe Uber is the reason for the declines, its just an easy scapegoat. Especially when the declines are highest in queens and the BX. An uber from those areas during rush hour are like 10x the cost of a subway ride, and not necessarily faster.

Alon Levy August 1, 2018 - 12:45 pm

That by itself is not a good argument, because the growth in taxi + TNC rides in the city is only about 1/3 in Manhattan. Most TNC rides are in Manhattan, but this was also true of most taxi rides before Uber, and the growth occurs all over the city.

BrooklynBus August 1, 2018 - 10:17 am

History certainly does repeat itself. Now we have a situation where declining service results in fewer riders and an increase in ride share services while the MTA blames these services for the passenger declines.

Back in 1975 when we had the fiscal crisis, the MTA decided they can save the most money not by reducing service on the least traveled routes, as they do today, but be cutting local bus rush hour service in half from every two minutes to every four minutes on every bus route in the city where there were two minute headways. These included the B2, B35, B41, B44 and B46. All these lines were jammed before the cuts were made. People has no way to get to work. Within a day private citizens were offering group rides to the train fo a fee.. Within a week, the “gypsy cabs” got in on the act. Several years later, rather than realizing their gross mistake, the MTA decided to further cut bus service in response to the lower ridership claiming the lost riders were no longer part of the demand. That is when the dollar vans began, in 1978, not after the 1980 transit strike as most believe. That just made their numbers grow even faster.

The MTA then blamed these dollar vans for the decreased ridership as they are blaming ride share services today. To this day, the MTA refuses to look at latent demand and only considers existing riders when increasing service. Nothing has changed in this respect. They will not reassess their fare policy that only allows for one bus transfer requiring some in outer Queens especially to pay double fares when three or four buses are required or a bus train and bus are required to complete one trip. They still believe it will be the end of the world if they allow a round trip for a single fare, which can be accomplished anyway with a little ingenuity.

On a side note, last month I was in Ocean City Maryland, and never used my car to travel within the city. That is because the bus which operates 24 hours a day and usually comes every five minutes, costs $3 for unlimited daily riding and is only $1.50 for seniors. On the first day we made 8 trips for $1.50.

Alon Levy August 1, 2018 - 12:47 pm

What was the B2 like then? The other four routes are four of the top five in the borough, but the B2 has very little ridership nowadays. Did it run the route of today’s B1 or something?

BrooklynBus August 1, 2018 - 4:56 pm

It was the same route as today.

A number of reasons did that route in. First, there was the cut in rush hour service from 2 minutes to four minutes. Then, off-peak service was cut, making it utilized only during the rush hours with under a half dozen on the bus at all other times. When Kings Plaza was built in 1970, it had a resurgence on the weekends. Then around 1980, the B9 was extended to Kings Plaza which gave many riders an alternative to the B2. But it really went downhill in 1993 when the B31 was moved from Avenue U to Avenue R to serve the Kings Highway Station instead of Avenue U. Service was cut by 50 percent. In 2010, weekend service was eliminated but brought back about three years later. Some passengers probably never returned.

Also, during the past 15 years Kings Plaza dropped in popularity for the locals with many choosing Gateway or Shopping Centers outside Brooklyn. Probably because of numerous well publicized crime incidents. Most of its clientele now come from northern Brooklyn. The losing of the movies theaters there also did not help. When it first opened, it really destroyed Downtown Brooklyn as Brooklyn’s go to shopping area. Downtown Brooklyn with is doing much better these days.

Alon Levy August 1, 2018 - 7:19 pm

Where did the B9 run before? Right now I’m mulling moving it from Kings Plaza to Bergen Beach, replacing the B41 branch on Avenue N.

How relevant is the B100 here? It’s closely parallel to the B2 but runs limited and is basically the fastest route in Brooklyn (except for a few low-frequency runs like the B39), but it doesn’t serve Kings Plaza. Is Kings Plaza just a weak destination? Evidently it’s not a big job center, unlike the various hospitals (Kings County/SUNY Downstate is the biggest off-Downtown job center in Brooklyn).

BrooklynBus August 2, 2018 - 11:07 am

Jeremiah already answered many of your questions so I won’t repeat.

I would run the B100 on the same streets as the B2 and B31 and shift to Fillmore at Flatbush. Kings Plaza is still a major destination especially on weekends. One thing I wanted to do is run the B46 via the B2 route between midnight and 6 AM. There is no need to serve Kings Plaza then.

Jeremiah Clemente August 1, 2018 - 8:44 pm

The B9 terminated at Flatbush Avenue and Avenue L, like it currently does before service ends for the night and for the first few buses of the day. It was extended to Kings Plaza in the 1980s. The moving of the B9 to Bergen Beach is not new. It was first proposed by BrooklynBus in his thesis in the 1970s about the bus routes in Brooklyn. He again proposed it in the 2000s for fixing southern Brooklyn bus routes when he was in the Committee for Better Transit. http://brooklynbus.tripod.com/id10.html. In Proposal F on the list of Proposals, he suggested this change again, while reducing the B41 Bergen Beach branch to rush hours only (makes sense since I see nobody using that route on the weekends anyway). Also to be added was an extension on the B11 from Brooklyn College to Georgetown Shopping Center via Avenue J and Avenue K. Fast Forward to 2018, I would this lovely proposal to have three services on the branch:

1.) B9 along Avenue N to Veterans Avenue and East 71st Street, all times except late nights. 5 minutes during rush hours, 8-10 minutes all other times
2.) B41 local along Avenue N to Veterans Avenue and East 71st Street, rush hours and late nights only. 5 minutes during rush hours, 30 minutes at night.
3.) B41 LIMITED along Avenue N to Veterans Avenue and East 71st Street, stopping only at Utica Avenue and Ralph Avenue, all times except late nights. 5 minutes during rush hours, 8-10 minutes all other times.

This would further improve service on Avenue N and provide for better access. Also, the B11 extension would be modified to operate on Avenue K in both directions, using East 59th Street, Avenue L and Ralph Avenue to turn around. Buses would then terminate 400 feet north of the current northbound B47 stop at Ralph Avenue and Avenue L.

What do you think?

Alon Levy August 1, 2018 - 10:47 pm

Just to put things in perspective, I’m trying to make it so that every single Brooklyn bus can come at worst every 6 minutes from 6 am to 10 pm (accomplished via speeding up the buses and eliminating duplicative routes). So a B9 along Avenue N, then Avenue M, then back to N, then 60th, gets 6-minute frequencies all day, and then some super-strong routes keep their current frequency, like the B41 on Flatbush (no Bergen Beach branch) at 2.5 minutes peak, 5 off-peak.

Doing what you’re proposing for the B11 is to some extent already in my current draft. What I’m debating is whether Borough Park has enough east-west demand to merit an interpolating route on 53rd or 55th; the area has high population density (even excluding children) and the B11 and B9 both have decent if not amazing ridership. But whatever happens in Borough Park, something should definitely be running on Avenue J and continuing east. The question in the east is then whether you want this route to take over the B103 to Canarsie rather than terminate at Georgetown…

BrooklynBus August 2, 2018 - 11:01 am

I am all for straightening routes, but sometimes routes turn for a reason. There is a reason why the B9 shifts to Avenue L at Ocean Avenue, which isn’t obvious from Google Maps. The portion of Avenue M between Ocean Avenue and Flatbush is actually much older than the portion where the bus operates and is much narrower. Two buses might not even be able to pass each other. The part between Coney Island Ave and Ocean was not built until the 1930s. The narrow part is probably 50 years older.

Are you planning your changes for Byford?

Alon Levy August 2, 2018 - 9:32 pm

I’m planning them independently, but I’m in contact with the MTA (not Byford himself, but intermediate planners and such).

And yeah, Avenue M is pretty narrow. Only reason to try stringing buses there is to serve Mount Sinai Brooklyn better.

Speaking of hospitals: is it at all feasible to reroute the B17 away from Utica and have it go west on Empire to serve Kings County Hospital? The argument against is that it wrecks the subway transfer to Utica (and the alternatives, i.e. Empire if you go west and Kingston if you go north, aren’t as good). The argument for is that Remsen is very close to other routes that do hit a bunch of subway stations on the 3, 4, and L, and the reroute would serve the hospital from the east.

BrooklynBus August 3, 2018 - 10:19 pm

Mt. Sinai is not a major hospital and the shift to Avenue L maintains an equal distance between the 6/11 and the 7/82.

Moving the B16 to Fort Hamilton Parkway to serve Mamonides which is a major hospital is more important.

I am not sure what you are proposing for the B17. That is one of the few routes in southern Brooklyn I would not change.

Look at my proposals to better serve Kings County Hospital. http://Www.BrooklynBus.tripod.com. I prefer primarily a grid system with some L shaped routes and would greatly avoid U shaped routes. There ahould be as few turns as possible while still maintains access to major institutions.

anon_coward August 3, 2018 - 9:41 am

people are working remotely from home. Some people i work with are in the office maybe 2 days a week and mostly only for face to face meetings. I’ve worked with people who’ve been remote for years sometimes hundreds of miles away.

Some jobs you can’t work remote, but more and more it makes financial sense since rent is so expensive in NYC. Especially support roles like IT.


Leave a Comment