The Washington, DC region is great >> and it can be greater.


Metro by the numbers, part 1

Last week, Ryan McNeely at Matt Yglesias's blog wondered how bad the DC Metro really is.

He introduced a few metrics, though mainly compared Metro to the transit systems in New York, Chicago, and Boston, which are systems from a different generation. Among the metrics were weekday ridership per mile, fares, span of service, and passenger fatalities since 1990.

McNeely concludes that the Washington Metro is the worst of the big 4. But he does make it clear that he doesn't think it "sucks."

However, there are far better metrics to use. Let's stop for a second. Think about the transit systems you've ridden. Pick your favorite. Now, reflect on why it's your favorite. I'm willing to bet that for most of you, the number of passengers per day per route mile is not one of the top factors. Most people like transit systems that get them where they need to go quickly and cheaply. They like clean, safe, and easy to navigate systems.

There are 13 heavy rail operators in the United States. I see no reason to restrict my analysis to the biggest four. Especially since the definition of "biggest four" can change depending on what you measure.

Since data speaks louder than words, let's take a look at what the National Transit Database has to say about the issue. Unfortunately, the NTD does not always give us the best data to evaluate the rider experience. So we'll have to use proxy measures to some degree. For instance, the NTD doesn't really show how much of a region is accessible by transit. But we can measure system mileage and number of stations to get an idea.

The chart below shows the ranking of the different systems based on different categories. All numbers are from 2007, the most recent year for which data have been released. The farther left on the chart a system appears, the better it is performing in that specific metric.

Heavy rail systems ranked by metric. Washington Metro is highlighted in red.

One note on the systems: Unless otherwise mentioned, the data refer only to the heavy rail lines in the system. For example, that means that Boston's Green Line (light rail) and other non-HRT modes are excluded from the analysis.

First, let's consider ridership characteristics. These measures are probably the best determinant of a successful system. And they probably also indicate whether the system is performing in a somewhat satisfactory manner.

Annual Unlinked Trips (HRT): Unlinked Trips is the count of the number of times a patron boards a vehicle. In this case, it's the annual sum of all passenger boardings on the heavy rail lines of each transit agency. As an example, someone traveling from Vienna to Dupont Circle would be making 2 unlinked trips: 1 on the Orange Line and 1 on the Red Line.

Washington is in second place. Only New York sees more annual unlinked trips on heavy rail. As Ryan noted, Chicago and Boston round out the top 4.

Annual Passenger Miles (HRT): This is the cumulative total of the miles each passenger travels. In this case, the number is only for the heavy rail lines for each agency. As an example, if 100 people ride 10 miles on a subway line, that's 1,000 passenger miles.

Using this measure helps to normalize for the amount of transit consumed. A trip from Metro Center to Gallery Place is not the same as a trip from Shady Grove to Metro Center. But in the previous category, unlinked trips, both trips count the same.

Again, Washington comes in second place behind New York. But BART and Atlanta move up into the top 5, while Boston drops to 6th place.

Unlinked Trips per Directional Route Mile: This is not directly measured in the reports available from the National Transit Database, but it is easily calculated using two of the other statistics. This is most similar to what Ryan analyzed in terms of ridership initially. It takes the number of unlinked trips and divides by the number of miles of revenue track (in each direction). It's not quite the same measure, but it's close. Note that the number of unlinked trips is per year, not per day.

In this category, New York still leads the pack. But New York is going to win in a lot of transit ridership categories no matter what they do. PATH, also in the New York area comes in second, with Boston, Washington, and Los Angeles rounding out the top 5.

Fares: Another one of Ryan's complaints is about how expensive it is to ride Metro. He also complains about having to swipe a farecard twice (once to enter, once to exit). So, how does Metro compare?

Metro uses a fare system graduated based on distance traveled. Of the other heavy rail systems in the US, only 2—BART and PATCO—use graduated fares. The rest all employ a flat fare. On the other hand, commuter rail systems use graduated fares. Since Metro (and BART and PATCO) is something of a hybrid between urban subway and commuter rail, this makes sense.

It's also positive from an equity perspective. Why should someone traveling 3 blocks (from Gallery Place to Metro Center, perhaps) pay the same as someone traveling from Franconia-Springfield to Shady Grove? Were Metro to change over to a flat fare, some of the longer trips would probably get cheaper, but then the shorter trips would get more expensive.

Let's consider fares on the low end. While there is a range of prices based on how far and when you travel (and starting August 1, the fare media with which you pay), short trips can cost as little as $1.60. Only 2 systems have a cheaper low-end fare—PATCO and Los Angeles. Baltimore's flat fare is the same as Washington's base fare. Tied for 3rd out of 13 sounds pretty good, no?

But we also need to consider the maximum fare. For the purposes of this exercise, I went ahead and considered the fare changes which will go into effect August 1. Starting on Monday, August 2, the highest price you could pay on Metro will be $5.45. That's if you travel all the way across the region, entering during the peak-of-the-peak period, and pay with a paper farecard.

That is pretty steep. It's almost 2.5 times higher than New York's base fare, for example. But the DC Metro is not the most expensive in the country. No, that honor goes to BART, where a trip from San Francisco Airport to Pittsburg/Bay Point will run you $10.90. That's twice the maximum Metro fare.

Next: Span of service, safety, size and efficiency.

Matt Johnson has lived in the Washington area since 2007. He has a Master's in Planning from the University of Maryland and a BS in Public Policy from Georgia Tech. He lives in Greenbelt. Hes a member of the American Institute of Certified Planners. He is a contract employee of the Montgomery County Department of Transportation. His views are his own and do not represent those of his employer. 


Add a comment »

I don't think you should even attempt to compare Metro with NYC subway system simply because what metro really is is a commuter railway/ Metro aka RER in Paris or Crossrail in London or S-bahn in Germany.

If you really want to assess Metro you really need to think about what its purpose is and how it was designed. I clearly see a system that goes 14 miles and more away from downtown and which relies a lot on commuter parkings such as those found in Shady Grove and Franconia.

To me that is not a metro system. Therefore it can't be compared nor to Chicago nor to NYC.

by Vincent Flament on Jul 28, 2010 3:16 pm • linkreport

@Vincent Flament:
I have argued that in posts right here on GGW before. However, most people do compare the Metro to New York. This post was written specifically to address the issue of Metro being compared to New York, Boston, and Chicago.

See also:

by Matt Johnson on Jul 28, 2010 3:18 pm • linkreport


Definitely. Whenever people demand that Metrorail revert to a flat fare I want to bash my head against the wall. The longest trip on the NYC Subway is half the distance of the longest Metrorail trip. The highest fare of $5 on Metro to go from Rockville to downtown D.C. is a bargain; a comparable-length trip in NY from White Plains to Midtown Manhattan costs $11 each way on the MTA Metro North commuter rail.

by Adam L on Jul 28, 2010 3:26 pm • linkreport

Not to nitpick, but the 2008 NTD data has been available since last October.

I was also under the impression that Metro counted passenger trips based on faregate entries. Even if based on entries and exits, they can't know if, for example, I took 2 (Orange-Green) or 3 (Orange-Red-Green) trains to get from Clarendon to Columbia Heights. On most rail systems you wouldn't be able to arrive at a figure for unlinked passenger trips without sampling. In general, FTA prefers actual data to sampled data.

by Esmeralda on Jul 28, 2010 3:35 pm • linkreport

The NYC Subway is a disaster. It is always late, crowded, filthy, broken, and none of the card readers work. However the DC Metro is so much more comfortable, clean, spread out, safe, cheap, accessible, and up to date (for the most part).

by Natalie on Jul 28, 2010 3:40 pm • linkreport

Most other regional transit systems have a much better mix of subway for in-city trips and more limited-stop commuter rail for longer trip.

Look at Boston's subway system. Their MSA is roughly the same size as ours, but their fixed-price subway system carries far fewer passengers each day. Yet their system overall is far more comprehensive, and covers many more suburbs, because it is fed at many places by commuter rail.

So comparing fares is unfair - since our system serves both functions. Many people pay a lot more than 5 bucks to get to
work in Boston because they start on commuter rail.

Likewise, the core of Boston's T system is less used every day than Metro even as it serves more passengers and a much wider area. This is because it is complemented by a commuter rail system that brings people directly into the city at a smaller number of stations.

Instead of building things like the silver line, we should be building limited stop light rail that will distribute ridership to more stations and create more points of entry to the system. We're facing a capacity problem even as Metro isn't nearly as useful as other regional transit systems, because we have focused only on one side of the transit problem.

by Jamie on Jul 28, 2010 3:50 pm • linkreport

A couple quick comments:

1.) The minimum fare in NYC is tough to judge considering you can get an unlimited pass. I pay about a dollar a ride based on usage.

2.) The A train runs just over 31 miles from 207th st in Manhattan to Far Rockaway. The longest line on Metro is the red, at 31.9 miles. That's not significantly longer.

by Matt on Jul 28, 2010 4:09 pm • linkreport

@ Matt:
thanks for putting that data together.

by Paul C on Jul 28, 2010 4:10 pm • linkreport

My biggest problem with metro is the inefficiency of our transfer points. For example, if I am at Farragut West, why should I have to continue on to metro center to transfer to red? They can't build a tunnel under Farragut Square to allow me to walk two blocks to the red line? The same applies between Metro Center and Chinatown.

by Kelsey on Jul 28, 2010 4:12 pm • linkreport

The two Farragut stations were originally supposed to be one station, just like Metro Center. However, the National Park Service would not allow them to build a station under Farragut Square because it would require removal of all the trees (and then replacement of them after construction).

A tunnel has been planned for some time, however no funding has been identified.

by Matt Johnson on Jul 28, 2010 4:14 pm • linkreport

@ Matt Johnson,

Indeed if I put your article in scope and context then it makes sense to compare Metro with other pre-war systems that you mentioned in the link you provided.

@ Adam L

Indeed again though a zonal fare structure could do the trick. Also no reason why a monthly pass couldn't be introduced. I travel from Woodley park to Rockville every day and it is indeed very expensive to travel. As a transport economist I know that the marginal cost of carrying me is near to 0 as I'm only using capacity that is anyway needed in the other direction.

@ Jamie

You are forgetting MARC and VRE. Granted the system is not as extensive as the one in Boston but nevertheless is there (at least in the morning and in the afternoon)

by Vincent Flament on Jul 28, 2010 4:14 pm • linkreport

The "transfer" problems are pretty inconsequential. For Farragut West/North, adds two stops to your trip to transfer at Metro Center.

For chinatown/gallery place, those stations are a half-mile apart. I don't really see all that many people making a ten minute walk through a tunnel to avoid the extra five stops and just transfer at Gallery Place.

Your comment is hardly the first time someone's mentioned these things, but it's more a "wouldn't that be nice" thing than a problem. If your destination happens to be the "other" stop from the line you come in on, then the only benefit would be to walking underground instead of aboveground if there were a tunnel. If not, it's just a few extra stops to change somewhere else -- and certainly less time than the walk would be.

by Jamie on Jul 28, 2010 4:17 pm • linkreport

Great post.

I'd like to see some research that polls transit users in each of these 13 cities. I get the feeling that no matter where you live, there is some serious 'grass is always greener' bias occurring on this topic.

by Rob Pitingolo on Jul 28, 2010 4:20 pm • linkreport

It is 1000 feet from the east end of the Red platform at Metro Center to the western end of the Red platform at Gallery Place. That's 1.6 times the length of a Metro platform. It's not really that long of a walk.

Of course, it's 300 feet from the end of the platform to the crossvault at Metro Center and almost 600 feet to the crossvault at Gallery Place. So that pushes the numbers up to about 3.1 times the length of a Metro platform.

And that has also been planned for a while. But again, no monies are available.

by Matt Johnson on Jul 28, 2010 4:22 pm • linkreport

@Vincent - I couldn't say I know what ridership is for MARC and VRE but I am sure it's nowhere near Boston's commuter rail ridership. And while these (two) lines do exist, there are only two places that they take you into the city (L'enfant and Union) and the service is far more limited and offers far less coverage.

They don't even run on weekends and even during rush hour the trains are infrequent. You can't really just use it as part of the system, it's more like Amtrak.

by Jamie on Jul 28, 2010 4:23 pm • linkreport


Sorry, should have clarified; I meant the length of a standard commuter trip, not the actual miles of track on any particular line. I also was looking at distance "as the crow flies", which makes the distance shorter that the route actually traveled.

by Adam L on Jul 28, 2010 4:24 pm • linkreport

Again I'm not saying that connecting these stations would be a bad idea. I just don't think it's making anyone's life miserable (for the reasons I state before) and I suspect the cost of digging the tunnel is enormous. We have bigger fish to fry...

by Jamie on Jul 28, 2010 4:25 pm • linkreport

Oh, in case anyone is interested.. VRE ridership

"VRE currently carries 16,000 round-trip passengers per day."

That's the same number that board metro at Rosslyn. Almost inconsequential.

MARC is 32,000.

Metro is something like what, 950K a day?

We have a lot of room for expansion in commuter rail, is there a reason people don't talk about this much?

I would think a VRE-like line that parallels the orange line (with more limited stops) and has it's own way to get to Metro Center would be a hundred times more useful than the silver line.

by Jamie on Jul 28, 2010 4:31 pm • linkreport

I agree with you on the Gallery Place/Metro Center walk. Every time I go to Verizon, I get off and on at Metro Center and just walk the, what, 3-4 blocks?

I'll walk from one Farragut to the other too, but I do wish I didn't have to pay to get back on (please WMATA, give us the "invisible tunnel")

by kidincredible on Jul 28, 2010 4:35 pm • linkreport

@kidincredible, but if you're going to just get back on at Farragut West/north, why wouldn't you just change at Metro Center instead? It would add 3 stops at most, which can't possibly take more time than exiting, walking, and re-entering.

by Jamie on Jul 28, 2010 4:39 pm • linkreport

@ Adam L

Well, these are numbers we can "get at," if only indirectly.

If we look at all of the people in NYC who commute to work by way of Public Transportation (these are people who live in the city. Yes they might be riding the bus), we find that the mean time of commute is 48.6 minutes. The NYC average subway speed (including stops etc) is 17.4 mph. This translates to roughly 14 miles/trip.

If we do the same for the people of DC, we find that it's 37.6 minutes and 33 mph including stops. That translates to approximately 20 miles/trip.

This doesn't include proximity to the station, as the commute time is door to door. People in NYC tend to be closer to the stations, but there aren't nearly as good of numbers on that. That said, the commute times are 20% shorter, even if the distance traveled is also 20% further. I'm still not sure that makes the case for or against a flat fare. I think there are other factors at work. It's also questionable whether NYC should have a flat fare, but the city has definitely been developed around the idea of one. Flushing would be in bad shape if the subway suddenly cost twice as much to ride from there.

by Matt on Jul 28, 2010 4:45 pm • linkreport

Fantastic Post. I rode the PATH for 4 years. Then lived in Manhattan for 5 years and rode the NYC Subway every day, and now ride the DC METRO every day.

From a customer comfort factor, I'd take METRO then PATH then NYC. The Metro has great A/C, including many stations! PATH is pretty good on this meansure too. Good seats & clean cars. NYC uses the hard seats, easier to clean soda, body odor, & human waste.

From a frequency / reliability viewpoint it's hard to beat the NY Subway, those trains are coming every few minutes all day long, and they seem faster. Oh plus they have express trains which are awesome.

Of the three, I prefer the METRO

by Mike Donnelly on Jul 28, 2010 5:10 pm • linkreport

@ Jaime:

You're right. I've often thought that there's incredible potential to expand MARC and VRE, without incurring the infrastructure costs that go with extending Metro lines. MARC's Penn Line (the only one of the five MARC/VRE lines that doesn't compete with freight rail for track access), in particular, could run more often on weekdays and run on weekends.

Perhaps as a result of Metro's hybrid subway/commuter rail legacy, I don't think this option occurs to people in the DC area as readily as it occurs to commuters elsewhere.

by Jimmy on Jul 28, 2010 5:16 pm • linkreport

The few times I've done it, it's been immediately at the start of service (e.g. the very first orange line train through Arlington). I looked at WMATA's trip planner and it seemed like walking would get me to Woodley Park quicker than going through Metro Center (and getting to Woodley by 6:00am was important).

95% of the time, yes, not a big deal for me to go through Metro Center.

by kidincredible on Jul 28, 2010 5:18 pm • linkreport

I've said this before: the price to get out to Dulles on the Silver Line is going to be very similar to what BART charges. $10 is not out of the question. Washington Flyer Bus is $9 just from Falls Church.

Once more reason NOT to metro to dulles.

by charlie on Jul 28, 2010 5:21 pm • linkreport

Feel free to not take the train to Dulles then.

I would point out that it is cheaper to take the train from Millbrae to Pittsburg/Bay Point than from SFO, and Millbrae is further. SFO is more expensive because there's a surcharge to pay for the extension.

by Matt Johnson on Jul 28, 2010 5:23 pm • linkreport

makes sense to charge airport passenger more: Their willingness to pay for a metro service to the Airport is way higher than an average commuter given the alternative on offer (Taxi or Bus)

It sounds unfair but it would be a good way to raise money: Flat fare to/from the airport from anywhere on the system - $12

by Vincent Flament on Jul 28, 2010 5:34 pm • linkreport


Even from Route 28 or Route 606?

by Mike on Jul 28, 2010 5:49 pm • linkreport

I just got back from a trip to Europe, and comparing the metro to systems there is somewhat difficult; as many people have pointed out, the metro is something of a hybrid between a RER/S-Bahn type system and an actual subway. But if you ask me, it does its job pretty damn well. I was in Paris and Vienna, both of which have regional rail systems (RER and S-Bahn) and actual subways (Le Metro, and the U-Bahn). The systems over there are great; I wish we had something more like them over here (Did I mention Vienna has a FABULOUS tram system? And they don't seem to care about spoiling the view with overhead wires...the opera house supports wires suspending the catenary for the tram). BUT, a subway system such as Paris's, or even Vienna's, with stations spaced approximately as far apart as Gallery Place is from Metro Center everywhere would never work in DC. The cost of constructing so many stations wouldn't come close to being justified, given the ridership they'd experience. Paris is nearly 6 times denser than DC. Additionally, so many stations would slow the trains down a LOT (the average speed of a Metro train in Paris is 12 mph...very, very slow) and mean that the subway would never reach as far as Falls Church, or Largo. DC doesn't need tightly spaced stations. But it does need a bunch of service to the suburbs, given that many people who work in DC also like to actually vote for Senators and Representatives. You could argue that we could have two separate systems: One, a subway, and one, an S-Bahn. But the stations on the subway in DC are already going to be reasonably spaced out, so why not simply extend those lines to the suburbs? It is much less annoying to (when the silver line actually gets built) take a one-seat ride from Dulles to Metro Center that takes close to an hour than it is to take a 25-minute commuter rail ride (where the headways are half an hour) to Wien-Mitte, transfer to the U-Bahn, and take a 10-minute ride to your hotel, while carrying all your stuff. Plus, we still have MARC and VRE, for people who live really far out. (Another note: Paris and Vienna do not really have Central Business Districts, so their systems aren't based on a core. But DC does have a downtown, so a hub/spoke system makes sense).
Comparing our system to others in the US, many of the same arguments apply. Boston (once you adjust for all the water that counts as part of the city) is about 2 and a half times as dense as DC, has a subway more like Vienna's (though not as ridiculously cramped together as Paris's) and an extensive commuter rail system. The only difference is that they are based around a downtown. New York is about 3 times as dense as DC, and, of course, much bigger. Same deal. Subway with closely spaced stops in the city (plus express tracks...cause it's New York) and then LIRR, Metro North, and NJ Transit.
Chicago is only slightly more dense than DC, but, like New York and Boston, has a much older rapid transit system. This means more frequent stops, though some of the lines do run quite far into the suburbs. And, again, it has a large commuter rail presence.
Each city has a system that suits its own needs. I would argue that the Metro fits DC better than the L or the T fit Chicago or Boston.
As far as basic things go (cleanliness, simplicity, reliability) DC beats pretty much every subway I've ever ridden. Compare transferring at any of the downtown stations in DC to pretty much any transfer in any subway system, and you will find that DC is much more intuitive to navigate. No labyrinthine passageways with more stairs than the Empire State Building. And you don't have to walk half a mile to get to the other line (Ahem, Paris). We're cleaner (duh) and, as consequence of having been built recently, more reliable, as well as faster. Everybody complains about how much the escalators break down, but the only other alternative is stairs. And when escalators break down, they're basically stairs. If people stand and block the escalator, you're moving at about the same speed as you would be if you were walking up/down stairs. If people stand to the right, you move twice as fast. What, exactly, is the problem?
DC's transport system may not be world class, but it gets the job done. And that's what it's meant to do.

by Nick on Jul 28, 2010 10:50 pm • linkreport

@Jamie: VRE is grand, but it's limited in how much capacity it can add by freight rail schedules. CSX owns the tracks, and their traffic gets priority because it pays the bills.

What we could use in DC is something along the lines of Paris's RER (regional rail), which could serve as something in between Metro and VRE. But no one has been talking about building more rail besides expanding Metro. It's a shame.

by Omar on Jul 28, 2010 11:13 pm • linkreport

@Natalie, RE: NYC subway issues.

I have to say, really? Metro, which in some ways is a theme-park version of a rail system, is very comfy. It is also unreliable due to a number of bad decisions when it was first being designed, relies on technology for safety that has not been maintained properly, incredibly unresponsive to consumer comments/criticisms, and not really a comprehensive service (on weekdays) outside of a standard commute due to long waits, an early close, and a late open.

NYC, on the other hand, is basically utilitarian. It is noisy, shaky, and significantly less comfy. It also works significantly better, and (someone back me up here with data?) safer as well. I think it is interesting that a system that operates 24/7/365 runs with fewer major disruptions than ours does. Every time I enter Shady Grove and there is no train sitting there, I wonder if anything is going on that is slowing down the system. In NYC the unexpected is much less of a concern than disruption caused by, say, weekend track work.

That having been said, what really disappoints me since moving here from NJ is the state of commuter rail. I would say that NJ Transit is a great model for how to run such a system. It is too bad the region is so focussed on Metro - I think that the red line should not have any stations further out than, say, Rockville and Silver Spring (if that!).

by egutin on Jul 28, 2010 11:29 pm • linkreport

@Mattjohnson; the problem is once the Silver Line is built the Washington Flyer bus and the 5A bus service will be cut or service greatly reduced. So we either have to suffer the Silver Line ride, or take a cab.

by charlie on Jul 29, 2010 8:29 am • linkreport

@Vincent: Given the cost of the bus service that it will replace with limited time savings, $12 is way too steep for a one-way fare to Dulles. With the myriad of surcharges, even BART to SFO isn't that high right now and when the surcharges got hiked SFO started employee shuttles to Millbrae to save employees money.

$12 for a one-way ride under the circumstances at hand would be totally counterproductive (in many ways) and would make the Silver Line a white elephant.

by Jason on Jul 29, 2010 8:53 am • linkreport

There are 13 heavy rail operators in the United States.

I am confused. What about MARC, NJ Transit, LIRR, Caltrain, etc., etc.? What is your definition of "heavy rail operator"? I think there are a lot more than 13.

by David desJardins on Jul 29, 2010 10:42 am • linkreport

@David desJardins:
Heavy rail is defined differently in the United States than it is in Britain. In Britain, MARC would be considered heavy rail. However, in the United States, it's called "commuter rail."

Basically, in the US "heavy rail" can be translated as "metro", "subway", "elevated", or "rapid rail/transit".

According to the FTA, heavy rail is defined as:

A transit mode that is an electric railway with the capacity for a heavy volume of traffic. It is characterized by:
  • High speed and rapid acceleration passenger rail cars operating singly or in multi-car trains on fixed rails
  • Separate rights-of-way (ROW) from which all other vehicular and foot traffic are excluded
  • Sophisticated signaling, and
  • High platform loading.
Additionally, consider the FTA definition for commuter rail:
A transit mode that is an electric or diesel propelled railway for urban passenger train service consisting of local short distance travel operating between a central city and adjacent suburbs. Service must be operated on a regular basis by or under contract with a transit operator for the purpose of transporting passengers within urbanized areas (UZAs), or between urbanized areas and outlying areas.

Such rail service, using either locomotive hauled or self-propelled railroad passenger cars, is generally characterized by:

  • Multi-trip tickets
  • Specific station to station fares
  • Railroad employment practices, and
  • Usually only one or two stations in the central business district.
It does not include:
  • Heavy rail (HR) rapid transit, or
  • Light rail (LR) / streetcar transit service.

by Matt Johnson on Jul 29, 2010 10:50 am • linkreport

@David: "Heavy Rail" is a term usually used for subways and similar systems that are powered by third rails and feature (usually) level boarding. The 13 operators are:

MTA Maryland (Metro Subway)
SEPTA (Broad Street & Market-Frankford lines)
MTA New York City
MBTA (Red, Orange, and Blue lines)
Miami-Dade Transit
LA County MTA (Red & Purple lines)
Cleveland RTA (Red Line)

Some debate can be given towards RTA as the Red Line is run similar to their light rail lines than any of the other systems that operate "heavy" rail. Also, some "heavy" rail cars (CTA) are lighter than some "light" rail cars (the Muni LRV's).

by Jason on Jul 29, 2010 10:53 am • linkreport

Maybe we should be building an actual subway in DC and return Metrorail to a commuter rail.

DC has a bus system and a halfassed subway/commuterrail we need a real subway in DC that is mainly for the residents to get around DC and not just commuters getting in and out of the city or we should be turning about half of the bus routes in DC to streetcars & lightrail.

Why wasnt Marc & Vre added to Metrorail in 86 and 92 it would have made a lot of sense to have all transit under one organization.

by kk on Jul 29, 2010 10:53 am • linkreport

According to the FTA, heavy rail is defined as:

A transit mode that is an electric railway with the capacity for a heavy volume of traffic. It is characterized by:
High speed and rapid acceleration passenger rail cars operating singly or in multi-car trains on fixed rails
Separate rights-of-way (ROW) from which all other vehicular and foot traffic are excluded
Sophisticated signaling, and
High platform loading.

You are making a pretty arbitrary distinction, right? NJ Transit fits this definition, too. The only difference is that someone has decided to call it "commuter rail" rather than "transit".

Some systems operate only in dense urban areas. Others extend into "suburbs". Some are in between. I understand the desire to categorize everything, but when you draw an arbitrary line at one point between "transit" systems and "commuter" systems, you discover that some systems like BART and Metro are on the boundary.

Metro serves some purposes that in other parts of the country are served by "commuter rail" systems rather than "transit" systems. If you're going to compare it to its peers, you have to consider those systems too.

by David desJardins on Jul 29, 2010 11:07 am • linkreport

@David desJardins:
I completely understand what you are saying. But before you accuse me of drawing arbitrary lines, read this post:

Now, if you would like to take issue with "arbitrary" lines, I would encourage you to submit comments to:
Peter Rogoff,
Federal Transit Administration
United States Department of Transportation
East Building
1200 New Jersey Avenue SE
Washington, DC 20590

Until such time as the FTA changes their definition of 'heavy rail', when I conduct an analysis of 'heavy rail' I will use their qualifications, despite my feelings that transit is not easily categorized.

But consider this:
New Jersey Transit commuter rail lines operate as far south as Trenton (from New York-Penn). They operate at speeds of over 100 mph. Stop spacing is much further apart than as it is for Metro. And for the most part, there is no urban circulation. (Manhattan's only NJ transit station is Penn Station).

Now, if the Metro ran from Wilmington to Richmond and only had one stop in the District of Columbia, I'd probably compare it to NJ Transit commuter rail more eagerly.

However, my time is not infinite. And not all commuter rail lines tend toward heavy rail. For instance, including MARC in the comparison would really only be appropriate with the Penn Line. But NTD statistics don't break down by line, only by mode. And that's "mode" as defined by the FTA.

Should I include Nashville's fledgling commuter rail line with New York's 6 Train?

I don't think so. But if you would like to make that comparison, I'm sure David Alpert would be happy to run your post.

by Matt Johnson on Jul 29, 2010 11:20 am • linkreport

@David desJardins:

The FTA makes the distinction. You have to draw an arbitrary line somewhere if you want to separate things. I would call NJ Transit commuter rail rather than heavy rail because it is centered on New York City but only has a couple stations right in the big dense centers that it serves.

Metro is like a combination of commuter rail and heavy rail in that it serves far-reaching areas but also has many stations clustered in the urban core. This makes it so that you can use the rail system to move around the urban core, something you usually can't do with commuter rail.

If you want to argue about a mode that's unclear, take a look at light rail.

Also, Matt forgot one heavy rail system: Staten Island Railway. Though I guess he's counting "operators" rather than "systems" but FTA counts it separately from the NYC Subway.

by MLD on Jul 29, 2010 11:22 am • linkreport

Fine. I don't mean to criticize, but it seems somewhat misleading to rank "heavy rail" systems by factors like "annual passenger miles". Because this just means the "commuter-like" systems will rank high, and the "transit-like" systems will rank low. It doesn't really have anything to do with service quality.

by David desJardins on Jul 29, 2010 11:31 am • linkreport

Matt, keep it coming man. I love your data spreads. I am also amazed that many people will bring in emotional arguments why the numbers do not make sense.

Folks, Matt is not manipulating numbers here like most lobbyists and politicians do. These are fairly straight-forward numbers compared. Whether you like it or not, the conclusions are easy to get.

by Jasper on Jul 29, 2010 11:42 am • linkreport

@David desJardins:
I apologize if I came down a little harshly. Please note that there are additional metrics coming (as evidenced by the title 'part 1' and the first graphic).

And also, note this paragraph, the 4th of the article.

However, there are far better metrics to use. Let's stop for a second. Think about the transit systems you've ridden. Pick your favorite. Now, reflect on why it's your favorite. I'm willing to bet that for most of you, the number of passengers per day per route mile is not one of the top factors. Most people like transit systems that get them where they need to go quickly and cheaply. They like clean, safe, and easy to navigate systems.
[Emphasis added]

You make some valid points. That's why this post includes 5 graphics. Ryan's (the post to which I was responding) only had 1.

But I'm not convinced that your conclusions hold up. If the Hybrid systems are so different from the traditional systems, then all of the modern systems would rank at the top in "annual passenger miles".

But they don't. New York (traditional) comes in first, and blows everybody out of the water. In fact, New York has more annual passenger miles than every other system added up. In fact, of the top 6, 3 each are traditional versus hybrid (although Boston's system is trending toward hybrid).

The reason is because while the hybrid systems do go further, fewer people make the long trip.

So if Modernville has a subway that is 20 miles long has has 100 passengers each day, it would have 200 daily passenger miles.

And if Traditional City has a subway that is 5 miles long but carries 400 passengers daily, it would also have 200 daily passenger miles.

But, I acknowledge that the statistic isn't perfect. That's why I normalized it by dividing annual unlinked trips by directional route miles. That's why I also analyzed annual unlinked trips.

by Matt Johnson on Jul 29, 2010 11:55 am • linkreport

@David desJardins

That's a little muddled, though. I mean, passenger miles are equal to passengers*miles traveled, right? So even if your system was only, say, a mile long, but a hundred million people rode it, you'd beat out a system that was two miles long but only 350k people rode it. Earlier we explored that trips on NYCT, the most "transit-like" system in the country (the subway very specifically doesn't leave the city limits!) are about 20+% shorter than trips on the Metro. It still beats Metro on this category due to the multiplier. I think the metrics we should consider moving towards, but that, at times, are more challenging to acquire (although not impossible, when I have time I'll pull some) are the % of trips in a metropolitan area made by transit. This speaks worlds about the usefulness of transit in the region as well as the costs of driving vs. commuting. If the Metro got you everywhere in DC and 24.7 more people would ride it. Similarly, if parking were free in NYC, there would be far more traffic and far less people on the subway. It's challenging to disambiguate these two effects, but that's what happens when you use a proxy.

by Matt on Jul 29, 2010 12:06 pm • linkreport

@ Jamie,

Though I see your point, I wouldnt't call VRE or MARC out of the system. People use it to go from home to work and vice-versa and a lot of people actually step off at Crystal City & L'enfant Plaza, therefore being part of the system

@ Omar

A RER or S bahn system would not work in DC simply because the demand and density is not there. That's why VRE only runs during the morning and evening with one single train mid-day. Unless DC becomes more dense it would be a folly.

@ Mike & Jason

Granted my extortionate fare could not work everywhere in every circumstances. Having a special smart trip card for airport workers aka disabled passes could also do the trick. And yes maybe have 2 sets of prices according to destination: ie $6 to/from Virginia. $12 beyond. Would still be cheaper on average than to fly from Reagan...

by Vincent Flament on Jul 29, 2010 12:10 pm • linkreport

The transfer between Farragut North and Farragut West as well as the transfer from Metro Center to Gallery Place can be accomplished by Smartrip. Connecting those stations with a tunnel is unnecessary and expensive.

by Interested on Jul 29, 2010 1:05 pm • linkreport

@Vincent Flament

The density and demand thing are an interesting argument. I would argue that there is the density. MD and VA around DC are not particularly less dense than Long Island, CT, and NJ around NYC, and yet those commuter rails carry thousands upon thousands more passengers per day. This leads us back to the economics of rail ridership. Truthfully, rail is an imperfect substitute for car travel. Even so, we would expect a rise in rail ridership at some ratio to an increase in the price of driving. It works with chicken and pork, and it will work here. I'm not sure there's a good study on the cross-elasticity of public transportation, though. I've only ever seen point elasticities here, and generally they are used to explain why a peak of the peak will raise revenue more than it will lower it (peak ridership is relatively inelastic). Were we to build a new highway, ridership would drop some unknown quantity. Similarly, were we to build a new rail line, ridership would increase some other unknown quantity. What might be more interesting, and certainly more readily manipulated would be the price cross elasticity. More specifically, what would the toll need to be on the beltway to maximize the ridership on the Metro given the current capacity. In some places, this might require lowering the price of Metro, in others, you might even be able to increase the price along with this hypothetical toll. Anyway, I would be very interested in a study of cross elasticity between transportation mode share, but I'm not sure the right numbers exist for something thorough to be done.

by Matt on Jul 29, 2010 1:27 pm • linkreport

@Matt Johnson: Matt, any chance you could figure out what the average cost per trip on Metrorail is - something like "total revenue from metrorail in 2009"/"total riders by in 2009"

by Dan on Jul 29, 2010 4:31 pm • linkreport


That's operating $ per trip. Check out part 2.

by MLD on Jul 29, 2010 4:37 pm • linkreport

Please put an asterisk ("*") on fare bars that include a surcharge. The BART / SFO surcharge is making the users of that connection pay for the planner's mistakes.

That airport (SFO) has an extended hours post office on West Field Rd. Had they run the people mover out to the Millbrae Inter-modal (BART, Caltrain [tri-county commuter rail], Samtrans [county bus system], employee shuttle buses, taxis) the locals could have had ready access to the terminal amenities and that post office. These days one has to know about the employee shuttle and how to connect to the people mover once at the terminal.

NB - The above is from personal experience. I've been gouged, shuttled, and have walked from both ends (avoid the South side - it's mostly a shoulder walk).

by Ted K. on Jul 31, 2010 1:37 pm • linkreport

I have to say, visiting a city is much nicer when there's a direct connection to the airport. Look at the usage at National.

by Omar on Jul 31, 2010 3:50 pm • linkreport

Matt, it's interesting to see this, even though I don't live in Washington. One problem is that transit--even rail transit--in most American cities isn't dominated by heavy rail. New York, Boston, Washington, Atlanta, and Chicago are pretty much the only American cities that are.

Here in the San Francisco Bay Area, BART only serves 4 of the 9 counties in the region, and has only 43 stations in a region of 7,000,000. More than twice as many passengers use the San Francisco Municipal Railway (Muni) every weekday as use BART. So assessing BART is interesting, but it only tells you a little bit about the state of regional transit.

As to inter-station walkways, I don't know the logistics or politics in DC. I would say that there are some very long connectors in New York and some even longer ones in London, so it is possible.

by Nathan on Aug 4, 2010 5:19 pm • linkreport

I've not traveled on the subways and light rail of the other cities mentioned but I do ride the metro every day - the red and green lines from Shady Grove to Anacostia. I find the trip fast, easy, and affordable. I always feel safe and the cars and stations are generally clean.

by mannie on May 23, 2014 10:46 am • linkreport

Add a Comment

Name: (will be displayed on the comments page)

Email: (must be your real address, but will be kept private)

URL: (optional, will be displayed)

You can use some HTML, like <blockquote>quoting another comment</blockquote>, <i>italics</i>, and <a href="http://url_here">hyperlinks</a>. More here.

Your comment:

By submitting a comment, you agree to abide by our comment policy.
Notify me of followup comments via email. (You can also subscribe without commenting.)
Save my name and email address on this computer so I don't have to enter it next time, and so I don't have to answer the anti-spam map challenge question in the future.


Support Us