Charlie Chieppo, call your office

The vaunted Pioneer Institute, that of free marketry and good government, has a contest out which solicits plans to “move people and goods forward.” (Your fearless blogger is preparing an entry.) They encourage many people to apply, and say

If you are a rider, driver, transit employee, or simply an observer, we invite you to share how you would improve your commute, from immediate issues around parking, communication, safety, and station repair and design, to big ideas that will reduce congestion, advance bus rapid transit, and double the number of commuter rail riders.

So imagine my surprise when a Pioneer acolyte, Charlie Chieppo (along with a professor at D’Amore-McKim … which I’ll abbreviate as “Dim” because that’s my impression of the intellect of anyone associated with this article), goes into the pages of the Globe to argue that, no, we don’t need Commuter Rail, because it will be replaced by self-driving cars in no time. Also, come on, Boston Globe, you’re above this. (Oh, wait, except for Jeff Jacoby, he of amazing Twitter ratios. The man has no shame, so that’s something.)

As for the article: where to start? This basically hits every trope of the “cars are going to save us” argument that people make without even the faintest clue of how self driving cars are (or aren’t) working, and thinking about, you know, what you do with cars when people get out of them. But this is so chock-full of nonsense it makes sense to critique it line-by-line. Grab a beverage and hang on, because here we go. (Original in plain text indented, my comments in italics)


Self-driving cars could make commuter rail obsolete

The could. Anything could. Unicorns and puppies could. Let’s see if you can make a cogent argument. I’m ready to put on my surprised hat.

The rise of shared electric self-driving cars and the transition from a world of ownership to one of consumers purchasing transportation as a service holds the promise of significant economic, environmental, and quality-of-life benefits.

Wait, this is being written in present tense? Because outside of a few very few minor prototypes, there are literally no self-driving electric cars. Certainly not operating on roads. Maybe they mean Teslas, which are electric, but are pretty far from self-driving, although they do have a tendency to self-drive their drivers in to stationary objects. Waymo, which is way ahead of Tesla on the tech front, has a CEO pointing out that not only are self-driving cars decades away, they may never be able to fully drive themselves. So let’s not pretend that this is something that is a couple of years away. Because it’s been a couple of years away for a few years now. So, maybe, use the future tense here.

But it will also pose an existential threat to public transportation in general and to commuter rail in particular.

Because … let’s see. Commuter Rail has been the least-affected by Uber and Lyft, which are basically self-driving cars with drivers. They have the same utility, and Commuter Rail ridership has gone up in the past few years (more on this in a moment). But, sure, go on.

The first recommendation in the December report from Governor Baker’s Commission on the Future of Transportation is “Prioritize investment in public transit as the foundation for a robust, reliable, clean, and efficient transportation system.” In broad terms, the commission is right. But maximizing potential benefits from the unprecedented disruption of surface transportation that lies ahead will also require fundamental change at the MBTA and a hard look at which transit modes are positioned to compete in a brave new world.

Do we have any idea of what this brave new world holds? No, we do not. And why not? Because fifteen years ago Segways were the future. Three years ago we were all going to be riding hover boards. No one knows the future. Sure, we should probably think about contingencies. But we should also invest in what we have now that, you know, exists.

The commission’s charge was to look at the Commonwealth’s needs and challenges over the next 20 years. But if that horizon is extended to 40 years,

But it’s not. It’s 20 years. Why? Maybe because no one can predict the future in 40 years. In 1890, would anyone have predicted the coming of the car by 1930? In 1930—when biplanes ruled the day—would anyone have foreseen the 747 in 1970? In 1970, when gas cost a quarter a gallon, would anyone have seen $4 gas and global warming, but no flying cars or Jetsons highways? The only constant from 1970 is that cars cause traffic and traffic sucks. And you’re proposing more cars. Go on.

station-to-station service to the suburbs is unlikely to be very attractive in a world where shared electric self-driving cars will offer much faster door-to-door service at a price that won’t be much higher.

Where to begin with this word-salad. Station-to-station works because it’s a nice straight line from the city to the suburbs. Door-to-door sounds good, especially if it’s much faster, but this relies on the completely fallible assumption that somehow cars are going to solve traffic. In a region which now has the worst traffic in the country. Self-driving cars are likely to increase vehicle miles traveled (just as Uber and Lyft have), and thus congestion. The reason people take the train is that it is faster than driving (as well as cheaper than parking). And don’t show me some perfect model, because the Occam’s Razor is “you put a lot more cars on the road and you’ll get more congestion.”

Drivers are normally the largest expense for any transportation business. It currently costs about 55 cents a mile to operate a vehicle with a single occupant. But it’s estimated that the cost could fall to 15 cents a mile for autonomous vehicles carrying two or three passengers, which would significantly reduce public transit’s price advantage.

Again, the cost of driving is generally not the only reason people take Commuter Rail. The time also is. And, yes, maybe the cars will make this faster. But the jury is long out on that. Also, should we not invest in Commuter Rail, which is much more efficient at moving people than cars, for 40 years with the hope that maybe cars will somewhat get more efficient even though they haven’t—space-wise—uh, ever? That doesn’t sound like planning for the future. Also, according to this, most of the efficiency of a car comes from sharing, since 45¢ split three ways is 15¢, so not much off of 55¢. And, hey, slugging works … in a couple of dense markets with very specific conditions, most notably carpool lanes which can bypass traffic (again, time ≥ money). Those aren’t a bad idea, but can’t really replace Commuter Rail. Also, many people want to be matched with three people, not two, for safety in numbers. This is a feature of slugging, but makes the matching harder and the trips take longer. Also, if you have door-to-door service, you necessarily have a portion of the trip with two people. This isn’t terribly easy to do.

Connected vehicles will also dramatically reduce human error, resulting in big increases in throughput thanks to variables like higher travel speeds, less space between vehicles and less frequent braking in response to accidents and other travel events.

Oh, here’s the second BINGO square they’re checking off. This is nonsense. First of all, connected vehicles will leave more space between cars than people do today, because drivers today generally follow too close. Second, they may deal with some of human reaction time, which is solvable. The problem is that at highway speeds, this accounts for about 15% of braking time. The other 85% is physics. That’s a little harder to solve.

In the future, agencies like the MBTA will probably subsidize trips that are currently taken on commuter rail rather than operate them. Even with the transportation transformation in its infancy, Florida’s Pinellas Suncoast Transit Authority, which serves the St. Petersburg/Clearwater area, eliminated some bus routes further from the urban core, after it experienced an 11 percent overall drop in ridership, and replaced them with subsidies for Uber and Lyft rides. Since then, over 25 US communities have established similar partnerships — and the disruption caused by ride-hailing services is minuscule compared with what is to come.

Ah, yes, the transit mecca of … Tampa-St. Petersburg? Which has 57% the population of Boston and 3% of the transit ridership? (NTD data here and here) That seems like a great comparison. We should aspire to be the same Tampa-St. Pete which is well known as the strip club capital of the world? World class city, indeed! 

MBTA commuter rail ridership has declined.

No. It hasn’t. This is based off a Pioneer Institute study which selectively chooses a start and end to argue that Commuter Rail ridership is down. But this is based on data which basically useless, the T has all but admitted that outside of a ridership report in 2012, they really have no idea how many people are taking Commuter Rail. So the authors could have pled ignorance last month. Except that a couple of weeks ago, a new report came out, an actual rider count, and the numbers are pretty clear: Commuter Rail ridership is up. Way up. Up on every line. Up 20% in six years. Up nearly 50% on the Worcester Line. So this is just false. I can’t wait until Pioneer comes out a report that argues that the sky is green, the Pope is most decidedly not Catholic, bears defecate anywhere but the woods, and Commuter Rail ridership is down.

Nonetheless, it will remain with us for the next couple of decades. It still needs to be improved, but massive investments in new lines like South Coast Rail or, even worse, Springfield, would be a fool’s errand.

It’s been with us since 1834, so yes, that’s probably true. And we can debate the intelligence of extensions to Commonwealth’s 3rd, 6th and 10th largest cities (and I’d suggest there are certainly better ways to do so for South Coast Rail and that service to Springfield is probably more of a political issue than anything else). But let’s not pretend that people going to Springfield are going to want to get in to a car with two or three strangers for a two hour ride to Springfield. If you sit next to a smelly UberPool passenger on your way across town, you can hold your nose. If you sit next to a smelly person on the train, you can switch seats. If you’re on your way to Springfield, once you’re on the Pike, you’re on the Pike. Whoops.

The biggest challenge for the future will be making transit work in congested downtown areas. One Boston traffic simulation model showed that while shared autonomous vehicles would reduce travel times and the number of vehicles on the road even as total miles traveled rose by 16 percent overall, downtown travel times would be 5.5 percent longer because the vehicles would substitute for transit use.

Correct. Although actually this is usually the type of relationship where congestion goes up faster than vehicle miles traveled. But, correct. Something about stopped clocks and blind squirrels.

Rising to this challenge will require focusing more investment in the urban core. But success will require something more: changing the MBTA’s top priority from providing jobs and pensions to serving its riders.

During a three-year exemption from the Commonwealth’s costly anti-privatization law, the T dramatically improved performance in areas such as cash collection and reconciliation and warehousing and logistics, and saved millions. Despite this success, there was nary a peep about extending the exemption or making it permanent.

Can’t be from Pioneer if you don’t have a non-sequitur about privatization. What that has to do with focusing on the urban core is beyond me. Also, where do you think the destination of most Commuter Rail ridership is? Also, isn’t Commuter Rail privatized? My head is spinning.

Few would argue that the MBTA is skilled at putting customers first.

Well, a stopped clock is right twice a day.

The question is whether — in the face of an existential threat to public transit and with far less margin of error — political leaders, bureaucrats, and unions can change the authority’s culture and begin to lay the groundwork that will allow the T to perform the way we’ll desperately need it to in the future.

We’re back to Commuter Rail. Is this about privatization or Commuter Rail? Was the first two thirds of the article about Commuter Rail so we could segue to an argument for privatizing something which is already privatized? I’m really confused. Also: does the Globe have editors?

Part of that culture change will be recognizing that commuter rail is poorly positioned to compete over the long-term.

This article has offered exactly zero evidence to support this claim. None.

When the Patriots win the 2060 Super Bowl, stories about a suburban rail network overwhelmed with riders are likely to generate the same reaction as when we tell our kids about having to get up and walk to the television to change the channel.

First of all, Tom Brady will be 82 when the 2060 Super Bowl takes place. Also, football may not be a sport we recognize. And the Patriots, without Brady or Belichick, will be like any of the other 32 teams, giving them a 3.1% chance of winning Super Bowl XCIV. So there’s a 3.1% chance that the Patriots win the 2060 Super Bowl. And, I’d say, a 3.1% chance we don’t have a Commuter Rail network then, either.


Here’s the rub: Commuter Rail (and transit in general) is really quite good at moving many people in to a small space. Cars are singularly bad at this. A lane of cars carries about 1600 people per hour, even with two or three people in each, it may be able to handle 4000. Commuter Rail, as it runs today, carries more than this, but Commuter Rail, as it’s run today, is woefully inefficient. Electrified rail service with level boarding platforms (relatively small investments, compared to the cost of building the rights-of-way which already exist) can increase this five-fold. A well-functioning Commuter Rail line should be able to carry 10,000 passengers per hour, more than a roadway can, no matter who’s driving (unless there’s a bus lane with a bus terminal at the end).

As Jarrett Walker likes to point out: transit is a question of geometry. You can make all the taxi-summoning apps you want, but cars still take up a lot of space. (Remember, at rush hour, Commuter Rail brings in about as many people to Boston as highways do.) Get rid of Commuter Rail, even if everyone carpools (and people who don’t take Commuter Rail already probably won’t want to) and you’ll increase cars coming in to the city by a lot. And then where do those cars go? In to garages? Not enough garages. (No, they won’t go back out to the suburbs to fetch more people; rush hour is over by then.) They’ll probably do what Ubers and Lyfts do: drive more miles without passengers. Congratulations, you’ve traded parking in the city for parking outside the city and more vehicle miles traveled.

But, yes, Commuter Rail is going away.

A quick look at the T’s privatization bus list

If you follow this page, you’ll know that I’ve gone on in the past about hyperbole about MBTA privatization. And today, with the list of routes out that are up for bid, a few comments. The idea, apparently, is to reallocate the resources from these routes to core services, so that there will be an increase in service by covering these few routes with other vehicles (the T can’t buy any more vehicles because it is out of room to store and maintain buses), except for late night. The T breaks it down in to three groups—low ridership local routes, express buses (most but not all) and late night routes—but I’d split the first group further in to routes under 200 (core MBTA service routes originally run by the MTA) and routes numbered above 200 (peripheral or suburban routes run by other providers).

If this is as far as privatization goes, it seems to make sense, especially reallocating the service hours and vehicles of large buses being used on low volume routes to other routes and times of day. The T shouldn’t spend time and money to maintain smaller vehicles, and for the few routes where smaller vehicles would work, as long as this expands service rather than replacing it (i.e., the buses are put to better use on other routes) it will help service and ridership. The fear, of course, is that in the long run the number of Carmen-operated 40-foot buses goes down, with, perhaps, a deterioration of service. I’m less sure about the express buses, where I think the idea might be to use coach-style buses for these trips, see below.

A couple of resources:
The 2014 MBTA Blue Book
Changes to transit service since 1964
Miles on the MBTA (a kid—a rising sophomore now?—who had ridden nearly every route the T operates, which is impressive)

Note: service hours are estimated, as are riders per hour and riders per trip. Ridership is from the Blue Book, cost is net cost from a separate spreadsheet from the T.

Core Routes

5. The 5 is a social service route; it completely duplicates the 10 except for the stretch along the McCormack Housing Project which would not have service during the midday. It operates with few passengers, most of whom could probably take the 10. (It’s not as bad as the 48, but similar.) It could be operated with smaller buses, easily, which would probably yield cost savings.
6.5 service hours per day
161 riders, 25 riders per service hour, or 12 per trip.
$0.91/passenger mile

18. The 18 is a local route that runs on Dot Ave from Andrew to Ashmont. Because it runs roughly parallel to the Red Line on 30 to 60 minute headways, most people would opt for the Red Line. So it is likely used for short hops by people unable to walk as far. It is likely similar to the 5 in that is is more a social service route than anything else, and could also be operated with smaller vehicles.
17 service hours per day
619 riders, 36 riders per service hour, 18 riders per trip
$0.65/passenger mile

68. The 68 is a lot like the 18: it parallels the Red Line, has few passengers and rarely-if-ever a full bus (except when the Red Line has a fault, but it is doubtful that enough people know about it to use it) and has the main purpose of getting people between Kendall, Harvard, Cambridge city offices and the library. Nearly all of its functions are duplicated elsewhere, and it didn’t even run between 1981 and 1998. It, too, could be operated with smaller buses to carry the passengers necessary, and the bus used for the 68 could be pressed in to service on Mass Ave where it would carry far more passengers.
12 service hours per day
468 riders, 39 riders per hour, 20 riders per trip
$1.00/passenger mile

99.  This one is harder to explain. The 99 is a subway feeder trip, and at rush hour runs every 20 minutes, which leads me to believe that if it has two or three heavy Orange Line trains empty on to it, it will need a full-sized bus. It’s not a route like the 5, 18, or 68: it has a purpose other than shortening a walk slightly for a few people. It has more than the ridership of those three routes combined. Could it be better integrated with the 106 to provide even headways where they share a route? Certainly. But other than cheaper operators, I don’t see how you get away with operating this with smaller buses.
30 service hours per day
1555 riders, 52 riders per service hour, 26 riders per trip
$0.49/passenger mile

201/202. These are two variations of what was Route 20 from Neponset to Fields Corner. The two trips now combine to run every 15 minutes (or so) on a variety of routes. The route is short and it seems that it has relatively low ridership per bus but high frequency; and most people who would take it instead walk to the Red Line (a station at Neponset on the Braintree Branch would obviate much of the need for the route). Depending on rush hour loads, smaller buses may suffice.
72 (?) service hours per day
1339 riders, 19 riders per hour, 9 riders per trip

Suburban routes

52. The 52 is an old M&B Route through Newton on Centre Street in Newton and through to the Dedham Mall. It has minimal rush hour feeder ridership since it traverses a not-so-dense part of Newton and most people can walk to the 57 or Green Line or Express bus instead. Whether its ridership merits a smaller bus depends on the loading at different times of day.
17 service hours per day
766 riders per day, 30 riders per hour, 23 riders per trip
$0.49/passenger mile

70A. This is really interesting. The 70 is not listed, because it’s a busy trunk route which could never have smaller buses. However, it is one of the most dysfunctional routes the T runs because … of the 70A! I’ve advocated in the past for splitting the 70A off as a feeder or transfer route to the 70. So in theory this would accomplish exactly that; I would assume that the privatized portion of the route (the 70A) will not run all the way in to Cambridge; otherwise you’ll have even less coordination between the two mostly-parallel routes with different operators. (If the T does do this, they should all be sacked.) I still think interlining the route with the 556 makes as much, or more, sense, but splitting the 70A off of the 70 could be very beneficial.

I have no idea what the 70A portion of the ridership is as it is not broken out in the Blue Book. But if the current two buses were allocated to the 70, it could provide 8 to 10 minute headways at rush hour and 15 minute headways midday, allowing it to provide a Key Route level of service, with even headways (something it does not do right now) and alleviating chronic crowding along the route. This could be a major bright spot in this privatization scheme, if implemented properly.

210 and 212 are two local routes in Quincy and Dorchester that parallel the Red Line, although they provide frequent feeder service at rush hour. The 212 is a slight variant and only runs a few trips at rush hour, and has far higher ridership per bus/trip than the 210, which has ridership spread more across the day. Since 212 is a feeder, it may require larger vehicles to cope with rush hour crowding, although 210 is more a Red Line parallel route (although it does provide direct service from Quincy to Dorchester that would otherwise require a transfer).
26 service hours per day (21: 210; 5: 212)
210: 736 riders per day, 35 riders per hour, 18 riders per trip, $0.31/pax/mi
212: 293 riders per day, 59 riders per hour, 30 riders per trip, $0.89/pax/mi

439. This is a great candidate for a smaller bus. It carries 97 riders all day, but is the only service to Nahant; it is one of few cases in the area where a town has transit service at rush hour and not at any other time of the day. Even assuming that all riders go inbound on two trips and outbound on the others, (a fair assumption) this is little more than a feeder to the Lynn Commuter Rail station (and a couple of trips to Wonderland) with 20 riders per trip, which could be handled by a smaller vehicle, and this bus could be put to better use.
6 service hours per day
97 riders, 16 riders per hour, 8 riders per trip.

451. Way out in the suburbs, this is a bus that is a Commuter Rail feeder to Salem, running at rush hours to and from Beverly. It also provides last-mile service between the Commuter Rail station and the Cummings Center. It could probably be operated with smaller vehicles.
9 service hours per day
163 riders, 18 riders per hour, 10 riders per trip

465. Most of the time, this bus runs between Salem and the North Shore and Liberty Tree malls (and you know how I feel about buses to malls; maybe the mall should pay for it). At rush hour, however, it operates like the 451, with more direct service to and from the train station. In fact, during the PM rush hour, passengers to the mall ride the commuter trip first, and can then stay on for the inbound which loops back to the mall. Which actually makes sense. Still, it has relatively low ridership, and a smaller bus may suffice. For a time, the 451 and 465 were through-routed, allowing one-seat rider from Beverly to the malls.
20 service hours per day
414 riders, 21 riders per hour, 16 riders per trip.

Express bus routes

I’m a bit more perplexed about the express routes shown. The 500-series routes to Newton, for example, are heavily used (often standing room only on 40′ transit buses) and operate frequently: the type of service transit buses excel at. The same goes for the Medford buses (320 series). The 505, for example, operates 20 inbound trips with 556 inbound riders, meaning that the average load is 28 riders per bus, but this includes a few trips in the 6 and 9 o’clock hours, so at peak times—when the bus runs every 9 minutes—the buses are full. And the bus runs like a city bus from Waltham to West Newton, so a low-floor transit bus is far superior to a high-floor commuter coach. (Many transit agencies do operate coach-style bus service, but it is usually from a park and ride to downtown, which Commuter Rail provides in Boston.) The 502 and 504 provide similar service and frequencies. So in this case you’d need a 40-foot bus, so I’m not sure what the savings would be, unless the successful bidder ran a feeder system to frequent rail service on the Worcester Line that could bypass traffic.

The costs per passenger mile for these express buses ranges from $0.17 to $0.32, with higher fares and longer distances.

The 351 makes a lot more sense, in theory. It is an outbound express bus from Alewife to Burlington and carries about 20 passengers on each of its four trips, some of which may serve as pull-outs for other routes. It also serves an area akin to those served by several of the 128 Business Council‘s shuttles, which serve similar office parks with similar schedules, in areas which otherwise have little if any transit service. It is a good example of a private service that leverages the existing transit network (rather than competing with it) in an area which would likely be hard for the T to justify serving. (Full disclosure, I work for another TMA which operates last-mile shuttles.)

But the issue with the 351 is that it might not operate in a vacuum: some of its trips may be a pullout which then turns as a 350 or a 352. The 352 is also on the list, and I would assume that these would be sold as a package deal. Similarly, it’s not surprising to not see the 170 bus because it serves basically as a revenue pull-out trip for 70 buses. (Although a later trip for the 170 during each rush hour would dramatically help that service; perhaps some of the 70A’s resources could go towards that.) This is the issue with trying to privatize parts of a network: you often wind up pulling on a string and unraveling the whole sweater. These seem to be surgical enough that they won’t dramatically affect it, but—to mix metaphors—it could be a slippery slope.

It will be interesting to see what kind of response the T gets. The union doesn’t seem keen on letting anything out of their grasp, and there would be some perceived risk in operating a trip privately and not running afoul of the Carmen. Unlike the lead-up to Pacheco, there are no proposals to privatize an entire garage, but rather a few small routes. It’s less likely to ruffle a lot of feathers, but the slippery slope argument holds. First they came for the 68, and I said nothing, because I did not drive the 68. Will they then come for the 28 or the 66?

It will be interesting.

The T’s bus maintenance costs are … generally in line with other large transit agencies

Shirley Leung, the Globe’s business columnist (and Olympics cheerleader) wrote a spurious article on Friday about MBTA privatization (we dissected that piece line-by-line here, and promised this longer post), where she leaned significantly on a recent study from the Pioneer Institute. Why Pioneer is taken seriously in regards to transportation is beyond me. Their numbers rarely stand up to the light of day, and when Very Serious Columnists are taking their analysis as gospel, it is a bad sign for the news media, and bad for local politics in general.

This post will look at some depth at the Pioneer Institute’s bus maintenance cost study, titled “The T’s Bus Maintenance Costs are Out of Control.” I’ve referenced the spurious “research” from the Pioneer Institute before, and mentioned this report; this is a full examination. Their assertion is that the T’s bus maintenance costs are “out of control”, some of the highest in the country, and that the T could save tens of millions of dollars a year if they just reformed how they maintain their bus fleet, asserting that their costs are twice as expensive as “peer agencies” (as defined by them, although we’ll see that their definition of “peer” is, well, suspect). There are numerous problems with this study and the numbers they use in it, to the point where their conclusions are drastically overstated and need to be fully reexamined. Since this is a very long post, I am going to break it in to two pieces: a shorter executive summary, and a larger jump deep in to the weeds of the data.

This post will discuss and dissect the following specious and disingenuous parts of the report:

  • The Pioneer Institute’s faulty definition of “peer agencies”; the metric they use to select agencies is one of few which by definition varies by agency policy.
  • Is “cost per revenue mile” the best metric for measuring the effectiveness of maintenance operations? Does a different number (cost per revenue hour) give us better data and better match how most bus operations account for bus costs, anyway?
  • Pioneer cites Minneapolis as a close analog. But they don’t account for dramatic differences in costs of living and maintenance facilities in the two cities.
  • Pioneer also digs in to the T’s salary database to pull out examples of highly-paid maintenance workers. Yet they look only at the top salaries, not at averages. This is simply hyperbole; their selective use of data serve not to inform people but to scare or anger them.
Overall, much of the supposed variation between the MBTA’s costs and other agencies that Pioneer cites are easily debunked, and they misuse data to misconstrue a problem to fit their agenda of privatization and overall cuts to transit. Instead of looking at appropriate data to try to see how the MBTA could best reign in costs, they use some of the worst data they can find to make hyperbolic statements. And when MBTA reform panels view these data as fact, they’ll make uninformed decisions.

In Brief: By Manipulating data, Pioneer Misrepresents Facts on the Ground

To make its case that the T’s costs are “out of control”, Pioneer first needs to find other agencies to compare the costs to. While there are many ways to use the National Transit Database to choose systems similar to the T, the Pioneer Institute takes an interesting approach. And I don’t mean interesting as in novel, I mean interesting as in suspect. The most logical idea would be to use a list of other large transit systems, but the Pioneer Institute uses miles between “failures” which the NTD explicitly points out in their definitions is subject to variation in agency policy.

By doing so, and by selecting agencies which carry one fifteenth as many passengers at MBTA buses alone (and in some cases as few as one fiftieth—or two percent—of the total number of passengers) they are really comparing apples to oranges. The T is being compared to sunbelt cities (no road salt, roads with less traffic and fewer acceleration and deceleration cycles) with many fewer passengers. These include systems which serve Palm Beach County, suburban Detroit and El Paso, for example, yet the report doesn’t compare the T to Philadelphia or Seattle, much better analogs. Comparing the T to its actual peers—other top-20 transit agencies—makes the costs go from 100% higher to just 40%. More than half of the supposed out of control costs are because of a false comparison.

The report also uses “revenue miles” while “revenue hours” would be a better metric. The T, which serves a compact, urban area, has a relatively high maintenance cost per revenue mile while its cost per revenue hour is more in line with other areas. It turns out that per hour (and many bus operation costs are per hour, not per mile) the T has some of the most efficient buses in the country, in the company of Chicago, San Francisco and New York, and far ahead of El Paso and Spokane. To put it another way: not all miles are created equally: a mile on the T will have more stops, more acceleration and more stress on the bus infrastructure, but also many more passengers.

Pioneer then cites Minneapolis as a good analog for the MBTA. While Minneapolis has a similar number of buses to the T, this analysis is somewhat fraught for a few reasons. First, Minneapolis carries only about half the bus passengers of the MBTA (and operates only a nascent rail system, so its overall transit ridership is only 20% of the T). Minneapolis also has much newer maintenance facilities (all have been built since the 1980s) as opposed to the T’s facilities, many of which date from the streetcar era. In addition, Minneapolis has the capacity to store nearly all of their vehicles indoors, a far cry from the T’s outdoor yards. Boston does have an advantage, however; it’s facilities are much more centrally-located, and it has only about half the deadhead requirements moving vehicles in and out of service that Minneapolis has, saving significant operating expenses.

Finally, Pioneer makes a hyperbolic statement about how highly-paid some T maintenance employees are, but it of course only cites the highest paid employees, and doesn’t look at any averages. The average worker is not paid in excess of $100,000; and the $70,000 maintenance salaries commanded by the skilled workers who keep the T’s rolling stock in place may be high, but are commensurate with the high cost of living. (House prices in Boston are twice what they are, for example, in Minneapolis.) The high salaries cited are due to overtime accrued (those lazy union members at it again, working 60 hour weeks), likely at times when many vehicles are in need of repairs to maintain service (the T has a relatively low spare ratio, so a problem which at another agency might be put off while a new bus is put in to service would have to be fixed at the MBTA). These cost differentials, and the fact that the MBTA has invested far less in capital facilities than Minneapolis (upgrading all of the T’s bus facilities would likely stretch beyond a billion dollars), bring the costs much closer together. Additionally, the T keeps a lower headcount (which saves on the number of benefited employees) and instead pays more overtime. This may actually save money, since for many employees the cost of benefits amounts to 30% or more of base salary, so having more employees working fewer hours may cost more.

Pioneer’s study is a textbook case of having an agenda and massaging data to best fit ones premonition. The problem is that under more scrutiny, the numbers mostly fall apart. The T’s maintenance costs are more expensive than several other large systems (although notably less than New York), and there are certainly lessons that we could learn from them. However, instead of beating the drum of the greedy unions and that privatization is the only answer, we should look at what works—and what doesn’t work—and how the T can take advantage of more efficient workplace and management practices. Relations between government and the T unions have never been warm, but specious threats based on hollow data and parroted time and again in the news media will not help the matter. And rather than trying to drive a wedge between management and the laborers, the T should give the unions a seat at the table: workers without the constant specter that their jobs may be outsourced are likely to be more productive, which may save money in the long run.

In Detail: The Numbers Behind the Curtain

What is a “Peer Agency” anyway?

In their report, the Pioneer Institute defines the “peer agencies” against which they compare the MBTA’s costs. Before we get in to exactly how they do so, below are three lists of cities (or in the case of larger statewide agencies, the city or region in which the agency operates). Two are what the Pioneer Institute used as their “peer agency” list. Which do you think is a better fit for the comparison against the T? (NB: I’ve colored systems which appear on two or more of the lists to show the amount of crossover between them.)

List A: List B: List C:
Washington, D.C.
Salt Lake City
San Jose (VTA)
Saint Petersburg
Baltimore (MTA)
New Jersey Transit
Orange County, Calif.
San Diego
Delaware Transit
Suburban Detroit
Palm Beach County
Suburban Chicago
Washington, D.C.
Salt Lake City
San Jose (VTA)
Saint Petersburg
San Francisco
Fort Worth
San Bernardino
Rochester, N.Y.
Providence (RIPTA)
El Paso
Washington, D.C.
Baltimore (MTA)
New Jersey Transit
San Francisco
New York City
Los Angeles
San Diego
Las Vegas
Any guesses?

List C is a list of the largest bus transit systems in the country by daily ridership. Lists A and B are the lists the Pioneer Institute uses. Does it make sense to compare the MBTA to cities like Spokane, Salt Lake City and Saint Petersburg? Or does it make more sense to use cities like Los Angeles, Philadelphia and Denver? I’d argue the latter. The Pioneer Institute, apparently, thinks the former. Note that of the Pioneer Institute’s two lists comprise 32 cities, of which just 8—25%—are among the top 20 transit systems. The MBTA, for comparison, is the 7th largest.

Before we go any further, let’s look at one piece of data for these lists:

Pioneer List A: Pioneer List B: Top 20 Agencies:
Mean Daily Ridership 157396 186315 483682
Median Daily Ridership 100596 67335 320815

They’re comparing apples to oranges! The average ridership for the top 20 bus agencies is 483,000. (The T carries 405,000 daily.) The average sizes of the agencies Pioneer is comparing the MBTA to are 32% and 39% as big. Looking at the median size (more important, actually, since Pioneer does not weight their averages), the median Pioneer agency is 16-25% as big as the T. Unless there’s a very good explanation as to why they chose these agencies, it seems that they weren’t chosen for a logical reason.

So, how did Pioneer define “Peer Agency”? They looked at two metrics for vehicle maintenance: miles between failures (List A) and miles between major mechanical failures (List B). Pioneer Institute took some lower bound for the size of a bus fleet or daily ridership (somewhere in the neighborhood of 100 buses or 25,000 daily riders; they don’t define this) and sorted it by the overall miles between mechanical failures metric (which they define as “maintenance productivity performance”) from the National Transit Database (NTD) to find the “most similar” agencies.

This is problematic for two reasons. First, choosing this lower bound means that they are going to be comparing a wide variety of systems very different from the T, as seen by the list above that includes several transit systems with one tenth or fewer the daily ridership of the MBTA. Six of the agencies have ridership under one tenth of the T—40,000—another 13 come in under 100,000, or less then a quarter the size of the T. This means that nearly half of the MBTA’s “peer agencies” carry 1/4 or fewer the number of passengers daily; few of these have interconnected rail systems either. One comparison, to Palm Beach Transit, is so risible that on a Sunday, the 28 bus in Boston carries more passengers than the entire Palm Beach transit system.

This is more of an issue, because it means that Boston is being compared to many cities which have far, far less transit service. Since the MBTA carries 2/3 of its passengers by rail (New York, DC and—believe it or not—Atlanta are the only other cities in the country with more rail riders than bus riders) it means that the MBTA carries fifty times as many passengers each day than the smallest “peer” agencies in cities where transit accounts for a tiny percentage of trips. I’m going to pause and say this again: the Pioneer Institute defines a “peer agency” as one which carries 2% as many passengers daily as the MBTA. That’s a rounding error! This is like trying to draw conclusions by comparing Starbucks and an independent coffee shop. It doesn’t make any sense.

Thus, most of these systems will have buses which operate in much different conditions than the T. They have longer distances between stops, far less acceleration and deceleration, fewer passengers per bus and much less time spent with crush-load capacities which put more strain on the vehicle’s physical infrastructure. In most cases, their routes often operate through less-dense city centers and not mainly as a feeder system for subway stations like the T does, with many fewer cycles between full and empty.

Second, and perhaps even more egregious, there’s the definition of system failures (italics mine):

(List A) Other Mechanical System Failure: A failure of some other mechanical element of the revenue vehicle that, because of local agency policy, prevents the revenue vehicle from completing a scheduled revenue trip or from starting the next scheduled revenue trip even though the vehicle is physically able to continue in revenue service. 

(List B) Major Mechanical System Failure: A failure of some mechanical element of the revenue vehicle that prevents the vehicle from completing a scheduled revenue trip or from starting the next scheduled revenue trip because actual movement is limited or because of safety concerns.

While most NTD numbers are rather concrete (“passenger miles traveled” or “revenue service hours”) these much more fungible based on different maintenance criteria for different agencies. Note especially that the first list is explicitly based on local agency policy. So, not only is Pioneer choosing cities with much smaller systems, but they are compounding this issue with a metric which varies by local agency policies! What’s more, some agencies, like the Maryland Transit Agency, don’t report “other mechanical system failures”, only major ones.

It would make sense to use a variable that is well-correlated for this analysis. Here are the r-squared values for agencies with at least 25,000 daily bus passengers:

  • Major failures and maintenance cost per revenue mile: 0.049
  • Other failures and maintenance cost per revenue mile: 0.018
  • Average weekday passengers and maintenance cost per revenue mile: 0.39

One of these things is not like the others. One of these things is correlated order of magnitude with the cost per mile than the others. And the numbers range more than one might expect: major failures range from every 1600 miles to once every 117,000 miles; and other failures, for agencies which report them, range from once every 2900 miles to once every 953,000 miles.

What might cause a bus to be removed from service in one system might not in another. There’s nothing wrong with that, but it makes it very suspect to use these numbers as a basis for further analysis, especially when better numbers certainly exist. It would make much more sense to choose the cities in the top 20 to compare to the MBTA. The problem is that such a list wouldn’t have supported the Pioneer Institute’s “conclusions” nearly as strongly.

Pioneer is measuring the “productivity” of a bus maintenance shop by the frequency with which the buses broke down without looking at any other data. This is like measuring the “productivity” of a doctor by the frequency with which his or her patients die. Which is all well and good—some doctors are actually more productive than others—except that a cardiologist is going to have more patients die than an optometrist, a gerontologist is going to have more patients die than a pediatrician, and a doctor that works in a high smoking state like  Kentucky is going to have more people die than someone working in Utah. If you’re not correcting for the age of the vehicle, variations in the kinds of vehicles used, and variations in the duty cycle, you’re going to come to some screwy conclusions.

I like charts, so here is one. It shows transit systems with more than 20,000 riders. Note how the Pioneer Institute’s “peer agencies” are mostly clustered in systems with much lower ridership than the MBTA. I threw in a best fit line for fun, and, yes, larger agencies have higher maintenance costs. Which is not surprising.

In any case, it comes down to a sort of Occam’s Razor question: the obvious definition of “peer agency” would be “other large transit systems.” Since that was not the metric used, we need to look in to why it was not. If there’s no good answer why it wasn’t—and there isn’t—then it certainly seems like the reason Pioneer used these criteria is to try to prop up their preconceived conclusion, rather than to make an honest argument.

I’ve only compiled the 2013 data (why? because I am doing this evenings and weekends and I don’t have a paid team of fellows and researchers funded by right-wing climate change deniers to go back through several years), but note the following:

Pioneer List A: Pioneer List B: Top 20 Agencies:
MBTA maintenance
cost % of average
208% 192% 153%

In other words, just by choosing a more representative data set, we’ve explained away half of the discrepancy between the T and other agencies.

But, wait, there’s more. Let’s say you only looked at the small group of agencies clustered near the T. There’s a big (50% increase) gap between Denver at 250,000, and Seattle, at 391,000, and another 50% gap between SEPTA and the CTA. If we compare Boston to Seattle, New Jersey, San Francisco, New York MTA Bus division and Philadelphia, the T averages just 140% of those other agencies, barely one third of the difference Pioneer’s “data” show. That seems like a much more representative sample than, say, Salt Lake, Sacramento, San Jose, Saint Petersburg and Spokane. I would wonder if the Pioneer Institute would deign to offer an explanation as to why they chose “peer agencies” in the manner that they did.

As I see it, there are three. One is that they randomly picked the metric to categorize this, and somehow stumbled on to the worst one they could since it’s really the only NTD data which varies by agency. I doubt this based on the Occam’s Razor principle that if you asked a student in a first year transportation statistics class what to use, they’d be hard-pressed to come up with what Pioneer did. The second is that someone doing research at Pioneer just isn’t that bright, and that’s a possibility. But the third is more nefarious: they came in to the study with a question they had a preconceived answer for, and then found the snippet of data which best supported their thesis. The problem is, well, the data.

Are all miles created equally?

Once the Pioneer Institute chose the wrong systems to compare the T to, they then chose to compare those systems based on a metric of “maintenance cost per mile.” On its face, this seems like a good analysis, and it’s not bad, but it ignores the issue that not every mile is created equally. For instance, which causes more wear and tear on a bus: a mile in stop-and-go traffic with a stop every 800 feet and 60 passengers on the bus, or a mile at 35 mph with two traffic lights and two passenger stops carrying 20 or 30 passengers?

Here’s another way to think about it: which is more stressful for your car? Driving with four passengers, a full trunk in stop-and-go traffic or a highway trip? The former is what buses in major cities are put through on a daily basis. But in many of the “peer systems” on the lists above, buses ply faster suburban arterials in much less dense areas. This means that they make fewer stops, and at those stops they pick up fewer people. That means less wear and tear on door mechanisms, air bags, transmissions, tires and even the vehicle frame. Wear and tear on a bus depends on its operational environment.

A better metric—one that accounts more for operational differences—would be maintenance cost per revenue hour. Bus costs are usually measured in hours (operators are paid by the hour, not the mile) and calculating revenue hours is a better way to account for different operating environments. For instance, a bus in City A might average 10 miles per hour and a bus in City B might average 15 mph. However the bus in City A might have more passengers over the course of that hour, more openings and closings of the doors, and a heavier average load. Cost per hour lets us better account for this.

So how does the T stack up on this basis? It’s still quite high, around $45.00, while the top 20 cities average to $30 and the Pioneer “peers” come out at $25. Instead of being 200% of its peers, the T is “only” 180%. So by using this metric, the T is still underperforming. However, there are more outliers in the same range. In the previous chart, the T was the only agency in the range of the New York systems; in this metric, it is joined by three others: VTA (Santa Clara County/San Jose), Pittsburgh and Detroit. Still, it is 50% higher than its peers, so there is some room for improvement. Perhaps Pioneer’s conclusions should not be “we need to privatize everything and eliminate the unions” but rather “let’s see what these other cities are doing and what we can do better.”

Is Minneapolis the right analog?

Much of Pioneer’s study focuses on comparing the MBTA to MetroTransit, the bus system in Minneapolis, Saint Paul and the surrounding area. After comparing the T to the two lists of 20 cities (one of which includes MetroTransit) the authors focus in on comparing the T to one such agency which is more like the T. (And in a stunning use of logic, they don’t choose, say, El Paso.) MetroTransit operates a similarly-sized fleet to the MBTA, and also operates in winter weather (which is not the case for many “peer agencies” above). So comparing the T to Minneapolis is not a bad choice.

But are there differences that probably account for a lot of the difference? Certainly. First of all, Minneapolis carries about 215,000 bus passengers, about half as many as Boston. The average trip is longer, as buses are the workhorse of the system (the two rail lines combine for only about 60,000 passengers per day), so there are fewer bus-rail transfers. But the number of trips per revenue mile in Minneapolis is still only 60% of Boston*—in fact, Boston has one of the highest number of trips per mile of any system in the country.

[ * This will probably be lower in coming years, as Minneapolis replaced one of its busiest bus lines—the 16 and 50, connecting downtown Minneapolis and Saint Paul and accounting for 10% of the total system ridership—with a light rail line in 2014. This has led to a significant overall increase in transit use (already nearing 2030 estimates with 36,000 daily riders), but an overall decrease in bus patronage and the number of passengers per bus. The 16 ran at 24 hour service and 10 minute headways at most times of day and was supplemented by the limited stop 50, especially at rush hours. Combined, they carried 24,000 passengers per day. ]

Pioneer examines several years of data (to their credit) and I’m looking just at 2013 (because this isn’t, you know, my job), so there are obviously fleet age questions, but when looking at a sample of this size (2 cities), there are all sorts of questions. Did numbers spike some years because many buses went out for mid-life overhaul? Were old buses give only very necessary maintenance in anticipation of retirement? Did new buses coming online require much less maintenance? I’m not sure how much difference this might make.

But what I do think makes a difference are two factors:

  • The Twin Cities have a much lower cost of living than Boston
  • While Boston uses a variety of maintenance yards, some of which date back well in to the streetcar era, the Twin Cities made large investments in large, modern and enclosed maintenance facilities.

Boston is an expensive place to live. Housing prices are skyrocketing and affordable housing is somewhat hard to come by. Working on buses requires the ability to commute to a maintenance facility, most of which are located in areas with high housing prices, and staff need to be paid accordingly. Minneapolis is far less expensive. Housing prices in the Twin Cities in 2014 were just 53% as high as in Boston, and the overall cost of living was 16% lower. Since personnel costs are a large part of maintenance costs, it is not surprising that it costs more to pay staff in Boston.

Then there’s the question of capital investment. One of the reasons the T is reluctant to add bus service is that it can’t expand its fleet. The number of buses is constrained by yard space, and new bus yards don’t come cheap. Recent discussions about rebuilding the bus yard at Arborway have quoted figures in the range of $200 to $250 million, and that’s one of more than half a dozen facilities the T operates. Rebuilding enough bus yards for the whole system would likely be a billion dollar project, or higher.

What’s more, the current locations of the T’s bus yards are worth a lot of money, and would help with the regional housing crunch. One sits at Arborway, and could be sold for $20 million (or leased to a developer as well)—and the city would also reap property taxes from a property which is currently untaxed. Another, Cabot, sits steps from the Seaport District and South Station, Albany is similarly situated in the South End, and the largest is the Charlestown Yard, in a decrepit-but-transit-rich part of Sullivan Square. So there’s the opportunity cost of not selling these off for development as well. But having well-located bus yards is an asset to the T, as it’s non-revenue (“deadhead”) mileage is some of the lowest of any large system (11%), it can’t afford to lose any facilities without replacing that capacity, and large plots of land in warehousing or industrial areas are not plentiful in Boston.

Minneapolis, on the other hand, has relatively new bus maintenance facilities. Every facility has been built since 1980. (And there are only six facilities, as opposed to the MBTA’s nine, lending economies of scale.) With lower land prices and much less density, most of the facilities are relatively centrally located, yet don’t take up valuable, transit-rich land. What’s more, every bus facility in Minneapolis is fully enclosed. In Boston, most bus facilities consist of an open yard with a maintenance garage. In the case of Arborway—one of the largest facilities—it is a small, temporary building. Buses are by and large stored outside. In Minneapolis, nearly all buses are stored inside. There’s adequate room to work, and the buses being stored don’t sit overnight in freezing temperatures.

Even still, the Twin Cities’s bus yards are not as centrally-located as Boston’s (T yards are often shoehorned in to small parcels; MetroTransit yards are usually larger and squarer). While Boston runs 11% of its mileage out of revenue service (deadhead), the Twin Cities run 24% (and as far as hours go, Boston is 9% deadhead and the Twin Cities 14% owing to a larger and less-crowded highway system in the Twin Cities). To access these newer, more efficient facilities requires significantly more operation expense. It would be as if the T built new bus facilities in Billerica, Southborough and Brockton: they’d likely be more efficient, but savings would be eaten up driving buses back and forth to and from them. Given the T’s salaries, if their garages were located such that they required 14% deadhead hours like the Twin Cities, it would add $5 million to annual operating costs.

That being said, MetroTransit has certainly invested money in vehicle maintenance facilities that the T has not. If the T had spent a billion dollars on a set of brand new bus maintenance facilities, it would likely spend less on maintaining its buses. Of course, if we assume that there is $1 billion in deferred facility capital outlay and that a bus maintenance facility has a useful life of 50 years, it would cost $20 million per year to build those facilities, money the T is currently spending on other projects. I’m not saying that this is a logical way to do business; the T should budget for and build adequate maintenance facilities. But comparing a system with new, enclosed facilities with the T’s antiquated mishmash of bus yards and small garages certainly needs some qualification.

The T’s maintenance budget runs $40 to $60 million ahead of MetroTransit each year. But, again, there’s much more to this than meets the eye. If the T required as many deadhead hours as MetroTransit, it would cost $5 million extra in operating costs just to get the buses to and from the garage. Given the difference in cost of living, the T’s salaries are higher than MetroTransit’s. If we only look at salary, the cost of living difference accounts for $6 million of that difference, and if we add in benefits it’s $11 million. And then, there’s the $10 to $20 million annually that the T hasn’t spent on new maintenance facilities. Better fleet renewal and capital outlay for maintenance facilities would probably help the T’s maintenance budget: maintenance is not siloed financially from the rest of the transit system’s operation.

In 2013 the T spent $104 million on vehicle maintenance to MetroTransit’s $45, so I’ll use those numbers. I will note that was a year before the T accepted a large order of New Flyer hybrid buses, which should require less maintenance than the 20-year-old RTS fleet they replace. In 2013 T’s bus fleet (average age: 9.4 years) was significantly older than MetroTransit’s (average age: 5.1 years). (I’m not even accounting for the lower capital costs the T has with an older fleet, although the fleet age obviously manifests as higher maintenance costs.) This could certainly be reanalyzed for other years; by using the year when MetroTransit had a much newer fleet, this probably paints the T in a worse light than is actually the case. According to Pioneer, during years in which the fleets for the MBTA and MetroTransit were similar ages, the T’s costs were only about double MetroTransit’s.

In any case:

Total Annual Expendeture $105m % of total
– Deadhead operation cost difference

– Cost of living difference (salary, salary+benefits)

– Facility construction







Adjusted total $69-$84m 66-80%
% of difference $21m-$36m 35-58%
† Adjusted total ($90m MBTA) $55m-$70m 22-39%
† % of difference ($90m MBTA) $20m-$35m 44-78%

† From 2006 to 2009, MBTA costs were about 200% of MetroTransit costs, and the bus fleets’ ages were similar. Assuming that if the T had a similarly-aged bus fleet to MetroTransit and double the costs, it would have a $90 million maintenance budget. If this is a good way to account for the fleet’s age difference, between 40 and 80% of the difference could be explained by deadhead operation expenses, cost of living and facilities.

Citing high salaries as a scare tactic

One of the favorite pastimes of the anti-union, anti-government folks is to find someone who works for the government and makes a lot of money and hold it up as an example of government waste (well, unless it’s a football coach; they’re fine with that). In their report, they go and find examples of people making a lot of money (a painter making $80,000! A machinist making $120,000! A car cleaner bringing home $70,000!) and use this to show how wasteful government is. But, of course, this is hyperbole. These are outliers: probably people who worked a lot of overtime—some 60 hour weeks—in order to make that much money. Since transit runs 24/7, there’s overtime to be had: a few extra buses break down, or someone calls sick and a worker picks up an extra shift. Are there some savings to be had from better oversight of overtime? Probably. But without showing the average numbers, these are pure scare tactics, a rhetorical device that doesn’t stand up to any scrutiny. Luckily, I have the T’s salary database sitting around (which is public record, of course), so we can do some such analysis.

Yes, all MBTA salaries are public information.
You can access the file here.

It’s unlikely the T will ever be able to use non-union employees, even if maintenance is privatized. Given the construction boom in Boston, these salaries do not seem far out of hand for union labor (union sheet metal workers, for example, make nearly $100 per hour, or $200,000 per year, although they don’t of course, have the job security a T worker does). Here are the base salary ranges for the major employment categories among the T’s bus maintenance personnel:

Fueler $34964 – $43014
Machinist $57262 – $77750
Car cleaner $60278
Painter/Carpenter $82035
Foreman $87672 – $100672
Sheet metal $89669
Pipefitter $94994
Wireperson $95617

On average, employees earned 119% of their 2015 base salaries in 2014. Why? As mentioned before, overtime. Could the T hire more employees and pay less overtime? Probably. But if they did, there’d be outcry from Pioneer about how even more workers were getting the platinum-level union compensation package and benefits. They want to privatize operations, and assure lower pay and benefits for everyone. Is that a way to treat public employees? I’m all for accountability, but it shouldn’t require these sort of threats.

To add it all together …

Pioneer’s top line statement—that the T could have saved $250 million in bus maintenance costs over 6 years—is specious. They massage and manipulate data, but it doesn’t stand up to any scrutiny. Does the T have some of the highest maintenance costs around? Yes, it does (although New York is higher still). Some of this is due to the operating environment, some is due to local costs of living, some is due to the lack of appropriate facilities and, yes, some may be due to the performance of MBTA employees. The T could probably reduce it’s overall maintenance expenditures with different management practices, but to insist that those are put in to place without any funding (for instance: to insist on better management while many repairs take place in outdated and/or temporary facilities) is just silly. To use blatantly and obviously manipulated numbers to make this argument is disingenuous, and, once examined, the Pioneer Institute’s argument falls apart. There is very little defense for this sort of sloppy “analysis,” and it certainly should not be relied upon to make policy changes.

Using several different—and more honest—metrics, we can explain away much of that difference:

  • 60% of the Pioneer Institute report’s difference in costs arise from their suspicious definition of “peer agency.” Of the $250 million that they claim the T could save, that accounts for $150 million of it.
  • Just by using maintenance cost per hour instead of maintenance cost per mile, the difference is 20% lower.
  • In comparison to the Twin Cities, we can explain 35% to 75% of the difference in costs between the two systems.
The Pioneer Institute had an agenda: to reduce the power of public employee unions and privatize all that they can. Their position may well be “why spend time and money figuring out how to reform our management when we can just outsource it and let the magic of free market competition figure all that out for us, via a contractor, now?” But without good data on other systems (Has any large system privatized and had major cost benefits? They don’t say.) this is a hollow sentiment. And the MBTA’s current contractors have been doing a less-than-stellar job of late: the publicly run subway operations recovered from the blizzards far more quickly than the private-contract Commuter Rail system.

Pioneer saw a well-compensated set of individuals, and tried to make a case that they are lazy, that they have a poor work ethic, and that compared to their peers in other cities, that they have poor work quality. It’s true that MBTA costs are higher than many other agencies, and the T could probably implement policies to bring costs down to some degree. Upon further examination, however, it is the “research” coming from the Pioneer Institute that is low quality, not the work of the maintenance department at the MBTA.

Jim Stergios is bad at math

Update: It turns out, the Pioneer Institute as a whole is bad at math.

Jim Stergios, of the Koch-funded Pioneer Institute (edit: NOT the author of the discredited absenteeism report; my apologies to them for that insinuation, it’s only that, you know, Baker ran the Pioneer Institute, and the report was used by his commission had similar issues to other Pioneer reports), uses a lot of numbers to try to set up his arguments in response to a piece by Jim Aloisi. The problem? He uses numbers which are very convenient for him, ignoring longer trends which make his argument far, far more flimsy, and in many cases, completely refuted. He starts with an agenda, he warps data to make it fit that agenda. The problem is, the data tell a much, much different story.

• First, he references that in 2011 and 2012, 8 million commuter rail passengers were inconvenienced by late or delayed trains. That’s a big number. 8 million! But he doesn’t have a denominator. A big number without a denominator is meaningless (which was what most of the hubbub about the absenteeism report was about.) How many trips are there, annually, on the T’s commuter rail? If it’s 16 million, that’s a lot of delays. If it’s 80 million, it’s not quite as much. The answer? In 2011 and 2012, there were about 75,000,000 trips on Commuter Rail. So 8 million represents about 10%. Could this number be improved? Certainly. But without a denominator, this is a scare tactic: a number with no context. He claims that this resulted in a loss of ridership and revenue. But without any other years to compare it to (Were delays better or worse in 2006? He doesn’t say.), those claims are specious, at best.

(Vertical lines show locations of system expansions.)

• Then there’s this:

Notwithstanding the fact that the MBTA added more commuter track miles than other major transit systems in the country over the past 25 years, quickly raised fares and continued substandard service led, remarkably, to a decline of 13 percent in commuter rail ridership from 2003 to 2013.

Can you spot the incongruity there? Stergios assails the T’s expansion over 25 years, but is only concerned about it’s ridership over 10. It turns out that before 2003, the T was the fastest-growing commuter rail system in the country. If you look at the period from 1988 to 2013, T commuter rail ridership more than doubled. Even if you exclude extensions, at stations open in 1988 and 2013, it grew 65%. I made the argument that T ridership is hampered by high fares, and stand by that conjecture. In fact, for trips between 18 and 27 miles, the T has the highest fares of any commuter rail system in the country (this will be explored in depth in a later post).

It’s almost as if the investment in commuter rail in the 1980s and 1990s paid dividends in ridership during that time. But for an institute that wants to cut investment in transit, those data are very inconvenient. So they choose to ignore them. Thus, his data are misleading at best, and borderline fraudulent at worst.

• So he’s stepped in it already, but then he links to an article from his policy “research” institute that allows him to step in it some more. That article goes to great lengths about how, between 2003 and 2013, the T was the only major commuter rail system to lose ridership. You know what, I can’t argue with that. During that time period, the T did lose ridership, while other agencies gained. Again, I contend that it’s due to fare policy, but we each have our ideas why. But notice how he again very conveniently picks 2003 as his start date, which was the highest ridership on record. By doing so without showing any other data, he suggests that the T has underperformed other commuter rail networks. Let’s see if that is actually the case.

Annual ridership for SEPTA, Metra, MBTA

The two most similar commuter rail networks—with multiple legacy lines feeding the city center—are SEPTA in Philadelphia and Metra in Chicago. They also (conveniently for me, this time) have ridership data back as far as 1980. (Here’s SEPTA, here’s Metra, which I estimated from a chart but is exact enough for this post. Unlike Pioneer I don’t obscure my data sources; all MBTA data came from the Blue Books available on the T’s web site.) And, yes, the MBTA has had stagnant ridership in recent years, while SEPTA and Metra have both trended upwards. (This is also the case with New York’s commuter railroads, as well as Caltrain and MARC.) So there is certainly a case to be made that the MBTA’s commuter rail networks has been a laggard in recent years. This is likely due to a variety of factors, including stagnant service levels (SEPTA has, in recent years, been adding service), increasing fares (both Metra and SEPTA have lower fares per mile) and equipment and trackage which has been allowed to fall in to disrepair (SEPTA has invested heavily in their physical plant recently, and Metra runs on freight lines which have kept their tracks in good working order).

But the chart above is only one way to look at these data. Another is to normalize everything by an arbitrary year. I used 1988 (left), because Stergios likes to look back that far (sometimes). But for fun, I also made a chart that goes back to 1979 (right), because that’s the first year I have data from for all three systems.

Feel free to click to enlarge. The 1988 chart shows how, in the past 25 years, ridership on the MBTA Commuter Rail system has far outpaced SEPTA or Metra, growing by more than double while the others grew at a much slower rate. Go back to 1979, and the T has more than quadrupled, while, after falling off in the early 1980s, Metra and SEPTA only recently surpassed gas crisis ridership levels. In fact, if you look back to most any year but 2003, you get a very different picture. But, again, Pioneer’s “research” is picking and choosing numbers to fit their narrative, but not to show what actually happened.

• But wait, there’s more. He also claims that the MBTA has added more commuter track than any other system in the country in the past 25 years. Here he’s not fudging numbers, he’s just plain wrong. The T operates 394 miles of commuter rail. Metrolink, in Los Angeles, operates 388 miles. However, Metrolink began operation in 1992 which—let me get out my abacus, carry the 2—is only 23 years ago. So in the past 23 years they’ve added 388 miles. I’m not sure to the decimal of the amount the T has added (it’s about 145 miles over that time, of which at least 20 is in, and paid for by, Rhode Island) but it is certainly less than 388; even in 1988 the T operated more than, say, a shuttle from North Station to West Medford.

• Stergios also references his bus maintenance study, but that study borderlines on laughable, and may also be the subject of a separate post. Of several data irregularities there, the most glaring are the comparisons that the Pioneer Institute draws from the most comparable bus systems. For example, the list of most comparable bus systems to the T’s includes many systems in warm climates with low living expenses and ridership 1/10th of the T. Are we surprised that the transit authority in El Paso or San Bernardino has lower costs?

They make a major comparison to MetroTransit in the Twin Cities, an agency that also maintains buses in a colder climate. But nowhere in the report do the point out that while they have the same number of buses, the T carries twice as many passengers, and therefore, twice as many passengers per bus. This means that the T runs many more buses at or over capacity. A bus crammed with 75 passengers on board carries about 30% of its total weight in passengers, putting much more stress on not just the motor, but the air bags, axles, struts, tires and other equipment. (Imagine loading a Toyota Corolla with five 180 pound people and 500 pounds more in the trunk and a roof box. That’s what the T asks much of it’s fleet to do several times per day.) Many of the T’s bus routes run at this capacity on a daily basis. Only a few MetroTransit routes do, and often over longer distances. For instance, the MetroTransit Route 5 has comparable ridership to the T’s #1 bus, but its route is three times longer, meaning that the bus is not full nearly as often.

It is opaque as to how the Pioneer Institute chose other comparable agencies, but they often talk about the 79 other “large” agencies. However, the T is one of the top 10 agencies, and comparing it to an agency with as many passengers a day in total as the #1, #39 and #66 buses carry makes no sense. Is there a correlation between bus maintenance costs and overall ridership? They don’t bother to find out, and continue with these false, apples-to-oranges comparisons. The only comparable agencies in their database of the 20 most-similar systems are WMATA in DC and Muni in San Francisco (and even this is not apples-to-apples; WMATA carries fewer passengers per bus while San Francisco has no winter weather). Those agencies’ costs perfectly bracket the T’s maintenance costs per mile: the T is $3.80, WMATA is $3.20 and Muni $4.40. Compared to those agencies, the T is about where it should be.

Stergios claims that if the T operated with the efficiency of the average of these “comparable agencies” it would save $40 million a year or more. But if it operated with the efficiency of the agency in San Francisco? The T would actually spend more money. This whole study comes apart if you pull any one of many loose threads. That it is even in the discussion shows how picking only very particular data can make pretty much any point. What’s sad is that the legislature and governor bought it hook, line and sinker.

• There’s the stuff that’s just plain wrong. He claims that:

MBTA Board of Directors inexplicably authorized $47 million to purchase the Pittsfield-to-Connecticut Housatonic line

Really? The T is buying rail lines in Berkshire County? That sounds a lot like a MassDOT project, and indeed it is. They’re related, certainly, but that’s not money coming from the T’s pot. This is just careless.

WGBH fancies itself a news organization, and, as such, should have a fact checking department. Most of Stergios’s article does not pass even the slightest sniff test. GBH should be ashamed for publishing this article full of half truths at best, and several outright lies. As for Stergios and the Pioneer Institute? Anything that comes from them is immediately suspect, and usually, when examined, mostly false. They should crawl back in to their hole until they can present data with a straight face.