« the metro as metaphor | Main | journalists! network restructuring stories can be interesting, and clear »


Feed You can follow this conversation by subscribing to the comment feed for this post.

James Wong

I think that there is a well balanced argument here, but disagree with both your proposal for the second funded pot of money for coverage/welfare routes. By identifying these separately, it creates a larger divide in the minds of riders and taxpayers. It allows for targeted service reductions or funding cuts where the impact will be much greater because of the bare-bones operations for many of these lines. (cutting 10% of the budget of a line with 2 buses operating may mean a 50% cut in service when bus 2 simply cannot run).

We must live in the reality of transit agencies today, but I'm nervous about this particular approach.


In some ways this bears a distinct similarity to how budgeting for subsidies works in the UK since privatization. I would very much like to see more systems produce more granular budgeting, and documenting what is self supporting and what isn't is definitely useful. That said, I do share the concerns of those who suggest that this approach would make systems more vulnerable to cuts and make the impact of those cuts larger.

On the other hand their is a definite potential for this kind of examination to provide real guidance as to how much service a community can realistically provide. I have seen a number of occasions where what I suspect are profitable services have been cut to shift funds to the "equity/mobility/coverage/whatever" network.

My bigger concern would be how much overhead is attached to this level of granularity in budgeting, especially when one does consider the fuzziness of determining borders between various types of services and the real "profitability" of individual services in a flat fare (or close) network. My suggestion is that for most communities it would be more appropriate to produce a one time (hopefully recurring, but on a longish timeframe) report studying the network as a whole rather than expecting an ongoing assessment of the profitability of individual services. This would, IMO be especially usual for determining things what you are actually doing with funding, answering questions like how much of the budget really goes to allowing unprofitable service and how much subsidizes lower fares.

Ideally such a package could even include something like a localized examination of the actual subsidy to other modes, helping to guide the funding decisions for transit in terms of equalization. This kind of comprehensive network study, while unquestionably a massive project for almost any agency would, IMO, fit well into the type of work that goes on in many official plans and similar documents, and could (for those places really unable to handle the scope of work) be well suited to significant work from consultants from outside the agency or community.

Stephen Smith

If we propose a free-market view in which transit should be breaking even, well, I'd like to see this as well in a perfect world, but that would be a world in which government isn't heavily subsidizing transit's competitor, the private car -- not just through road expenditures but more massively through such interventions minimum parking requirements and petroleum-based foreign policy.

Two things about this:

1. Is American foreign policy really a subsidy to the automobile? It seems like this is only true if you believe that the US is a stabilizing force in the world, which I do not believe is true – and I suspect you don't either. Sure, right now in the short-term it might be stabilizing (all hell would break lose tomorrow if the US announced it was withdrawling from Iraq and Afghanistan and reaclling its 5th Fleet from Bahrain), but in the long term this doesn't seem to be the case. If the US had pursued a completely non-interventionist role in the world after WWII, the price of oil would almost surely be lower right now, given that the high prices seem more to do with global instability than a growing shortage of oil.

2. It seems like we do actually have two instances where automobiles are not massively subsidized: Europe and Japan. Both have very high gas prices to cover the cost of the roads, and neither have really bad pro-parking policies (in fact, Europe has a lot of anti-parking ones). And of these two – Europe and Japan – which is doing better? I'd have to say Japan, which also happens to be the one where

3. Keeping in line with number 2, it seems to me that in Europe and Asia, the problem isn't so much subsidized cars, but too many restrictions on railroad companies/agencies and the density they need to thrive. In Europe this plays out with vehemently anti-growth old city cores and the protectionism that keeps them away from lighter Japanese train technology, and in Japan it plays out in fare regulation and trade/labor protectionism that keeps railroads away from cost-saving European automation technology. (Of course Japanese rail for the most part actually isn't subsidized, but the crowding on both trains and housing is bad enough that we might be able to say that something isn't quite right.)

4. …which brings me to my last point: the so-called "welfare" lines, in the case of Japan, can actually be anti-rail in that they support regions that otherwise would not survive without links to the Tokyo-Nagoya-Osaka megalopolis. In the US, cities might die if we severed the link between city and suburb, but in Japan, it would almost certainly be the other way around.


Stephen Smith

Er…I guess that was 4 things. I just couldn't stop myself!


I would further divide (3) into the following:

- 3a: Low-ridership bus routes which run parallel to and within walking distance of more important bus routes (categories 1 and 2). Cutting these routes will result in longer walking distance to bus stops.
- 3b: Low-ridership routes which run in isolated areas where there is no alternative route within walking distance. Cutting these routes will result in areas having no transit service.

If funding is scarce than type 3a routes should be subject to service cuts (reducing frequencies, reducing span of service, or eliminating entirely) before type 3b routes.

David M

It's an interesting discussion. In BC, as Jarrett is aware, we're having a similar discussion about the way our ferries are funded. Our current right-thinking government created rules that prevent the profitable routes from subsidising the money losing routes. Over the past 8 years, it has resulted in dramatic fare increases, lower ridership and currently it is threatening the viability of the ferry system - there's more to it in regard to funding.

I think ultimately, we have to or improve how we look at transportation and land use together. Buses and rail transit are only part of the picture. Roads are there too. We don't close a road because it's under used, or charge more to use it. We know that road feeds to another road creating a network. Transit is the same. The profitable routes for the most part probably won;t be profitable if they don't have the support of the feeder and links to remote locations.

Finally, I think it depends on what you want your transit agency to be. Is it simply to move people who decide to use it (the UK outside of London privatised model), or is it to shape the urban fabric and support higher densities and a living environment where the car is not king (the Vancouver model).


Have anyone considered running peak service by using commuters as drivers? It could be a way to enhance rush hr frequency and reduce the need for split shift. Professional commuter bus drivers could run frequent mid day express service to/from airports.

Another way is for the likes of Coach USA and Greyhound to run a hybrid service:
Rush hr commuter service.
Mid-day/evening short distance intercity service.

Jarrett at HumanTransit.org

Al.  Yes.  See the "worker-driver" services of Kitsap and Mason Counties, Washington.

Stephen Smith

Our current right-thinking government created rules that prevent the profitable routes from subsidising the money losing routes.

Japan has similar rules, and I wanna say that France (and maybe the UK?) does too, but I'm not sure about that. Seems to work pretty well for them.

Andre Lot

@David M, Stepehn Smith: Italy also adopted these distinctions.

I think some people are afraid that a mere accounting exercise to calculate costs of each service is made. It's a bit sad to read some comments that are playing into the political side on the blogosphere, instead of the technical ones when it comes to issues like these.

There is a case, on a merely financial budgeting service, to segregate services that are offered on "welfare" or "universal coverage" grounds, and the regular network operating and designed under efficiency principles.

What can't be forgotten is that if budgets are to be stretched out to provide everyone in a given area with some lifeline service without enough money to even cover it, the likely result will be an inefficient network that in a 180min span will take you anywhere in a - say - 25-mile radius, but will be so fragmented and useless except for the very poor, retirees etc. that it will become a "transit of last resort", as usually referenced here.

So I think it is positive to discuss costs of providing universal coverage on a short-term basis, instead of loading agency budgets with these costs. Ideally, a good process would be:

1. Define a basic level of coverage service for a given area in terms of accessibility, frequency, distance to nearest station/stop, average travel time for a set of pre-defined locations in the area.

2. Design an efficient, "breaking even" network without regard for #1, one that can operate and pay its costs through fares or long-term funding like sales tax transit add-ons.

3. With the "breaking even network in place", calculate the marginal cost of providing basic service coverage (#1) for areas not services properly by the network designed on #2. Calculate the costs for this.

4. If marginal costs of coverage-based services are greater than what the municipality/government wants to pay, reduce the requirements of coverage and redesign #1, check coverage provided by #2 and re-calculate costs for #3 until the available "coverage budget" is met with estimated costs.

A second set of financial problems for many transit agencies are special fares and concessions. Student and senior discounts concessions (or free passes) should be financed, annually, out of the welfare budget, not thrown into the agencies as lower income.

My rationale is that, in most cases, discounts and concessions are given to those less likely to have easy access to cars. Full fares for junior high-school students would certainly not mean they'd be jumping on cars they can't legally drive yet. At most, they will stay put. But if they really need to travel, they can pay (or their parents) full fares.

Tom West

If we apply this to transit, we shoudl apply it to roads.
When a car drives down any given road, it uses fuel, on which the user has paid fuel tax. Therefore, every road generates a certain amount of tax revenue, which is (roughly) proportional to traffic levels. (Congested roads will generate more tax per vehicle!)

We can apply Levinson's three categories to roads - those which are profitable in their own right, those which make others profitable, and those which loose money.
Carrying his logic through, those unproftiable roads will probably be rural roads and cul-de-sacs in low density residential neighbourhoods. It may also include collectors and arterials, depending on your local fuel tax levels.
Aren't those "welfare" rather "transporation" services?

... and yet most people accept that roads should be funded out of general revenue, because they know how utterly vital a transportation network is to the economy. Transit is a means of transportation, and done right, it results in lower costs to the economy than cars.


Having worked in transit finance in the US (and gone through Mr. Levinson's institution), I like the Levinson model. Public agencies in the US are torn between an efficient network and a politically palatable network, and often make concessions to political considerations to stay in the good graces of legislators and executives who direct funding. I would have loved to present political leaders with a bill for subsidy based on some empirical measure of the cost of service, but as a political organization, the agency lumped all costs together and presented the budget as a whole. I wholeheartedly dislike the "welfare" connotation of low-use feeder services, though. These lines are almost always political animals created to appease vocal and focused groups of constituents, not arbitrary social service offerings.


Levinson is a U of M civil engineering professor who writes the Transportationist blog, in addition to his scholarly output.

Somehow, I suspect that he wouldn't be making similar arguments about the public street network. His use of the word "utility" to describe transit is an interesting framing; while utilities are meant to be utilized (duh) and thus are presumably useful; there's a longstanding bias that these things aren't deserving of public subsidy, whereas roads presumably are.

The problem with his suggestion--and you alude to it but don't come right out and say it--is that it invariably leads to a transit death spiral. By the same token, it wouldn't make sense to build a single paved road in isolation and expect this one road to meet a rigorous cost-benefit analysis, if nobody had cars to drive on them, and there wasn't a large number of garages, filling stations, and dealerships for buying, fueling, and fixing cars.


Many "social service" lines aren't really meant to serve a constitutency of (pardon the phrase) transit-dependent riders; many exist instead to meet some minimum-service guarantee or expand/preserve the service footprint for tax reasons.

Alexander Craghead

While I disagree with Levinson's premise of profit as the primary decision making method, I have to note that his comments about the social welfare or equity routes may have merit but for different reasons. Because they are essentially poor quality service for the sake of token equity -- equity routes are rarely frequent -- they are the dogs of a transit agency. They are the mostly empty "ghost buses" that skeptics and non-riders see and thus brand all of the transit agency with. They make good transit agencies look disdunctional and bad.

They also fall into a fixed route trap that smaller transit agencies often get stuck in. Man a small town has tried to establish fixed route bus lines when they are far far too light density to make such routes work, and when they instead should have been using a paratransit model.

Rethinking equity service methods would be a huge step towards both improved transportation quality and improved image. Running almost entirely empty buses on equity routes serves no-one well.

david vartanoff

Both SF Muni and AC Transit have done this sort of analysis. In Muni's case the Transit Evaluation Project spent 18 months plus discovering that the major trunk routes were the most heavily used and proposed gutting the lightly used routes to re-deploy the resources on busier routes. What happened, however, was the light use lines were abolished and none of the touted improvements (were they window dressing?) were implemented. In AC Transit's case, too, the last decade has seen a continuing series of cutbacks. While AC published specifications for 3 classes of routes, current service patterns do not meet the declared standards and entire swaths of the "service district" are now abandoned.

Jarrett at HumanTransit.org

@David Vartanoff.  This stuff happens because of unexpected financial collapse.  US transit agencies are mostly very exposed to financial volatility because they rely on volatile sources like sales and payroll taxes.  Given the sudden need to make massive cuts, isn't it better that SF and AC had SOME policies in place to help them prioritise?  SF did not abandon large areas, btw.  It was done mostly by dropping routes that were within a short walk of other stronger lines.  See here: http://www.humantransit.org/2009/11/san-francisco-cuts-for-effectiveness.html 

Chris, Public Transport

I believe Los Angeles Metro is perhaps the prototypical agency who is actively retrenching to become profitable. Over the past thirty years many suburban routes have either been abandoned or given to other municipal operators to run, until today only a few bus routes exit the city limits of Los Angeles. As a result Metro has become one of the most efficient American bus systems, despite a very low fare structure.

I suppose an assumption would be that after these transit agencies retrenched that people who need to or wish to use transit would relocate to the area where transit service still existed. However, sometimes those above people need to access places that are controlled by persons who have no desire to ever use transit. If this retrenchment happened, how will transit riders be able to access suburban shopping malls and office parks unless this retrenchment happened simultaneously with zoning laws that required businesses to locate near transit lines?


Transportation is ultimately a discussion about cities.

Limiting the issue to the profitability of transit lines instead of comparing them as an alternative, at least, to roads, is, as Tom West indicates, a bit myopic.

What freeway or arterial has demonstrated a profit? There are obvious benefits to an economy and urban efficacy with a decent road system, but then again there are pitfalls, one of them, being the utter failure to account for their terrible, all-consuming legacy on human health, ecosystems, urban land base and public finances.


LA County as a community has sufficient critical mass to support a transit system all on its own--LACMTA doesn't need to provide services to (or collect taxes from) Orange County, etc., to have a viable operation.

That isn't true for many other urban transit agencies. If many employment opportunities are in the suburbs, and/or a significant part of the tax base, than busses and trains will need to run there. And if the suburbs are predominantly sprawlville, than the services in question are going to have low ridership and suck by pretty much any efficiency metric.


The advantage of providing coverage routes, is that they make it more viable for people to live without owning cars at all. I may rarely ride a particular route, but I won't need my own vehicle if I know it's there when I need it.

Terry LW

I had this argument with Wendell Cox about 15 years ago and still it goes on! My personal point of view is that transit provides both a transport and social function and that these need to be recognised and funded appropriately. In Australia we call these unprofitable routes Community Service Obligations (CSO), and we provide a marginal-cost subsidy to the transit operator from general revenue to cover their costs.

This system came from a dated view that transit is for people who can't afford anything else, while the able drive cars. While this is a little outmoded in thinking, because now we view transit as the superior economic and environmental option, the CSO idea is still valid. That is, if we want to effect mode shift towards transit to manage congestion, local pollution, city amenity and the like, then it is not the core job of service operators to provide a level of service that achieves that change while incurring a loss. Because, the car is the refuge of those who need mobility during the day, not just on the commute. And the largest subsidy is for between and off-peak services.

But, I have to note, that in Sydney moderate frequency, high coverage services all day are proving popular throughout the day and into the evening.

What now needs to be done is the analysis to see where break even is reached, and then just pay a CSO for marginal costs where a political and social decision is made to run additional services for wider public benefit. And I do understand that concept of loss-leading to build a market, but transit is rarely sufficiently profitable for the service provider to take a sustained hit.

For those interested, it is worth noting that operators in most of Australia are private, the system design and timetable is mandated by State Government agenices and there are a variety of operating contract models to subsidise the private sector providers.

In Sydney there is an argument for fine tuning the model to squeeze more service per dollar and to have a more effective incentive and penalty system based on performance (customer service as well as financial). But, on the whole, it is a pretty good system.


@david vartanoff

SF MUNI, which has taken the "Profitability doesn't matter" philosophy to its logical extreme, currently pays a a six figure salary to the average operator. There isn't a single MUNI efficiency improvement proposal that isn't seen as a grab bag for the union. If they can cut costs by 100 percent, they will take 99% and give 1% to the public. Worst transit agency ever.


@Andre Lot

"A second set of financial problems for many transit agencies are special fares and concessions. Student and senior discounts concessions (or free passes) should be financed, annually, out of the welfare budget, not thrown into the agencies as lower income."

This doesn't apply to free passes, but reduced fare passes can be seen as a good idea from a commercial standpoint, because they encourage travel for those with a lower reservation price. If a senior makes a trip at a fare of $1, but wouldn't at $2, that's $1 that the agency wouldn't have made without the discounts available.

Good evidence of this is in the UK, where legislation prevents direct subsidy of fares for passengers aged 16-64 of bus services that are provided commercially. In West Yorkshire, the discounted passes for university students and the unemployed are a commercial product of the passenger transport authority, who provide them in agreement with bus operators that participate because they believe that providing those tickets is commercially sound.


There's also a chicken-and-egg perspective: dense development requires transit and transit requires dense development. Sometimes it might be a good idea to subsidize a money-losing transit line if it enables dense development, which may make that money-loser profitable in the future.

Of course, this doesn't apply in most cases where there is no discussion of development patterns at all. In most cases, it seems the only question is whether to provide unprofitable bus service to a low-density area, or not to provide the service.

As such, I think that people should focus their energies on improving the service that a lot of people use already, and on getting rid of car-promoting policies like parking requirements in dense areas. If that's done, parking will become rarer and more expensive, while the pedestrian environment improves (as parking lots are developed for more profitable purposes like housing and commerce). As this (voluntary) process happens, the core lines become more important relative to cars, and even more heavily used, which can lead to increased frequencies and better finances. Finally, this has spillover effects to the feeder and even equity lines, as people going to and from those areas find the bus service cheaper or more convenient when traveling to or from the core area.

In the long run, the feeder service may develop into a core service, and the process may repeat. In short, allow the city to grow organically by providing transit where it makes the most sense, and by not stifling it by filling it with mandated parking lots.

Stephen Smith

Kind of disappointing to see all the "I bet David Levinson wouldn't apply these standards to roads!" – you're throwing up a straw man. If you have any evidence that David wouldn't say the same things about roads, then present it, but don't just make things up because it fits your preconceived notions of what a professor who advocates transit profitability thinks.


I think the same thinking should definitely be applied to roads. Some level of barebones road network should be in place in rural areas, but we should stop building new roads or bypasses or widening highways in rural areas where few people actually use the roads.


Professor Levinson responds here.


"They should be considered public utilities..."

Right then, let's talk rural electrification. Electrical utilities, public or private, are regulated to ensure among other things that electricity is provided at relatively equal and affordable rates to both people who live right in the city and require relatively little infrastructure-per-consumer and people who live far out in the country and require relatively a great deal of infrastructure-per-consumer. Universal availability of electricity is viewed as a societal good and a basic requirement of civilization. Apparently public transit (like health care in the US, though I range off topic...) not so much.


From Prof. Levinson's blog:

"I didn't think equity, welfare and redistribution were derisive terms, though I am sure there are more politically correct terms of art to talk about these things to avoid bruising egos..."

Egos are not the point. You cannot simply use any term you want without recognizing its political implications, or sneer at anyone who assumes you meant it in the sense it is commonly understood - at least not without some explanation. Well, you can, but I don't think it, or snide put-downs of those who raise questions, helps get your point across.


"Egos are not the point. You cannot simply use any term you want without recognizing its political implications, or sneer at anyone who assumes you meant it in the sense it is commonly understood - at least not without some explanation. Well, you can, but I don't think it, or snide put-downs of those who raise questions, helps get your point across."

His point was well made, using words that have actual dictionary meanings. That some people are ubersensitive and read into things that don't exist is not his fault.


Re: electricity. Isn't/wasn't electrical service subsidized by government in some places (Tennessee Valley Authority, etc.)?

As far as rural transit (in the US): some places have got it, some don't...there are a few county, state and/or Federal subsidized systems in rural areas, but not everywhere. And then we get into Amtrak, Greyhound, airlines (whether subsidized through the Essential Air Service or not) and we get a picture of rural transit as a hit-or-miss affair...no real top-level planning.

As for urban/suburban transit....many times, routes are retained for political reasons, not necessarily because of ridership needs. See: http://la.streetsblog.org/richards-on-metro-board-agenda-item-54-92211/


This does sound like similar thinking to UK practise. Here, around 80% of bus mileage is operated commercially without direct local council subsidy. It is 25 years this month that deregulation took place (26 Oct 1986). On that day subsidies ceased and bus companies started to operate their commercial networks, registered in the run up to what was known as 'D-Day'. Local councils had to tender additional services to fill the gaps. By focussing subsidies on the routes that actually needed them local authorities no longer were faced by annual blanket subsidy demands by the various subsidiaries of the then state owned National Bus Company and Scottish Bus Company to maintain large networks. These cash savings were a bigger driver for the UK Dept of Transport for deregulation than competition for its own sake.


Transit agencies can be profitable. The question is only if public transit is - or should.

In some cases here in Germany a city itself (like Frankfurt) decides what lines should be operated and in what frequency/quality/with what rolling stock. The same is done with regional rail service (for Bavaria this is done by the BEG, Bavarian Rail, a government owned institution). The transit operators are bidding then to run the service on one or several lines saying how much money they would need. In that way, the commuter rail lines for example in Berlin or Munich are generating profits. Or the transit operators in Frankfurt.

Is that what is paid to the operators by the city (minus the farebox revenue) subsidies? Or simply a price that is paid for a public service? Would you say the police, military or firemen get subsidies or must be profitable?

In other cities like Munich the operational costs and a lot of the investment costs are covered purely by farebox revenue. A high ridership in a dense city with high quality service makes this possible. Most bigger cities in Germany are above 75%-90% of cost coverage by farebox revenue. A profit could be generated if the prices for the tickets could be set by the agencies and not by political decision makers. So if is possible to have profits in public transit and not to have some is a political decision to make tickets cheaper.

In some cities public transit is a part of the municipal owned public utilities company providing also water, electricity, public swimming bathes etc. - and the losses in public transit covered by profits in electricity, gas and so on.

It is all in the eye of the beholder if money paid for public service is a subsidy:
Generally the society, represented by its politicians, decides what they want to reach. If public transit is seen as a service making it possible for a society to work, making cities more liveable, giving relief from congestion, giving access to all, reducing pollution or providing services for a lower price for social reasons then it is not a subsidy but a price you pay for that service. A price a society pays for it. Like you pay for police or military to get safety. (In the US the politicians are even ready to ruin the whole county by total overspending on military and security following the bad example of the Soviet Union but do not want to spend money for infrastructure). So you could say you pay a certain amount of money to have a center that is not collapsing by car traffic, and you spend this on public transit. A subsidy or a way to keep a city from collapsing?

Of course you could expect that public transit is financed by its riders and not care about the positive effects public transit can have for a society. As 400 years ago it was common that armies were financed by the war activities and had to generate profits from looting enemy territory. Or you could shut down police and let private secutity take it over, who want safety has to pay for it. A concept that is popular now in countries like Iraq.

Transit agencies are profitable. The question is only what a service a society wants and in what way the bill is paid.

But what certainly would help is a certain efficiency. Often you can get more service with the same money if you have some smart planning.

And no, service cuts are no general solution. In our expirience here in most a cases a line running every 20 minutes is ignored by most potential riders. When you have a minimum frequency of 10 minutes the ridership usuallly explodes. The same is true for better ifrastructure, rail projects, bus prioritization and so on.

The comments to this entry are closed.

the firm

Jarrett is now in ...

Related Posts Plugin for WordPress, Blogger...
Related Posts Plugin for WordPress, Blogger...