The High Cost of Parking at Caltech

Like Pasadena as a whole, Caltech’s population is growing, but we cannot expand geographically. This means both Caltech and Pasadena must increase density by building vertically or packing our buildings more closely together. Pasadena, much to the dismay of some long time residents who fondly remember the days when Orange Grove Blvd. actually passed through orange groves, now has six story live-work “transit oriented developments” sprouting up around the major business districts, within walking distance of the light rail. Similarly, Caltech has a new Chemistry building appearing between BBB and Noyes, a new Astronomy building where there used to be a surface parking lot next to Keith Spalding, and a new CS building rising up between Facilities and Avery House. Those new buildings will mean more people, and probably more cars, coming to Caltech every day.  They have to go somewhere, and our neighbors have made it clear to the City that parking them on the street is unacceptable.  Those new commuters will largely be parking in the recently completed subterranean garage under the athletic field.  However, this kind of solution to our parking demand has a cost, and I think we need to understand just how large it is in order to have a reasonable discussion about whether it’s the best solution going forward.

How much does a parking space cost?

According to the Caltech Master Plan and John Onderdonk in Facilities, the costs associated with building and maintaining our four current parking structures look like this:

Construction Costs:
Holliston + Wilson 1 + Wilson 2	$13.8M
California (beneath athletic field):	$16.3M
Operational Costs:
Typical Electrical Utility Cost:	$120,000/yr
Typical Cleaning Cost:	$20,000/yr
Opportunity Costs:
Footprint of above grade structures:	120,000 ft2
Neighboring property Value:	$120/ft2
Estimated Replacement Cost:	$14.4M

The construction (or capital) cost is the money we had to lay out up front to build the structures.  Operational costs are ongoing expenses associated with using and maintaining the structures.  The opportunity costs are a measure of what we have forgone in exchange for the structures.  We could have built laboratories, student housing, or administrative buildings where they stand.  The opportunity costs listed are only for the above ground Wilson and Holliston parking structures, since we still get to use the athletic field with a parking structure underneath it.  In practice it might be difficult to purchase adjacent land at any price.  In exchange for these investments, the Institute has gotten a total of 2,009 parking spaces:

Parking Spaces:
Wilson 1:	437
Wilson 2:	443
Holliston:	440
California:	689

An unreserved Caltech parking permit currently costs $40/month, or $480/year.  Reserved spots cost more, and should, because they can’t be dynamically re-allocated when you aren’t using them, but for the moment let’s assume all spots are unreserved.

If we want to compare capital (up front) costs to cash flow (operating expenses, and revenue from parking permits), we need to know something about the time value of money.  We can either convert future cash flows to their net present value, or we can convert the capital costs to an equivalent break-even cash flow.  In order to do either of these conversions, we need to know the discount rate, which is a measure of how much more a given lump sum of cash is worth now than at some time in the future.  If you’re borrowing money, you can think of the discount rate as the interest rate you’d be paying on the loan.  If you’re spending your own money, the discount rate is the rate of return you would have made on that money if you’d invested it instead.  To first order, all you do is multiply or divide capital cost by the discount rate, depending on which direction you’re converting.  So for example, $20,000 today is equivalent to $1,000/year if the discount rate is 5%.  This is not strictly true, but it’s a good approximation for relatively small discount rates, and things that have long lifetimes, like parking structures.  For the moment, let’s call the annual discount rate D, and the average annual cost of a single parking spot in one of the structures on campus C:

C = (D*(capital costs + opportunity costs) + annual expenses)/(# of spaces)

C = (D*(4.7e7)+1.4e5)/(2009)

Assuming C=$480, we can solve and find D=1.75%.  This rate is so low as to be implausible, suggesting that Caltech is subsidizing parking.  Digging through Caltech’s financial reports from the last few years, one finds that the construction of our parking structures has largely been financed using bonds issued in cooperation with the California Educational Facilities Authority (you can even see what CEFA staff had to say about Caltech’s application).  Debt financing was probably used because inglorious infrastructure is a hard sell to our noble benefactors.  Can you imagine the Gordon & Betty Moore, or Arnold & Mabel Beckman… Parking Lot? More importantly, Gordon, Betty, Arnold, and Mabel would probably all realize that a parking structure does not directly support Caltech’s core scientific and engineering mission. It is something that we might as an institution do well to do without, if it means taking resources (real estate and cash) away from our core mission.  The interest rate you’ll pay on debt (especially as a tax exempt organization) is also much lower than the expected rate of return on investments, so given that drumming up alumni funds to pay specifically for a new parking structure was unlikely, it certainly made more sense to take out a loan than to spend endowment money (at least if you think that stock markets go up over time…).

The tax-exempt bonds that Caltech issued through CEFA paid between 4.25% and 5.00% interest.  Let’s call it 4.5% interest on average, so that’s D, implying that the cost of providing a parking spot at Caltech is about $1122/year, more than twice what Caltech charges.  One might legitimately ask why California offers a tax exemption for “educational” bonds used to construct parking structures.  Without that benefit, Caltech probably would have had to offer more like a 6% bond yield, making the true unsubsidized cost of providing a parking spot nearly $1500/year.  The Institute was able to do the equivalent of refinancing the bonds in 2006, resulting in an effective interest rate of only 3.6%, and at D=3.6%, a parking spot still costs $943/year (The refinancing was done using an interest rate swap, as outlined in this cartoon of the deal.  Let’s hope that the Bank of New York remains solvent!).  However, for the purposes of this discussion I’m going to stick with D=4.5% for a couple of reasons.  First, when Caltech had to make the decision about allocating its funds, it was willing to make a commitment to the 4.5% interest rate.  Ostensibly at the time $1122/year accurately reflected Caltech’s valuation of having an additional parking spot, otherwise we wouldn’t have built the structures.  Second, historically 3.6% is a very low rate.  Looking forward, especially given present uncertainties in the credit markets, we should not be planning on getting that rate the next time we need to build a parking structure.

How much parking do we need?

Accepting that a parking space costs $1122/year to provide, and that Caltech only charges $480/year for a permit, what we have in effect is a net $642/year or $53.50/month subsidy to drivers.  It is instructive to compare this to the monthly subsidies offered on other methods of transit:

Mode	Subsidy
Vanpool	$50
Metrolink Train	$50
Metro Pass (bus/light rail)	$25
Bicycling	$0
Walking	$0

You’ll notice that driving is actually the most subsidized mode of transportation, though vanpools come close to competing with driving on an even playing field.  Metrolink doesn’t really count since it only gets you as close at Union Station in downtown LA.  The odd thing is, these subsidies are supposedly meant to incentivize the use of alternative transportation options.  In reality, the biggest financial incentive Caltech provides is for the thing that it doesn’t want you to do: drive, and the smallest incentives ($0) are for walking and biking, which require virtually no investment on the institute’s part, and don’t generate congestion or pollution.  Additionally if the point of the subsidy is to make one option more attractive than another, it doesn’t make any sense to subsidize everything.  Doing so makes it impossible economically to differentiate one option from the others.  Furthermore, offering more money for the more expensive options hides the real relative costs, which are then being borne by the subsidizing party.

The obvious thing to do if we actually wanted to discourage people from driving, or even if we just want to stop wasting money providing subsidized parking, is to charge the full cost of the parking spots, ~$100/month, and let people adjust their transportation decisions with full knowledge of the costs involved.  If you’re one of the many car-loving southern Californians, you might question the aim of discouraging driving — fair enough, but charging $100/month wouldn’t be punitive.  It would just be a removal of our present incentive, and I’d be happy to accept abolishing the public transit subsidies Caltech offers as well, just to see what transportation decisions people make.  It seems like this unperturbed state should be the baseline we start from in designing any incentives, or deciding how much parking to provide in the first place.

Unfortunately, it turns out this isn’t how parking works at all.  How much parking Caltech provides is not Caltech’s decision.  It’s law, codified in the Caltech Master Plan, which is actually a city ordinance.  In 1986, prior to the construction of the second phase of the Catalina apartments, the City performed a parking survey, and came up with the following parking ratios (PR = required spots/person) for various parts of the Caltech community, and applied them to our predicted future growth assuming that 60% of graduate students would live on campus and including a 10% buffer for visitors and vacancy.

Group	PR	Pop.	Demand
Off campus grad students:	0.44	480	209
On campus grad. students:	0.67	720	480
On campus undergrads:	0.4	845	338
Faculty and Staff:	0.5	2430	1215
Visitors/Vacancy:	—	—	224
Total:	0.55	4475	2466

Should we expect that the driving habits of the Caltech community in 1986 represent a good estimate of our behavior looking forward to 2030?  In 1986 there was no Gold Line, and no Pasadena ARTS bus.  Adjusted for inflation, nationwide gas prices in 1986 averaged about $1.75 versus $3.00 so far in 2008, while median household incomes and vehicle fuel efficiency remained nearly unchanged ($45,000 vs. $50,000 and 22.0 vs. 20.2 mpg, respectively).  James Hansen had not yet testified to Congress about the dangers of climate change, Pasadena hadn’t labeled itself a green city, we had no Bicycle Master Plan, and Caltech did not have a Sustainability Council.

But probably more important than any of those changes is the fact that, in 1986, parking at Caltech was free.

What would global demand for crude oil be like if OPEC gave it away for free?  Would you take a ride to the space station if it were free?  If electricity was free, we’d probably use it to heat and cool our homes and never bother with any insulation, as was very nearly the case in the Pacific Northwest, after construction of the Grand Coulee Dam flooded the region with cheap power for which there had not previously been demand.  Unsurprisingly, the demand materialized.  Price is paramount when you’re trying to estimate demand, and that goes for parking just as much as anything else with a non-zero marginal cost of production.  Nevertheless, in the Master Plan it states:

Estimates of Caltech’s population growth, in conjunction with these ratios, indicate that the demand for parking might reach 2,500 spaces within the next 15 years. Analysis of existing parking and car ownership habits validates the demand ratios used to generate this number.

However, the usage of on-street parking could not be quantified during the surveys of parking lot utilization. This unknown factor suggests that the parking strategy should allow for the construction of more parking than might be suggested by these estimates, to account for the anticipated loss of all on-street parking.  For that reason, the proposed parking strategy will allow for the creation of approximately 3,500 parking spaces on-site, 1,000 more than the projected need.

So the plan of record was to provide as much as a 40% oversupply of a resource for which demand was estimated assuming a price of $0.  This kind of resource allocation depresses prices, and expectations of prices, which is why the $100/month price tag probably seems high to you.  The cost doesn’t go away, it just moves into the background, and you don’t have the option of making decisions based on it.  In a retail context, the cost of free parking is rolled into the prices of the goods and services you buy.  In a residential complex it increases the price of a unit of housing.  In Caltech’s case, it increases our operational overhead.  This hidden cost is real, and impossible to recoup if parking is underpriced, even if as an individual you choose not to own a car or drive. (These ideas and the data behind them are explored in more depth in UCLA professor Don Shoup‘s book The High Cost of Free Parking.)

If instead we charged users the true cost of parking, we’d probably need less of it. Much of the student demand would vanish, since if you live in Caltech housing, you don’t actually need to drive anywhere, and renting a car a couple of weekends a month to get away would be cheaper than paying for a parking spot (even before you add in the avoided costs of maintenance and depreciation).  Students would have an incentive to learn how to use our surprisingly good public transit network, or to use a bike to get around, or to get a smaller vehicle – maybe a Vespa.  A higher proportion of Caltech’s staff would end up being sourced from nearby, avoiding the time and expense associated with a long commute.  Or they would decide to take advantage of organized car and vanpools and public transportation.  We would exceed our obligations to the South Coast Air Quality Management District, and Pasadena’s Trip Reduction ordinance by wide margins.  We would have more real estate and more money to dedicate to our core mission, of expanding human knowledge and benefiting society through scientific research and education.  It would reduce the traffic congestion, greenhouse gas emissions, and other pollution resulting from the operations of the Institute.

You might reasonably point out that the construction and opportunity costs of our parking structures are unrecoverable.  What’s the point of charging everyone $100/month if all that will do is create a bunch of empty parking spots that don’t do anyone any good?  I did an informal survey of the four parking structures and found the following vacancy rates at 11am on a Tuesday during the academic year, suggesting that even at $40/month we may have surplus parking with our current campus population.  Of course, remember we have yet to add all of the people who will eventually inhabit the new Chemistry, Astronomy, and Information Science buildings:

Structure	Vacancies	Total Spots	Vacancy Rate
Wilson 1:	204	437	46%
Wilson 2:	65	443	15%
California:	154	689	22%
Holliston:	85	440	19%
Total:	508	2,009	25%

According to the Master Plan, we still have several parking structures yet to build: a 2 story, 250 space subterranean lot for visitors, to be constructed underneath the gateway plaza that will eventually serve as the main entrance to campus at the northern edge, off of Del Mar, another 400+ space above ground structure on Holliston between the fire station and the power plant, and another 2 story subterranean lot beneath the Athaneum tennis courts.  Yet, according to the Caltech website we already exceed the projected campus population from the Master Plan by a significant margin.  Altogether there are 6114 students, faculty and staff (not including JPL).  I was unable to find an exact number of parking spaces currently available on campus, but even if we are providing the entire proposed allocation of 3,500 spaces our current population would imply an overall PR of 0.57, which is very close to the observed overall average of 0.55 from the 1986 survey.  This makes me wonder if maybe we’ve already adjusted the campus population to take advantage of the 1,000 “extra” parking spaces that were written into the Master Plan.  The fact that at least 500 of the spaces were unoccupied means we’ve in the meantime reduced our average PR from 0.55 to 0.49, suggesting that we might be able support a larger ultimate campus population with less of a parking subsidy, if we continue to raise the price of a parking permit.  This assumes of course that we would prefer as an institution to have more students, researchers and faculty, and that growing furthers our mission.  The Master Plan has the entire northern margin of the campus slated for re-development eventually, and that has to mean a lot more people.  Wouldn’t it be more fair and more financially responsible for drivers to pay the entire cost of their parking?  Wouldn’t it be to our advantage to avoid constructing tens of millions of dollars worth of new parking at least until we know what the demand for it would be at its full cost?

A few modest proposals

What if over the next five to ten years, before building any new parking structures, we were to allow the price of a parking permit to float upwards along with demand, until it’s equivalent to the true cost of adding a new parking space?  In conjunction with detailed annual parking surveys, this would give us information about the elasticity of demand for parking within the different segments of the Caltech community, and would allow us to determine the allocation of resources between parking and academic facilities that results in just enough parking spaces, provided at cost, to accommodate our final population when the campus is entirely developed.  What would this be worth?

If we assume that the current plan is to continue subsidizing parking at $642/year per space, and that with this level of subsidy we can expect parking demand will require us to build the additional ~1,000 parking spaces listed in the Master Plan, for a total of ~4,500 spaces, that’s ultimately an operating loss of nearly $3 million per year.  Assuming a 4.5% discount rate, that loss is equivalent to a lump sum today of about $64 million.  Conversely, if charging the full price of parking on our existing spaces reduces demand enough that we can avoid having to build any more parking, we don’t have to spend that money, and we preserve about 40,000 square feet of real estate on campus that can be used for a new academic facility.  Just for reference, $3 million a year is equivalent to full scholarships including living expenses for 60 undergraduates, every year.  It’s roughly 40 grad students, or 30 post doctoral fellows, including benefits, every year, or it’s enough to cover 10 faculty positions, with some institutional start-up funds, every year.  It’s a new world class Beowulf cluster somewhere on campus every couple of years.  Alternatively, a lump sum of $64 million is more than the entire project cost of the recently constructed Broad Center for the Biological Sciences.  Should we really be willing to forgo these kinds of things in order to continue subsidizing parking?

What are the likely objections to market priced parking at Caltech?  The current recipients of the parking subsidy would probably complain loudly, because it would be such a departure from normal Southern California policy, but would they have a more substantial objection than their desire not to pay?  Would students who had been admitted to Caltech really decide not to come here because parking seemed expensive?   Or would they decide to live off campus nearby, or consider doing without a car?  Would potential new faculty reject a Caltech position because of parking?  If so, rather than offering them a “free” parking space, wouldn’t it be better to explicitly offer them some portion of the full cost of a parking space as a straight addition to their salary, and then let them decide if they really think the parking spot is worth it?  That way they would still have an economic incentive not to create demand for parking on campus, and any time the parking issue didn’t come up in negotiation, we’d avoid the demand for free.  In any case, a large proportion of students and faculty already end up living very close to Caltech, and whatever the social expectation may be, do not actually need to drive to get here.  The non-faculty staff would probably be the most significantly impacted, especially the least well paid among them, for whom $100/month would be a significant portion of their compensation.  In effect, we would be stating a hiring preference for those who live nearby and do not need to drive, or who are willing to come to campus via a carpool, vanpool, or public transportation.  Especially for most easily substituted employees, is that preference not in our best interest? Isn’t it also in the best interests of the person being hired?  If having a job at Caltech means one fewer cars in their household, they would save a lot more money than the price of a parking permit, and that makes a larger financial difference to someone working as custodial staff, or an entry level administrative assistant, than to a higher level manager.

It’s not our responsibility as an employer or institution to try and change the social norms surrounding car ownership and transportation in Southern California, but neither should it be our responsibility to actively subsidize those norms, which is what we do today.

I think the more significant objection to market priced parking would be likely to come from the City of Pasadena.  One plausible result of charging the full cost for parking would be that some people would drive anyway and avoid paying by parking in the surrounding neighborhoods.  This is unacceptable to nearby residents, and thus to the city government.  We could argue about whether that’s justified, but it is largely beyond our control.  In 1996, Pasadena enacted a parking permit regime in the area surrounding Pasadena City College to deal with a similar problem there, and it has reportedly worked well.  The cost to administer it was estimated to be $12,000/year, which is a far cry from the $3 million/year we may ultimately end up paying to subsidize parking on campus.  Alternatively, we might address the City’s concerns by working out an arrangement in which we are responsible for aggressively enforcing the 2 hour parking restrictions that already exist on many of the nearby streets, but which are currently poorly policed.  This would also cost much less than $3 million/year.  I’m sure there are other ways we could address overflow.  The point I’m trying to make is that we should be more than willing to work with the City to figure out what they might be, and that even if the solutions have significant costs associated with them, the freedom to stop subsidizing parking on campus is potentially very valuable to us as an institution.  Eventually, if we’re able to show that demand for parking on campus is fairly elastic, we will also need to work with the City to draw up a new Master Plan, which accommodates that elasticity into our development.

In the meantime, so long as we continue to subsidize parking, we should also be willing to explore other parking demand reduction strategies.  If it costs less to prevent someone from needing a parking space than it does to provide them the subsidy for that parking space, we should be willing to take the preventative measures.  Not only might demand reduction be cheaper, it is generally much less of a commitment.  If we buy someone a metro pass, or help pay for the lease on a vanpool van, we’ve committed to that expenditure for a month, or a year.  If we build someone a parking space, we’ve committed to it for 40 years, and we have no idea what Southern California’s transportation infrastructure and expectations will look like in 40 years.

For instance, in the late 1960s, Copenhagen, Denmark was as automotive a city as any in Europe.  Today its daily trips are split evenly between cars, public transit, and bicycles, at a latitude and in a climate that is dark, and cold, and wet for half the year.  Closer to home, the first light rail line in Portland, Oregon only opened in 1986, and the city only started taking cycling seriously after the Bicycle Transportation Alliance (BTA) sued them in 1993.  We should hedge against parking, because it is both expensive, and long term, whereas measures to reduce demand may be cheaper, and even if they are similarly priced, they are much more limited commitments.

A few other examples of possible demand reduction strategies:

Currently we use the same parking permits, at the same prices, for scooters, motorcycles, and cars.  Instead we should offer a separate class of less expensive permits for scooters and motorcycles, since many more of them can be fit into the same physical space, creating an incentive for people to make more efficient use of our resources.

We should consider establishing a car sharing program like the ones administered by Zipcar for Pomona, USC, UCLA, and UC Santa Barbara.  Zipcar estimates that 15 cars are taken off the road for each Zipcar.  If that held true for Caltech, at our current parking subsidy rates we could provide almost $10,000 per year toward each Caltech Zipcar vehicle, and still break even financially, while avoiding the 40 year commitment to those 15 parking spaces.  This option would likely be popular with graduate students living in on campus housing if the cars were made available in the Wilson parking structures, and the 1986 parking survey showed that on campus graduate students had the highest PR (0.67) of any segment of the Caltech population.

We could significantly improve Caltech’s bicycle facilities by installing more ground-level bike racks next to buildings on the interior of campus, and by adding a block of secure covered bike parking within each of the parking structures (maybe right next to where the Zipcars park…).  We might consider partnering with Long Beach based Bikestation, and providing a bike mechanic to work in one of the parking structures, charging normal bike shop rates, as is common near many transit hubs in Germany.  This allows commuters to conveniently have their bikes serviced during the day while they’re at work.  It would also help to have consistent removal of abandoned bicycles from on campus bike racks.  To encourage new bike commuters, and help people use their bikes safely in traffic, we could provide organizational and financial support for bike safety, handling, and commuting workshops like those provided by CICLE.  These improvements could partly be funded by instituting a modest campus bicycle parking permit, which would also facilitate registration of bikes with the police, aiding in their recovery in case of theft, and would give a mandatory point of contact between the Institute and Caltech cyclists, allowing us to more accurately track bike use, and integrate it into our transportation planning.

Transitioning to market based parking rates, or if we must subsidize driving then at least making a point of providing equitable support of all transportation choices, will not be easy.  However, I suspect that if we did so, the transportation decisions of the Caltech community would be different than they are today.  At the very least, these changes would make our transportation policies more equitable, and at best, they would represent a significant financial gain to the Institute.  Beyond those tangible and self-serving benefits, it would be an important demonstration that even here in Southern California, the Car might not be King were we to stop unthinkingly paying his ransom.