The Wall Street Journal reported last week that New York has scrapped its plans to privatize parking meters.
It’s no secret that I’ve generally been a critic of long term privatization deals for on-street parking. The reason is simple. Parking meters are not a capital asset like a toll bridge, nor are they a traditional city service like garbage collection. Rather, parking meters are an urban planning tool we use to manage access to precious on-street real estate. As the way we will want to manage this real estate over time will change, it’s a bad fit for a long term privatization deal. In fact, parking meter privatizations are effectively a lose-lose-lose proposition that involve high interest off balance sheet borrowing, selling off a ground lease on the streets of our cities (i.e., the most important component of our public space), and ceding some degree of de facto urban planning control over our streets to a private entity.
Privatization – and in this case I am referring to items like long term asset leases, not shorter term operating contracts – can actually be a great thing. For example, Governing magazine just ran a piece on the effectiveness of public-private partnerships in British Columbia and elsewhere. They highlight having a professional staff that is able to properly evaluate deal structures.
But another thing is to make sure we understand what it is we should potentially privatize. I won’t claim to have the entire answer, but it’s something that every city and state should be thinking through. What attributes make something a good candidate for privatization versus a bad one?
I will share a few considerations.
1. Standalone nature of an asset. The more standalone an asset is, the better a candidate for privatization. For example, a freeway may have had a huge neighborhood impact when constructed, but after it is in place, additional impacts from tolls, repairs, etc. are more limited (though not non-existent). Contrast this with on-street parking, which impacts everything that happens on that street intimately.
2. Bespoke nature of the asset. If an asset lends itself to only one type of use, then it’s a better candidate for privatization. A highway, hospital, or airport are difficult types of items to reuse for other purposes, whereas something like on-street parking is simply general real estate that could easily be repurposed to other items. The more like general purpose real estate or buildings something is, the worse of a candidate for privatization it is.
3. Public visibility of the asset. I saw someone quoted one time who suggested that sewers were a good thing to privatize but not water, because people care about what comes out of their tap but not what goes down their toilet. This is exactly backwards. We want to privatize particularly where there is huge public visibility and demand for proper performance. This is ultimately the only way to have accountability on the contract. Where a service is delivered off the radar, it is a horrible candidate for privatization. That is why I am resolutely opposed to any type of private prison, for example.
4. Likelihood of disruptive change. How likely is the basic need likely to change over time? Think about change in the 20th century. Could we have privatized a road or airport 75 or 99 years ago and had the contract still be relevant today? No. Standards and needs have changed radically. I would assume we’ll see similar radical changes in many things over the course of this century. So we need to have our eyes wide open about what changes in needs means, particularly as it in effect creates a built-in option for the vendor to renegotiate and extract even more money when the public needs change. For example, we are on the cusp of a revolution in parking policy where we will increasingly see shifts to more dynamic pricing versus fixed rate, X quarters per hour type pricing.
These are just some initial thoughts. Feel free to add your own.
A transaction that: (1) erases a liability, e.g. the bonds issued to build the bridge or expressway and (2) moves the maintenance, upgrading, modernizing, etc. liabilities to the new owner.
Carl Wohlt says
Fascinating and timely question/post.
Coincidentally, this was posted on Counterpunch earlier today:
“John Kenneth Galbraith argued, in American Capitalism (1956), that as the economy grows, its crises become increasingly severe and government remedial action greater in scope. We may draw the appropriate conclusion: increasingly harsh crises will prompt the owning class to demand increasing control of the State. Let’s not mince words. We are talking about the creeping -now galloping- privatization of the State. In fact, domestic policy is now entirely shaped by Timothy Geithner and Ben Bernanke.
The masters of capital have taught the working class a priceless lesson. You will not get what you want unless you mobilize in order to capture State power, i.e. power to turn the State into one whose dominant objective is to further the interests of the working population. It’s not impossible; it only looks that way.”
I can think of two other situations where privatization might be appropriate, although neither are as clear-cut as your examples.
The first would be where political realities mean that an asset can’t be run in a fiscally responsible manner. I.e., where as the expenses of operating an asset increase, political pressure makes it impossible to increase the costs by the same amount.
This is tricky, though, because sometimes the appropriate political resolution is a public subsidy. But other times, particularly where the benefit is localized but the expenses generalized, this may be part of a reason to privatize an asset.
I think this was one dynamic (although far from the only one) operating with the Indiana Toll Road lease.
The second case for privatization would be where short/medium term privatization would obviate the need for a significant capital expenditure. This is easiest to demonstrate with prisons.
Articles discussing the cost of incarceration tend to use the “average cost of incarceration” as a touchstone – i.e., it costs an average of $45,000 (or whatever) to incarcerate an inmate for a year. While this figure is approximately correct, and useful for some purposes, it is not really helpful to really understand the costs of incarceration.
The costs of incarceration consists of a large upfront expense to built a prison (roughly $100 million for a new 1,000 bed maximum security facility), plus some relatively low fixed staffing costs. This means that the $45,000 average per bed consists of a very high sunk cost, plus a fairly low marginal cost for each additional prisoner (I’ve seen the figure of around $6,000 annually for each prisoner).
IOW, once you’ve built the prison, the actual (marginal) cost of incarcerating each new offender is about $6,000 – not the $45,000 you might expect. Moreover, the growth of inmate populations is due almost entirely to increasing the penalties for existing crimes, not to creating new crimes.
This means that, initially, the costs of increasing the penalty for a crime are 0 (while the political benefits of doing so can be large).
I.e., let’s assume that the average time served in prison for manufacturing meth is 10 years. A legislator from an area with a meth problem may introduce a bill to increase the average time served for manufacturing meth to 15 years. As discussed above, the actual budgetary expenses of doing this per inmate are 5*$6k; not 5*$45k, so it is fairly inexpensive. But – more importantly – this cost doesn’t even kick in until 10 years have passed, since under the old law the meth makers would be serving 10 years anyway.
This leads to a situation where there are concrete political benefits to increasing the penalties on certain crimes, with the costs of doing so being fairly small and also distant.
But this dynamic only works until you fill the prison. Once the prison is full, the marginal cost of incarcerating the the next prisoner jumps from $6k to $100,000k.
So while I’m not generally in favor of private prisons, either, I think that they would be appropriate as a temporary measure to avoid constructing a new facility while the state looks for ways to reduce the overall prison population and eliminate the need for new prisons and private prisons.
Sorry this was so long. I assume that there are also examples of this being useful involving government office buildings…I’m just more familiar with prisons. (And I do think it’s useful to discuss the economics of them, since I think they are generally not understood well).
Aaron M. Renn says
@PeterW, thanks for sharing. Great thoughts.
Ah, the old marginal vs. average cost debate. Been through a few of those cost accounting wars. In this case though, it’s really an artifact of cash budgeting by governments. All the more reason for them to produce real GAAP type financial statements.
In any event, whatever the merits of private prisons, the fact that the inmates have very little ability to protest their treatment and so little visibility from the outside world, I consider private prisons beyond the pale at this point in time. I’ve already read way too many stories of abuse to believe my concerns are only theoretical.
Great piece. Could you elaborate on why privatization is a “high interest, off balance sheet borrowing” endeavor and therefore a bad idea? I am not getting the high interest comment.
Here are the scenarios I can think of:
If company X purchases public parking spaces from city Y, then this is akin to a city Y selling a parking garage to company X. City Y receives payment upfront. No issue.
If company X enters into a operating lease for public parking spaces from city Y, then city Y receives predictable, non-seasonal/non-volatile rents from company X. If company X breaks lease, then the city steps in and the only issues are friction and interim care in this event.
If company X enters into a long-term lease for public parking spaces from city Y, then city Y receives predictable, non-seasonal/non-volatile rents from company X. This is not considered a capital lease because the land does not depreciate retaining economic value. If company X breaks lease, then the city steps in and the only issues are friction and interim care in this event.
Again, great piece and thought provoking.
Aaron M. Renn says
@VGS, a long term parking meter lease is functionally debt. In a traditional loan, you borrow a fixed sum X and repay it with interest with periodic payments. In a parking meter lease, you are still getting the lump sum up front (at least in Chicago’s case), and you are still repaying it with periodic payments. In this case, the payment is in the form of the quarters people are putting into the meters. The city gave that entire revenue stream to the leaseholder.
Whether this results in a high interest rate in the Chicago case is a matter of debate. In the Indy case I think it was pretty clearly high interest. The vendor invested about $7 million in new meters + gave the city $30-40M (I don’t remember the final value), but is getting over half the future revenue stream in return. It’s like a payday loan.