Sunday, November 15th, 2009
I’ll conclude my series on privatization with a series of guidelines for action or best practices you should look at to determine if privatization is right. I take a pragmatic view on this. The private sector and government are always going to work together. We just need to make sure we do it right. I had a list of considerations I had developed it, and supplemented it some insights from Indianapolis blogger Paul Ogden. Ogden is a strident populist Republican who used to favor privatization but is now mostly skeptical. Even if you don’t go for his politics, I think he made some good points, so I included them.
- Understand why you want to do the transaction. There are lots of potential motivations and we should understand the ends we are going for. This is often independent of the entity in question. For example, when Indianapolis Mayor Steve Goldsmith privatized the water utility management, the motivation was typical “good government” efficiency and a belief that a private operator would do it better and cheaper. Today, water and other privatizations tend to be more about jackpot payouts.
- There should be clear, obvious, and compelling value, and transparency how it is being generated. Given the risks inherent in any transaction, it should only be undertaken if the value is clear. Back to my post on value levers, we should understand which ones we are pulling.
- One time funds should be spent on one time expenses or put in the bank. Any operating tails created should be fully funded.
- There should be a universe of multiple qualified bidders. If there is one or only a couple firms who can provide the service, this exposes the government to the risk of bad deal with a near monopoly and no alternatives. When looking at contracting out to non-profits, this is also a risk. For example, Community Development Corporations tend to be territorial monopolies. Given the size of most of those contracts, the financial risk is not high, but certainly there is risk from having a single or handful of organizations consistently do everything.
- The transaction should be as purely financial as possible to maximize bid fairness. Think about highway construction projects. These are subject to a sealed bidding process where every contractor has access to the exact same specs and the lowest qualified bidder is required to be picked for the job by law. The Indiana Toll Road and Chicago Skyway leases were similar. The terms were known and the competitors were merely putting out dollars. Not not every service is this way. Professional services are generally not, because skills are so different, the deliverable often is not sufficiently spec’ed to give a true apples to apples comparison, and relationships are very important. But that does render them more subject to outside influence.
- Avoid contracts of excessive length. In a large scale private outsourcing, the benefits are usually higher a few years down the road. It takes time to centralize, offshore, etc. In fact, doing that can even cost more money in the short term. So to get the client to buy, the vendor has to generate current year savings for the client by effectively fronting some of the benefits. They hope to recover in this, along with their investments in efficiencies and transition costs, in the later years. That’s one reason why durations can be longer, such as 5-7 years. The vendor needs to have a long enough time span locked in to recover those up front costs. So it can be reasonable to have longer term contracts, but that should be based on things like recovery of investments and pre-funded benefits. We shouldn’t just sign a long term deal just because.
I always wondered why infrastructure leases are so long. The Skyway and Toll Road leases are 75 years. I’ve never seen an investment made on a 75 year business case. For the IT investments we used to make, we expected maximum three year payback. When you start discounting back the revenue from 75 years in the future, the present value of the money can’t be that high in the out years. I haven’t figured it out completely, but one thought is that vendors want the long term lease so that they can force the government into an advantageous renegotiation at some point.
Consider the toll road leases. I noted that part of the value of these lease in the hedge created. The lessee is taking on all of the risk about revenues, repair costs, traffic volumes, etc. If something takes a turn for the worse, as with the current recession, the concessionaire is left holding the bag. But conceivably down the road some provision of the lease might trip up the government as well. For example, highway leases often contain non-compete clauses that prevent the government from improving parallel roads. If the government ever wanted to change something, they lessee is now in a strong position. The longer the lease, the more likely something like this will happen. Think of it as a hedge for the hedge.
Whatever the case, we should understand exactly why the length of the deal is what it is and it should make intuitive sense.
- Termination for convenience. The government should be able to terminate the deal with some appropriate notice period and the payment of a known in advance and reasonable termination fee. We should not be stuck with an under-performing vendor and the government should have the right to get out of the deal without having to sue for breach of contract. It is clearly reasonable to expect to pay a penalty to do so, but this should be a reasonable penalty, negotiated in advance, and declining over time. Possible there could be a variable element in things like Toll Road leases to protect the vendor from the government canceling the deal just because it would look advantageous to rebid it in the current market.
- Avoid impairing the right of future governments to change public policy. This can be very subtle. The non-compete clause in a highway lease is a good example. That one is probably ok, as long as it is not overly broad. But vague wording could be disastrous. Outsourcing water utilities shouldn’t do anything to prevent various water usage reduction initiatives the city might want to take on, or cause problems with other regulations related to the use of the systems.
- The privatized service should be one that is visible to the general public. Someone in a Chicago newspaper said that privatizing sewers would be better than privatizing water because people care about what comes out of their tap but not about what goes out of their toilet. My view is just the opposite. Perhaps sewers are politically easier to privatize it, but we should want to privatize things where the public itself will provide an extra layer of vigilance. Things like roads, parking meters, and water are used by a broad section of the public and it is very visible whether or not they are being managed well or poorly.
- Do not privatize services consumed by vulnerable populations who cannot effectively protest against poor performance. Ogden criticized the Indiana social services outsourcing contract as an example of how not to do it. However, I think this one actually worked. That is, when it was discovered that the contract was not working, the state backtracked. Those in need of social services are, unfortunately, fairly numerous. Also, there are established advocacy organizations that serve as watchdogs. So while this particular contract didn’t work out, ultimately the public oversight function worked. I would agree with Ogden, however, on privatized corrections. This seems to be a pretty popular function to outsource, and it is easy to see why. People who are incarcerated probably can’t vote, have little ability to influence policy, and are publicly unpopular. I can’t imagine a scenario in which I believe privatizing corrections is a good idea. That’s not to say all vendors and contracts are performing poorly. But there is just no possibility of public oversight.
- Vendors in privatization contracts should be subject to open records laws related to those contracts as if the government provided the service. Privatization of services should not involve moving government behind closed doors. However, there are probably some viable differences here. For example, with a vendor on a fixed price contract, it probably isn’t necessary to be able to find out what they paid subcontractors for materials and services. That’s legitimately sensitive competitive information. But clearly when operating in the public sphere, businesses need to be accountable to the public. If they don’t want to be, that is there choice. They can simply not bid on the contract.
- Avoid asset leases or sales in bad markets. Right now would not be an ideal time to try to, for example, lease a toll road. Credit it tight and the economy is tough. Indiana and Chicago leased their toll roads at the peak of the bubble. You can’t always time the market, but anyone can tell that right now is a tough time to get deals done and the terms probably aren’t the best.