Hacker News Reader

Agreed. The government using private contractors to maintain the infrastructure is fine. In that instance, the government can still account the resource in a holistic fashion. Where it makes sense for end-user-fees to completely pay for the infrastructure, then fine (sometimes it will). But where it makes sense to minimise end-user-fees so as to maximise the positive externalities (and recoup the costs out of taxes), then they can still do that. Whether the workers on the infrastructure are managed by a public or private manager, in this case, doesn't make a lot of difference.

The problem occurs when a private entity must make a self-contained business case off the back of the infrastructure. Since they'll never be able to capture any of the positive externalities, they have to take everything out of end-user-fees, externalities be damned. This can be very, very bad for the city as a whole.

And yes, agreed about cable internet being in this category too!

> Sure but rather than directly hiring N people to run the water works and plumbing crews, the city could contract out the maintenance of those things to a private company.

I know this can work, but in addition to the usual misalignment of goals between public and private sector, I still see more inefficiency with this than just letting the government do it.

For example, the government has to pay all the workers (indirectly, at least). On top of that, it has to pay to negotiate contracts, and an entire layer of oversight of the contractors (which also have their own HR and bureaucracy that you are paying for). On top of that, they will likely have built-in some extra profit to give them some flexibility in case they don't get the next contract. Plus money for lobbying in many instances.

If you find the contractors aren't doing their job, now you will probably have to get lawyers and whatnot, and have expensive, time-consuming litigation. Then, if the government wants to change companies at some point, you have a large expense in finding a new one, with bidding, etc. And that is if there is another contractor available in your area. And then you have lobbying, cronyism, and other political issues.

The entire reason for hiring another company to do this is to get around public pay and pension issues. If this is all just a workaround for that, fix the source of that problem, rather than trying to abstract away responsibility.

> If you find the contractors aren't doing their job, now you will probably have to get lawyers and whatnot, and have expensive, time-consuming litigation. Then, if the government wants to change companies at some point, you have a large expense in finding a new one, with bidding, etc. And that is if there is another contractor available in your area. And then you have lobbying, cronyism, and other political issues.

Business contract law is much simpler to deal with than employment law. Plus it wouldn't be the governments job to deal with individuals at all. Only the end result of the service provided. It'd be the private companies job to deal with their own employees and contractors.

> The entire reason for hiring another company to do this is to get around public pay and pension issues. If this is all just a workaround for that, fix the source of that problem, rather than trying to abstract away responsibility.

That's exactly the point! Because no private company would be stupid enough to enter those types of agreements with their workers because they can't maintain their end of them. The government can't do it either (without bankrupting and/or eroding it's tax base) so the solution is to prevent government from ever having a chance to enter those agreements. Otherwise some weak willed politician will hand out defined benefit plans like it's candy to get votes for one election, only to doom their children's children for years to come.

That's an interesting way of looking at it. Not sure I completely agree or that there isn't a better way, but it's certainly a point of view I didn't see. Thanks

>Sure but rather than directly hiring N people to run the water works and plumbing crews, the city could contract out the maintenance of those things to a private company. It would still own the pipes and be on the hook for the maintenance, but it wouldn't be directly employing the workers. If company X doesn't work out (say due to crappy service), then they could award the maintenance contract to company Y. In this model, the private company doesn't get a local monopoly on the resource, only a contract to provide the service for some limited duration (say N years).

This American Life had an example of this with Colorado Springs city services, act 3 of http://www.thisamericanlife.org/radio-archives/episode/459/w...

A problem with this approach is often you see corruption or crony-capitalism seep in. These contracts are then given in very favorable terms to certain bidders, or they are locked in for a very long time that can't be broken without paying huge penalties to the service contractor, no matter how terribly they are performing. You also can't adjust the goals of the service under contract over time - say you set your KPIs for the contract one way at the start but realize later actually this isn't working the best for the community and we need to manage it differently. Screwed.

> You also can't adjust the goals of the service under contract over time - say you set your KPIs for the contract one way at the start but realize later actually this isn't working the best for the community and we need to manage it differently. Screwed.

The answer to that is to have shorter contracts and/or having clauses in them that allows you to break the contract. The would naturally lead to higher costs as there would be additional business risk but it'd probably be worth it.

In fact, a more general law I'd like to have in place is that a given government cannot create obligations that extend past N years unless there's a damn good reason for it. Short of distributing the cost of building a bridge over 20-30 years, there's no reason for governments to hand cuff their citizens like that (besides croneyism and kickbacks). Maybe have the max duration of the obligation be tied to the percentage of vote required. 51% max would be 1 year, 60% max would be 3 years, etc.