Public vs Private is actually less important than the competition vs monopoly dimension. Both Ontario and Alberta privatized DMV-type services. In Alberta anybody can start a registry, they get paid a processing fee for each service $10 say to to renew a license. In Ontario they contracted out the operation of the existing offices to one company (incidentally this company also runs a lot of private prisons). In Alberta you can renew your drivers license in 5 minutes and there is probably a registry within a 5 minute drive from your house. In Ontario, you're driving across town to the one location and it'll take you all day.

I haven't looked into it, but I'd bet private prisons aren't much better than public prisons. However, if you'd let anybody open a prison, let prisoners choose which prison they went to and paid these prisons a flat fee per prisoner, private prisons would guaranteed be better. Come to think of it, that might not be such a bad idea.

Expand full comment

Yes, having choices often does make the system better. Where that's a feasible alternative it's something the govt tries to do, or should try. Like you say, in most cases incentives and org design matter so much more than the ownership structure. In public instances competition itself isn't feasible of course, but designing better KPIs or transparency or incentive systems is which often compensates well.

Expand full comment

I'm not a big fan of empiricism in the "dismal science" precisely because the argument above can seem impossible to solve and economics till 20th century was not empirical even when it was possible to be empirical. If you think from first principles (praxeology) you should be able to resolve the problems quite easily. The best form of management is the "benevolent dictator" format which the entire corporate world runs on. You have a good leader who then hires subordinates, gives them responsibilities, who in turn hire more subordinates and so on. Leaders can listen to subordinates, or be a complete dictator but really that is just comparing management styles. Governmental Organisations however have a bureaucratic management style. Here there are nominal heads, who don't get to appoint most of their subordinates or fire them. Decisions are made not top down, but bottom up rule based. The lowest bureaucrat tries to decide on a certain rule, if there is ambiguity it goes up the food chain, to the end reaching the leader who is basically a guy that resolves ambiguity and sets precedent. If he's wrong, the courts of course can write the final law of the land. This problem also falls when the government tries to provide a service like the UPS. UPS is a rule run organisation, if the UPS wants to change it needs to change some rules that will be applied equally to the entire organisation. When the UPS encounters a problem, it solves it in a process based way. It checks the rule book, tries to find the process that it needs to follow given the problem and follows it to a T. It is an accountability free organisation, where if everyone follows their rules to a T, then everything is fine, no yearly targets, no goals nothing like that. Amazon is goal based, when it encounters a problem it empowers its employees to solve it, not solving it can get you demoted, transferred etc. Interestingly as corporations grow old and die they become increasingly process driven, (as you might have noticed if you worked in a large Fortune 500 corporate like IBM or ITC) and lesser and lesser goal based. It is less likely for a corporation to die by a leader choosing the wrong goals, more likely when the leader tries to avoid accountability to the board by setting up a process for everything. The market is useful in letting corporations die when they become sclerotic and paralysed, something government agencies do not have the fortune off. So this is the praxeological/ deductive argument based on the first principles, let us see what I can predict based on this:

NASA has become a process driven organisation and will never innovate compared to a company like Space X. NASA in its founding was in fact goal based and more like a corporate with lots of money and a non profit motive (goal was to reach the moon). To see evidence of this, there is the book "Apollo" by Charles Murray that shows the culture between early NASA and later NASA, with many of the early engineers lamenting what happened to it.

The fact that most of our modern inventions have become from DARPA does not surprise with the above argument. Defense is actually run command and control based and goal based, not rule based. DARPA also has deep pockets which is necessary to pursue risky ideas. To see this: you can check the culture of DARPA in this in depth article by Benjamin Reinhardt: https://benjaminreinhardt.com/wddw

DARPA actually goes above and beyond and follows the structure of some of the most innovative organisations in the world. This is described in the book "Dynamic Economics" - by Burton Klein, where he lists principles that he found in the most innovative organisations of its time, and its clear that DARPA follows most of them. This Blog Post: https://scottlocklin.wordpress.com/2021/02/17/planning-of-invention-part-1-burton-klein-and-dynamic-economics/ provides a useful overview in case you can't read the book.

Singapore of course provides excellent service because it is run in a command control goal based manner. It pays the highest salaries in a bad to attract the best talent and also empowers the employees. In fact it seems like the entire country of Singapore runs like a corporation and it is a democracy in name only. People have some voice but they can't openly criticise the government and they vote for the opposition to keep the ruling party on its toes).

Government money isn't inefficient, that doesn't make sense. So when the government just gives money to private organisations, its not bad if the organisations are run well. Government can this way fund a lot of innovation (this is very different from giving money to the NSF). Solyndra occurs but also Tesla and Space X occurs so giving money is fine, unless you choose an overly bureaucratic way of giving money.

I suppose the final conclusion is what can we do here to improve our government:

1- Retire all Government Employees (from most organisations). You can provide them good pensions, good severances but they need to go. A stage comes when a corporate has become so paralysed under it's weight that it kills innovation and requires an exceptionally incredible leader to turn it around. Markets fortunately automatically kill such corporations, governments are not subject to markets, so the government should kill it on its own. One should preferably do this at a regular time scale (say every 50 years?)

2- Recreate all the organisations that are necessary from scratch with new talent and a command and control structure. Hire good leaders and ensure they can lead the organisations for at least 10+ years. Changing the leadership every 5 years with a new presidency is a very bad idea, I also don't advocate for allowing the president to fire employees as that means every 5 years entire organisations can be diluted, which isn't great.

Im actually just recreating the conditions of the New Deal. They had a command and control structure, they had the best talent and were stable for around 12 years and even after that, the heads weren't changed all that regularly. I don't doubt the new new deal (or the green new deal) will again over time 70+ years, become incompetent sclerotic and inefficient. No modern corporation has lived long, the only corporation that lives long are family run businesses that are too small in scale to compare. The goals of the agencies recreated can be decided by the president, also I don't think this is politically possible and I think America is going to hell, but if theres a way out, we need a new FDR.

Expand full comment

I feel you've given the qualitative side of my argument above :-) Govt has inefficiencies, so does sclerotic private sector companies (hello Oracle!). Private sector has great companies, but so does govt (hello Singapore!). Even in sclerotic public sector environments, like in the US, plenty of govt funded orgs function really well, much more than what would be predicted purely on the basis of a first principles approach. FWIW I'm also skeptical of first principles approaches that don't lend themeslves to empirical proof!

To me therefore the 2x2 above kind of shows where things need to converge towards. At the corners, govt has an advanage, and towards the middle, there are innovative movements from the private sector/ individuals.

I'm sympathetic to your conclusions however, because I do believe that govt, and indeed most private sector orgs, regularly require a cleanout and "defragmentation" so that they can rediscover their elan vital! None of what I said indicates "things are great" or "we don't need to substantially improve" by the way, it's just a bit of empirics against the simplistic "govt is inefficient, so we need less of it" sloganeering.

Expand full comment

Makes sense. I think empirics works when you need to find new first principles and it works when empirical data is reliable. The problem with economics is that reproducible experiments are rare and impossible when you want it at a large scale. Pure empirical correlations are simple not useful for highly complex systems (and also a problem with modern Machine Learning that I think won't go very far in recreating the human mind, contrarian but more people seem to be coming around these days). Austrian Economics makes the same claim, (and it isn't anarcho-capitalism or radical free markets advocacy, it is an apolitical science (like maths/philosophy though many Austrians are radical-libertarian).

Expand full comment