One of the more interesting and accessible modern economists (published int he 30s, died recently) is Ronald Coase.

A few years ago (before his death) he interviewed on econtalk where he complained about economics teaching "Coase theorem" with the exact opposite conclusions to the ones he intended. I was taught "coase theorem" so I went back and read his old papers.

His perspective is all about examining transaction costs. He starts with absurdities, that exist but shouldn't given standard economics. If free markets and incentives are so important to economies, why don't they exist in companies. Companies are centralized bureaucratic organizations. Incentives are highly distorted and the structure is non dynamic. No internal exposure to "market" results, no competition, no creative destruction.

Why don't contracts and dynamism yield an emergent company-like-thing where individuals contract with eachother? His answer is transaction costs, and the friction inherent in free market transactions.

A side effect of the theory is that there is a tremendous amount of potential wealth lost to inefficiency and transaction costs.

Charles Koch has a management theory called "Market-Based Management" that claims to run the internals of organisations as markets with internal competition:

https://en.wikipedia.org/wiki/Market-Based_Management_Instit...

[NB I'm in the middle of reading "Dark Money" so don't take this as a positive recommendation.]

You might be interested in his recent interview on Freakonomics:

http://freakonomics.com/podcast/why-hate-koch-brothers-part-...

Just finished it. Much like Koch would say, I walked away with some good things and bad things.

I agreed with the ability to not shackle workers with arbitrary social crimes but, in the end, fell flat.

It seemed like there is no stance too severe that will cause the Koch brothers to not work with you if you support less regulation and restrictions on trade and production.

I had hoped that the podcast would bring up the idea that maybe there is a higher cost being incurred by their activities and relate it to tragedy of the commons. I assume Koch would have dismissed it as speculation and that you can't pin their work to causing environmental damage.

His point of view shouldn't matter. He's work is not a philosophical text, it's an empirically based model that explain a part of a real system. You do see physicists complaining around that, an interpretation of quantum mechanics is a bad one, but it's always in a "it does not really matter" context, the same should apply to economics.

Anyway, some times I wonder if economists shouldn't use nonsense made-up names, like physicists do. Coarse's "transaction costs" include way more than contract settling and money transfer taxes. The use of that meaningful term confuses the hell out of everybody, and we get people talking about simplifying contracts in contexts where the highest costs are around acquiring shared knowledge.

Any more on this subject for further reading / listening?

His original papers are as accessible as popular press. The big ones are “Nature of the Firm” and “The Problem of Social Cost.” For the 2nd one, Coase complained it is used to reach the opposite conclusion than he intended.^

http://onlinelibrary.wiley.com/doi/10.1111/j.1468-0335.1937....

http://home.cerge-ei.cz/ortmann/UpcesCourse/Coase%20-%20The%...

^Essentially, Coase argues that in the absence of transaction costs, abcd. He meant this as a way of highlighting the importance of transaction costs, with abcd being obviously false in reality. Economists tend to assume efficient markets in their theory, so many read that statement to be boilerplate economist caveats.

I have a theory that all really successful, long-lived software projects are either maintained by committed individuals or by institutions (in the most general sense of the word - an institution could be a really long-lived team, or an org like Google). There is no middle ground. Software projects that bounce around between teams, contractors or whatever are dead already.

It seems like your usage of the word "institution" in the general sense is just a group of "committed individuals" perhaps with documentation in place for knowledge transfer as individuals leave and join the group. I think that joining/leaving at points is orthogonal to an individual's commitment while a member of the institution/group.

Yes, basically an institution is a long-lived collection of people, with at least some parts of it with extremely slow turnover of people. This is where the deep knowledge lives that can sustain large projects. An institution also implies to some extent a culture too - a set of shared practices unique to that group that foster a (hopefully) good environment for the project to develop.

It is interesting to note that, broadly speaking, the world is becoming de-institutionalised. Talk to any older academic, engineer, doctor, etc., and they'll tell you that less and less people are interested in settling into a job for the long-term - for maybe their entire careers. The gig economy is one end of the scale, but it's happening across the board, and it's having many detrimental side-effects (talk to any older academic, engineer, doctor, etc.). Many of the more successful software companies today foster an institutional environment - one with a strong identifying culture, making it a more difficult place to leave (e.g. Google).