I really appreciate your insight that different political structures work better for different scales and different situations. I suppose that's intuitively obvious but I hadn't ever considered it.
You might wanna check out the story of Edward Lampert (the CEO of Sears). Ayn Rand inspired him to pit departments against one another. Sears devolved into a battle royale and profits dropped like a rock.
I listened to the first earnings call when Lampert took the helm at Sears, and he said to think of Sears as a real estate company. In other words, I don't think he thought a lot about Sears potential as a retailer.
> Plagued by the realities threatening many retail stores, Sears also faces a unique problem: Lampert. Many of its troubles can be traced to an organizational model the chairman implemented five years ago, an idea he has said will save the company. Lampert runs Sears like a hedge fund portfolio, with dozens of autonomous businesses competing for his attention and money. An outspoken advocate of free-market economics and fan of the novelist Ayn Rand, he created the model because he expected the invisible hand of the market to drive better results. If the company’s leaders were told to act selfishly, he argued, they would run their divisions in a rational manner, boosting overall performance.
Which is odd since he only became CEO in 2013, "In January 2013, it was announced that Lampert would take over as chief executive officer at Sears after Louis D'Ambrosio stepped down due to family health matters, which took effect in May 2013."
There's some classic economic work on "the theory of the firm" by Coase that really gets into this question. Within companies, it's not an internal market but some kind of autocracy. Why? Coase makes an argument about transactional and informational costs.
It would be interesting to have the departments of a company interact based on capitalism.
Sales would invest in RnD, and buy services from IT. HR would would need to focus on high quality services, not dictating unwanted routines.
Profit could be extracted as a tax, and all departments could be required to just break even (with an allowed buffer for investments spanning multiple years).
It would be interesting to have the departments of a company interact based on capitalism.
I've worked at such companies and it kind of sucks. Everybody optimizes for their local optima without any concern for what is good for the company. People end up a lot less willing to lend a hand to a 'rival' groups, and projects that would be good for the company as a whole end up scrapped because the individual departments involved cannot agree on how to spread the profits/risks internally. Also a lot of work ends up duplicated because it is 'cheaper' to get someone in your department to do it rather than to 'pay' a rival department to be allowed to use their solution.
Basically you end up losing many of the advantages of scale that a big company should provide without gaining most of the advantages of flexibility and nimbleness that a small company should provide
What about the opposite? What if every team department had a pool of tokens representing their services that they had to give away in a certain time period. I suppose tokens are the same thing as money but I like the shift from "Gimme, Gimme" to "Hey, I've got plenty, let me give you some." Each department would also be accountable for how many tokens they collected from other teams.
I'm not sure how you would work out how many tokens a department has to give. My hope is that there would be a way to incentivize going outside your team/department to get help.
Maybe I'm naive, but if you remove any disincentive to do so I find that most people, on an individual level, are more than happy to lend a hand to other departments and many people rather enjoy the variety and novelty of the projects that this can entail. I don't think it needs to to be actively incentivized as much you just need to remove as many barriers as possible for doing so.
OK, so it's not automatically a disaster, but companies have attempted to do this with "virtual org" and clouds of freelancers. It only works for certain types of piecework and small organisations. For larger things it's simply impossible to write the required contracts. This is part of why the London Underground privatisation failed: contract documents were heading into the millions of words and still inadequate to describe the working relationships.
A family member worked for IBM.
Her department would buy non-IBM PC's because it was cheaper than buying from their own company. Helping out "the competition" within your own company doesn't help you meet your department numbers even if it would help the company as a whole.