« Money & happiness: an experiment | Main | Johann Hari: idiot »

February 01, 2007



Is the slow growth of established companies versus new companies not just a factor of differently aligned incentives around risk taking? A small company has a limited downside in taking a large risk that may pay off or may bankrupt the company. In a large company, by definition, the downside is larger. Large companies should therefore be more risk shy than smaller ones (even on risks that have a positive expected value), which should mean that smaller companies grow faster on aggregate.


ChrisA raises an important question.

Another is how are we measuring "innovation"? It's not at all apparent that we have a good measure of innovation inside the Inland Revenue. For medicine, the evidence is rather mixed. Most innovations between 1950 and 1980 came from within either large hospital groupings (including the NHS) or large private companies.

I actually don't disagree with Chris' idea, as the evidence is that innovation in large organisations tends to occur in patches where the hierarchy is weak. But we shouldn't ignore that innovation also tends to feed off "slack" which can sometimes require the resources of a larger organisation to provide. You can do a lot of innovations with software and the "slack" that is a the willingness of a bunch of twenty-four year olds to code all night.

But, the computer revolution is also built on microchips, with ever decreasing nm size tracks. Those tracks are only possible because of innovations in lithography from those tiny little startup lithography firms, Nikon and Canon.


What does it mean to say that "the west’s economic performance and innovation been so good"?

All we know is that the economic record of the not-West has been less good. And those countries are even more hierarchal.

Which suggests that hierarchy is bad for the economy.

Kevin Carson

Anon beat me to it: "so good" compared to *what*? All we know is the existing state capitalist system as it actually developed, not what alternatives it may have crowded out.

I'm not sure it's fair to judge cooperatives by their innovative performance under the current incentive structure. The way they behave in this economy, barely hanging on by their fingernails in a general environment of hostility from the financial industry, is not necessarily the way they'd behave in an economy where the producer co-op was the norm. For that matter, most of the "labor-saving" innovations of mainstream capitalist enterprises are pretty suspect in terms of absolute efficiency, as opposed to "efficiency" at overcoming the agency costs of supervising labor. In an economy where most productive capital was worker owned, and it was a given that workers would internalize all the benefits from increased productivity, the incentives might be considerably different.

Eric H

I'd add - why limit the argument to the superior economic performance? Simply because that's something that can be measured?

Yes, we all like to think this is a much better life because of longevity and iPods and what-not. However, I wonder whether we get much more enjoyment out of it. Prisoners have been found to outlive cohorts outside the prison, but that doesn't mean they enjoy it more.

18 years of school, followed by 30-50 years in a cubicle, and then finally getting to retire at an age so old you worry more about your breaking a hip than enjoying life probably sounds like torture to someone who grows up in a "primitive society" where they get to spend most of their days in blissful repose with their friends and families. Sorry to romanticize the savage, but I'm mindful of Jared Diamond and now reading The City in History and wanted to add something useful after Anon and Kevin appropriated the best point(s).


I'm not sure I agree with this Chris. Henry Ford didn't invent the motor car, but he did invent mass production of the motor car. Ford Motor started off as a small company (by definition; you can't start a large company except in oddball restructuring situations), but in order to become Ford Motor, it was written into the plan that it would become a big company and a hierarchy.

In general, the work of an "innovation" is not done when the idea is had. Scaling up production to the industrial level is rarely a trivial task, and the people responsible for it almost always end up using a hierarchical management structure as part of the plan.

I also don't see how "competition mitigates hierarchy" in any important sense. Toyota is a hierarchical organisation. They have a suggestion box scheme and kaizen, but Toyota still has managers (and Nissan is very similar to Toyota but ended up having to import French-style command and control management to get them out of a hole).

Show me a company which has gone from 23 layers of management to 6 and I'll show you a company with 6 formal layers of management and 17 informal ones. I really don't think that IBM isn't a hierarchy.

Finally, I don't think that your statement about governments is a truism at all. Over the last hundred years, governments have invented universal education, mass vaccination, the welfare state, atomic energy, put a man on the moon and created the Internet. In terms of really important basic science from the private sector, it's the transistor and surprisingly little else. And Bell Labs was the hierarchy from hell.

Glenn Athey

I think that we kid ourselves that enterprises in the market are all brilliantly run and organised, when there's a big number of middling or mediocre performers, and perhaps only a handful of fantastically vertical and successfully organised ones.

If the hiearchical ones were the worst, and the market could drive them out of business - the shake out of crap firms just takes time, and there's all sorts of barriers and distortions to the market speeding this up such as protectionism, barriers etc.

And some pretty dire companies manage to survive, and some companies you'd imaging are pretty hot, are in fact, direly hierarchical and awfully run.


I'd suggest (like everyone else, it seems) that Co-ops are not very good at all at innovation. Flat structures may not necessarily translate into flat pay differentials, but they are flatter than normal - which means that in-demand skills can't be retained, and salespeople rarely get very good performance bonusses like their commercial counterparts.

They are good at taking over an established service and running it more efficiently. As they can share cost-savings, staff are encouraged to be more efficient, and competition and the real threat of losing your job motivates you to work on client-retention.

As Greenwich Leisure showed, taking a public-sector workforce and turning them into a co-op can work very well.


D2: IBM strikes me, whenever I have any dealings with them, as pleasantly egalitarian, or at least meritocratic. Once you pass a certain level of geekhood there, your job title is just "Senior Technical Staff Member", whatever you are doing.


Are you sure of that? I've not had many dealings with them (mainly with the consultancy arm) but they seemed to me to have vice-presidents, EVPs and directors more or less like any other consulting firm. In the days when they were inventing Computers As We Know Them, they were much more of a hierarchy; EDS exists today because someone decided that they couldn't give Ross Perot a bonus that owuld have taken him out of his pay grade.


[Co-ops are not very good at all at innovation]

This rather points up the need to be total clear on what's meant by "innovation". Oxford, Cambridge and Harvard Universities are all three of them mutual socities owned by their Fellows. They haven't changed their teaching methods in the last hundred years (actually, I think Harvard has) but they have produced the odd useful invention or two in that time. Organisational form matters a lot less IMO than Chris thinks it does and the specific competence of the actual administrators a lot more. The fundamental attribution fallacy isn't a fallacy in business.

The comments to this entry are closed.

blogs I like

Blog powered by Typepad