Organizational stupidity

One of the great curiosities of the UK economy is that productivity has stagnated. Output per hour has fallen in the last four years and has grown by only 1.1% a year over the last 10 years.
This is curious because it seems to contradict the principle of learning by doing. I spend about 10 hours a week playing guitar, and I’ve gotten much better in the last four years. Why, then, have the 900 million-plus hours per week we devote to work not yielded productivity improvements?
Here’s a theory - which I’ll overstate for expositional reasons. It’s that companies are institutionally stupid. They are incapable of learning.
Three pieces of evidence are consistent with this:
- 90% of the growth in total factor productivity in UK manufacturing in the 1980s came from plants entering and exiting (pdf) the market, rather than from incumbents improving productivity.
- The statistical distribution of corporate deaths is very similar to that of species’ extinction. This, says Paul Ormerod, suggests (pdf) that “firms have very limited capacities to acquire knowledge about the true impact of their strategies.”
-  Alex Coad has found that corporate growth is largely random and unpredictable. This is consistent with the possibility that there is no “corporate intelligence“ that leads to growth (Or it could be that such intelligence is uncorrelated with observable variables - but why?)
But why don’t firms improve with practice in the way that individuals' musical or sporting performance improves? Here are four possible differences:
1. Within firms, there’s no mechanism for translating individuals’ learning, or incremental knowledge, into corporate knowledge. As Hayek said, hierarchies are terrible at using fragmentary, tacit, dispersed knowledge.
2. Job turnover means that job-specific human capital gets lost.
3. Bosses are selected for overconfidence. But overconfidence militates against learning.
4. In companies, the feedback that’s necessary for improvement gets warped by adverse incentives or ego involvement. If I play a phrase or chord badly, my ears tell me to practice it more. But if a company gets some adverse feedback - falling sales, say - no-one has an incentive or desire to say “I screwed up: I‘d better improve.” And formal efforts to generate feedback, such as performance reviews, often backfire.
What I’m saying is  what every methodological individualist knows: companies are not individuals writ large. The differences between them can mitigate against learning by doing.
And herein lies the cost of the banking crisis. Because productivity growth comes from entry and exit rather than firms’ learning on the job, anything that retards the entry process - such as a lack of finance - will retard aggregate productivity growth, and hence economic growth.
Another thing: What’s true of firms is probably true of other organizations such as government departments. Organizational stupidity might well contribute to Baumol’s cost disease.

Comments

[Mil]

I think this idea of a dispersed knowledge which glues everything together - and really adds value - is spot-on. I also get the feeling large companies are poor at learning as a whole because they choose to pit their workforces against each other via a kind of competitive internal market. More important to divide and rule than bring together and share.

One final observation: maybe productivity is stagnating simply because it's not being entirely measured. Lots of stuff we do in our leisure time - open source activity for example - might add real value to our processes and national economy, but can't be properly taxed or valued by the state. Just a thought.

[Metatone]

Wait:

"90% of the growth in total factor productivity in UK manufacturing in the 1980s came from plants entering and exiting (pdf) the market, rather than from incumbents improving productivity."

But what about (for example) German, Japanese and South Korean manufacturing?

[Metatone]

Also, have to say the Ormerod paper is almost a parody of the problems Dave Snowden has articulated with the use of simplistic agent modelling to draw conclusions about complex systems.

[PeterM]

Your theory is even more poignant in the world of information technology. It just takes one over-confident, ego-centric idiot to completely crush the very next important strut. Because this is not manufacturing, this is not nuts and bolts but a mercurial substance close to thought itself, and easily blown away by idiots. Which, as you say, is the norm in corporations. The larger they are, I have found, the more irrational and incapable they are of producing software. Cannot imagine how Apple have managed. They must create small internal silos.

[Metatone]

(Caveat: Godwin)

There's a paper out there somewhere which I can't find which relates the structure, culture and practices of German and UK armed forces, traced through WW1 and WW2 to German industrial and footballing pre-eminence of the UK.

There's plenty of evidence that many of the problems of the British economy are local - and ironically not because the workers are lazy (the usual refrain) but because UK management seems to have specific blindspots and weaknesses overall.

[Metatone]

gah - pre-eminence OVER the UK

[Left Outside]

I'd be interested to see how much of the productivity stagnation can be explained by people moving from fast growing productivity industries to slow growing productivity industries.

It happened in Latin America, and rarely happens in advanced economies, but it could (is probably?) a factor. I wrote on this sort of thing last week.

[Andrew]

You get better at guitar, Chris, but why should the UK in aggregate be getting better at guitar?

[Andrew]

Additional point: improvements in societal knowledge and skills (ie science and technology) don't necessarily result in price rises (ie productivity). In fact, all other things being equal, a given amount of useful work should get cheaper and cheaper in an era of expanding technology.

Since productivity is price/hours, why should it increase in general in such an environment?

Aren't we conflating productivity with improvement and optimisation of work? Not the same thing at all.

[Lee T]

The last ten years has seen internetz (tm) go mainstream. Computers are now tools of leisure instead of/as well as tools of productivity. So where once a worker could only use new tools to work, he can now also use them to play.

Technology has always driven productivity and efficiency.. but now it also drives distraction. Putting an internet/email enabled computer in front of someone at work can be akin to putting a television in front of them. The industrial revolution would have taken way longer to change the world if the spinning jenny had come wth lolcats.

Where people do use computers to try and increase their productivity (independently of institutional change) it's often counterproductive anyway because most people don't have the level of IT literacy required to make suitable advances. I think that we'll see improvements over the coming years as people who have grown up with computers displace those who have not - both in management and the workplace in general.

[Andrew]

To clarify the above: I am talking of the price of the production and "hours" is shorthand for labour input costs, assuming a simple case where other input costs are negligible.

So if productivity is output/input costs, and if labour hourly rate is unchanged, why should getting more done in an hour increase productivity if everyone is doing more?

Everyone still has the same wage, so aggregate demand is constant and the price of an hours production must therefore remain constant.

If money supply keeps up with the increased production so that the price of an hours increased production is commensurately increased, then wages must similarly be increased.

So productivity stays the same.

Productivity can increase if there is local advantage. It can increase is labour price or other costs are suppressed for whatever reason. But there is no reason why it should in general increase with corporate or societal "practice".

Productivity does NOT need to increase for living standards to increase.

[Andrew]

Sorry, I have been rather confusing. I stick with my first comment.

[WH]

In my company (software) people tend to stay a few years, then leave to do the same job for an investment bank at much higher pay. Generalising, couldn't you have three companies (say), A, B, C, where A employs inexperienced people, who move to B when they become more experienced, and then to C when they're most productive, before retiring after that. If you looked at the productivity of A, B or C over time, you'd not see that much in the way of improvement, even though the individuals involved, as they progress along their careers, become greatly more productive. Silly idea?

W

[Ralph Musgrave]

Chris, There is a fundamental flaw in your opening premises / paragraphs. You say that productivity should improve because of learning on the job. I suggest it won’t.

There is a constant steam of experienced people retiring and a constant stream of inexperienced youths starting work for the first time. Given constant technology, AVERAGE levels of experience and skill etc will never change.

[Richard]

Ralph, your thesis relies on the assumption that the number of experienced people retiring is commensurate with the number of inexperienced people starting work. I question the validity of that assumption.

[Richard Gadsden]

@Ralph, that presumes that there is a steady-state age profile (OK, experience profile) in the workforce.

Given an increasing population, there should be more inexperienced people joining than experienced people leaving the workforce - so experience should result in a productivity fall over the medium term.

However, presuming the original thesis, it does raise the question of whether industries with little turnover of firms will have greater productivity stagnation than ones with stronger turnover.

For instance, very few law firms are less than a century old.

[georgesdelatour]

Isn't this the place to bring up the Tyler Cowen "Great Stagnation" argument? Or maybe not...