We spoke to professor Klepper about this research. This is what he had to say:
I have specialized more broadly in looking at the evolution of industries from conception through to maturity. And I ask a few big questions.
the biggest unanswered question though was why do markets differ so much in the fraction of market owned by a leading firm?
I asked almost from the beginning of my career as a student of the economics of industry – the whole field was predicated on market structure, what percentage of a market was owned by a firm, which determined decisions around R&D, marketing, etc. – the biggest unanswered question though was why do markets differ so much in the fraction of market owned by a leading firm?
So I focused on about 50 new industries over the long term, successful products for example autos from 1895, TV from about 1946, but pretty much only in the US.
What I plotted out was a log of entry to these markets. For example in the early 1900′s there were over 200 auto producers in the USA; that then is followed by long periods of shake-out in autos over a 40 year period.
So the question I ask is why does this happen so regularly? In autos the number of firms plummeted down in the 1930s when there are only about 30 manufacturers left (and by 1941 only 9). Yet there is growth of output of about 15 – 20 % per annum through the start of the Great Depression in 1929.
Growth and a shrinking number of firms – why is this so common in innovative industries? What determines market structures so that a small number of firms dominate in the long run?
A second question I have been asking is why are successful firms so incredibly regionally clustered in some industries, for example Detroit in autos and semi-conductors in Silicon Valley?
Silicon Valley is an extraordinary engine of new company creation and economic growth as was Detroit. So why did they emerge? Why are some industries incredibly regionally concentrated when often there are no obvious geographical advantages?
The third question is where do great companies come from? What breeds them? Who are they? What’s the underlying process and why does the US generate so many?
There is a fourth question. In a bunch of the industries I looked at they became so successful and were all US based – autos, types TV receivers – they all followed a similar pattern. After formative periods in the US they all collapsed. All these industries have shown horrible performance in modern times. In fact there is now no indigenous TV receiver company in the USA, and only one tyre domestic producer left, and this is in the context of industries where innovation had set the winners apart.
In pursuit of these questions I developed a kind of nano-economics approach as distinct from micro. I go into the detail and identify every company that enters these industries. Where did it come from? How long did it stay? Who were the founders? Why did they set it up? I study all the innovations, all patents in relation to innovation and reconstruct the evolution of an industry over a long period of time.
What I’ve tried to argue is there is an involuntary birthing process taking place. Existing companies give birth to new ones in the same industries as employees leave and start their own company. This tends to happen to the best companies that have their best employees leaving to create new companies, which in turn have the best performance.
I’ve argued that these companies, these spin-offs, tend to be exemplary performers in a number of industries. The second generation company tends to overtake the first generation leaders. It’s almost as if you need to have them to overtake the parent.
This is the essence of how Detroit became the center of autos and how Silicon Valley was founded.
Often the reason that the employees leave is that the parent is unable to see the value of what they have internally. The new value takes off in new companies founded by talented individuals who get very frustrated when they try to get their ideas working in the parent company.
This is the essence of how Detroit became the center of autos and how Silicon Valley was founded. They tended to be driven by very high rates of spin-offs.
One odd thing about Silicon Valley and semi-conductors is initially there were no semi-conductor companies in Silicon Valley. Initially the centres of electronics in the USA were New York, Boston and LA, with the main firms producing transistors in the 1950s in these centers. Not one was in Silicon Valley until William Shockley opened a factory in 1956 – Shockley was later to win the Nobel Prize, and was with Bell labs in New Jersey. He recruited what turned out to be great talent, eight of whom went on to found Fairchild Semiconductor which was immensely successful until it ran into its own problems. A lot of people left there to found their own companies in Silicon Valley.
The point is Silicon Valley grew out of dysfunction. Texas Instruments in Dallas was a very well run company and very few people left to form spin-offs and as a result Dallas did not become a Silicon Valley.
Another lesson is you don’t see clusters because it’s good to have a lot of companies together in one region. Economists argue the security in numbers thesis. That’s not how they emerge. They emerge from very good people leaving firms out of frustration.
Part of what I argue is that whole countries can have policies that unwittingly prevent this happening – for example Japan because its success is predicated on life time employment. One cost of that is not a lot of firms are formed. Even in the USA we have a lot of talented people who signed non-compete covenants that if enforced would prevent them from starting their own firms in the same industry. The State that does not allow companies to enforce a non-compete covenant is California.
I am always amazed how when I look back how companies miss the boat so much. The number of times people try to get an idea going inside their own company and leave when they can’t. The people who are in power over what ideas to pursue are often very narrow in their backgrounds. People who then go out of their companies meet others with broader backgrounds and can give them an estimate of how the future will work out.
The problem in companies lies with the ability of those in power to evaluate ideas. But the great thing about capitalism is that it allows and facilitates you to leave your company. It is a tricky business – it is not something that companies want to have happen. They don’t want talented people to leave and begin recruiting other talented people.
To the question then why do markets become dominated by so few firms? This is to do with innovation. It is very difficult in most industries to protect innovative ideas. When you innovate you only have a brief window where you can make money. If you are not a big firm and can’t develop a big output in a short space of time, you won’t make money.
Bigger companies have a much bigger chance of exploiting innovations, better products and better productive processes. So markets have a natural tendency to evolve into being dominated by a smaller number of firms. Does it always occur? No! Does it always occur quickly? No! What slows it down are industries with a tendency to grow niches.
For example integrated circuits began as logic devices and evolved into memory devices. There are whole new uses in this process. An industry with niche opportunities invariably brings new firms with them. This slows down the process of dominance but eventually this tends to grow out.
The relationship between dominance and innovation is a paradox. That part of it, the industrial sclerosis, I don’t have a theory but you can see it.
Take Ford Motor Company. In 1908 the company innovated with the Model T and it was not long before they were selling many, many cars. That gave them an incentive to start tinkering with the process and so they unleashed people to tinker with the production line. That more than halved the price of the Model T from $800 to $ 350 by 1915. This was incredible and wealth creating.
Imagine if Honda came out next month with a car that consumes 100 miles per gallon and reduced the price to $10,000. It would generate incredible economic growth. That is where Ford was at. But over the last 40 years it has been the opposite. They have been among the slowest.
What is striking is how conservative these companies have become. In the 1980s GM and Toyota set up a joint venture NUMMI. Toyota stood to benefit through entry to the US market. GM stood to gain from learning about Japanese productivity.
The production system they set up was unlike anything in the USA and it required labor to be much more flexible. Even the United Auto Workers began to like it. It was very well received in fact. But GM killed it. They never instituted it outside the first plant. They had in their hands the key to the future. But even though the Unions liked it and it introduced flexibility, GM did not want it.
It would have required a revolution in the way GM ran its company. You ask yourself would Henry Ford have gone for this in 1910. Of course he would. He lived for creating low cost motoring.
All I can say is when you look at these behemoths – nobody really owns them. A CEO would have to mobilise the owners…. would he take the risk? No, this type of ownership is not conducive to innovation.
GM in the 1980′s is the key to US success. It is one of the crown jewels of the USA. And basically it defaulted on being progressive and down goes the USA. It’s really that simple. If GM had instituted the Toyota system Ford and Chrysler would have to follow suit. But your crown jewels became your handicap. Companies like GM, RCA, Intel, Ford are unbelievable and spurred heroic economic growth, Dial forward!
By Haydn Shaughnessy