Companies still crowd together despite modern communications. Why?
Published: May 22 2001 19:06GMT
As commuters leave their small, barely affordable
houses, push on to crowded trains and head towards their offices in the expensive business district, they may
be forgiven for wondering why companies across the world still squeeze themselves into cities and industrial
clusters. Aren't the costs of communicating across distances supposed to have virtually disappeared?
Unfortunately for these long-suffering workers, while
many of the original forces for agglomeration have faded, new incentives to
group together are springing up - and, so far, they seem to be just as powerful.
A paper by Peter Swann* explains how, in theory, a continuous fall in the costs of
transport and communications could result in a U-shaped pattern of development for cities and clusters. In medieval times, high transport costs and
a lack of communications made large-scale conurbations a practical
impossibility.
As countries industrialised, transport costs fell and economies of scale became
increasingly important. Companies also needed to be near to factors of production such as coal. As the benefits started to outweigh the costs, so towns
and industrial regions formed.
Now, the theory goes, advances in technology are making agglomeration
unnecessary. Companies no longer need to locate themselves near a coalmine
to get a good supply of power. E-mail and videoconferencing mean that distance
is no barrier to communication.
As the reasons for companies to locate themselves near each other disappear,
so expensive and congested cities and industrial clusters should wither. Eventually, the result could be a new "global village", where economic activity is
once again more or less evenly dispersed around the world.
But there is a problem with this theory: the evidence does not bear it out. The
pace of urbanisation has not slowed. As if to prove the point, the very companies
that are driving the communications revolution have populated one of the world's
most important industrial clusters, Silicon Valley.
To understand why low communications and transport costs do not spell the
end for agglomerations, we need to look more closely at the reasons why companies like to locate close to each other.
For one type of company, the answer lies in attracting business. Retailers of
similar goods group together because they want to maximise the chances of a
customer finding them. This produces streets of antique shops, boutiques or
electronics retailers.
Overwhelmingly, though, the reason for clustering is that for most
developed-world companies, employees have replaced capital as the most important factor of production. Companies no longer need to be near to supplies
of natural resources but they do need to be where there is a good supply of
suitable employees who can be used to best advantage. There are two reasons
why agglomerations meet these conditions.
First, industrial clusters and large cities have a pool of labour that can be tapped
by new companies moving there. Companies can set up safe in the knowledge that they will be able to attract specialised workers.
Second, in spite of the advances in communications technology, the most
effective way to communicate is still face to face. The economic literature
distinguishes between information, such as the share price of Vodafone, which
can be easily be "codified" or written down; and knowledge, which is vague and
difficult to communicate.
Information can easily be posted on a web page, faxed or e-mailed. But
knowledge tends to be revealed in more informal ways. New ideas are better
discussed at the pub, in a corridor or over lunch than through more formal
channels. People who know each other personally are more likely to swap ideas
than those whose only contact has been via e-mail.
So while the transmission of information is not sensitive to distance, the
transmission of knowledge is far easier when individuals are near each other.
This effect is hard to measure directly. But an indirect way of testing it is to look at
whether companies that specialise in knowledge-intensive industries have more
of a tendency to group together.
This is precisely the question tackled in research conducted by David Audretsch
and Maryann Feldman**. They classify 210 industries into four different stages of
the life cycle. They find that industries that are at an early stage of development -
and therefore characterised by high innovation - are relatively strongly clustered.
[top]
But as industries mature, the disadvantages of congestion start to mount. Rising
rents, overcrowding and pollution drive companies to disperse. The researchers
add that the very proximity that was so stimulating for an industry in its infancy may, later on, become intellectually stifling and hinder the diversification of
companies into new areas.
The importance of knowledge transfer also helps to explain why so many
clusters, such as Silicon Valley, Boston's Route 128 and Cambridge in the UK,
are found near universities that have a strong research base.
The question of where to locate your company is still a trade-off and changes to
the costs of communications inevitably affect the balance. But advocates of the
global village view forget that even the most sophisticated communications
technology cannot replace the value of a chat over a cup of coffee. Hard as it may
be to believe when you are stuck on the 8.17 train into town, man really is a
social animal.
* The Internet and the Distribution of Economic Activity: G.M.P. Swann, in S
Macdonald and J Nightingale (eds), Information and Organisation, 1999
** Innovative Clusters and the Industry Life Cycle: D. Audretsch and M. Feldman,
Review of Industrial Organisation, April 1996
[top]
