Bold strokes - A strong economic stylist wins the Nobel prize - 經濟

Bold strokes - A strong economic stylist wins the Nobel prize
The Economist Oct 16th 2008
http://www.economist.com/finance/displaystory.cfm?story_id=12429411


WHEN Paul Krugman won the Nobel prize in economics on October 13th, the news
was greeted with nostalgia as well as congratulation by some of his fellow
economists. Since 1999 Mr Krugman has written a twice-weekly column for the
New York Times, in which he has devoted himself to attacking the Bush
administration and all of its works. The nostalgists feel these jeremiads
have distracted him from the cutting-edge research that secured his
reputation. The polemicist, they feel, has buried the theorist.

And yet the old Krugman is still recognisable in the new. Indeed, the arts
of the columnist are not so far removed from Mr Krugman’s style as an
economist. In his most celebrated academic papers, Mr Krugman paints with
bold strokes, striving to render his insights as starkly as possible. Like a
good columnist, he cuts to the quick of a problem, stripping it of clutter
and encumbering nuance. The result is a revealing caricature: what economists
call “models”.

Mr Krugman won the prize for his models of international trade and economic
geography. Both belong to the same grand project he confidently launched just
a year after earning his doctorate: “Before my 25th birthday,” he has
written, “I basically knew what I was going to do with my professional life.
” In 1978 he realised that a model of “monopolistic competition”,
published a year earlier by Avinash Dixit and Joseph Stiglitz, could help him
introduce economies of scale into trade theory and beyond.

Economies of scale had long posed awkward problems for theorists. If bigger
firms face lower costs, then in principle one firm should supply the entire
market, thereby enjoying the lowest costs of all. But in the Dixit-Stiglitz
model, this monopolising logic is offset by a countervailing force: consumers
’ taste for variety. People prefer to spread their custom over different
versions of the same good. The market is therefore carved up among competing
firms, each offering a product bearing its own distinctive stamp. The model
is highly stylised. Nonetheless it gave Mr Krugman, as he put it, “a tool to
open cleanly what had previously been regarded as a can of worms”.

Mr Krugman used this tool to save economics from an abiding empirical
embarrassment. According to one of the discipline’s founding doctrines,
countries gain from specialisation and exchange, concentrating on what they
do best and importing the rest. The theory explains why the Portuguese might
sell wine in exchange for English cloth. But it cannot explain why similar
countries, blessed with similar ratios of capital, labour and land, should so
vigorously trade similar goods back and forth. This is not a small blind
spot. According to the World Trade Organisation, 52% of Germany’s exports to
France are things France also produces and exports to Germany. But the
Dixit-Stiglitz model, with its subtly differentiated firms competing for
variety-loving consumers, lent itself to explaining why Germans might import
Renaults, even as the French imported Volkswagens.

Mr Krugman’s model showed that when trade barriers fall, firms gain access
to bigger markets, allowing them to expand production and reap economies of
scale. But openness also exposes them to competition from rival foreign
firms, paring their margins. Some firms may go out of business. But between
the domestic survivors and the foreign entrants, consumers still have more
goods to choose from. Thus the gains from trade arise not from
specialisation, but from scale economies, fiercer competition and the
cornucopia of choice that globalisation provides.

Scale economies also allowed Mr Krugman to give economics for the first time
a sense of space. In a 1991 article, he notes that night-time satellite
photos of Europe reveal the distinctive contours of economic activity: bright
lights cluster around metropolitan centres, shining particularly brightly
around the triangle of Brussels, Amsterdam and Dortmund.

Before Mr Krugman, economists found these images difficult to square with
the rest of their body of theory. They were accustomed to assuming that firms
face constant returns to scale. But if that were true, then every peasant
could build a small smelter or assembly line in his backyard. There would be
no need for an economy to divide into a farm belt and an industrial belt.

Geography lessons

In Mr Krugman’s model, by contrast, big factories benefit from lower costs
of production. Manufacturing firms might therefore cluster near to a large
market, leaving behind a sparsely populated hinterland, in order to make the
most of scale economies and minimise the cost of transporting goods to their
customers.

Earlier theorists had instead assumed that firms herd together to benefit
from some kind of “spillover”. Perhaps firms pick up tricks of the trade
and other know-how from their neighbours. However plausible, these
explanations were nonetheless unsatisfying. Because economists could not
measure spillovers or delimit their scope (“How far does a technological
spillover spill?” Mr Krugman wondered), they could invoke them to explain
just about anything.

Mr Krugman’s models instead identified a less elusive benefit of proximity.
He pointed out that a firm’s decision to locate in a district is a gift to
other firms in the area, because in attracting new workers it also brings new
customers. Unlike a technological spillover, this gift would in principle
leave a paper trail, showing up in local firms’ sales figures.

In neither contribution did Mr Krugman claim great originality for his ideas
or great realism. His achievement was to formalise insights that many people
had previously had informally. Ideas that had fluttered in and out of people’
s grasp for decades, he pinned down like a butterfly on display. Sometimes a
good economist, like a good columnist, succeeds not by making a point before
everyone else, but by making it better than anyone else.


--