more than 80 per cent. The cost of a three-minute phone call from New York to
London fell from $250 in 1930 to few cents by the 1990s and with the internet
it is now close to zero (O’Rourke and Williamson, 1999; Wolf 2004).
Trade liberalization: the debate
There has been an extensive theoretical and empirical debate about the impact
of trade liberalization. The general trend over recent decades has been towards
more support from academics and policy-makers for trade liberalization.
There are important pre-conditions for trade liberalization to be good for
growth: a sufficient time lag for the economy to adjust; sufficient infrastruc-
ture; trade liberalization considered credible; liberalization not squeezing out
other more beneficial reforms and not leading to a serious squeeze on govern-
ment finances.
The first step towards the contemporary popularity of trade liberalization
was the accumulation in the 1970s of empirical evidence against trade protec-
tion and import substitution. In the late 1960s the OECD sponsored a study of
seven developing countries which were summarized in an influential volume
by Little, Scitovsky and Scott (1970). These studies found that the use of tariffs
had raised the relative price of industrial output and so had motivated a shift in
investment from agriculture to industry. These newly emerging industrial
sectors were discovered to be very inefficient as tariffs had simultaneously
removed the pressures to compete against imports by improving quality or
price competitiveness. The use of over-valued exchange rates was a common
tool to promote import substitution by making the imports of capital equip-
ment and inputs for industry cheaper. This tended to made exports less compet-
itive so made export pessimism self-fulfilling, reduced the incentive to
produce such capital equipment domestically and biased domestic production
to the use of imported and capital-intensive production methods. This latter
feature led to slow employment growth as employers imported machines
cheaply to carry out tasks in factories and farms rather than employ people,
reduced the progress of poverty reduction and led to continued import depend-
ence. The use of tariffs, quotas and licenses to allocate resources replaced the
market with a bureaucratic form of allocation and opened up opportunities for
corruption of the political and administrative systems.
The second step was a growing belief that governments, however well-
intentioned, would be unable to implement good trade interventions. Theory
and evidence offered no guide as to what ‘good trade policy’ actually consti-
tuted. How, for example, were policy-makers in the 1950s and 1960s supposed
to distinguish an infant industry? How much protection was necessary to stim-
ulate dynamic externalities? In practice many countries granted trade protec-
tion as soon as production became feasible and the policy contained no
provision to measure whether learning was occurring or externalities being
stimulated or for protection to be reduced after a specified period. Even had the
International Trade, Openness and Integration 275