Stability:
The costs of relying on imports for a significant portion of domestic supply are most obvious
during a period of worldwi~e shortages. Import
prices move above domestic prices (6, p. 11-8).
Moreover the recently released study for the American
Iron and Steel Institute by Pifer, Marshall, and Merrill
(PMM) (23, p. 21), claims to have found evidence that:
. . . the Japanese have aggessively manipulated
export pr ices in order to sell steel in the U. S.
market. In times of shortage, very sharp premiums
have been extracted from U.S. customers; in times
of surplus capac~ty, prices have been reduced
precipitously tb iincrease export volume.
-226-
The PMM study made a fundamental error in its cost
estimation procedure that invalidates its results. The
authors inappropriately compare their estimated costs 'of
making carbon steel products, with the prices of all steel
products. This leads to an overstatement of Japanese
costs in relation to prices by an amount which may dominate
any of the wdumpingW effects they allege. 2l/ Nonetheless,
.---. -.-
the PMM study is another one that expresses concern about
the cyclical pricing of Japanese and European exports.
The essence of these arguments is that the Japanese and
Europeans rely on exports,
by cyclically varying their export
prices more than their domestic prices; to smooth out cyclical
fluctuations in demand., Thus, it is alleged that these
countries have a dual pricing structure with exports priced
higher or lower than domestic products, depending on whether
there is a domestic boom or recession; i.e., there is a
cyclical pattern to the dumping.
The Friden study (7), cited above, purports to have
found evidence of a dual pricing structure. However, there
are a number of limitations with respect to using the
Friden data for the hypothesis that is being examined here:
(1) Friden's data terminate in 1967; (2) no statistical
analysis or tests on the data were preformed; and, most
76/ For the details
chapter.
df this argument, see the appendix to this
¡
-227-
importantly, (3) Friden did not separate the time periods into
boom and recession per iods so that relative export pr ices might
be examined in relation to their hypothesized values. Thus,
while Friden founó limited evidence of dual pricing structures
in countries other than the United States, he did not find
evidence of a cyclical pattern in the dual pricing structure.
To elaborate on point (3)
, this study designates the
..~ .
export price of a steel commodity divided by its domestic price
as the relative export price. The cyclical dual pricing
structure hypothesis asserts that relative export prices rise
in booms and fall in contractions. Consider as an example the
years 1974 and 1975. Since 1974 was an international boom year
in steel demand, and 1975 was an international recession year,
the cyclical dual pricing structure hypothesis asserts that
relative export prices would be higher than average for 1974,
and lower than average for 1975. If in a contraction year,
such as 1975, we found relative export pr ices above average,
that would be evidence tending to disconfirm the cyclical dual
pricing structure hypothesis. It is, therefore, necessary
to examine first whether a country is experiencing above or
below average steel demand in a given year. If steel demand
is above (below) average, then the cyclical dual pricing
structure hypothesis asserts that relative export prices will
be above (below) aver age.
For the purpose of analyzing the cyclical dual pricing
i
hypothesis, a new var iable for each country and each product
-228-
was developed. For a boom year, this variable is defined as
the relative export pr ice for that year minus the sample
average for aii years. For a contraction year, it is defined,
as the average for all years minus the relative export price
for tha t year. The cycl i cal dual pr ic ing hypothes i s al leges
that these numbers are positive for boom years and contraction
Years. Take Japanese carbon steel bars as an example.