Competition and Trade in the Global Semiconductor Industry
CASE B (1986-1991)
Coming to an Agreement
1986, Japan became the world’s largest producer of semiconductors as
well as the largest market for semiconductors. The trend was felt in the
United States as nine of the 11 U.S. DRAM producers left the market,
resulting in Japan’s world dominance of this largest market for
semiconductor products. The EPROM producers in both Japan and the United
States suffered great losses as well. At this time, there appeared to be
no signs of relief for U.S. semiconductor producers.
numerous previous attempts to reverse the negative American trends that
had become the focus of the complaints made by the U.S. semiconductor
industry had not achieved the desired results. But the Section 301
petition filed by the SIA, the EPROM and DRAM dumping suits filed by
three U.S. chip makers, and the self-initiated suit by the U.S.
Department of Commerce had finally created an atmosphere that demanded
the attention of Japan. The U.S. Department of Commerce suit was
particularly alarming as Japanese semiconductor firms were declared to
be selling DRAMs at levels far below the cost of production and dumping
memory devices in the U.S. market by up to 34.5% of cost. With all of
these pressures on Japan, a formal agreement was finally initialed by
MITI, DOC, and USTR to address the problems of dumping and market
U.S.-Japan Semiconductor Agreement that was formally signed in September
1986 had three basic conditions. The first condition was that Japanese
firms would stop dumping in all world markets, not merely the U.S.
market. This was a precedent setting condition since it was a bilateral
agreement that governed behavior in third-country markets. In addition,
the Japanese firms were to maintain detailed cost records that were to
be the basis under which each firm would set a “fair market value” (FMV).
The FMV became the selling price for the firm, as it was based on the
total cost of production plus an 8% profit. Each firm could sell at any
price as long as the price was at or greater than its “fair market
value” (FMV). Thus an efficient Japanese firm could sell below its
U.S. or Japanese competitor’s prices so long as its price was above
the calculated FMV for its product.
second condition of the U.S.-Japan Semiconductor Agreement was that
Japan would encourage and expect that foreign semiconductor firms
achieve an increased share of the Japan market. A
specific target market share for foreign firms of 20% was included
in a side letter to the agreement. Though the side letter was considered
to be a secret, in time it would become the SIA’s quantitative measure
of compliance by Japanese firms .
third condition of the agreement was that the U.S. government would
suspend the antidumping
duties estimated to be as high as $1 billion. Consequently the agreement
balanced Japan’s promise to cease dumping and to open its market in
return for a suspension of the hundreds of millions of dollars in
penalties. It was expected that full compliance with the first two terms
of the agreement would take the Japanese firms some time to achieve.
the interested third parties observing the negotiations and the
agreement were the semiconductor producers of Europe, who
were highly critical of the agreement. The DG12, the EC’s
directorate of information technology, wanted to ensure that the
European chip producers would have equal opportunities to increase their
access to the Japan market. At that time, their share was less than 1%
of the total. Specifically, the European chip producers were concerned
that the bilateral agreement would be discriminatory in granting the
United States primary access to the Japanese market. The European
producers filed a complaint with the GATT on the basis that the
U.S.-Japan agreement violated the GATT principles of multilateral
agreements. A GATT panel would eventually find that the export and
production controls imposed on Japan were contrary to GATT principles
but that the agreement did not provide preferential market access to the
chip producers were not alone in their criticisms of the agreement. U.S.
consumers of semiconductors were alarmed that, as fair market value
prices began to take effect, the prices of DRAM chips had risen to
several times the pre-agreement prices. They were concerned that more
expensive chips would make their computer and electronic products
noncompetitive. And even though continued high growth was predicted for
the semiconductor market, U.S. chip makers too were not satisfied. They
believed that now, six months after signing the agreement, dumping had
still not ceased in third-country markets and that access to the
Japanese semiconductor market had not significantly improved. Only
American semiconductor production equipment producers seemed unaffected.
On March 27, 1987, after listening to strong complaints from SIA¾and
to the surprise of many in Japan¾President
Reagan placed sanctions of $300 million on Japanese imports for failure
to comply with the U.S.-Japan agreement. The SIA supported the president
but decided to refrain from taking a position with regard to products
that would be sanctioned. The sanctions amounted to customs duties of
100% of the value of Japanese imports of televisions, computers, power
tools, and other products for which U.S. consumers had alternative
country sources. Although this action did not apply to Japanese
semiconductors, it did result in a backlash from the Japanese producers
of consumer goods against their country’s chip makers who had not
complied with the agreement.
the sanctions were effective in encouraging the Government of Japan to
act, the SIA and chip consumers were unhappy with the uniform floor
price that MITI had imposed as a result on Japanese chip exporters to
satisfy the conditions of the agreement. The SIA implored the Reagan
administration not to lift the remaining sanctions until the MITI price
monitoring became company-specific. After MITI provided assurance that
production export controls had been eliminated and dumping in third
countries had ceased, sanctions related to dumping were removed at the
end of 1987. The Japanese buyers also began to recognize that purchase
of foreign-made chips was necessary to meet the market access terms, but
those sanctions related to market access remained, pending full
compliance with the 1986 agreement.
the advent of 1988, several new events appeared to bolster the fortunes
of U.S. chip makers. The SIA obtained government approval in 1987 for a
new organization, SEMATECH. This consortium of 14 U.S. chip makers was
expected to usher in the next phase of U.S. semiconductor industrial
development. SEMATECH’s goal was to support new manufacturing
technology development through the use of U.S. industry funds and
matching government grants. The ultimate result to be expected was
superior American products. Another event, the combination of a rise in
demand for chips and a rise in value of the Japanese yen relative to the
U.S. dollar began putting further economic pressures on the Japanese
producers to raise their export prices, thus favoring U.S. competitors.
These events seemed to reduce the intensity of the conflict.
some easing of tensions between the U.S. and Japanese industries in
1988, the SIA invited the Electronic Industries Association of Japan (EIAJ)
for a meeting in Monterey, California. The overriding goal of the
meeting was to explore ways in which U.S. chip makers could increase
their opportunities in the Japanese market, but SIA’s specific
objective for the event was to promote long-term relationships between
U.S. producers and Japanese users. At the meeting, U.S. industry
representatives advocated the benefits of designing in foreign-made
semiconductor components in the production of finished goods in Japan
and of broadening the base of users and suppliers in the market access
effort. Following this meeting the EIAJ formed the Users Committee of
Foreign Semiconductors (UCOM). UCOM would hold seminars to promote
market access and review market access improvements. This meeting was
followed by the establishment of a Distributors Association of Foreign
Semiconductors (DAFS) that would become the link for distributors of
U.S. products in Japan to increase market share there. In conjunction
with the SIA’s Japan chapter, the foreign suppliers, Japan users, and
foreign distributors were organized and able to begin the most extensive
series of market access activities in history. Because of the 1986
Semiconductor Agreement, these activities covered a range of
applications from televisions to cellular phones to automobile engine
By 1988, the Reagan administration and Congress appeared to have
developed a succinct set of a policies for dealing with the complaints
of U.S. semiconductors manufacturers about Japan business practices. In
cooperation with the Congress, the USTR, and the Department of Commerce,
the Reagan administration carried out policy actions to retaliate
against alleged dumping in the U.S. market and restrictive business
practices ostensibly limiting market access for U.S. firms in the
Japanese market.. In addition, in 1988 Congress approved formation of
the National Advisory Committee on Semiconductors (NACS), a committee of
high-level government and industry officials. This committee would
become the platform through which proposals could be articulated for
further development of the national strategy for the U.S. semiconductor
industry. But then came a new cause for concern regarding trade policy.
Inauguration of George Bush as president and the inception of his
administration in 1989 forced all parties to reassess the semiconductor
agreement and its impact.
SIA began its appraisal of the position of the new administration by
sending its officials to Washington to request that the White House
preserve the integrity of the Semiconductor Trade Agreement (STA) by
keeping the heat on the Japanese to comply with the agreement. In
addition, the SIA began to respond to renewed criticisms from
journalists that the agreement was allowing Japanese producers to gouge
chip consumers. The SIA replied that prices in the DRAM market rose
because Japanese producers were able to drive out U.S. competition by
selling their products below cost. It insisted that the terms of the STA
were not intended to harm U.S. consumers of chips since prices were to
be based on FMVs of each company plus an 8% margin for profit. In
addition, the SIA issued a report by a business research group that
concluded that DRAM prices were expected to rise, regardless of STA, as
a result of dumping that occurred in the mid-1980s. The report also
claimed that prices were consistent with the long-term projected figures
for the overall time period beginning in 1984 and ending in 1989. U.S.
journalists, however, clung to anecdotal evidence suggesting that the
STA gave Japanese manufacturers a license to cut production of 1K DRAMs
and then boost up the prices.
DRAM markets also saw prices rise worldwide in the first half of 1989
with 256K DRAMs prices remaining higher than their lowest point in 1986.
Higher prices meant that $3 billion to $4 billion in additional profits
on sales were being made. Half of Toshiba's 1989 profits came from its
semiconductor division. Prices finally dropped in the second half of
1989, a result believed to be partially due to the introduction of new
four megabit DRAM devices. By 1991, however, these were priced so low
that the costs of building production lines for them could not be
recovered. There were two reasons for this new price decline: a cyclical
slump in demand, and emerging new competition from Korea. Samsung, a
Korean chaebol had appeared on
the scene as a new supplier, and Korea's global market share of sales in
DRAMs climbed to 18% by late 1989.
this time, most major U.S. chip consumers reported that they were able
to obtain adequate supplies at competitive prices, and the gap between
spot and large contract prices had disappeared. These changes spelled
the end of Japan's total dominance in DRAM production. Its global share
fell from 80% in 1987 to 57% in 1991, while its North American share
also fell from 61% in 1987 to 52% in 1991. Furthermore, it was not only
Japan’s global share in the DRAM market that fell, but also its share
in the overall semiconductor market, which dropped further in 1991.
However, the prices of DRAMs remained high enough for the Japanese
producers to remain profitable through the end of 1990 during which time
prices remained above individual FMVs. Two American DRAM producers,
Texas Instruments and Micron Technology, also gained during the increase
in prices. They were the only two who were not squeezed out of the
market during the abrupt decline in U.S. producer share of the global
DRAM market in 1985. After the U.S.-Japan agreement, this decline halted
and the United States actually won back a small increase in global
market share from 17.9% in 1987 to 19.8% in 1991. However, the U.S.
share of the North American DRAM market continued to decline, from 31%
in 1987 to 27% in 1991.
effects of EPROM pricing were not quite as dramatic as those of DRAM
pricing. No EPROM consumers complained about their price increases or
pointed an accusatory finger at the trade agreement. U.S. producers also
had a much different and better experience with EPROMs than with DRAMs.
In 1991, U.S. EPROM producers had 64% of the North American market and
25% of the Japanese market.
By 1989, the dumping had clearly abated, and new sources of DRAMs
were available. Many considered that the crisis in dumping was over.
With regard to access, U.S. firms had taken the initiative to increase
their sales in Japan by setting up offices and centers in Japan
beginning in 1986. With the assurances provided by the U.S.-Japan
agreement, U.S. suppliers added 30 sales offices to the 42 previously
there. They also opened 16 new design centers, which was a fourfold
increase from before the agreement. Six new testing centers were added
for a total of 18, and 4 new failure analysis centers were added for a
total of 15. Total additional sales offices and other facilities
amounted to one per month during five years of the agreement. The sum
total: U.S. firms' expenditures for personnel in Japan increased by 32%,
for capital 162%, and for sales offices, 86% between 1986 and 1990.
Despite these efforts, SIA provided evidence to the Bush
administration in 1989 that no progress had been made on gaining access
to the Japanese market. The U.S. government pressed Japan for action,
which led the Japanese system’s firms finally to begin developing
closer relationships with U.S. chip producers. This may have been due to
a fear on the part of Japanese producers that Washington might accept
the SIA's recommendation that semiconductors continue to be a top
priority under Section 301. As a result, there was a significant
increase in U.S. market share between early 1989 and the end of 1990.
This increased market share translated into additional annual sales of
over $1 billion for U.S. companies.
As the agreement neared its end in 1991, however, there was no
further increase in market share despite the efforts made by U.S. firms.
The market share of foreign producers was still well below 20%. With an
agreement whose terms were yet to be completely met, the SIA pressed for
and the U.S. government agreed to campaigning for another semiconductor
agreement with the Japanese government.
problems could you expect based on the situation as it was in 1990? What
would be the likely points of contention between the United States and
Japan in formulating the next agreement? How would it be different from
the first agreement? How strong was the possibility that the United
States and Japan would no longer be primarily facing each other in these
debates? How much of a threat were the emerging semiconductor producers
from third countries in obtaining world market share? What kind of role
would the SIA and U.S. government play in the next round of
negotiations? What about EIAJ and Japanese government actors? What could
you predict to be the future market share that the U.S. would hold in
Japan for semiconductors? What about third country share in Japan? And
Europe’s share in Japan?
WORLDWIDE DRAM MARKET 1976 - 1991
B.2 FOREIGN FIRMS SHARE OF THE
JAPANESE SEMICONDUCTOR MARKET
1973 - 1991
B.3 FOREIGN MARKET SHARE IN
JAPAN, (by quarter 1986 -1991) AS DEFINED BY THE 1991 U.S. - JAPAN
B.4 WORLD SALES BY REGION AND
BY TOTAL 1984 -1991
B.5 WORLD MARKET SHARE U.S., JAPAN, & OTHER 1982-1991