The lost 30 years: The decline of Japan’s semiconductor industry

Global industrial competition has entered an era in which “whoever controls semiconductors controls the future.” According to a Gartner forecast, the global semiconductor market is expected to reach a record high of $733 billion in 2025, and semiconductors are at the core of a variety of cutting-edge technologies, including AI, electric vehicles, space development, and quantum computing. Semiconductors are no longer simply components; they have become “strategic materials” that form the foundation of national security, industrial infrastructure, and technological supremacy.

Japan was once at the forefront of this global competition. From the late 1980s to the early 1990s, Japan’s semiconductor industry held more than half of the global market, ushering in a golden age known as “Hinomaru Semiconductor.” In 1986, Japanese companies dominated the top three spots in the Gartner’s semiconductor rankings, with NEC in first place, Hitachi in second, and Toshiba in third, and six companies in the top ten.

However, this glory is rapidly fading due to flaws in institutional design and structural rigidity. In Gartner’s 2024 forecast, not a single Japanese company was ranked in the top ten.

To learn more about out why Japan’s semiconductor industry has fallen into decline, we spoke with Kazukazu Sekiguchi, one of Japan’s leading IT journalists and chief researcher for the Japan Public Relations Association’s special study group, the “Lost 30 Years Verification Study Group.”

Japan dominated the global semiconductor market in the 1980s. What was the reason for this?

Kazukazu Sekiguchi: The background to Japan’s ability to lead the world in the semiconductor industry was its advanced manufacturing technology and a management style based on a general electric company model. Semiconductor technology was originally developed in the United States, but in Japan, general electric manufacturers such as NEC and Toshiba entered the market in earnest. A major advantage of Japanese companies in particular was that they had in-house product groups with clear uses (exits), such as home appliances and personal computers. This clarity of exit strategy made it easy to forecast demand and enabled efficient operation through a vertically integrated management structure, from design to manufacturing and sales.

Another factor in our success was that the scale of semiconductor investment in Japan at the time was still small, allowing flexible and swift investment decisions to be made at the business division manager level. In addition, the emergence of highly competitive semiconductor manufacturing equipment and material manufacturers in Japan allowed us to establish a system that could handle everything from design to manufacturing and sales, enabling us to establish an international advantage. At the time, semiconductors were manufactured by general electronics manufacturers in Japan, and we were fortunate that we had in-house suppliers of semiconductors for home appliances and personal computers.

Why did Japan’s semiconductor industry, which dominated the world until the early 1990s, decline? What are the fundamental causes of this? Technology, policy, or corporate culture?

The loss of competitiveness in Japan’s semiconductor industry is not due to a single cause, but rather to a complex interplay of multiple factors. The biggest turning point was the policy failure symbolized by the Japan-US Semiconductor Agreement. Initially, the US held an overwhelming share of the semiconductor market, but Japanese companies rapidly improved their technological capabilities and production capacity, strengthening their presence in the global market. However, for the US, semiconductors are also the core of the national defense industry, and from a security perspective, there was growing wariness about Japan taking the lead. In response to this growing public opinion, the US government began to strongly urge Japan to open its market.

On the other hand, Japan, with its general electric management style, promoted mass production driven by internal demand. This led to an expansion of market share, but intensified price competition led to what was perceived as dumping. In addition, Japanese companies had a strong tendency to be self-sufficient and were reluctant to actively adopt American-made semiconductors in their products.

It was under these circumstances that the first phase of the Japan-US Semiconductor Agreement, concluded in 1986, included a “side letter” (confidential document) in which the Japanese government promised to raise the market share of foreign-made semiconductors to 20%. This was effectively market intervention and had a major impact on the domestic industrial structure. Furthermore, the second phase of the agreement made this market opening share target explicit, intensifying pressure for structural transformation of Japan’s semiconductor industry. This gradually weakened the competitiveness of Japanese companies and led to a decline in their position in the global market.

Were there any technical biases or corporate culture issues?

In the late 1980s, Japan accounted for over 50% of the global semiconductor market. However, much of that was DRAM, and the vertically integrated model meant Japan was slow to shift to logic used for logical operations and ASICs (application-specific integrated circuits). As a result, profit margins became harder to obtain, and the business lost its appeal. Furthermore, as DRAM capacity increased, the scale of capital investment also ballooned. While investment decisions were initially made at the division manager level, as the scale grew, company-wide decision-making became necessary, and decisions were delayed when the top management was not an expert. The strengths of a general electric company actually became a hindrance. Meanwhile, overseas, there was a shift toward a horizontal division of labor that separated design and manufacturing, leading to the emergence of giant foundries (contract manufacturing companies). Japan’s insistence on vertical integration meant that it was no longer possible to make timely investments as investment amounts increased, significantly undermining its competitiveness.

Some have pointed out that Japanese companies had difficulty adapting to the silicon cycle. What do you think?

The silicon cycle is highly volatile, and for a general electric manufacturer, the impact on business performance is too great. That’s why they tried to separate the semiconductor business. Takashi Kitaoka, who was then president of Mitsubishi Electric, made this decision at an early stage, and there was a lot of criticism at the time. However, I think it turned out to be the right decision in the end.

Hasn’t the rise of overseas semiconductor manufacturers, including Samsung Electronics, also accelerated the decline of Japan’s semiconductor industry?

I think the emergence of new players, symbolized by South Korea’s Samsung Electronics, was a major factor. Samsung is owner-managed, so it makes decisions quickly. It can also make large-scale investments quickly. Japan has a strong culture of vertical division and horizontal alignment, which prevented it from shifting to a horizontally separated model. The vertical integration model was effective in the initial phase of semiconductors, but as the scale expanded, it lost flexibility and was unable to respond quickly, which I think was the problem.

Why has Samsung become so powerful so quickly?

I think a big factor was the collapse of Japan’s bubble economy after the 1990s, which led Japanese manufacturers to refrain from investing. Also, restructuring and retirement age systems at Japanese companies forced a large number of engineers to retire, which led to Korean companies hiring them at high prices. As a result, there was an outflow of Japanese technological brains. This has in part led to a decline in the competitiveness of Japanese manufacturers and an increase in Korean competitiveness.

The rise of overseas companies was probably also due in large part to the strong yen at the time.

After the Plaza Accord in 1985, the yen rapidly appreciated. This was also the time when the Japan-US Semiconductor Agreement was concluded in 1986, and Japan entered a recession due to the strong yen, making it difficult to sell goods overseas. Furthermore, in 1990, the real estate bubble burst due to quantitative restrictions, and corporate investment appetite suddenly cooled. This had a major impact on the semiconductor industry. The previously “aggressive, aggressive investment” was no longer possible.

The Ministry of International Trade and Industry (now the Ministry of Economy, Trade and Industry) tried to restructure the semiconductor industry in order to revive it, but it seems that it did not work out.

The Ministry of International Trade and Industry (MITI) used the reorganization of the mainframe era into the three major computer groups (Fujitsu + Hitachi, NEC + Toshiba, Mitsubishi Electric + Oki Electric) as a model for grouping the semiconductor industry. Subsequently, the government led the formation of “Hinomaru Alliances,” resulting in the creation of companies such as Elpida Memory, which merged the DRAM businesses of NEC, Hitachi, and Mitsubishi, and Renesas Technology, which merged the system LSI businesses of Hitachi and Mitsubishi. However, because these were alliances formed by spinning off companies from the original companies, decision-making was slow and responsibilities unclear. In the end, Elpida was acquired by Micron Technology of the United States, and Japan’s semiconductor industry was swallowed up by global competition. A similar failure occurred in the LCD field. When parent companies are too concerned with face, flexible reorganization becomes impossible.

While most of Japan’s major vertically integrated semiconductor manufacturers have disappeared, material manufacturers and the equipment industry are still thriving. Doesn’t this mean that Japan’s semiconductor industry will be reborn?

Semiconductor manufacturers require huge investments in domestic manufacturing processes, but equipment and material manufacturers can expand into overseas markets, so even if the domestic market is deteriorating, business can be sustained as long as there is demand. However, it is also true that manufacturers in countries such as Korea have gained strength as a result of supplying these equipment and materials overseas for many years. While Korea has strengths in finished semiconductors, it still relies on Japan for many of its parts and materials. On the other hand, however, it could be said that the birth of Lapidus and the entry of TSMC into Japan, which hold the key to the future rise and fall of Japan’s semiconductor industry, were made possible precisely because there were excellent equipment and material manufacturers remaining in the country.

What does Japan need to do to revive its semiconductor industry in the future?

The impact of COVID-19 has led to a global shortage of semiconductors and disrupted supply chains. As a result, even semiconductors for industrial machinery like automobiles cannot be supplied reliably. Until now, Japan has relied on overseas markets, following a procurement policy that said, “Buy parts wherever they’re cheapest.” This global procurement approach is probably the right decision when the global economy is functioning well, but it won’t work in times of crisis. It is essential to maintain a certain level of manufacturing capacity domestically. Semiconductors are not just parts; they can be considered the nervous system of the entire industry. In order to protect Japan’s industrial competitiveness, I believe it is extremely important to rebuild Japan’s semiconductor manufacturing and supply system domestically.

This article originally appeared on CIO Japan.