Book review

Divided Sun

MITI and the Breakdown of Japanese High-Tech Industrial Policy,

1975-1993

by Scott Callon

Stanford University Press, 1995; ISBN 08047-2505-5

reviewed by John Boyd

Japan's electronics industry celebrated a landmark event in 1985. This was the year Japanese semiconductor manufacturers overtook their US rivals to become the world's largest producers of silicon chips. Not since the 1957 launch of Sputnik by the USSR had US pride been so up-ended. In a technology American firms had pioneered, perfected, and parlayed into the most important new industry since automobiles and aircraft, Japan was now top gun.

Japan's success has often been attributed to the Ministry of International Trade and Industry's VLSI program. This was MITI's oft-praised effort of the 1970s to get domestic semiconductor manufacturers to pool R&D resources, and work together to catch up with a US industry several generations ahead in the art of silicon integration.

In Divided Sun, author Scott Callon does not dispute that the VLSI Consortium helped Japanese firms become world leaders, particularly in DRAM memory chips. What he does discredit is the myth that Japan's success was a result of firms laying down their competitive arms, under orders from MITI, and cooperating to beat the foreign devils at their own game.

"Although the Japanese VLSI effort has been proclaimed by Western and Japanese analysts to be a triumph of Japanese joint R&D over splintered US efforts, the reality is that an estimated 85 percent of VLSI research and development was done separately by the Japanese firms in their own labs. And much of the remaining 15 percent of the funds that was devoted to joint R&D went to a particular technology bet... which did not pay off."

In other words, Japan caught up with and overtook the US in semiconductor manufacturing not through cooperation, but through the individual efforts by Japanese conpanies generously fueled with MITI money.

The truth of the matter, says Callon, is that "Japanese firms do not like to cooperate under government mandate, and despite all MITI's efforts to try to force them to do so, they largely do not." This is putting it kindly, as Callon reveals in his examination of three other consortia: Supercomputers, Fifth Generation, and TRON.

Perhaps the nadir of intercompany antipathy came during the Supercomputer Consortium, MITI's attempt to coerce major electronics companies to jointly "build a better Cray-1" (the pioneering supercomputer). Hitachi, NEC, and Fujitsu each worked separately on different components of the supercomputer, but when it came time to integrate them in a Fujitsu factory, Fujitsu -- fearful of accidentally revealing company secrets -- went to extraordinary lengths to keep the visiting engineers isolated. They were not permitted to ride Fujitsu's commuter bus to the outlying factory, so had to take taxies daily; they were prevented from eating in the Fujitsu cafeteria; and, most comical of all, they were forbidden "even to see the main computer that they were working on [and so] had to ask Fujitsu employees to test their subcomponents and come back and tell them what had happened."

Callon interviewed nearly 100, mostly anonymous, participants in the four consortia and has delved into official and commercial publications to piece together similar wretched stories of wasted talent, effort, money, and time. Both Japanese and Western readers can learn much from this gold mine of detailed information that sheds light on the workings of Japan's high-tech consortia. But be warned: The book's heavy academic style forces readers to dig through a convoluted presentation in order to find things precious.