Japanese Companies Losing Their Ability To Exploit New Technology, Part 2

Dearth of PhDs At Companies, Shutdown of Industrial Labs

Source: https://tinyurl.com/4by66ktu 

[FLASH JAPANESE ELECTIONS: I’m going to wait for the dust to settle a bit before musing about the possible longer-term implications of the LDP's big defeat. Just a couple of things worth noting. Within the LDP, “Most of the 46 members who were punished for their role in the financial scandals were conservative members of the now-disbanded Abe Faction, and of the 46, only 18 will be returning to the Diet.” Secondly, this was a vote against the LDP, but not for the leading opposition party, the Constitutional Democratic Party (CDP) as documented by Tobias Harris. While the CDP gained 50 seats to reach 148, the number of votes it got was only a smidgeon higher than in 2021. It did not bring out the “floating voters” as Junichiro Koizumi did in 2005 and the old DPJ had done in 2009. In fact, the overall turnout was the third lowest under the current electoral system (see chart). The LDP has stayed in power, not because voters are happy, but because so many who don’t like it see no alternative and stay home. The biggest question is this: Was the repudiation of the LDP caused mainly by the ongoing financial scandals? Or, were these scandals just the trigger that brought to a head long-simmering frustrations over the LDP's failure to revive the economy.]

Source: https://www.statista.com/statistics/1263233/japan-voting-rate-house-of-representatives-elections/

In a world where new science and technology can easily cross borders, the fastest-growing rich countries are not necessarily the ones creating the most new inventions. Rather, it’s the countries that figure out how to apply inventions to the broadest range of new or improved products, plus more efficient processes for making existing products. Japan used to be a master at this, but, as I detailed in a previous post, they’re not as good as in the past, nor as good as their counterparts in other rich countries.

            I recently asked several Japanese experts to ask how this deterioration happened. While Japan’s leaders are focused on overcoming its relative decline in basic research—a worthy goal that I discussed before—these conversations convinced me that this won’t solve the problem unless companies regain their past ability to transform scientific advances into economic value.

A Tale of Two Technologies

In 2019, Akira Yoshino shared the Nobel Prize in Chemistry for his work in the 1980s on the rechargeable lithium-ion battery. Two predecessors had laid the foundation, but it was Yoshino’s breakthrough that, in 1991, enabled SONY, a joint venture between Toshiba and Asahi Kasei, and later Panasonic, to turn an enticing notion into a lucrative product. The L-I battery is a so-called “general purpose technology,” i.e., an invention that, like electricity or the Internet, can be applied to a myriad of products.  In 1985, Toshiba created the world’s first mass-market laptop, and a decade later, laptop producers switched to the L-I battery because it was lighter and could run longer on each charge. In 1995, NTT Docomo developed the world’s first mobile phone using the Internet. It, too, was powered by the L-I battery.  Finally, in 2010, Nissan came out with the world’s first mass-market electrical vehicle, also powered by the L-I battery. Without the L-I battery, none of these invaluable products could have triumphed.

            Even had the L-I battery been invented abroad, Japanese companies could still have pioneered these applications, just as SONY created its radio around the American-invented transistor. Certainly, commercial is just as indispensable as technical prowess. However, unless a company has personnel who really understand what a given technology can do, their business imagination will be for naught. You see, none of these pioneering products could use the L-I battery until the producers’ own top-notch chemists and metallurgists modified it to make it lighter, longer-lasting, cheaper, and less likely to catch on fire. Some observers mistakenly dismiss this process as mere “incremental” improvements (see this post).

            And that brings us to the second tale, this one concerning quantum computing, destined to become one of the “big new things” in digital technology. It will be critical to drug discovery, long-term weather forecasting, finance, and other data-heavy areas, while classical computers will remain more suitable in other areas.

            Japan has top scientists who are helping to advance the science of quantum computing. In the lead on this are Tokyo University and the government-affiliated RIKEN Center for Quantum Computing. They sometimes collaborate with private companies like Fujitsu and IBM. Tokyo University hosts Professor Yasunobu Nakamura, who created the basic element of superconducting quantum computers, and Professor Akira Furusawa, who has achieved impressive results in developing the optical quantum computer. Yet, Japan’s private companies lag behind others in developing commercial applications that will emerge within the next five to ten years. It’s as if Japanese companies had failed to pioneer commercial uses for the L-I battery.

            The reasons for this were explained to me by two experts. One is Koichi Akaishi, a veteran of the Ministry of Economy, Trade, and Industry (METI), who, before his retirement, was seconded to the Cabinet Office to be Vice-Minister for Science, Technology and Innovation. The other is Takehiro Ueyama, a veteran academic who is now Chief Executive Member at the Cabinet Office’s Council for Science, Technology and Innovation. I met them together.

I met him together with Akaishi.

PhDs and Industrial Labs

“I believe Japan’s shortcomings in basic science and in company absorption of that science are not only equally important, but they reinforce each other,” Akaishi explained. One significant factor, he said, is that, “Too few Japanese companies hire PhDs, and so most of their technologists can’t understand the science well enough to envision the possible applications and what would have to be done to make such applications practical.” In Japan, only 14% of PhDs work at private companies; in the US, the share of PhDs working in private companies doubled from 24% in 2002 to 48% in 2022. Europe appears to be between these two poles.

 The dearth of PhDs in companies creates a vicious cycle hindering both basic science and commercial innovation. “Even when Japanese companies hire PhDs, they don't understand the value of these experts. They pay them only a tiny more than a normal graduate. That gives university graduates little incentive to acquire a PhD. The ones that do get a PhD have a career at a university rather than in private industry. Rather than thinking about commercial innovation, they write academic papers. That, in turn, gives companies little expectation that universities can come up with good ideas for commercial innovation, and so they have little incentive to fund university research.”

            Ueyama pointed to a parallel problem. In the 1970s and 1980s, several large corporations had created industrial labs (also called central research institutes) that focused on profound issues of science and technology, now called “deep tech.” They were akin to institutions like AT&T’s Bell Labs, Xerox’s Palo Alto Research Center, and others in the US. Such labs have a freer hand than the standard R&D Departments to work on ideas that may not lead to products for a long time, or often not at all. However, the rare successes, like Bell Labs’ transistor, can change the world.

            Unfortunately, with the onset of the lost decades and the bad debt crisis, said Ueyama, one after another of these research institutes was closed. In the US, he explained, when such labs were closed, “various industrial policies were adopted to encourage the excellent researchers who had worked there to start up new companies. One example is the SBIR (Small Business Innovation Research) system, which was created in the US in 1980 to help finance innovation by small and medium-sized enterprises. Japan's biggest industrial policy failure was when the government turned its own SBIR system into a policy to bail out small and medium-sized enterprises. This resulted in the survival of many ‘zombie’ companies. No policy was implemented to create startups from researchers at Japan's central research institutes.”

Why Not Recreate Industrial Labs

Japan’s leading companies certainly had the cash to recreate these labs once the debt crisis was overcome. The fact is that Japan’s 5,000 biggest companies now generate more cash flow (retained earnings plus depreciation) than they know what to do with. Each year, their cash flow exceeds their capital investment by more than 4% of GDP (see chart below).

Source: https://www.mof.go.jp/english/pri/reference/ssc/historical/all.xls

            If they can easily afford to invest in basic research, then why don’t they? Part of it is the same reason they don’t use their cash hoard to raise wages enough or pay higher dividends. The trauma of the debt crisis still rules the minds of executives who had endured it in their younger days. They are hoarding cash “in case of a rainy day.” When Covid happened, many said this showed the wisdom of their “rainy day” posture.

            The greater the company's trauma back in the 1990s, the lower their investment, whether in plant and equipment or in R&D, especially R&D with no obvious early payoff. Profs. Yaichi Aoshima (whom I also met) and Hideto Koizumi looked at 700 large companies (sales up to ¥100 billion or $725 million), and they divided them into two groups. A “treatment group” suffered net losses in the early 1990s but have been profitable since then. A “control group” suffered no unprofitable years in the early 1990s, just big declines in profits. The treatment group reduced their annual investment so much that they showed a small decline in tangible assets (land, buildings, equipment) from the mid-1990s through 2021group (see chart below)! Moreover, they’ve not increased their spending on R&D since 2007. By contrast, the control group invested enough to keep expanding their tangible assets, and they grew their R&D spending by 20% from 2007 to the pre-Covid period.

Source: Yaichi Aoshima and Hideto Koizumi. From Impact Of Economic Shocks On Investment Preferences Of The Japanese Firms, a conference paper. Reprinted with permission

            Among all Japanese corporations (large, medium, and small), R&D had more than doubled from 1987 to 2007. After 2007, by contrast, it increased just 0.2% a year (see chart at the top of the blog). What a short-sighted decision.

            It’s not just about failure to invest enough in R&D, but what kind of R&D they do and how much long-term benefit companies get for each dollar of R&D. I’ll discuss that in the next installment.

Also In the Next Installment: Why Japanese firms hire fewer PhDs and how that hurts them; immobility of star scientists and engineers across firms; lack of collaboration (including across borders)