Takaichi Wasting Billions On Illusory Technological Magic Bullets, Part I

A 21st Century Industrial Policy: Diffusing Technology from Leading Companies to Laggards

Source: Mizuho Research Institute at https://share.google/3j1XvH5C8T16Jx9cn Note: log growth scale -0.3 is around 30% relative decline; chart is explained in the text

For decades, Japanese policymakers have searched in vain for technological magic bullets to end 35 years of lost decades. Fifth generation computers, superconductivity, lasers, nanotechnology, etc. etc. Under Sanae Takaichi, the magic bullets du jour lie in 17 special sectors where she intends public-private investment. They range from pioneering technologies like AI, perovskites (thin mineral films for solar power), and quantum computing to pedestrian sectors like shipbuilding and port logistics. Some are supposed to jumpstart growth, while others posit independence from China as a reason and/or a pretext.

Her efforts are as likely to fail as past gimmicks. Undoubtedly, some of her choices are worth abundant government investment in basic research and, at some point, sometimes even subsidies for producers and consumers. Perovskites are a sorely needed supplement because they expand the flat space for solar power on windows and sides of buildings in land-scarce Japan. However, their high cost and short lifespan make it a fantasy to think they will allow Japan to dispense with conventional solar panels. If security vis-à-vis China is the actual motivation, it might be more effective to partner with India to help expand the latter’s growing solar panel capacity as a partial alternative to China. I discussed the illusions regarding Rapidus in a previous blog.

More importantly, her whole posture is based on a fundamental misconception. Technology does not produce economic growth. It is only when companies transform technology into economic value that growth is ignited. While Japan’s scientists have often pioneered advances in basic science and technology, the country suffers from the notorious “death valley” between the laboratory and commercialization, as do many other countries.

Even when Japanese companies successfully commercialize or import technology, its use is all too often limited to a thin layer of corporate giants. It fails to spread to the median company. As a result, productivity growth remains hobbled at companies that employ more than 90% of all Japanese workers. This is the dreaded “High-Technology/Low-Productivity Trap.” The severity of Japan’s trap is among the worst. It also seems the case in China, a topic I’ll examine down the road.

Part I begins to analyze the problem. In the final part, I will recommend concrete solutions.

Growth Requires Diffusing Technology To Ordinary Companies

What counts for a country’s growth is not how much technology it generates but how much it uses, and how well. That, in turn, depends on the strategy and management smarts of its companies, as well as the skills of its employees. A thin layer of companies, known as Global Frontier (GF) corporations, exploit technology—no matter where invented—to fashion new products and processes, boost company productivity, and grow their market share. SONY did not invent the transistor, but its little radio created far more economic value than the transistor itself. By contrast, technological brilliance, when not married to commercial genius, is economically sterile. Look at Nissan’s failure to make the Nissan Leaf EV live up to its potential.

Normally, technology is first commercialized by GF companies, then spreads to national frontier (NF) companies, and finally to ordinary companies within a country. When it spreads “from the best to the rest,” the great mass of companies creates economic value many times that created by the original innovators. Consider all the companies using the Internet.

The chart at the top of this post compares GF companies with the leading companies on stock markets in the US, Japan, and the rest of the OECD. GFs here are defined as the 50 companies in each sector with the highest productivity. We care about productivity because, without it, living standards cannot improve. Productivity in this case includes the combined productivity of labor and private capital, e.g., factories, offices, stores, etc., which is known as Total Factor Productivity (TFP). TFP shows how much of an increase in output you get for every 1% hike in capital and labor inputs. In the chart, the productivity of GF firms is set to zero, and a negative value indicates that non-GF companies fell behind the productivity growth of GF companies between 1999 and 2009. Here’s a simpler way to think about it. Let’s use 1999 as the base year, with an index of 100 for all companies. Suppose the productivity of GF companies doubled to 200 by 2009, but only to 160 in Japan, or just 80% of the GF level. In that case, even though Japan’s productivity grew, it fell behind the GFs by 20%.

As the chart shows, America’s NFs kept pace with the GFs. European productivity grew, but more slowly, so that Europe fell almost about 18% behind the GF moving target (a log growth scale of -0.2 equals an 18% decline). Japanese companies experienced even slower productivity growth and thus fell further behind the GFs by 36%.

The Great Divergence

In the last few decades, the world has moved to a “winner take all” economy when it comes to productivity growth. In the past, by contrast, the gap between frontier companies and the average firm was a positive leading indicator of growth in per capita GDP. The bigger the gap—i.e., the more that leading companies advanced the technological frontier—the faster the diffusion of that technology would improve productivity in the average companies. Since the frontier is a moving target, this never-ending catch-up process fueled growth via productivity. It is part of why countries like Japan, Korea, and China were able to grow so fast in their catch-up periods.

Now, however, things have changed, a phenomenon the OECD calls “the great divergence.” We can see this in the chart below. In both manufacturing and non-financial services, the labor productivity of GF firms has soared over a dozen years, by around 33% in manufacturing and 45% in services. Among all the other OECD companies, productivity progrssed at a snail’s pace, just around 7% in manufacturing and 5% in services.

Source: http://dx.doi.org/10.1787/0748e0bc-en

The typical Japanese company fared particularly badly during 1994 to 2003 (I’ve not found more recent data). The top 10% of Japanese manufacturers, 14,000 corporations, enjoyed an 80% hike in TFP productivity. At the typical manufacturer, TFP rose just 30%, and in the bottom 25%, just 10%. In non-financial services, the top 10%—74,000 corporations—increased productivity by 30%, far less than in manufacturing. Firms around the median saw a 20% hike, and the lower tenth, a meager 5% improvement (see chart below). (This does not include all the unincorporated tiny firms with even lower productivity.)

Source: http://www.jstor.org/stable/40587821

Japan’s services productivity need not have performed so badly. In fact, except for Finance and Insurance, Japan performs substantially worse than the other Group of Seven countries in TFP productivity growth in services. Even negative in some substantial segments (see chart below). Hence, as Japan becomes more of a service economy, overall productivity suffers.

Source: https://stat.link/fak8zd

Arteriosclerosis Sets In Beginning in Mid-1970s

Let’s see why Japan is underperforming. A country’s productivity rises in four ways:

1) Incumbent firms become more productive

2) Superior incumbent firms take market share away from inferior competitors

3) Capital and labor move from low-productivity sectors to high-productivity ones

4) Uncompetitive firms exit the marketplace, and new superior companies replace them in the same or different sectors.

In this post, I’ll focus on the first method and, in the next post, method three.

During the high-growth era (1953-73), Japan was a highly fluid economy, with both capital and labor moving to more dynamic sectors and to superior companies within each sector. Moreover, companies took more chances because, if they failed, high growth made it much easier for executives and workers alike to get a new job at a good company, despite the lifetime employment system. When the oil shocks of 1973 and 1979 halved GDP growth rates, Japanese policymakers chose to keep less efficient firms and even “zombies” alive to maintain employment. Bank financing for new potential innovators became tighter. This eventually led to the lost decades. (See details in my 1998 book, Japan: The System That Soured, Chapter 7.) During the lost decades, arteriosclerosis worsened, as I detail in Chapter Eight of The Contest for Japan’s Economic Future.

Consider one facet. These days, in rich countries, R&D is often a stronger contributor to a firm’s growth than investment in tangible assets like machinery. Moreover, the growth gap between innovators and non-innovators is even stronger for younger, smaller companies. It often requires a firm’s own R&D to determine how best to adapt and utilize innovations developed by frontier companies.

Unfortunately, expensive R&D is not without risk, especially if the financial system is less willing to finance newer firms, especially those taking chances and having less collateral. As a result, smaller firms investing in R&D and innovation are more likely to fail than those that don’t try. But those that survive become much bigger and enjoy higher productivity. Consequently, a flawed banking system is one of the main reasons the share of patents held by Japan’s medium-sized enterprises (20–250 employees) is the lowest in the OECD: just 27% compared to 60% in a typical country.

The digital era has changed how innovation occurs. In 1981, 71% of the business R&D in the United States was conducted by giants with at least 25,000 employees; by 2014, the share had halved to 36%. Conversely, the share carried out by firms with fewer than 100 employees had risen to a stunning 8%. Similar changes have occurred in Europe, but less so in Japan. In 2015, only 7% of Japan’s business R&D was conducted by firms with less than 500 employees, compared to 17% in the US and 33% in France and the UK. In 2018, just 10 Japanese giants conducted almost half (43%) of all business R&D in the country.

Worse yet, Tokyo actually retards the R&D shift seen elsewhere. 92% of all Japanese government subsidies and tax breaks for R&D go to large firms, the highest share among OECD countries That’s another reason why it’s so much harder for SMEs to catchup to the technological frontier. See chapter two of The Contest for details.

And now Takaichi’s equivalent of DOGE is discussing a reduction in subsidies for R&D. How is that supposed to aid innovation?

That technological lag is another reason it’s so hard for productivity in Japan to catch up to the NFs. In Japan, the productivity ratio for companies with 10 to 250 employees compared to those with at least 250 employees is the third lowest among 21 countries (see chart below). It’s also the third lowest among companies with 50-249 employees. Since a third of employees work for firms in the 10-249 range and even more at those with fewer than ten, no wonder it’s so hard to diffuse technology and productivity throughout the economy.

Source: http://dx.doi.org/10.1787/888933563474

Next: Capital, labor, and the share of GDP don’t move to companies and sectors with higher productivity

To Support the Blog

If you feel you’ve gained insight from this blog, ever restacked it, if you ever subscribed to my previous publication, The Oriental Economist Report, and certainly, if you or your firm have gained insights that helped guide your investments, please support the blog with a subscription or by “buying me a cup of coffee.” You can buy a cup or two on a one-time basis, or once a year, or once a month.