METI's “2025 Digital Cliff,” Part I
Shortfall in ICT Pros Stifling Hoped-For Digital Transformation
Source: METI at https://www.meti.go.jp/policy/it_policy/jinzai/gaiyou.pdf
Note: Assumes that labor productivity continues to rise at a meager 0.7% per year and demand for IT services grows at either a baseline forecast of 2.7% per year or at 4.4%
Japan’s only real resource is the skills of its people. And yet, when it comes to using these skills to generate economic value for its people, Japan is wasting this most valuable resource. On the one hand, when it comes to math and science tests, Japan’s high school students regularly score at the top, or very close to it. In 2019, Japan scored second only to Korea in the share of its students who ranked among the top global performers in math and/or science. And yet, as I’ll detail in part II, it scores below average in the share of all students who move into careers focused on digital technology and near the bottom of those who choose science-based careers.
As a result, Japan suffers an ever-growing shortage of professionals in Information and Communications Technology (ICT). METI (Ministry of Economy, Trade, and Industry) calculates that the shortage of ICT professionals stood at 300,000 in 2020 and will grow to 450,000 by 2030 (see chart at the top). Japan would need 1.58 million ICT professionals but have only 1.13 million. That assumes a baseline annual growth of 2.7% in demand for ICT services. Should demand rise at a 4.4% pace, the shortfall in 2030 would hit 800,000. Unless this situation is remedied, one of Japan’s latest policy buzzwords, DX (short for digital transformation), will come to naught.
In 2018, METI issued a report warning of a “2025 Digital Cliff.” If not prevented, GDP would be lower than it could be by as much as ¥12 trillion each year, an amount equal to more than 2% of 2022 GDP. This could be an exaggerated number, given that real GDP grew an average of just half that amount during 2010-19. Still, there is no question that the consequences for growth are dire.
People and Legacy Systems: The Key Bottlenecks
The key bottleneck is not simply those who know how to develop better computers or write new software. It’s the far larger number of those who know how to use digital technology to improve performance at non-ICT companies. From making cars to performing brain surgery (e.g., the use of augmented reality) or enabling airlines to figure out what price to charge to fill more seats, digital technologies comprise an increasingly vital part of every industry. Companies adept at digitalization consistently show higher productivity growth than those who do not. But how can firms take advantage of the latest innovations if they cannot hire enough skilled personnel?
What makes the shortfall even worse is the approach to computer systems followed by so many leading Japanese companies. While they buy large computer systems off the shelf, up to the mainframe level, they have had their in-house personnel custom-design the software to work with them for decades. As the equipment and software age, they add this or that update. To overhaul this practice and use the same systems used by others, as is done elsewhere, would be very costly, not only in terms of money but in terms of the personnel required. Hence, most of these companies don’t change their ways.
METI cites the Japan Institute Users Association of Information Systems as saying that, if current trends continue, by 2025, 60% of Japan’s large companies will be operating core systems that are more than 20 years old. Would anyone today use a 2005 PC? Would today’s software and communications even operate on it? Why, then, are so many companies doing the equivalent with their larger, more complex computer systems? Partly it’s a mindset problem, but it’s also a result of the personnel shortage.
In a 2021 report, the Ministry of Internal Affairs and Communication wrote: “ICT investment in Japan has been on the decline from the peak in 1997. Furthermore, there remain many conventional systems (legacy systems) as exemplified by the fact that 80% of its ICT investments are used for maintenance and operation of the current businesses. It is said that Japan is stuck in the old way of thinking.”
Already, says METI, data losses, software defects, hardware failures, security issues, and other problems of these legacy systems cause economic losses adding up to almost ¥5 trillion per year. The ¥12 trillion annual economic loss noted above assumes a multiplication of these losses as the personnel shortfall soars and maintenance of these legacy systems becomes more difficult.
The Dire Results
The personnel shortfall is greatly exacerbated by a big fall in productivity growth within the sectors producing all the ICT services. Back in 1965, Japan’s growth in output per ICT professional came in fourth behind the US, Germany, and France, at 2.4%. While productivity slowed in the US and France, it slowed even more so in Japan. In the 2010s, productivity growth stood at a truly dismal 0.7% per year (see chart below). Given METI’s expectation that demand for IT services will grow at 2.7% per year, productivity would have to rise to 3.5% per year to eliminate the personnel shortfall. An even bigger problem is the drag on productivity of all the non-digital companies that need to consume the latest systems and software.
Source: METI at https://www.meti.go.jp/policy/it_policy/jinzai/gaiyou.pdf
It is worth noting that the OECD shows Japan suffering an actual drop in output per employee in ICT business services from 2005 through 2021, its ICT productivity having peaked out in 1999. By contrast, Korea’s productivity grew 41% in the same period. Some of this may be due to the hiring of part-time workers, but most is not. Unfortunately, Japan, like the US, does not produce a measure of output per workhour, which would have eliminated this distortion. Whichever measure we use, Japan’s performance is alarming because poor productivity exacerbates the consequences of the personnel shortage.
Source: OECD https://stats.oecd.org/BrandedView.aspx?oecd_bv_id=pdtvy-data-en&doi=data-00687-en
Worse yet, what if METI’s estimate of Japan’s actual need for ICT services is way too low? The fact is that Japan has a huge digital divide between its large companies and the small and medium-sized ones (SMEs) that employ 70% of Japan’s workers. While large companies devote 10% of their total investment to software, the ratio is just 4% at firms with fewer than 300 employees. Only one in four SMEs invested in ICT equipment or software in 2017. But one cannot run 2022 software on a 2012 computer. Back in 1993, the average age of equipment was nearly the same at large companies and SMEs, just 4.4 and 4.9 years, respectively. By 2015, the average age had risen to 6.4 years at big companies and 8.5 years at SMEs.
Why do SMEs use so much less digital technology? When METI asked these SMEs, the biggest reason given, at 43%, was “a lack of personnel who can introduce IT.” A close second at 40% was “The effects of introducing IT are unclear or are not sufficient.” Part of the personnel shortage is a dearth of consultants who could show these SMEs how much they could gain via ICT.
Suppose, however, that these SMEs began to realize what they were missing and therefore the demand for ICT services grew at, say, 4.4% per year instead of the base case of 2.7%? In that case, says METI, Japan would need 1.92 million professionals but have only 1.13 million, a shortage of 800,000 (see again chart at the top).
In Part II, I’ll delve into some of the reasons for this shortfall.
Very interesting to see similar developments in Germany and Japan regarding ICT service output productivity per hour. Could be that SMEs who are world leading in manufacturing in their tiny niche find it difficult to go outside their core competencies of hardware manufacturing and digitalize their business? Would be interesting to see how various industries compare, e.g. with the UK becoming more de-industrialized and service focused, how are the remaining manufacturing SMEs doing compared to similar industries in other countries? Also updating/unifying legacy systems is HARD, there are large companies who couldn’t unify legacy systems of their subdivisions and basically gave up on it… It may change via no-code software development solutions/customization simple enough for small companies to pull off without a dedicated IT department. Or whoever finds the holy grail of easily customizable software with all the diverse functions needed across industries would make a killing selling this solution… The fact nobody has yet does not really spark optimism though.
Great stuff! Suggests that the Japanese Government has taken its eye off the ball, just like its Western allies...