How Many Jobs Does 1 Billion Won Actually Create?

Every time the government announces it will pour trillions of won into the AI industry, one phrase never fails to appear: "job creation." But is that actually true? To answer that question, we first need to dust off an old economic indicator.

Employment Inducement Coefficient

The number of wage workers directly and indirectly created — in a given industry and its related industries — when that industry generates 1 billion won in final demand. Calculated by the Bank of Korea using input-output tables, it's the most intuitive yardstick we have for how many livelihoods a given industry actually supports.

As of 2023, the average employment inducement coefficient across all industries in Korea is 6.2. That means every 1 billion won in demand brings roughly 6.2 jobs along with it.

The catch is that this number varies wildly from industry to industry.

Semiconductors: 2 Jobs per Billion Won. Education: 13.

There's an unbridgeable gap between services and manufacturing. Labor-intensive service sectors like education, healthcare and social welfare, and food and lodging all post employment coefficients well above 10. Construction is much the same. Wherever an industry needs human hands, money going in quickly turns into jobs coming out.

Advanced manufacturing sectors like semiconductors and electronic components tell a different story. Capital-intensive and heavily automated, they generate only 2 to 3 jobs per billion won invested. Software and IT services absorb far fewer workers than traditional service industries too, since their whole structure is built around wringing high output from a small headcount.

The AI industry pushes this tendency to an extreme. A handful of GPU clusters and a small team of engineers can generate hundreds of billions of won in revenue. The ratio of employees to revenue at AI companies is barely comparable to that of traditional industries. According to data the Bank of Korea released in December 2025, Korea's pool of dedicated AI professionals stood at roughly 57,000 as of 2024 — a strikingly small number given an industry now attracting trillions of won in investment.

This Is Already Happening

Let's set theory aside and look at what's actually happening. A report the Bank of Korea's employment research team published in October 2025, titled "AI Diffusion and the Contraction of Youth Employment," backs this up with numbers.

In the three years since ChatGPT launched (July 2022 to July 2025), 211,000 jobs held by young people (ages 15-29) disappeared. Of those, 208,000 — 98.6 percent of the total decline — were in industries with high exposure to AI.

The picture sharpens once you zoom in. Youth employment fell 11.2 percent in computer programming and systems integration fields like software development. Publishing dropped 20.4 percent, information services 23.8 percent. Even professional services like law and accounting shed 8.8 percent.

What's striking is what happened among workers in their 50s. Over the same period, jobs held by that group grew by 209,000, and 146,000 of those were in AI-exposed industries. The Bank of Korea calls this "seniority-biased technological change": as AI takes over junior employees' routine tasks, the organizational judgment and management skills that come with seniority become, paradoxically, more valuable, not less.

Put simply: the AI industry itself creates few jobs. At the same time, AI's spread through other industries is shrinking existing jobs there — and pushing out young people trying to enter the labor market first.

Measuring a New Era with an Old Ruler

And yet government industrial policy still leans on the old formula: investment leads to growth, and growth leads to jobs. Trillions of won for AI chips, trillions more for AI data centers, hundreds of billions for nurturing AI startups. These investments will almost certainly lift GDP. But hanging a "job creation" banner over them is a stretch — the employment coefficient is simply too low.

The coefficient itself has real limits, too. Because it's derived from input-output tables built around fixed industry categories, it can't properly capture AI's ripple effects as they cross those boundaries. When AI raises manufacturing productivity and cuts jobs there while simultaneously spawning entirely new occupations somewhere else, the existing coefficient simply has no way to measure it.

In other words, if we want to measure the employment payoff of national industrial investment, the traditional employment inducement coefficient alone won't cut it.

We Need a New Ruler

Employment policy for the AI era needs at least three supplementary metrics.

First, an "employment displacement coefficient." We need to measure how much investment in one industry reduces or reshapes employment in others. If a billion won invested in AI creates 3 jobs within the AI sector while wiping out 5 jobs in traditional services, the net employment effect is negative. We need to abandon the habit of calculating an investment's employment impact solely within its own industry.

Second, an "employment quality index." We need to look at the wage levels, job security, and skill requirements of the jobs being created, not just their count. If AI-sector jobs are few but well-paid and stable, that changes the calculus beyond a simple headcount comparison. Conversely, if jobs displaced by AI are simply resurfacing as platform work or gig-economy labor, that isn't job "creation" — it's job degradation.

Third, a "generational employment impact assessment." As the Bank of Korea's research shows, AI's employment shock cuts in opposite directions depending on generation — an opportunity for workers in their 50s, a barrier for people in their 20s. If investment policy ends up blocking a specific generation from entering the labor market, that's not merely a failure of industrial policy. It's a social risk.

What Needs to Change Isn't Where We Invest — It's How We Measure It

None of this is an argument against investing in AI. Falling behind in the AI technology race would put the future of the entire industrial base at risk. That much is clear.

But measuring the success of AI investment by "how many jobs it created" is now an anachronism. Slapping the label "job creation" on an industry whose employment coefficient is just 2 or 3 does a disservice both to the public and to the policy itself.

It's time to honestly redesign the metrics we use to judge AI investment policy. Put contribution to productivity growth, industrial competitiveness indices, and technological self-sufficiency front and center, and separate the employment question into its own distinct policy track. A more realistic approach would be to design a structure that redistributes some of the value AI generates into fields with high employment coefficients — education, caregiving, healthcare — instead.

How many people go to work when you pour in 1 billion won? If this old question still matters in the age of AI, then we need to change how we go about answering it. You can't measure a new era with an old ruler.