Demographics explain 121,000 of the 193,000 jobs Korea's 20-somethings lost

South Korea's 20-something population averaged 5.55 million a month over those eight months, down 3.5% from a year earlier ^[5]. That puts the year-earlier cohort at about 5.75 million and its employment rate at 60.4% ^[1]. Shrink a cohort by 3.5% and hold the rate steady, and you lose about 121,000 jobs ^[2]. The reported loss was 193,000 ^[1], so roughly 72,000 of it, about 37%, comes from the employment rate itself falling ^[3].

Almost all of that residual sits in the early 20s. Employment among 20- to 24-year-olds averaged 924,000 a month, down 114,000, and their employment rate fell 2.6 points to 41.3%, the weakest January-August reading since the series began in 2000 and a tenth of a point below the 41.4% of 2020 ^[7]. Those aged 25 to 29 fell 78,000 to 2.355 million, with their rate down 0.8 points to 71.1% ^[8]. The younger group holds about 28% of the jobs and supplied about 59% of the loss ^[4].

Jung Ji-woon, a senior research fellow at the Korea Research Institute for Vocational Education and Training, and researcher Lee Hae-yang found in June that the point at which young adults settle into the labour market has moved from the 20s to the 30s, and that as hiring comes later, the age at which people begin working in earnest is rising ^[10]. The 30s figures fit that, partly. Their population rose 83,000 and their employment rose 103,000 ^[9], so at the prior year's 80.4% rate, population growth alone would have produced 5.579 million against the 5.615 million reported, leaving about 36,000 jobs that came from the rate ^[6]. Credit every one of those 36,000 to someone who would once have been hired at 24, which is generous, and half of the 72,000 shortfall is still unaccounted for ^[7].

A cohort staying in education longer shows up as lower participation with the unemployment rate broadly steady. The 15-29 unemployment rate was 5.4% last month, the highest in four years ^[13].

The cyclical case is real: total employment fell 40,000 in May, then rose 63,000 in June and 108,000 in July ^[12], and youth hiring lags an aggregate recovery. In my view the deterioration is mostly not cyclical, because the early-20s employment rate is now below its 2020 pandemic reading ^[7] while the total is adding jobs at the fastest pace in five months ^[11]. What would change that is the 20-24 rate recovering its 2.6 points over the next two prints while the cohort keeps shrinking. One comparison note: the 193,000 is an eight-month average, while the earlier declines of 91,000 in 2023, 101,000 in 2024 and 177,000 last year are full years ^[4], and the 193,000 is also bigger than the 142,000 of 2009 and the 111,000 of 2020 ^[3].