This study applies K-means clustering to classify food manufacturing firms in Busan, South Korea, based on labor and capital productivity, using establishment-level data from the Mining and Manufacturing Survey. Three-cluster and four-cluster configurations are examined and compared against national benchmarks to identify structural characteristics unique to Busan's food manufacturing sector. The three-cluster analysis shows that Busan has a higher share of small-scale, capital-efficient firms than the national average, largely driven by its dominant seafood processing industry. However, the four-cluster analysis reveals that the largest group in Busan consists of capital-intensive firms that fail to generate adequate returns, pointing to capital misallocation as a widespread structural problem. Given Busan's economic scale, its food manufacturing sector has the potential to grow into a larger economic driver, but the current structure suggests this potential is not being fully realized. Policy should focus on adopting food tech, particularly smart production technologies, to improve productivity across the sector.
Seo et al. (Thu,) studied this question.