한계 소상공인에 영향을 미치는 요인 분석: 다수준 로짓 모형을 활용한 접근

Analyzing the Factors Influencing Marginal Small and Micro Enterprises: A Multilevel Logistic Regression Approach

Purpose: This study aims to identify and analyze the factors affecting insolvency among small and micro enterprises in Korea. Unlike prior research focusing mainly on financial indicators, this study considers both individual business characteristics and regional environmental factors to understand the risk of insolvency among small businesses. Research design, data, and methodology: Utilizing a multilevel logistic regression model, this research analyzes 417,273 small and micro enterprises across administrative districts (eup/myeon/dong) in 2020. The study incorporates two-level hierarchical data: level 1 variables reflecting business owner and store attributes (such as gender, age, creditworthiness, income, assets, delinquency status, business longevity, and average sales), and level 2 variables reflecting regional characteristics (population density, average age, number of households, single household ratio, income level, number of businesses, and rental fee). Data sources include Shinhan Card, Korea Credit Bureau, Statistics Korea, and the Environmental Big Data Platform. Results: The findings reveal substantial regional variance in insolvency occurrence, with an overall insolvency rate around 11%. At the individual level, male owners, younger age, lower credit score, higher income and asset levels, presence of unresolved delinquencies, longer business experience, and lower average annual sales are associated with increased insolvency risk. Industry differences are observed, with manufacturing, wholesale, transportation, and accommodation sectors presenting higher insolvency probabilities than retail. Regionally, lower population density, lower single household ratios, higher business density, and higher rental fees significantly increase insolvency likelihood. Implications: This study highlights the importance of integrating both micro-level business factors and macro-level regional characteristics to effectively identify and support financially distressed small enterprises. Policymakers should consider these multidimensional risk factors when designing targeted intervention and support programs, especially in light of external shocks such as COVID-19 and rising economic pressures. Monitoring insolvency indicators in a context-sensitive manner can enhance the efficiency of resource allocation and contribute to healthier small business ecosystems.