맹지의 도로 접면 전환이 지가에미치는 인과효과:동일필지 반복거래 패널 이중차분 분석

초록

This study estimates the effect of gaining road access on land value using a repeat-sales panel difference-in-differences design. Combining Korea’s land transaction records with officially assessed land price characteristics for 2020–2025, we construct a panel of 144,368 parcels transacted at least twice nationwide. The treatment group comprises 1,702 parcels sold while landlocked and resold after gaining road access; the control group comprises 38,801 parcels that remained landlocked throughout. Controlling for parcel fixed effects and district-by-year interaction fixed effects, the net effect is 5.72 percent and statistically significant. A before-after comparison of the treatment group without a control group yields 19.87 percent, more than a threefold overestimate. Stratifying by land category conversion, parcels whose category remained unchanged show a 6.37 percent effect, whereas parcels converted to building lots show a 35.23 percent price change. This gap indicates that appreciation is tied less to physical accessibility than to the change in permitted use that road access enables, consistent with case law limiting rights of way to current land use and with the statutory road-frontage requirement. The results hold under heterogeneity-robust estimators, covariate balancing, and alternative outlier rules. However, parallel trends hold only immediately before treatment, the building-lot estimate is conditional on a post-treatment outcome, and the treated sample represents a selected subset of landlocked parcels with active transactions and visible development.

키워드