Title Model-Fit Analysis of the Public Design Fee Estimation System and Improvement Framework
Authors 배병윤(Bae, Byungyun)
DOI https://doi.org/10.5659/JAIK.2026.42.8.169
Page pp.169-177
ISSN 2733-6247
Keywords Public Design Fee; Model Fit; Multiple Regression Analysis; Stage-Based Framework; Actual-Cost-Plus-Fixed-Fee Method; Design Service Contract Data
Abstract Public design fees in Korea function as institutional outputs shaped by budget logic, task scope, procurement, and contract conditions, yet practice relies on construction-cost ratios that fail to capture actual workload and complexity. This study examines the gap between budget-planning and contract-formation logic and proposes a stage-based framework from empirical evidence. Of 201 contracts, 87 valid cases with complete records on contracted fee, construction cost, service type, fee method, duration, and procurement were analyzed via multiple regression. The baseline model using only construction cost yielded R²=0.138, while the extended model with institutional variables yielded R²=0.374. The actual-cost-plus-fixed-fee method (β=0.712, p=0.011) and comprehensive evaluation bidding (β=0.902, p=0.003) were significant, while construction cost lost significance (p=0.272) once these variables were controlled, indicating that contracted fees are governed less by cost magnitude than by procurement and fee structures. A three-stage system is thus proposed: at budget planning, construction cost and facility type set preliminary fee bands adjusted for complexity and procurement; at design procurement, scope, BIM, permit burdens, and manpower are reflected via a refined actual-cost-plus-fixed-fee method; at contract adjustment, added services from client changes or extended duration are recorded and compensated, reframing design fees as multi-layered institutional outputs.