Spatial misallocation is an essential reason for the low utilization efficiency of construction land. Optimizing the spatial pattern of construction land allocation can improve the efficiency of economic operations and resilience to food crisis and climate change challenges. This study constructs a quantitative measurement method for the spatial misallocation of construction land (SMCL), proposes a new government market society support (GMSS) analytical framework for the contributing factors with interlocked relationships, and conducts empirical research in Shandong, China, which is a typical area with a rapid development of construction land and significant regional disparity. It is concluded that the SMCL ensues through the interaction and coevolution of the GMSS system, which plays a key role in adjusting the construction land use sequence, structure, and efficiency under the regulation of the upper government. Effectively using the estimation method based on the equal marginal output principle, the SMCL in Shandong is established as a downward trend, with evident temporal and spatial differentiation characteristics and spatial morphological mode changes, although most sub-regions are significantly approaching the adaptation interval with fluctuation. Furthermore, the empirical results of the regression model indicate that there are different effects and intensities on the SMCL among the contributing factors under the GMSS framework, wherein the local government force has an aggravating and the greatest effect, the market forces have a dual and second-ranking effect, the social forces play a positive but still weak role, and the support system has a differentiated improvement effect. However, the impacts of various dimensional factors on the SMCL also have heterogeneity in the development stages and different regions. Generally, in the low-level development stage and underdeveloped areas, the effect of local government intervention is stronger, the market forces’ importance is lower, and the social forces and support systems remain insufficiently robust.