A big data approach to mitigating the MAUP in measuring excess commuting

Abstract Excess commuting, defined as the inefficiency resulting from spatial mismatches between residential and employment locations, poses significant challenges for urban planning and transportation systems. This study uses big data from individual vehicle trips collected in Tampa, Florida, to quantify excess commuting more accurately than traditional zonal approaches. Through the application of Linear Programming (LP) and Integer Linear Programming (ILP) models, this research measures minimum and actual commuting patterns across different spatial scales—census tract, block group, and individual trip levels. The findings reveal a clear scale effect associated with the Modifiable Areal Unit Problem (MAUP), as smaller spatial units consistently yield shorter minimum commuting distances and times and the ILP model at the individual trip level yields the least. By directly analyzing actual trips rather than simulated data, this approach provides a more precise and realistic assessment of excess commuting. The results underscore the values of methodological improvements and individual-level data in refining our understanding of excess commuting and supporting more efficient urban planning and policymaking.