Changes in the determinants of travel demand for the Washington DC Metrorail system after COVID-19: Evidence from a replication study

• ODLURM models are estimated for the Washington DC Metro rail system and previous methods are replicated from 2014 and compared to results for 2022. • Due to pandemic-induced behavioral changes, the 2014 model does not explain ridership in 2022 as well and many factors are less associated with ridership. • Replications are useful for highlighting important system changes and informing planning decisions. The COVID-19 pandemic influenced many changes to daily lifestyles, including public transportation ridership. To accommodate changes in ridership patterns, it is important to consistently update knowledge of the factors that most influence a rider’s decision to use public transportation. Origin-Destination Direct Ridership Models are one of the methods used to accomplish this, as they examine the sensitivity of transit demand to changes across a series of explanatory variables. This study replicated a time-of-day, multi-level Origin-Destination Land Use Ridership Model (OD-LURM), previously developed using 2014 ridership data, with 2022 SmarTrip card data from the Washington, D.C. Metrorail System. Directly comparing model results from 2014 and 2022, the magnitudes of the effects of various determinants on Metrorail ridership were examined, considering three different groups of variables at the level of origin stations, destination stations, and OD-station pairs. The effects of the total number of jobs within a half-mile radius, parking users, parking capacity, bus line counts, and median household income on ridership in the AM peak, PM peak, and off-peak time periods were reduced compared to the previous study, sometimes drastically. The results also indicated a weaker model fit using the updated data. Overall, this replication study provides new context behind changes in post-pandemic ridership, and the results suggest persistent changes in travel behavior since the pre-pandemic period. The findings of this study could be used to inform transit-oriented development, fare structures, and service levels in the D.C. metropolitan area post-pandemic.