The present project studies the pattern of spatial variations of employment accessibility in the Mexico City Metropolitan Area. The question is to assess differences in employment accessibility according to transportation mode (car or public transit) used and between the formal employment sector and total employment (formal + informal sectors). Two indicators were explored at Traffic Analisys Zones (TAZ) level: gravity-based job accessibility (GBM) and the indicator developed by Shen (1998). We explored two data sources of travel times: the 2017 Household Origin Destination Survey and the region’s travel demand model TRANUS. Thus, there are four aspects of comparison in this evaluation of job accessibility: sources of travel time data (Tranus vs HODS17), methods (Shen vs GBM), type of employment data (total vs formal) and transportation modes (Car vs Transit). Notice that when doing each comparison, the rest of the aspects are kept constant. In the first two, absolute values of accessibility are not comparables. Both sources of travel time data followed different approaches in gathering the data, while both methods have different units as well as upper and lower limits. For this reason, when comparing job access results between sources of time data or between methods the objective is to determine the consistency and robustness of the results in base of the job access ranking among TAZs. Thus, the Spearman Rank Correlation (SRC) is the appropriate indicator to check the consistency in the accessibility estimations.Then, in the other two aspects of comparison (Total employment vs formal employment; Car vs Transit) absolute values and rankings of accessibility are relevant. These two aspects represent direct variables embedded in both accessibility equations and therefore offer insights about how these factors impact its estimation. Overall, the purpose of analyzing the importance of these variations was to select those estimations with the highest consistency between travel-time sources but included further differences according to employment-type data and transportation modes. This information offers insights into the nuances of these aspects in the disparity of intra-metropolitan accessibility.Our exploration of accessibility using the GBM shows an important variation in the metropolitan pattern according to employment data, travel-time sources and transportation mode. As a general description, jobs-rich areas in the inner city have the highest accessibility with a decrease in accessibility with increasing distance from the urban center, however this negative relationship is not as clear as in the Shen´s type model. This is an expected result with GBM since this model focuses on the supply side of the jobs market, i.e. employment urban cores are predominant areas of accessibility. Total employment dramatically increases accessibility and gives more consistent results between travel-time sources than does formal employment, probably due to the reinforcement of the role of land use in the estimation. Accessibility is always higher for car drivers than for transit users with TRANUS, while for HODS17 this remains true for the most part but with a few exceptions.The Shen´s indicator shows a more consistent spatial pattern of accessibility (spearman correlations close to 1) regardless of travel-time data, and transportation mode choice, demonstrating the robustness of the method. In general, the spatial pattern of accessibility in relation to the urban center is a line with a negative slope. The resulting Shen´s accessibility landscape was compared with the urban structure cited in the literature. Results show that areas with the highest employment accessibility are within the central agglomeration and the associated corridors along main highways at its perimeters, according to the urban structure reported by Suárez and Delgado (2009). The disparity in terms of location means that access in the TAZ with the highest accessibility record is 26% higher than the metropolitan accessibility average. As expected, the inclusion of total employment increases accessibility in comparison with only formal employment. Commuting by car reduces travel time, and although this increases accessibility overall, the increase is negligible when comparisons are made with the increment of accessibility from formal to total employment, or with the difference between the higher and lower ends of job access by public transit. These comparisons show that as opposed to transportation mode, locations of residence, in direct relation with its closeness to the urban employment centers, is the main factor affecting access to employment. As en example of how results can be used to guide goverment interventions we identified priority areas for accessibility improvement as those TAZs with the worst accessibility index and highest marginalization in both the State of Mexico and CDMX.
Read full abstract