In this paper, we seek to determine the effect that a change in transportation supply can have on the poverty level of a neighborhood. In particular, the evolution of the unemployment rate by developing a model that takes into account differences in the dispersion of skilled and unskilled jobs as well as the location strategies of the employed and the unemployed. We begin by presenting some stylized facts representing the distribution of jobs, employed and unemployed in some French urban units. The location of these different agents seems to follow a constant pattern. Skilled jobs and skilled inhabitants are more concentrated near the city center, while the unskilled are more dispersed. Within the same category, the unemployed are the least dispersed agents. For jobs and employed workers, it depends on the urban units. Jobs may be more concentrated than the employed, or vice versa. We propose two models to account for these distributions. The first one is based on the idea that the worse the accessibility to jobs, the higher the level of unemployment. The location of unemployment then depends entirely on the distribution of jobs and inhabitants. The second approach integrates a search-matching mechanism. Unemployment is no longer a consequence of the distribution of jobs and inhabitants, but the result of location and job search strategies on the part of the unemployed. It turns out that this second model provides the closest results to those observed empirically, because it explains both the greater concentration of jobs and qualified inhabitants, and the greater concentration of unemployed.