published journal article
published journal article
working paper
Recent survey validation studies suggest that measurement error in earnings data is pervasive and violates classical measurement error assumptions, and therefore may bias estimation of cross-section and longitudinal earnings models. We model the structure of earnings measurement error using data from the Panel Study of Income Dynamics Validation Study (PSDIVS). We then use Rubin’s (1987) multiple imputation techniques to estimate consistent earnings equations under non-classical earnings measurement error in the PSID. Our technique is readily generalized, and the empirical results demonstrate the potential importance of correcting for measurement error in earnings and related data, particularly during recessions.
conference paper
published journal article
published journal article
published journal article
This paper generalizes the simple two-city work-from-home model of Brueckner et al. (2022) by adding a group of non-remote workers, who must live in the city where they work. The results show that the main qualitative conclusions of BKL regarding the intercity effects of WFH are unaffected by this modification, with WFH yielding the same aggregate population and employment changes in the two cities and the same house-price and wage effects as in the simpler model. Even though the aggregate population effects are the same, the population relocation of non-remote workers is in the opposite direction to that of remote workers, which matches the direction in BKL. These conclusions are useful because they establish the robustness of BKL’s highly parsimonious model. The paper also contains material surveying other theoretical research on WFH as well as empirical work in the area, including BKL’s empirical findings in support of their model.
published journal article
Peer-to-peer (P2P) ridesharing is a recently emerging travel alternative that can help accommodate the growth in urban travel demand and at the same time alleviate problems such as excessive vehicular emissions. Prior ridesharing projects suggest that the demand for ridesharing is usually shifted from transit, but its true benefits are realized when the demand shifts from single-occupancy vehicles. This study investigated the potential of shifting demand from private autos to transit by providing a general modeling framework that found routes for private vehicle users that were a combination of P2P ridesharing and transit. The Los Angeles Metro Red Line in California was considered for a case study because it has recently shown declining ridership trends. For successful implementation of a ridesharing system, strategically selecting locations for individuals to get on and off the rideshare vehicles is crucial, along with an appropriate pricing structure for the rides. The study conducted a parametric analysis of the application of real-time P2P ridesharing to feed the Los Angeles Metro Red Line with simulated demand. A mobile application with an innovative ride-matching algorithm was developed as a decision support tool that suggested transit-rideshare and rideshare routes.
conference paper
A multicommodity discrete kinematic wave model that possesses the theoretical rigor and computational efficiency inherent in the kinematic wave theory is proposed for simulating network traffic flow. In this model, fluxes through boundaries and junctions are computed systematically under the supply-demand framework. In addition, traffic is modeled by commodity type so that the effects of geometric characteristics of a road network on traffic dynamics can be captured. Although traffic is not ordered down to the vehicle level as in existing kinematic wave simulation models, the noncompliance with the first-in-first-out property in this model is still of the order of At, the time increment. Hence travel times in the average sense can be defined from cumulative curves. Finally, the evolution of traffic dynamics in a sample road network is shown to demonstrate the stability, numerical convergence, and soundness of the proposed network kinematic wave model.
published journal article
Recently, road scene-graph representations used in conjunction with graph learning techniques have been shown to outperform state-of-the-art deep learning techniques in tasks including action classification, risk assessment, and collision prediction. To enable the exploration of applications of road scene-graph representations, we introduce roadscene2vec: an open-source tool for extracting and embedding road scene-graphs. The goal of roadscene2vec is to enable research into the applications and capabilities of road scene-graphs by providing tools for generating scene-graphs, graph learning models to create spatio-temporal scene-graph embeddings, and tools for visualizing and analyzing scene-graph-based methodologies. The capabilities of roadscene2vec include (i) customized scene-graph generation from either video clips or data from the CARLA simulator, (ii) multiple configurable spatio-temporal graph embedding models and baseline CNN-based models, (iii) built-in functionality for using graph and sequence embeddings for risk assessment and collision prediction applications, (iv) tools for evaluating transfer learning, and (v) utilities for visualizing scene-graphs and analyzing the explainability of graph learning models. We demonstrate the utility of roadscene2vec for these use cases with experimental results and qualitative evaluations for both graph learning models and CNN-based models. roadscene2vec is available at https://github.com/AICPS/roadscene2vec.