Addressing the joint occurrence of self-selection and simultaneity biases in the estimation of program effects based on cross-sectional observational surveys: case study of travel behavior effects in carsharing

Gouri Shankar Mishra (), Patricia Mokhtarian, Regina R. Clewlow () and Keith F. Widaman ()
Additional contact information
Gouri Shankar Mishra: University of California, Davis
Regina R. Clewlow: University of California, Davis
Keith F. Widaman: University of California, Riverside

Transportation, 2019, vol. 46, issue 1, No 6, 95-123

Abstract: Abstract We estimate the effect of carsharing on travel behavior (specifically, household vehicle holdings, frequency of transit usage, and frequency of biking and walking) using data from the 2011–2012 California Household Travel Survey (CHTS). The effect of carsharing on vehicle ownership is a dynamic process that plays out over a period of time—past ownership influences enrollment decisions, which in turn influence holdings in a later period. Representing this process using cross-sectional data conflates causal effects with simultaneity bias. Further, members and non-members differ in various observed and unobserved ways—demographics, built environment of residential and workplace location, and attitudes—raising the potential for self-selection bias in comparing travel behavior between the two groups given the observational nature of the data. Drawing on established methods for dealing with each bias individually, we develop a method to help control for this joint occurrence of self-selection and simultaneity biases. Restricting the analysis to employed respondents residing in the San Francisco Bay Area, we find that 80% of the observed difference of 0.9 units in average vehicle holdings between carsharing non-members and members may be explained by the biases listed above. The remaining difference of 0.17 units reflects the estimated effect of carsharing, which is the equivalent of shedding one vehicle by about one out of every six households whose member(s) are enrolled in carsharing. The effect on transit usage and walking and biking frequency is positive, albeit small and statistically non-significant. We identify factors that may affect the internal and external validity of our results. Our methods cannot completely replace randomized experiments or panel data. However, the methods used here provide a way to help control for the joint occurrence of self-selection and simultaneity biases, and provide a ballpark estimate of causal effects, for large-scale, general-purpose, cross-sectional datasets such as the CHTS.

Keywords: Future mobility; Program evaluation; Omitted variable bias; Treatment effects; Vehicle ownership; Selection on unobservables; Emerging mobility; Sustainability (search for similar items in EconPapers)
Date: 2019
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (5)

Downloads: (external link)
http://link.springer.com/10.1007/s11116-017-9791-1 Abstract (text/html)
Access to full text is restricted to subscribers.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:kap:transp:v:46:y:2019:i:1:d:10.1007_s11116-017-9791-1

Ordering information: This journal article can be ordered from
http://www.springer. ... ce/journal/11116/PS2

DOI: 10.1007/s11116-017-9791-1

Access Statistics for this article

Transportation is currently edited by Kay W. Axhausen

More articles in Transportation from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().