Cover Image

Not for Sale



View/Hide Left Panel
Click for next page ( 151


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 150
A N I N N O VAT I V E M E T H O D O L O G I C A L F R A M E W O R K 139 attitudes, or other characteristics, may search for loca- but these methods require tedious computations to rec- tions with high residential densities, good land use mix, ognize the sampling variation in the predicted value of and high public transit service levels, so they can pursue the endogenous BE attributes to obtain the correct stan- their activities using nonmotorized travel modes. If this dard errors in the main equation of interest. Ignoring the were true, urban land use policies aimed at, for example, sampling variance in the predicted values of BE attri- increasing density or land use mix, would not stimulate butes, as done by Boarnet and Sarmiento, can lead to lower levels of auto use, but would simply alter the spa- incorrect conclusions about the statistical significance of tial residence patterns of the population based on motor- the effects of BE attributes. ized mode use desires. Ignoring this self-selection in residence choices can lead to a spurious causal effect of neighborhood attributes on travel, and potentially lead to Using BeforeAfter Household Move Data misinformed BE design policies. The literature that has considered the self-selection The third approach is to examine the travel patterns of issue (also referred to as the residential sorting issue) in households immediately before and after a household assessing the impact of BE attributes on travel choices relocation. The potential advantage of examining the has done so in one of three ways. same household in two different neighborhoods is that one can ostensibly control for the overall travel desires and attitudes of the members of a household, so that the Controlling for Decision-Maker Attributes beforeafter relocation changes in travel behavior may be attributed to the different BEs in the two neighbor- The first approach is to control for demographic and hoods. The idea in this approach is to consider the relo- other travel-related attitudes and perceptions of decision cation as a treatment, with the associated travel behavior makers that may impact the neighborhood type individ- changes being the response variable. The assumption is uals choose. This can be accomplished by incorporating that relocating households are in equilibrium in their decision-making characteristics as explanatory variables pre-move neighborhood in terms of BE attributes, and in models of travel behavior. This is a creative, and sim- moved because of factors unrelated to their preference of ple, way of tackling the self-selection problem, but its use BE attributes (such as to upgrade the physical housing is limited because most travel survey data sets do not col- stock in response to higher incomes or a change in life- lect attitudinal data. Further, it is unlikely that all the cycle). While such an approach can alleviate the self- demographic and travel lifestyle attitudes that have any selection problem, the relocating households are substantive impact on residential sorting can be collected themselves a self-selected group, and may have moved in a survey because it is difficult to gauge how close the because of dissonance in the pre-move neighborhood BE estimated BE effects are to the true causal effect. attributes. Instrumental Variables Approach PROPOSED MODELING FRAMEWORK The second approach to alleviate the residential sample This section addresses some of the challenges discussed selection effect is a two-stage instrumental variable in the previous two sections. In particular, the authors approach in which the endogenous explanatory BE propose a modeling framework that (a) accommodates attributes are first regressed on instruments that are differential sensitivity to BE and transportation network related to BE attributes, but have little correlation with variables due to both demographic and unobserved the randomness in the primary travel behavior of inter- household attributes and (b) controls for the self- est. The predicted values of BE attributes from this first selection of individuals into neighborhoods based on regression are next introduced as independent variables travel preferences. The framework can be used to con- (along with other demographic attributes of the individ- trol for residential self-selection for any kind of travel ual) in the travel behavior relationship of interest. A behavior variable and provides the correct standard problem with this approach, however, is that it is not errors regarding the effect of BE attributes. It is geared applicable to the case in which the travel behavior equa- toward cross-sectional analysis, recognizing that almost tion of interest has a nonlinear structure, such as a dis- all existing data sources available for analysis of BE crete choice or a limited or truncated variable. There are effects are cross-sectional in nature. Unlike earlier stud- control functions and related approaches to deal with ies, the methodology also models the residential location the case of endogenous explanatory variables in the con- choice decision jointly with the travel behavior choice. text of discrete choice and other nonlinear models (see The results of applying the model formulation to an Berry et al. 1995; Lewbel 2004; Louviere et al. 2005), empirical analysis of residential choice and car owner-