DOES COMMUTING DISTANCE MATTER? COMMUTING TOLERANCE AND RESIDENTIAL CHANGE

Final Report - Summary

William A.V. Clark

UCLA Geography Department

1255 Bunche Hall

Los Angeles, CA 90095-1524

wclark@geog.ucla.edu

University of California Transportation Center Year 13 (2000-2001)

 
DISCLAIMER
The contents of this report reflect the views of the authors, who are responsible for the facts and the accuracy of the information presented herein. This document is disseminated under the sponsorship of the Department of Transportation, University Transportation Centers Program, in the interest of information exchange. The U.S. Government assumes no liability for the contents or use thereof
Acknowledgements
The research on which this report is based was carried out for the University of California Transportation Center (UCTC). Funding for the Center comes in equal parts from the US Dept. of Transportation and the California Department of Transportation (Caltrans.)  The University of California also provides support though reduction of overhead charges and the donated time of many faculty, students, and administrative staff.  The researchers of the UCTC are grateful for all of this support.

OVERVIEW

The project used a longitudinal data set from the Seattle Washington metropolitan area to develop and test a model of commuting responses to changing residences. The empirical analysis showed that one worker households beyond a threshold distance of about 8 miles from the workplace have an increasing probability of moving closer to their work place when they change residences. The same threshold for two worker households was about 12-16 miles. Thus, the results show that the occupants of two worker households have a higher threshold before they move closer to the workplace. The model was able to capture the probability of moving closer to the workplace as the separation between workplace and residence increased.

PROJECT PURPOSE
The study of commuting in this project was designed to examine and test the extent to which individuals and households (two workers) minimize commuting distances when they change residences. The concept of commuting tolerance can be invoked to ask at what point do commuters become resistant to further increases in commute time. Although empirical observations have suggested a tolerance zone of about 45 minutes there is little hard data on the points at which commuters begin to make behavioral adjustments to the separation between home and work.

HYPOTHESES

The research was organized around two hypotheses:

1. Ceteris paribus, households who move, will move closer to the workplace. The larger their initial separation between residence and work place the greater the likelihood of moving closer to the workplace.

2. In two worker households women will have shorter commutes but those commutes will, on average, increase with residential changes.

DATA AND METHODS
A panel data set for the Puget Sound region (Seattle, Washington) was used to test, empirically, the extent to which changes in residence impact the commuting distances and times of individuals and households. The data set is a longitudinal sample of approximately 5400 households within the Seattle labor market. Demographic and commute characteristics for each household member were collected for the period 1989-1994. In addition, a sub-sample maintained a travel diary. For the purposes of this study the important data is that on residential location (tract and zipcode) destination of the workplace (tract and zipcode) purpose of the trip, mode, time and distance of the trip. In addition, the survey collected both residential and job location change.

KEY FINDINGS

 The findings indicate that both one- and two-worker households with greater separation between workplace and residence make decreases in distance and time. Overall, as other studies have shown, women commute shorter distances and are more likely to minimize commuting after a move than are men. The probability model fits the likelihood of decreasing distance with greater separation and provides a more exact specification of the connection between residence and workplace than previous analyses of this relationship.

In the aggregate more households, whether with one or two workers, reduced their commutes after moving. Analyzing the results by the pre-move commute reveals a distinct pattern in which households with longer commutes before the move, almost always reduced their commuting distance and time. The breakpoints, that is the distance or time at which more households reduced rather than increased their commutes occur at about 8 miles for one worker households and between 12 and 16 miles for two worker households.

Generally, the households that increased their commutes were those that had short commutes, either in distance or time, before they moved. This finding is especially true for one-worker households. For the very longest commutes before a residence change there is clear evidence of a tendency to maintain or reduce the commute with the residential relocation. A large number of households increase the commute time when they had relatively short commutes prior to moving. Clearly, changing jobs and houses has a significant effect on the amount of commuting. In terms of our first hypothesis there is substantial support for reduced commute distances and times with residential relocation.

 In general, women have shorter commutes than men, especially when there is no change in workplace. The findings for dual earner households who changed both workplace and residence had a large number of women with increased commutes in distance or time. Amongst women who had commutes which were under 27.5 minutes, or less than 12 miles, more than twice as many increased as stayed the same or decreased their commutes, after changing jobs and residences. There is certainly an indication in the data that women who are in two-worker households and where both the residence and the job changes, are impacted in their commutes. Even so, most of the increases for women as for men were not large. 

The conceptualization of links between residence and workplace which incorporates direction and distance can be structured as a two parameter model in which the move is a vector that has length and direction and the distribution of moves is a joint distribution of move lengths and move directions. The effect of this formulation is to discard information about the city structure and to focus on the dynamic relation between the residence and the work location.  It is possible to use this model to specify a value (k) which measures the likelihood of moving closer to the workplace with increasing separation between workplace and home. The analysis of k values for one and two worker households shows that, as expected, two worker households are less affected by workplace location than are one worker households.

 

Papers Produced with This Grant

W.A.V. Clark, Youqin Huang and Suzanne Withers, Does Commuting Distance Matter? Commuting Tolerance and Residential Change Forthcoming  Regional Science and Urban Economics. 2002

William A.V. Clark, Youqin Huang and Suzanne Withers, Does Commuting Distance Matter? Commuting Tolerance And Residential Change, Presented at the Western Regional Science Meetings in Palm Springs Februay, 2001.

Suzanne Withers and William A. V. Clark An Analysis Of The Spatial And Temporal Independencies Of Dual Earner Households Presented at the Association of American Geographers Annual Meetings, New York, March, 2001.