Is telecommuting sustainable? An alternative approach to estimating the impact of home-based telecommuting on household travel

ABSTRACT

Telecommuting is regarded as a sustainable travel-demand management strategy. In investigating how much telecommuting reduces travel, studies typically adopt one method of the following three: (1) multiplying the telecommuting frequency by the round-trip commute distance, (2) measuring travel changes using quasi-experiments, and (3) determining the marginal effect of telecommuting on travel using econometric analyses. However, each approach has methodological limitations, and previous studies fail to take into consideration various aspects of household travel other than the telecommuter's commute. Against this backdrop, this study presents an alternative approach to estimating the impact of the household head's telecommuting on total household travel other than his or her commute travel using data from the 2006 Household Travel Survey in Korea. Our series of stratified tobit regression analyses show that when the household head does not commute, the usual compensatory travel mechanism induces person-kilometers traveled of 2 km. This rebound effect doubles when the household head telecommutes, although not on regular commuting days. The effect suggests that the benefits of telecommuting are significantly less than anticipated. Consequently, planners and policy makers should consider this counteracting effect when predicting the travel reduction effect of telecommuting or determining the level of telecommuting required to achieve environmental policy goals for urban sustainability.

KEYWORDS:

• Household travel
• rebound effect
• Seoul metropolitan area
• telecommuting
• telework
• tobit regression

Funding

This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science, and Technology (2012 R1A1A2009216) and also partially supported by the Architecture and Urban Research Institute. I am grateful for the useful comments from the three anonymous referees and the editor. I would also like to thank Hyungkyoo Kim for his excellent suggestions for improving the manuscript.

Notes

¹ This study uses the dataset used by Kim et al. (2012). While Kim et al. (2012) analyzed the relationship between telecommuting and residential location, this study investigates the relationship between telecommuting and household travel. This study incorporates regional job accessibility, which was used as a dependent variable by Kim et al. (2012) as an independent variable to control for location-specific characteristics that may affect household travel.

² Both studies are based on the same survey data and geographical boundaries as this article, but the research questions and specific data and methodologies applied are different.

³ In Korea, the conventional meaning of a household head is the oldest man in the household regardless of his job or income. In the case of household without male members, the oldest woman is regarded as a household head, and only 5.4% of the household heads in the sample were female (Table 1).

⁴ All official surveys administered by the Korean government strictly protect respondents' personal information by not disclosing home and workplace addresses and travel destinations. Therefore, in measuring travel distances, I used the Euclidian distance that is based on the distance between centroids of the travel analysis zones in which origin and destination are located, instead of the network distance between exact x-y coordinates of them. This may generate shorter distances than actual, but the internal validity is secured by measuring all travel with the same method.

⁵ Kim et al. (2012, p. 1171) suggested four general conditions of defining telecommuters: “telecommuters should (1(1) ) be salaried employees; (2(2) ) work at a location that is either at home or closer to home than their regular location; (3(3) ) reduce commuting partially or completely; and (4(4) ) be in contact with their office using ICT equipment.” They introduced the fittest definition to these conditions, given the nature of the SMA HTS data.

⁶ This refers to a change in population density from 11,339 to 30,822 people/km² and in regional job accessibility from 165,877 to 450,900.