Several travel indicators were compared between telecommuting (TC) days and non-telecommuting days for a sample of 72 center-based telecommuters in California. Distance traveled decreased significantly on TC days, with average reductions of 51 person-miles (58%) and 35 vehicle-miles (53%). When weighted by telecommuting frequency, average reductions of 11.9% in PMT and 11.5% in VMT were found over a five-day work week. Person-trips and vehicle-trips increased slightly (but not significantly) on TC days. This was due to statistically significant increases in commute trips by telecommuters (who more often went home for lunch on their TC days), partly counteracted by decreases in non-commute travel. The drive-alone mode share increased on TC days, both for all trips, and for commute trips in particular. Walking and biking shares also increased modestly on TC days, whereas shares of transit and ridesharing declined. Despite the increase in trip rates, TC-day reductions were found for all pollutants analyzed: 15% for total organic gas emissions, 21% for carbon monoxide, 35% for oxides of nitrogen, and 51% for particulate matter. The reduction in VMT more than compensated for the marginal increase in number of trips (and consequently, cold starts) on telecommuting days.

The study of temporal patterns of telecommuting is essential in understanding the adoption of telecommuting and, hence, the impacts of telecommuting on the demand for equipment and services as well as the demand for travel. This research examines, in the context of center-based telecommuting, how often individuals telecommute, the duration of their telecommuting participation, and causes of attrition among telecommuters. It also presents related findings from previous studies of home-based telecommuting. Attrition at the telecenters studied was relatively high, with 50% of all telecommuters quitting within the first 9 months. The average telecommuting frequency across the sample was 22% or about 1.1 days per week. Nearly 64% of the participants telecommuted less than 1 day per week on average. The relationship between frequency and duration appears to be complex, with partially counteracting trends. The results suggest that there is a stable segment of the sample (stayers) who are committed higher-frequency telecommuters, but that within the segment having a propensity to quit, there is a slight but statistically significant tendency for higher-frequency telecommuters to quit sooner. The motivations of participants for quitting the program were investigated. The most frequent type of reason given was job-related (cited by more than a third of all quitters). Other important reasons were supervisor-related (16%) and closure of the center (12%). No one cited dissatisfaction with telecommuting as a reason for quitting, and most quitters expressed a desire to continue telecommuting from the center.