Abstract:Many regional agencies model travel demand without feeding assigned travel times back to the trip distribution step. This method saves time and money but is likely to be biased in favor of build alternatives because it would underproject trip lengthening induced by the added capacity. Emissions and travel costs for the new roadway projects would be consequently underprojected. We wanted to compare outcomes under the two simulation methods. Our methods of modeling travel demand are outlined and then the results are presented and discussed. With full feedback, building new freeway carpool lanes appears less favorable than doing nothing or than expanding light-rail transit, in terms of induced travel. Light-rail-system expansion has lower emissions than does building new carpool lanes, with partial feedback, and full feedback increases these differences substantially. Better modeling methods, to be used in extensions of this research, are outlined.