Date of Award


Level of Access Assigned by Author

Campus-Only Thesis

Degree Name

Master of Science (MS)


Resource Economics and Policy


Mario Tiesl

Second Committee Member

Gary Hunt

Third Committee Member

Kathleen P. Bell


Increased consumption of Light-duty vehicles (LDV) in the U.S has contributed to increased air pollution thorough criteria emissions and green house gas emissions. Marketing efforts such as eco-information programs can alter this inefficient and environmental damaging behavior. The eco-marketing campaign conducted by the School of Economics of University of Maine in collaboration with the Maine Department of Environmental Protection, the Maine Auto Dealers association, and the Natural Resource Council of Maine in 2005 is one such attempt to encourage eco-friendly (greener) consumption habits among Maine vehicle buyers. A survey done by Noblet (2006) evaluated the reaction of consumers to emission information presented by the campaign. Bacani (2008), using market data, suggested that the eco-marketing campaign did not positively affect the eco-composition of the market. Even though these studies assess the effectiveness of the campaign, neither shows how the campaign affected the likelihood that a buyer would purchase a greener vehicle. This research focuses in understanding the impact of the campaign on actual purchasing behavior. A wide range of LDV makes/models including their substitutes is available to consumers; rational consumers make choices based on their preferences and constraints. Among the large number of vehicle makes/models, a vehicle is more likely to closely relate to a subset of make/models than it is other subsets of vehicles. The ratio of choice probabilities of two vehicles is more likely to be influenced by the other substitutes available in the market, but not by the unrelated alternatives. We model this consumer choice using a nested logit specification that represents the subsets of related vehicle make/models by nests. This model permits correlation among alternatives in a given nest while maintaining the uncorrelated error terms across nests. Further, we used the SmartWaySM definition of EPA to set up the 'greener/non-greener vehicles' nest structure for our model. The data structure for the analysis was created by adopting a random sampling procedure to assign choice sets for each consumer. For the analysis we used a database of 46870 vehicle registration records from 2004 to 2007 that contain vehicle and environmental attributes, socio-demographics and gas price information. A two level nested logit model consistent with utility maximization revealed that this small-scale short-term eco-marketing campaign had short-term positive effects on the choice of new greener passenger cars. Also the results indicate that greener car buyers come from highly educated communities that mostly contain people who are employed in white collar jobs. Further, age has a significant positive, but non-linear effect on the greener vehicle choice.