Four Essays on Using Economic Analysis to Evaluate Canadian Public Policy

• Main Author: Chandler, VINCENT
• Other Authors: Queen's University (Kingston, Ont.). Theses (Queen's University (Kingston, Ont.))
• Language: en; en
• Published: 2014
• Subjects: Labor; Public; Evaluation
• Online Access: http://hdl.handle.net/1974/12213

The first essay shows the presence of discrimination in the allocation of doctoral scholarships using a unique dataset containing the 1901 recipients of 2004-2005 doctoral awards from the Social Science and Humanities Research Council (SSHRC). I provide evidence that evaluators give higher scores to candidates in their own discipline and then proceed to distinguish between two types of discrimination: taste-based and screening. This paper is the first to find support for screening discrimination: better candidates have higher scores when they are assessed by evaluators from their discipline, while there is no significant effect for weaker candidates. Moreover, the weakest candidates, those close to the funding threshold, actually have a lower probability of receiving a large award when there is an evaluator from their discipline in their evaluation committee. The second essay models the optimal distribution of government spending when the electoral benefit accrues to the district incumbent and not to the party in power. The model shows that, under certain parameters, more money is spent in core support districts. I test this prediction using the distribution of projects undertaken in the scope of the 2009-2010 Canada Economic Action Plan. The third essay assesses the deterrence of demerit points on the behavior of drivers. To address driver heterogeneity, I use the expiration of points as quasi-exogenous variation in the number of points. I can then compare the probability of a traffic violation for drivers who had the same number of points but now have a different number of demerit points. I find that a 3-point decrease in the number of demerit points through the expiration process increases the probability of a violation in the following two months by 40 to 50 percent, but only for drivers close to the threshold. The last essay studies the impact of a non-refundable tax credit for public transit introduced in July 2006 in Canada. I find no evidence that this tax credit increased the number of trips done using public transit. There is, however, some suggestive evidence that it did induce commuters to purchase monthly passes. === Thesis (Ph.D, Economics) -- Queen's University, 2014-05-29 11:18:01.817