UWSpace

UWSpace is the University of Waterloo’s institutional repository for the free, secure, and long-term home of research produced by faculty, students, and staff.

Depositing Theses/Dissertations or Research to UWSpace

Are you a Graduate Student depositing your thesis to UWSpace? See our Thesis Deposit Help and UWSpace Thesis FAQ pages to learn more.

Are you a Faculty or Staff member depositing research to UWSpace? See our Waterloo Research Deposit Help and Self-Archiving pages to learn more.

Photo by Waterloo staff

Communities in UWSpace

Select a community to browse its collections.

Recent Submissions

  • Thumbnail Image
    Item type: Item ,
    Monetary policy regimes and beliefs
    (University of Waterloo, 1999-06) Andolfatto, David; Gomme, Paul
    This paper investigates the role of beliefs over monetary policy in propagating the effects of monetary policy within the context of a dynamic, stochastic general equilibrium model. In our model, monetary policy periodically switches between low and high money growth regimes. When individuals are unable to directly observe the current regime, they will assign some probability to the low money growth, low inflation regime base don observed money growth rates. We show that for an empirically relevant money growth process, beliefs evolve slowly in the wake of a regime change. As a result, our model is able to capture some of the observed persistence of real and nominal variables following such a regime change.
  • Thumbnail Image
    Item type: Item ,
    Immigration control and the welfare state
    (University of Waterloo, 1997-01-28) Myers, Gordon M.; Papageorgiou, Yorgos Y.
    We examine immigration policy and its redistributive effects using a model of a rich country which must spend on boarder control in order to regulate immigration from a poor country. There are owners and workers in the rich country, and a public sector which makes redistributive transfers from owners to workers. We first consider the case where illegal immigrants have access to the public sector, a situation currently observed in many countries. We show that as border control becomes more expensive inequality in the rich country increases, redistributive transfers may increase or decrease, some immigration is permitted and foreign aid may be used by the rich country in order to reduce the migration pressure along its border with the poor country. Because of nonconvexities, we also show that a small decrease in the aversion to inequality or a small increase in the poor country's population can lead to illegal immigrants from the public sector (e.g. California Proposition 187). We find that the possibility of collapse vanishes and that the rich country takes the toughest official stance on immigration but does not enforce it with border controls.
  • Thumbnail Image
    Item type: Item ,
    Technology diffusion and aggregate dynamics
    (University of Waterloo, 1998-01) Andolfatto, David; MacDonald, Glenn M.
    This paper develops and analyzes a macroeconomic model in which aggregate growth and fluctuations arise from the discovery and diffusion of new technologies; there are no exogenous aggregate shocks. The temporal behavior of aggregates is driven by individuals' efforts to innovate and/or make use of others' innovations. Parameters describing preferences, production possibilities and learning technologies are estimated using post-war U.S. data. The model delivers predicted aggregates that grow and fluctuate much like the data. The key features of post-war growth are explained by new technologies that differ in terms of the magnitude of their improvement over existing methods and the difficulty of acquiring them. The model implies a negative trend in technological dispersion, and that the generally lower growth witnessed during the last two decades is the result of new technologies offering comparatively minor or less broadly-applicable improvements. Data on the growing and fluctuating share of engineering Ph.D.s support the model's technological interpretation of the growth facts, and data on patent applications and adult schooling are consistent with the notion that newer technologies are more specific and proprietary.
  • Thumbnail Image
    Item type: Item ,
    Evaluating LLM Robustness Under Adversarial and Conflicting Evidence
    (University of Waterloo, 2026-07-23) Amirshahi, Shakiba
    Large language models (LLMs) are increasingly used in applications that rely on externally retrieved evidence, including health question answering, scientific claim verification, and retrieval-augmented generation (RAG). A fundamental question underlies these systems: do LLMs genuinely reason over the evidence they receive, or do they primarily follow the stance expressed in the provided documents? This thesis investigates this question through two complementary empirical studies that examine model behavior under harmful, adversarial, and conflicting evidence conditions across health question answering and claim verification tasks. Study 1 evaluates RAG robustness in the health domain using expert-annotated collections from the TREC 2020 and 2021 Health Misinformation Tracks. Across six LLMs, eight document types, and three query framing conditions, results show that retrieved evidence strongly shapes model behavior regardless of its reliability. Helpful documents drive ground-truth alignment to near-ceiling levels, whereas adversarial documents generated from scratch can reduce alignment to near-zero. Even a single helpful document within an otherwise adversarial retrieval pool substantially improves robustness, highlighting retrieval composition as a key factor in RAG performance. Models also demonstrate greater robustness on COVID-19 queries than on general health questions, suggesting that resistance to misleading evidence may vary across domains. Study 2 extends the analysis to explicit claim verification, evaluating five LLMs across two domains: Check-COVID, a scientific verification benchmark, and Emergent, a journalistic rumor dataset. Under both single- and paired-document settings, models frequently reverse their verification decisions when evidence stance is flipped, struggle to maintain stable judgments under conflicting evidence, and exhibit sensitivity to document order. These vulnerabilities persist across both scientific and journalistic domains, suggesting that evidence-driven behavior is not domain-specific but a broader limitation of current verification systems. Across both studies, adversarial documents generated from scratch are consistently more damaging than naturally occurring harmful content. Taken together, the findings show that strong benchmark performance does not necessarily indicate robust evidence reasoning. Helpful evidence can mask differences between models, whereas adversarial evidence exposes substantial variation in robustness. These results highlight the need for evaluation protocols that explicitly test model behavior under misleading and conflicting evidence and motivate future evidence-grounded systems that assess evidence credibility rather than simply reproducing its stance.
  • Thumbnail Image
    Item type: Item ,
    Climate change and optimal rotation in a flammable forest
    (University of Waterloo, 2001) Stollery, Kenneth R.
    This paper builds a Faustmann-based model to investigate the effects of increased climate-induced fire risk on the optimal rotation period in a commercial forest. Simulations using species of trees prevalent in North America forests indicate that both the commercial and socially optimal rotation ages decline as the risk increases. This occurs despite the fact that the inclusion of carbon sequestration benefits in society's maximand means that the socially optimal rotation length exceeds the length that is commercially profitable. The increased fire risk as the climate warms also has important implications for the ability of forests to act as absorbers of carbon. The arguments of the 'Umbrella Group' of countries who desire to use their forests' carbon-absorbing ability to offset their need for fossil fuel emission reductions will have increasingly less force as the climate warms. Because the heightened fire risk dramatically reduces the ability of living forests to act as carbon sinks, alternative proposals for storing carbon by 'pickling' wood in cold lakes look increasingly attractive.