Academic Work

This dissertation examines the implications of prioritizing those worse-off in policy analysis, being the first study to construct a Prioritarian Social Welfare Function (SWF) using life satisfaction as a welfare measure. The study begins with a comprehensive theoretical analysis, justifying a focus on the improvements for those worse-off based on the claims-based fairness concept. It advocates for Prioritarianism over Egalitarianism, particularly in light of the levelling-down objection, and ultimately justifies life satisfaction as a welfare measure as it aligns most closely with democratic principles.

Building on this theoretical foundation, I analyzed the impact of Victoria's second COVID-19 lockdown (July-October, 2020) on life satisfaction, utilizing Australia's major household panel survey (HILDA) and a difference-in-differences approach to isolate the policy's effect from the general pandemic by comparing changes in life satisfaction over time in the treatment group (Victoria, August–November, 2020) with changes over time in the control group (all other Australian states, August–November, 2020).

The results indicate a highly significant negative impact, with a decrease of 0.0303 points on a 0-10 Likert scale (95% CI -0.0486 to -0.0120, p = 0.006). Further analyses reveal a disproportionately large effect on individuals with lower life satisfaction, poor mental health, and younger age (15-19 years), highlighting the need for targeted complementary wellbeing interventions in the event of future lockdowns.

Ultimately, the difference-in-differences design was used to construct a Prioritarian SWF, contrasting the modeled social welfare of Victoria in 2020 under the lockdown scenario (13,775,625 Prioritarian WELLBYs) to a Business-as-Usual scenario (52,366,212 Prioritarian WELLBYs). In conclusion, while the lockdown aimed to protect public health, the analysis suggests that the wellbeing decline outweighed the benefits of reduced premature deaths. This research offers guidance for prioritizing those worse-off in future policy analyses and adds vital evidence to the global pandemic discourse.

In his book Valuing Health (2015), Daniel Hausman argues that well-being is too person-related to be accurately measured for policy purposes. He contends that existing measures fail to account for the heterogeneity in how objective goods contribute to individual well-being while simultaneously enabling comparisons between individuals on a population level. However, this problem of heterogeneity can be mitigated by focusing on the well-being of specific kinds of people within specific circumstances.

This essay first examines the challenges of measuring well-being for policy purposes, particularly the issue of heterogeneity as highlighted by Hausman. It then outlines how Anna Alexandrova's contextualism (2017), which uses a variant concept of well-being, can be a potential solution, and ultimately, argues for adopting a differential realization view with life satisfaction as an invariant concept to ensure comparability in well-being policy analyses. It explores different philosophical theories that have inspired various concepts and measures of well-being, such as hedonism, subjectivism, and objective list theories and then addresses Hausman's critique that well-being is not measurable for policy purposes.

To address this challenge, the essay proposes adopting a differential realization view with life satisfaction as an invariant concept for policy evaluations. This approach enables comparisons across different contexts by maintaining a stable concept of well-being while allowing for contextual variations in the factors that contribute to life satisfaction.

The essay concludes that while measuring well-being for policy purposes presents challenges due to individual heterogeneity, the differential realization view offers a practical solution by using contextual indicators to account for individual differences while maintaining a consistent comparison.

This essay examines self-defeating behaviors through the lens of dual-process models, which distinguish between autonomous, unconscious (System 1) and reflective, conscious (System 2) processes. The paper conceptualizes self-defeating behaviors into a theoretical framework based on self-control issues and negatively biased self-reflecting, analyzing how these behaviors manifest through the interplay between impulsive and reflective processing.

Drawing on established dual-process theories (Evans & Stanovich, 2013; Kahneman & Frederich, 2002), the essay explores how System 1 processes function as rapid defaults shaped by previous experiences, while System 2 processes, though requiring more cognitive effort, can intervene and adjust for biases. This framework is applied to analyze cognitive vulnerabilities in behaviors like overspending, substance abuse, and task avoidance influenced by low self-esteem.

The paper further examines how self-control challenges arise from the dynamic between "the doer" seeking immediate satisfaction and "the planner" optimizing for long-term outcomes (Thaler & Shefrin, 1981). Additionally, it explores how negatively biased self-reflective processes can lead to avoidance behaviors that prevent progress toward goals.

Building on this theoretical foundation, I propose cost-effective interventions categorized into other-deployed approaches (like defaults, taxes, and bans) and self-deployed strategies (such as mindfulness meditation). The paper concludes with ethical considerations regarding the balance between enhancing autonomy through System 2 strengthening versus reducing autonomy by targeting System 1 processes to increase overall well-being.

This essay investigates the causal effect of governmental cash transfers on self-reported political support for the incumbent government in Uruguay. Using a Regression Discontinuity Design (RDD), the analysis examines how eligibility for the PANES program affected political support for the government.

The research utilizes a predicted income score as the running variable and includes McCrary density tests and placebo regressions on covariates.

The main results indicate a consistent positive effect of the treatment on government support in 2007 and 2008, with coefficients ranging from 9% to 13%. The study also finds significant positive effects on confidence in the president, confidence in the Ministry of Social Development, and interest in politics.

Extensive robustness checks, including donut hole regressions and local polynomial analyses, confirm the stability of the treatment effect. The research concludes that eligibility for the PANES Program in Uruguay is associated with an increase of around 10% in self-reported support for the incumbent government.

The essay concludes with limitations in external validity and potential confounding factors, while suggesting policy implications and avenues for further research using alternative quasi-experimental methods.