Work in Progress
Education and Risk Willingness over the Life Cycle
Discrimination in Credence Goods Markets
with Redzo Mujcic, Robert Stüber, and Michael Challis.
Artificial Companionship
with Tsz Kwan Emmy Wong.
Essays
Education, Conditional Aspiration, and the Lifecycle of Risk Willingness
MRes-2 Thesis. Supervisor: Prof. Daniel Sgroi.
This paper shows that education does not shift risk willingness uniformly: it rotates its lifecycle profile. Using 373,351 SOEP person-years, this paper separates between-person profiles from within-person change. More highly educated respondents report lower risk willingness in early adulthood but higher willingness later; within individuals, the higher-education profile declines more rapidly before approximately age 40 and is broadly parallel thereafter. A phase–amplitude model is proposed to summarise this pattern: moving from 10.5 to 14 years of education advances the fitted profile by 5.3 lifecycle years and increases its amplitude by 39 per cent. Conditional aspiration is defined as the standardised residual from fiveyear-ahead life-satisfaction expectations after flexibly conditioning on current life satisfaction and predetermined background. Its education gradient turns positive in the mid-twenties, its persistent component predicts subsequent risk willingness, and it updates asymmetrically: disappointment produces a larger revision than an equal positive surprise, with substantially less asymmetry at higher education levels. Two schooling reforms provide complementary local anchors. Among respondents in the locally high-Gymnasium-propensity group, G8 assignment advances academic transitions. Among those in the locally highest-Hauptschule-propensity group, C9 assignment raises completed schooling, later-life risk willingness, conditional aspiration, and its strictly prior history; under instrumental-variable assumptions, 2SLS indicates a positive local effect of schooling on later-life risk willingness. The paper’s central contribution is to map risk willingness, education, and aspiration over the life cycle using both pooled and within-person estimates. It offers an integrated account that combines panel dynamics, expectation updating, and distinct reform margins.
Children’s Cognitive Ability and Behavioural Difficulties: A Static and Dynamic Analysis from the UK Millennium Cohort Study
MRes-1 Thesis. Supervisor: Prof. Sonia Bhalotra.
This study examines the relationship between cognitive ability and behavioural difficulties in Generation Z, leveraging the UK Millennium Cohort Study. We employ a multi-method analytical sequence that progresses from static regressions to dynamic panel fixed effects, which distinguishes between-child from within-child variation, and finally to a Cross-Lagged Panel Model to establish temporal precedence. The results reveal a robust unidirectional relationship, where cognitive ability predicts subsequent behavioural difficulties, but not vice versa. This relationship is characterised by a non-linear, asymmetric association: while higher cognitive ability generally predicts lower subsequent difficulties, the pattern inverts for children performing far from their mean, associating greater cognitive ability with increased behavioural problems.
Welfare Effects of UK State Pension Age Reform
MSc Dissertation. Supervisor: Prof. Johannes Spinnewijn.
This paper analyses changes in the fiscal revenue of the government to evaluate the welfare effects of the UK State Pension Age Reform by the sufficient statistics approach. I identified five treated cohorts experiencing varying degrees of state pension delays due to the UK SPA reform. I constructed pseudo panels using EFS and LSFS data and estimated TFE, MFE, and BFE to assess the fiscal effects. Utilizing the sufficient statistics approach, I calculated the reduced form estimate M and compared the marginal benefits across cohorts and ages. I found that M averaged between 0.53 and 0.62 for the different cohorts. The average M increases slightly with longer delays but remains relatively consistent overall. Cohorts with shorter SPA delays exhibited higher M at age 60, attributed to the stickiness in employment conditions, resulting in lower negative BFE in terms of employment taxes and tax credits.
Consumer Behavioural Anomalies and Reaction Dynamics on Hong Kong Consumption Voucher Scheme 2021
UG Senior Thesis II. Supervisor: Prof. Matthew Shi.
Evaluating the effectiveness of Hong Kong Consumption Voucher Scheme (HKCVS) is an important study to improve future social policy design. Extended from my previous track survey study on spending behaviour reactions on HKCVS, it is observed that the voucher does not only play the role of the voucher money itself but acts as a nudge to boost extra consumption. The motivation of this paper is to provide a new point of view through the lens of behavioural economics on how and why HKCVS is effective in stimulating consumption, by utilizing the comprehensive transaction-level AlipayHK data to consolidate the conclusions beyond self-reported data. This paper 1) studies comprehensively the dynamic consumption patterns of the first wave of Phase I consumption vouchers to understand consumer impatience, 2) compares the consumption patterns between age, gender, and different payment channels groups to study reaction heterogeneity, and 3) compares the propensity to consume between the two waves vouchers in Phase I. Results go in line with the hypothesis that voucher receivers react the most when they just receive the vouchers. In precise, this study finds that compared to the average spending in the sixth-to-end weeks, consumers tend to spend around 450 HKD more in the first two weeks, and significantly less from the third week onwards (around 200 HKD more per week). Other results on reaction heterogeneity and the comparison between reactions on two waves will be added in future.
Saving Behaviour Reaction on External Shocks: A Natural Experiment on the Release of Hong Kong Consumption Voucher
UG Senior Thesis I. Supervisor: Dr. Vinci Chow.
This paper conducted a natural experiment by distributing two tracking surveys before and after the release of the second consumption voucher in the Phase II Hong Kong Consumption Voucher Scheme on 1st October 2022, to study if saving behaviour of young people will be significantly affected after the release of voucher. By comparing the two sets of data, paired t-tests were applied, voucher receivers under age 20 have found significantly saving less than before the release of voucher, and shown significantly lower anxious level when overspent. This change can be largely explained by mental accounting. Also, non-receivers in this age group have also shown significantly lower saving percentages in the second survey. Besides, by the OLS regression approach, a negative interaction effect between patience and receiving voucher has also been found on saving behaviour: after receiving the voucher, personal patience plays a less important role on affecting a person’s saving behaviour. Lastly, all respondents showed higher consumption on durable goods and services after the release of voucher.