Derrick Xu 徐明垚
I am a PhD candidate in Economics at the University of Bristol, supervised by Profs Sarah Smith and Yanos Zylberberg. I am an applied microeconomist working in the field of public, environmental, and behavioural economics.
Drawing upon my background in computer science, I have a strong interest in creating innovative datasets from unconventional sources, such as training deep learning models to analyse text and images collected from social media.
You can access my CV here.
Contact: derrick.xu[at]bristol.ac.uk
Working Paper
This paper provides novel evidence on the spillover effect of negative reputation shocks affecting a charity on donations to other charities. I identify negative reputation shocks to large charities by using sentiment analysis on the universe of news articles in the United Kingdom. I then link these large charities to all other charities by constructing measures of textual similarity between charities’ mission statements. I find that a negative reputation shock to a charity increases donations to other charities that share similar objectives. This finding is consistent with the charity market being a differentiated market in which donors care about charities’ missions and substitute across charities with similar objectives.
How directly do people need to experience climate change to change their actions? Using real donation records spanning a decade from 90,000 donors in England, I show that people are more likely to donate to environmental charities after direct exposure to a flood affecting their own postcode. However, the same is not true when a flood affects their close neighbors, even those within 200 meters. This localized response also extends to public support for the Green Party. Underlying the behavioral effect, I show that people tend to reassess their own environmental effort as not enough after direct flood experience. These results suggest an "only in my backyard" phenomenon, where people act only when personally affected by climate consequences. This offers a possible explanation for the widespread perception of global warming as a severe threat in the UK, yet still insufficient public responses.
Research in Progress
Response to Emotional Appeal
Emotional appeals are widely used to persuade people, and charitable giving, which is largely driven by emotions, provides an ideal context to study their effects. I collected 3 million images from 2,000 charities' Twitter posts and trained a deep learning model to predict the emotions they evoke. I find that responsiveness to negative appeals is tied to core values, such as universalism, and support for policies like EU integration and immigration. Further, while negative emotional shocks are effective, their repeated use shows diminishing returns. The results suggest that emotions might affect people’s intrinsic values, but key dynamics are at play. This project will add to the understanding of the strategic use of emotional appeals and their real-world effects.
The Impact of War-on-woke on Progressive Values and Prosocial Behaviour (with Sarah Smith)
Covid and Charitable Giving (with Kimberley Scharf and Sarah Smith)
Teaching
Data Analytics (postgraduate)
Economics Data (undergraduate)
Mathematics for Economics (undergraduate)
Econometrics II (undergraduate)
Mingyao Xu Mingyao Xu Mingyao Xu Mingyao Xu Mingyao Xu Derrick Xu Derrick Xu Derrick Xu Mingyao Xu Mingyao Xu Derrick Xu Derrick Xu Mingyao Xu Derrick Xu Mingyao Xu Derrick Xu Mingyao Xu