03 October 2019

When Economists Study Beauty in China

Dr Xinchen Dai dedicated two chapters of her PhD thesis on Chinese development to a topic as yet less studied: the relationship between body-mass-index (BMI) and physical beauty, and the relationship between beauty and individual labour market earnings. She found that beauty matters more for women than for men in the labour market, but the lack of data and the ensuing methodological challenges have proven quite daunting. Nonetheless, Dr Dai pursues her research work jointly with Jean-Louis Arcand, her PhD supervisor, on a topic that she has heard some call “trivial”. Read her account of her doctoral research.


Evolutionary psychologists have studied human physique attractiveness for quite a long time. They usually focus on the relationship between the body shape and the judgement of physical attractiveness, for example, how do body mass index (BMI) and waist-hip ratio (WHR) affect the perceived female attractiveness? Furthermore, what is the role of physical attractiveness in mate choice and sexual selection? Recently, social scientists started to investigate the influence of physical attractiveness on people’s social life. Both experimental and observational datasets are used to study the association between facial traits and political corruption, political promotion, and school and labour market performance.

Studies using data from developed countries predominate in this branch of literature. As a development economist, I worked extensively with Chinese household surveys during my PhD under the supervision of Professor Jean-Louis Arcand. We noticed that in two nationally representative Chinese household surveys, the interviewers rate the physical attractiveness of their respondents at the end of the interview. This provided us with an opportunity to enrich the literature with evidence from developing countries.



Economists have found the existence of a “beauty premium” (more attractive people tend to earn better salary) in the labour market. However, there are still methodological issues that have not been well addressed. First, as for most empirical research carried out every day by economists, one of the difficulties in our work is to establish the causality between physical attractiveness and wage. This requires the exclusion of unobserved factors that can influence wage through their correlations with physical attractiveness, and the reverse causality going from wage to beauty. Second, whether the beauty premium is universally present across different types of jobs or whether it is related to the productivity enhancement – which means that it mostly exists in the jobs that require frequent interpersonal interaction – is still an open question. If the latter is true, we should expect to see that attractive workers sort themselves into these kinds of jobs. Previous studies usually dealt with only one of the two issues. In my thesis, I tried to study the beauty premium by applying two methods which could solve both problems simultaneously. I identified all the parents-children pairs in the sample and used parental physical attractiveness as the instrumental variables of the adult children. Furthermore, I estimated the hazard rate of being employed in beauty-related fields for each individual and used it as the control variable for sorting.



We found that the beauty premium is more prominent for females than for males, but we were not able to find statistically significant support for such sorting in our sample. Our results seem to suggest an attractiveness-based discrimination in favour of women on the employer side but due to the data limitation, currently we cannot further explore the heterogeneity among job types or education levels. Moreover, the instrumental variables we used to establish causality are only arguably valid and the introduction of these variables forced us to work with a specific sample. Therefore, I regard my thesis work as a starting point for studying the role of physical attractiveness in broader settings. I am now an assistant professor in a local university in China. Together with Jean-Louis Arcand, we are working on subjective wellbeing represented by happiness or life satisfaction in China. One of our subprojects is to bring my thesis work into this topic and examine the interspousal association between physical attractiveness and subjective well-beings and how the two interact with individual and household income.

*  *  *

Xinchen Dai defended her PhD thesis in International Economics in November 2018. Professor Martina Viarengo presided the committee, which included Professor Jean-Louis Arcand, thesis director, and Professor Xavier Ramos from the Graduate School of Economics, Universitat Autònoma de Barcelona, Spain.

Full citation of the PhD thesis:
Dai, Xinchen. “Three Essays in Chinese Development.” PhD thesis, Graduate Institute of International and Development Studies, Geneva, 2018.

Good to know: members of the Graduate Institute can download Dr Dai's PhD thesis via this page of the Institute’s repository.

Edited by Nathalie Tanner, Research Office.
Banner image: except from a picture by KaptureHouse/