skip to content

Keynes Fund


Summary of Project Plan

In many developed countries, public opinion on many important societal issues is increasingly divided along political and ideological lines (Mason, 2018).

Furthermore, recent evidence from the US suggests that this division is greater among individuals with higher levels of education (Drummond and Fischoff, 2017). Building on these findings, our preliminary work documents an education-attitude polarization gradient in several countries (Section 2.1): both the strength of agreement or disagreement with key policy issues and the difference in attitudes of liberals and conservatives towards these policies increase with education. This education-attitude polarization gradient seems to contradict the commonly held notion that by increasing knowledge and promoting critical thinking, education helps individuals with disparate beliefs appreciate the merits of alternate viewpoints and build consensus.

This project explores the relationship between educational attainment, behavioural biases, and attitude polarization. Theoretically, education could affect attitudes by changing the way that individuals seek and process information, which in turn affects the strength of an individual’s attitudes towards an issue, but this important mechanism has received little attention in the literature. To this end, we will conduct a large-scale study to examine whether the education-attitude polarization gradient can be explained by educated individuals being more prone to interpreting information as confirmation of one’s existing beliefs (confirmation bias). Our focus on attitudes towards policies complements existing studies, which primarily focus on behavioural biases in attitudes towards scientific facts (e.g. Kahan et al. 2015, 2017). Using an online experimental setting, we also investigate the effectiveness of relatively costless behavioural interventions aimed at reducing this polarization gap, such as prompting individuals to consider alternative arguments or articulate causal mechanisms to explain their attitudes towards certain policies (e.g. how immigration increases the unemployment rate).

This project makes one methodological contribution and two empirical contributions, all of which have potential applications in other contexts: we (1) introduce a novel survey tool that obtains a direct measure of confirmation bias, in contrast to existing approaches that use indirect or proxy measures, (2) use a demographically diverse sample and controlled experimental setting to examine how confirmation bias varies across socio-demographic groups, (3) investigate how responsiveness to simple behavioural interventions varies with educational attainment and other socio-demographic characteristics.

The findings from this study can benefit the fund’s mission in several ways. First, information asymmetries are an important source of market failure. Standard models predict that providing individuals with information will “correct” their beliefs and behaviour. However, our preliminary results suggest that if people interpret information in ways that confirm their prior beliefs, then information provision alone cannot correct market failures. Second, the documented education-attitude polarization gradient has ramifications for real economic decisions, such as hiring decisions (Gift and Gift, 2015) and purchasing decisions (McConnell, 2018). Third, if more-educated individuals are more selective in the way they search and process information, this suggests education may amplify our cognitive biases instead of reducing them.



Dr. Ines Lee


Ines Lee is a Junior Research Fellow at Homerton College, University of Cambridge. Her research interests are in Applied Microeconomics: Healthcare, Education, Labour.


Eileen Tipoe


Eileen Tipoe is a Senior Lecturer at Queen Mary University of London. Her research interests are in Applied Microeconomics, focusing on Revealed Preference and Welfare Analysis.


Share this Project