PhD thesis

The Political Behaviour of Social Change: Political Accountability in an Ageing Britain

RQ: How does political accountability change when the demographic composition of the population changes?

I study the effects and consequences of an ageing population on political accountability in the United Kingdom. More specifically, I measure electoral responses to two key governmental strategies — pension and immigration reforms. I combine high-quality sources of data from national household panel surveys, population censuses, freedom of information requests, electoral results and other administrative data-bases, and apply causal inference methods to reveal how demographic changes shape the political sphere.

Accountability and time lags: The electoral consequences of pension reform in the UK


Can voters hold politicians accountable when policies are implemented and information is disseminated with a considerable time lag? This paper focuses on prospective and retrospective evaluations of changes in the State Pension Age (SPA) in the United Kingdom. Combining secure-level panel data from two national household surveys with administrative data, and applying a difference in differences research design, I analyse the variation in political preferences before and after (a) the implementation of the policy and (b) the time when voters receive written notification of the policy change (15 years later). I find that in the immediate aftermath of the reform, affected voters became more disillusioned with politics and slightly more more empathetic towards smaller parties. Surprisingly, the notification treatment has no effect on political preferences although there are still some weak effects on political disillusionment. The findings are of particular relevance to understanding how policy-makers communicate with the public and how the public responds to policies that have long-term implications.

Migration, electoral systems and the far-right: Evidence from London

[with Joachim Wehner]


When and how does the electoral system condition how voters and parties respond to immigration? We bring together ward-level census data and electoral outcomes to answer this question, exploiting the unique variation in changes to the density and composition of migrants from the European Union (EU) across London at the start of the millennium. Crucially, a new system of government for London was set up in 2000. Since then, Londoners are tasked every four years to elect their Mayor and London Assembly members. All voters receive three ballot papers, each vote being counted according to a different electoral system’s formula. This setting allows us to examine electoral and partisan responses to migration in a context where the same voters cast votes under three different electoral system (supplementary vote, first past the post and proportional representation) at the same time. The analysis highlights the importance of supply-side constraints, as extreme parties struggle to field candidates in constituency elections. We also find some voting patterns that are inconsistent with standard accounts of strategic voting.

The citizenship market in post-Brexit Britain


With dual citizenship comes not only freedom of movement, but also the right to have a say in national politics. Since migrants are not evenly distributed geographically, the sudden surge in applications for British citizenship in the aftermath of the Brexit referendum could mean that, in certain geographic areas, the median voter profile may match the profile of an immigrant. Thus, if the new citizens become active voters, a new electoral map could emerge; one where urban constituencies become even more salient in politics, and where new political entrepreneurs will start targeting this previously unenfranchised group. Studying the political socialisation trajectories of individuals who join the electorate by naturalisation is a key component in understanding the political behaviour of social change and making better electoral predictions. The paper identifies the most prevalent socio-economic characteristics of successful applicants for British citizenship and connects them with the corresponding voting behaviour literature and survey data in other to develop hypothesis about their political preferences and their participation in future elections.

Other Projects

Early work in progress:
Low Cost Flights and Far Right Votes: How Hypermobility Shapes the Electoral Geography of Migration [with Joachim Wehner]

Abstract: Coming soon…

Migration and the economic value of family [with Selina Hofstetter and Ben Wilson]

Abstract: Coming soon…

Age and electoral participation: The effect of lowering voting age in Japan [with Airo Hino and Jaehyun Song]

Abstract: Coming soon…

Past projects:
When should we cross the line? Conditions for American public approval of torture [with Rob Johns and Graeme Davies]

Abstract: Surveys of the US public reveal that around seven in ten people believe that torture can be justified under certain circumstances. This raises the obvious question: under which circumstances? In this paper, we report two survey experiments on the use of torture, embedded in a US survey (N=1,200) fielded via Amazon’s Mechanical Turk. A number of contextual factors are considered including the imminence and scale of the threat, the personal characteristics of the suspect, and the likelihood that he or she has useful information. Ramping up threat levels does surprisingly little to generate approval for torture, decisions seeming instead to crystallize on the suspect’s characteristics and affiliations. However, the experimental manipulations had on balance only a limited effect and those prepared to consider torture were therefore prepared to consider it in a wide range of situations. Given the growing consensus that public opinion constrains leaders’ foreign policy behavior, these results may help to explain why democratic countries such as the US are willing to flout international conventions by using ‘enhanced interrogation’ techniques.

How does Canada’s imported labour force perform on the labour market? [with Wouter Zwysen and Neli Demireva]

Abstract: Canada’s approach to attracting and integrating foreign labour is based on the principle of mutual understanding and respect. While this principle is effective for societal integration, labour market incorporation is still lagging behind. Using data from the Equality, Security and Community survey, we find the gap between labour returns of first-, second-generations of migrants and natives is best explained by a combination of factors – cohorts, geographical origins and educational qualification. Our findings suggest that, ceteris paribus, only first-generation migrants from outside Europe and North America are significantly less likely to be in employment and to earn less than native Canadians. We find that this sub-category of first-generation immigrants are severely penalised by employers. Even those with higher education (BA, Master or PhD) have a very small labour market advantage to Canadians with low or no education. These striking findings suggest that greater efforts are needed for the successful integration of foreign-born workers in Canada. The more positive findings of the study indicate that, by the second generation, those with migratory backgrounds succeed in closing the labour gap and become statistically indistinguishable from natives in terms of labour market integration.