Work in progress

Place

and political geography

Progressive nurture. Early-life exposure to urbanisation shapes long-term individual outlooks. With Z. Wei (Cambridge), N. Lee (LSE), A. McNeil (UCL).

Urban-rural polarisation of values is a key societal cleavage of our time. However, there is still debate about whether the differences observed across places are contextual or purely compositional. We provide novel causal evidence showing that exposure to urban density influences individual outlooks beyond compositional effects, but this influence is primarily driven by the environments where individuals spend their early years. We do this by using household panel data from both the UK and Switzerland, which allows us to combine Two-Stage-Least Squares and ‘Movers’ research designs. We show with two different approaches that exposure to urban density before age 20 causally increases support for immigration and gender equality, highlighting the lasting influence of exposure to urban density early in life. Our results are not driven by ethnic diversity or income effects, suggesting that other mechanisms such as peer-to-peer socialisation and schooling may play a role.

Urban design interventions to reduce fear of crime in public spaces have non-linear and gender-sensitive effects. With P. Navarrete (U. Chile) and T. Schlesser (Cambridge). Nature Cities: Revise and resubmit.

Addressing crime and fear in public spaces is a priority in global urban development agendas. A family of approaches under the umbrella of Environmental Crime Prevention (ECP) suggests that modifying the built environment can reduce crime rates and improve safety perceptions. However, it remains debated whether these interventions are equally effective across genders, and what are the optimal type and intensity of ECP interventions needed to maximise safety perceptions while optimising costs. We evaluated the impact of varying intervention levels on perceived safety in neglected urban areas through a randomised experiment run in Chile using photo simulations and an adapted Positive and Negative Affective Schedule. Participants assessed three popular interventions: blind wall modifications, graffiti removal, and the integration of green infrastructure, each applied at different intensities across four types of public spaces. Based on 18,150 affective perception ratings from 2,028 respondents, our findings show that nature-based solutions targeting subconscious physiological responses to fear are more effective than interventions addressing conscious fear of crime, especially at higher treatment levels. The responses are highly gender-specific. Interventions were more effective for males than females, highlighting that environmental changes alone do not fully alleviate women’s perceived insecurity.

The global geography of climate attitudes, North and South. With F. Block (Camrbidge).

Climate change is both materially and politically uneven, with impacts and perceptions shaped by geography. While subnational spatial divides in climate attitudes are well documented in high-income democracies, much less is known about whether these patterns hold in the Global South, where policy choices will be critical for both national development and global mitigation efforts. We combine a systematic review of the spatial drivers of climate opinion with an analysis of a new dataset covering around 40,000 individuals in 20 countries. We test whether urban–rural divides, distance to core cities and local economic structures influence climate perceptions. Our findings reveal both shared and divergent geographies of climate opinion between the Global North and South, highlighting why policy approaches developed in industrialised democracies cannot be assumed to translate seamlessly elsewhere.

Political economy

of spatial divergence

Ballots, budgets and bricks: Brexit and the polarisation of individual economic behaviours. With Z. Wei (Cambridge) and P. Kuang (University of Macau). Link to WPLink to older WP. Journal of Public Economics: Revise and resubmit.

Does political polarisation influence economic expectations and actual behaviours? Using British nationally representative surveys and administrative data, we show that the Brexit referendum triggered stark divergences in individual micro and macro expectations between Leave and Remain supporters. These diverging beliefs influenced major financial decisions. Leavers became more likely to purchase durables and engage in housing transactions, and areas with higher proportions of Leave voters experienced increased housing transaction volumes and rising prices. Our findings highlight how issue polarisation, beyond partisanship, can influence both economic expectations and real-world decisions.

Local wealth inequality linked to political protests across the Global South. With Z. Wei (Cambridge), N. Lee (LSE), Rodríguez-Pose, A. (LSE).

The political consequences of inequality have become a matter of mounting concern. Relative deprivation theory posits that frustration arising from comparisons with proximate reference groups can fuel protests. Yet research has concentrated overwhelmingly on the Global North, notwithstanding the acute wealth disparities experienced by the majority of the world’s population in the Global South. To address this gap, we assemble novel, fine-grained estimates of relative wealth at the 2.4 km × 2.4 km grid level to derive measures of local wealth inequality, which we link to over 645,000 georeferenced protest events recorded from 2014 to 2018. Exploiting variation within subnational regions, we document a robust, positive association between local wealth inequality and protest incidence. We show that the inequality–protest relationship is anchored in individuals’ immediate surroundings. As the spatial radius defining ‘local’ expands, the association weakens, highlighting the primacy of experienced inequality. We also find that the inequality-protest link is mediated by national characteristics. The study advances the social sciences by furnishing novel empirical evidence on the micro-geographic underpinnings of political instability, demonstrating that inequality’s political effects hinge on proximity, local class structure and national institutional environment.

Policy

evaluation of place-based and other spatially targeted actions

NextGenerationEU lands: Assessing the territorial distribution of Italy’s Recovery and Resilience funds. With F. Scotti (Politecnico di Milano) and C. Caporali (GSSI). Link to WP1, Link to WP2.

In response to the COVID-19 crisis, the European Union introduced NextGenerationEU, its largest stimulus programme to date. We examine its territorial allocation in Italy, the programme’s largest beneficiary. We analyse the determinants of fund allocation from the central government to municipalities. Although centrally coordinated, most expenditures are awarded through calls and implemented by local governments, making the allocation process critical for evaluating spending efficiency. We assemble a municipality-level dataset on funded projects and estimate a two-stage Heckman selection model to account for local participation and award intensity. Results show that funds are disproportionately allocated to Southern urban municipalities with stronger administrative capacity and prior experience with EU Cohesion Policy. We then test whether allocations support structural change in local economies and find mission-specific logics. Inclusion and Cohesion investments follow a convergence rationale, directing more resources to areas with weaker pre-existing sectoral conditions. In contrast, Digital, Education, and Healthcare investments follow a specialisation logic, concentrating funds in municipalities with existing sectoral strengths. Finally, project delays are significantly more likely in areas characterised by lower administrative efficiency and weaker local governance. However, we do not see a clear North-South divide. Together, these findings clarify how institutional capacity and pre-existing territorial features shape the distribution and implementation of EU recovery funds.