Research

Abstract: I study the salience of dynamic incentives provided by the welfare system, as revealed by labour supply responses to foreseeable reductions in benefit income. I show that claimants fail to anticipate a large lump-sum reduction in benefit entitlement, arising predictably from children ageing out of eligibility for the UK’s Child Tax Credit. I show also that the salience of the rules increases with experience. I then develop a structural life-cycle labour supply model incorporating potential non-salience of eligibility rules. The model estimates suggest that 82 percent of claimants initially fail to anticipate the benefit reduction. The resulting optimisation errors have substantial welfare costs---equivalent to a 14 percent reduction in income from the programme, with no offsetting benefits to the government. The findings reveal a previously undocumented source of inefficiency in the welfare system, arising from non-salient policy features with significant financial consequences.

Abstract: We develop a measure of labour market opportunities for heterogenous types of worker, exploiting information on their suitability different jobs encoded in historical patterns of worker mobility. We provide a theoretical foundation for our measure, which features naturally in a general random search framework. Our measure is flexible in the sense that it admits general definitions of worker and job heterogeneity, and is easily implementable empirically with data on worker mobility and labour demand. We apply our measure to high-quality data on labour demand in the UK, based on the universe of 104.7 million job adverts posted online from January 2015 to June 2021. We demonstrate the utility of our measure with an analysis of worker prospects throughout the Covid-19 pandemic. First, while the direct impact of lockdown policies was concentrated on relatively few industries, labour demand fell much more broadly. And, as our measure highlights, the full effects were broader still because of the disruption to usual career progression, even for those in less-affected sectors such as healthcare. Second, despite aggregate labour demand returning to pre-pandemic levels by June 2021, 25% of the workforce faced new job opportunities more than 10% below pre-pandemic levels. This is because of a change in the composition of vacancy postings (towards lower-paying occupations) which our measure of labour market opportunities is sensitive to. Finally, the majority (64%) of unemployed workers faced at least 10% more competition for jobs from unemployed jobseekers than before the pandemic.

Abstract: We provide empirical evidence on the labour market impacts of covid-19 in the UK and assess the effectiveness of mitigation policies. We estimate the relationship between employment outcomes and occupational and industrial characteristics and assess the effects on consumption. Seventy percent of households in the bottom fifth of the income distribution hold insufficient assets to maintain current spending for more than one week. Finally, we compare the effectiveness of the UK’s Coronavirus Job Retention Scheme to Economic Impact Payments in the US. The EIPs are more effective at mitigating consumption reductions as they have full coverage, depend on household structure and are higher for low-income workers. 

Abstract: Saving for retirement is arguably one of the most important decisions people make over their working life. Pension schemes are often complex and provide infrequent (and delayed) feedback, making it difficult to learn about their features. In this paper, I exploit the substantial heterogeneity in pension schemes in the UK to estimate the labour supply response to similarly sized incentives arising from different features of the pension system. If people are equally well informed of the different features, the labour supply responses should be similar. Using detailed data on pension rules, wealth and labour supply for a representative sample of older people in the UK, I estimate option-value style models of retirement which allow for different responses to different types of pension. I find that (i) there are systematic differences in responses, and (ii) responses are more muted in general among workers with multiple pensions (a measure of pension complexity). The findings point to differential salience of pension incentives.