working papers

Abstract: Universal Credit (UC) was a large-scale reform of the UK welfare system reform, replacing six existing benefits. UC aimed to simplify claims and encourage more claimants into work. We identify its effect on mental well-being, treating the phased rollout from 2013-2018 as a natural experiment. We estimated differences across well-being outcomes associated with UC exposure across Local Authorities, using not-yet-exposed areas as controls. We included working-age (18-64) respondents of the Annual Population Survey in Great Britain from 2012-2019 (n=245,658), living in low-income households. We tested for differential effects by markers of vulnerability.

UC was associated with per-claimant decreases in Life Satisfaction (-0.66; 95%CI -1.01 to -0.30), Happiness (-0.41; 95%CI -0.77 to -0.05) and Life Worthwhile (-0.73; 95%CI -1.03 to -0.42), and increases in Anxiety (+0.79; 95%CI 0.30 to 1.27). These changes were two to six times the effects of the COVID-19 pandemic. Several subgroups experienced greater effects, especially increased anxiety amongst disabled people (+0.19; 95%CI 0.12 to 0.27), single people (+0.13; 95%CI 0.06 to 0.21) and people aged under 25 (+0.27; 95%CI 0.15 to 0.39).

The introduction of UC had adverse effects across all four measures of well-being. Vulnerable groups typically experienced greater harms, reinforcing calls for health prioritising reform.

[ read ]

Abstract: One challenge presented by population aging is how to adjust public support for social care in a way that achieves desired quality of life outcomes without compromising budget sustainability. This study uses best-practice methods of economic analysis to explore projections for care and related public policy between 2020 and 2070 in the United Kingdom (UK). The UK is an interesting case study, as diverse social care provisions are adopted in the four constituent countries. Projections indicate that the number of people in need of care will approximately double over the prospective half century, with informal carers playing a key role in meeting the growing burden. Policy counterfactuals contrast the budgetary implications of closing the social care gap, particularly in England and Northern Ireland, and of easing poverty among informal carers.

Abstract:
Background

Universal Credit (UC), introduced in 2013, has led to acute health harms among unemployed people, but the policy’s impacts longer-term and on broader claimant groups remain unknown. 

Methods

We exploited the geographical variation in UC rollout to investigate impacts on population mental health (SF-12 Mental Component Summary) for up to four years following implementation for a larger cohort of claimants, including employed people. We linked 108, 247 observations (29,528 individuals) from the UK Household Longitudinal Study between 2009-2019 to administrative Local Authority district data. We used standard and novel difference-indifferences approaches to estimate the average effect in the follow-up period and at different time points post-introduction, comparing a working age UC eligible population (treatment group) to an alternative benefits population (comparison group).

Findings

UC was associated with mental health declining by 0·70 units (95% CI -1·24 to -0·15), a 1·5% relative reduction. Effects were larger during the first year of exposure (-1·01, 95% CI -1·93 to -0·10) without returning to baseline levels. Between 2013 and 2019, an estimated 111,954 (95% CI 35,497 to 182,948) additional people experienced depression and/or anxiety after UC’s introduction, 27,115 of whom may have reached diagnostic threshold for common mental disorders. 

Interpretation

UC led to a sustained deterioration in population mental health, particularly marked in the first year of implementation. Reforms to UC are warranted to mitigate adverse mental health impacts.

Abstract: This paper investigates the behavioural dynamics of the take-up of social benefits in the UK. Utilising data from the first nine waves (2010-2019) of the UK Household Longitudinal Study (UKHLS) and eligibility simulations based on the UKMOD tax-benefit calculator (UKHLS-UKMOD), the study finds that there is a significant state dependence effect once initial conditions and unobserved heterogeneity are considered. While economic factors are found to play an important role in explaining the take-up of social benefits, personality traits and cognitive skills do not exhibit a strong and direct influence on the take-up of social benefits. The study concludes by discussing policy implications.

Abstract: Caring has its most obvious effects when it is actually required. Yet the effects of care are likely to extend to other periods of the life course. People may anticipate the need to provide informal care, either as part of their fertility decisions, or in response to deteriorating health of loved-ones. Similarly, a reason given for high savings rates among the elderly is the desire to self-insure against the needs consequent on adverse health shocks, including the need for (expensive) formal care. Furthermore, both informal care and incapacity demanding care can have effects that persist well after the actual episodes of care are past, for example, due to labour market scarring and/or depleted savings. This study uses current best-practice methods of economic analysis to explore these phenomena. Focussing on the channels of employment and savings, the study considers how the effects of care vary over the life course, and with the time that episodes of care are encountered.

Abstract: The purpose of this note is to discuss the distributional effects of some a priori identified UBI schemes, as an illustration of the different trade-offs involved, and to foster the debate around UBI in France. All these schemes share the feature of retaining some limited form of conditionality - in particular discriminating access to the scheme and generosity of the benefit by age. The first three schemes consider a different UBI for adults of working age, children, and pensioners,  are characterised by the elimination of child benefits, unemployment benefits, and other working age benefits, and differ with respect to the increase in tax rates envisaged. On one extreme, we assume no increases in the marginal tax rates: in this case the UBI is funded only by a reduction in other expenditures. We then allow funding to include an increase in revenues, coming from an increase in the marginal tax rates and a decrease in the personal income tax allowance. We find that the option funded only by the elimination of selected targeted benefits is regressive on average, and that only by funding the UBI with a significant increase in general taxation, we are able to obtain a reduction in poverty rates. Finally, we also consider a more limited basic income scheme targeted to young adults between the age of 20 and the age of 24 (included) only, funded by an increase in all marginal tax rates, with no changes to the personal tax allowance). No benefits are eliminated in this case. Results point to drastic improvements for the targeted group and their families, while the costs remain manageable as they are shared by the whole population of taxpayers. 

Abstract: LABSim is a rich dynamic microsimulation model of individual and household life course events, which means that it simulates individual units over time and allows for individual characteristics to be changed according to the processes specified within the model. One of the key innovations in LABSim is its linkage with EUROMOD, a static tax-benefit microsimulation model used to evaluate the immediate distributional impact of policy changes (the “morning after” effect). The static model allows ex-post evaluation of policies, as well as ex-ante evaluation of hypothetical policy changes. When combined with the dynamic model, the policies are applied to an evolving population and evaluated over time. 

Abstract: This paper proposes a new method for imputing taxes and benefits in dynamic microsimulation and agent-based models, which draws upon existing third-party tax-benefit calculators (e.g. EUROMOD). The suggested approach has the advantages of adapting to the degree of related household sector detail described by a model, mitigating computational burden associated with inter-model communications, and permitting taxes and benefits to be imputed directly from commonly available micro-data sources. A practical application of the proposed method is reported for a dynamic programming model designed to reflect the contemporary UK policy context. Reported results indicate that the proposed method for projecting taxes and benefits can imply a comparable computational burden, and generate qualitatively similar results, to a detailed functional description of fiscal policy.

Abstract: This paper analyses the distributional effects of the COVID-19 pandemic and related tax-benefit measures in 2020 in a cross-country comparative perspective for five African countries: Ghana, Mozambique, Tanzania, Uganda, and Zambia. We first estimate the impact of the crisis on disposable incomes, how effects vary across the income distribution, and in how far tax-benefit policies stabilized earnings losses. We then evaluate the impact on income-based poverty and inequality and the contribution of discretionary tax-benefit policies in alleviating the shock. Our analysis shows modest increases in headcount poverty rates and inequality, and somewhat larger effects on the poverty gap due to lower relative earnings losses of the poor population at the early stage of the pandemic analysed here. We find very limited stabilizing power of tax-benefit policies overall and automatic stabilizers in particular. This illustrates gaps in coverage for the large informal sector and a general lack of income-related means-tested benefits. Except for the Emergency Social Cash Transfer in Zambia, discretionary tax-benefit policies adopted in response to COVID-19 have had limited impact. Pausing a large school feeding programme in Ghana during lockdown has in turn put additional pressure on households with school-age children. 

Abstract: Using UKMOD, the UK tax-benefit microsimulation model, we analyse the impact on Londoners of the Covid-19 crisis, of the emergency policies put in place since March 2020 and of some counterfactual policy options, including the continuation of the £20 weekly uplift in Universal Credit and Working Tax Credit. Our main results can be summarised as follows: 1) The emergency measures introduced by the Government in March 2020 proved very effective at protecting the incomes of Londoners, particularly some of the most disadvantaged groups. 2) Withdrawing the £20 a week uplift to Universal Credit and Working Tax credits will put 130,000 more Londoners in poverty, with some of the most disadvantaged groups, including 60,000 lone parents, being disproportionately affected. The number of Black Londoners living in poverty would grow by 8% compared to 6% for all ethnic groups. 3) Less well-off Londoners will be worst affected, with the poorest 10% experiencing an 8% decrease in their incomes. 4) While Londoners will be more affected than the rest of the country by the withdrawal of the £20 uplift; across the UK, the poorest 10% of people will see their incomes cut by 5%. 5) On the other hand, keeping the £20 a week uplift to Universal Credit and Working Tax credits and going further by removing all the Benefit Caps would cut child poverty in London by 10%, keeping more than 70,000 children out of poverty.

Abstract: We suggest a new individual-based measure of economic insecurity where expectations about the future evolution of individual life courses are derived from a dynamic simulation model. This allows to take into account risks over many dimensions including work, family and health. On the earning side, we summarise the uncertainty over future income trajectories with a monotonically increasing concave function of income, which penalises income volatility over time and over possible individual trajectories while assigning higher value to higher levels of economic resources available. On the expenditure side, we take into account different household characteristics by means of appropriate equivalence scales.

Abstract: Economic (in)security is attracting a growing interest in the social policy debate. However, there is a lack of consensus about how to measure it. In this paper we review the existing measures of economic insecurity, with a particular emphasis on objective measures. Our assessment is that all the measures are to some extent arbitrary, and that no clear consensus has so far emerged in the literature. This stems from some analytical confusion that still surrounds the concept, and from the lack of clear normative foundations. We also identify and discuss some characteristics that ideal measures of economic insecurity should have.

Abstract: We provide a clarification of scaling issues in simulation models, distinguishing between sample size determination, discovery of emergent properties involving a qualitative change in the behaviour of the system at an aggregate level, and ‘true’ scaling, the dependence of the quantitative behaviour of the system at any given level of aggregation, to its size. Scaling issues arise because we want to understand what happens when we run one billion agents, without actually having to run one billion agents. We discuss how we can use the Buckingham Pi theorem, a key tool in dimensional analysis, to provide guidance on the nature and structure of scaling relationships in agent-based models. 

Abstract: The objective of this report is to provide an assessment of labor market regulations in Liberia, in particular employment protection legislation (EPL), minimum wages, and social insurance schemes. This is particularly relevant as a new labor code, the Decent Work Bill, has just been legislated with a twofold increase in the minimum wage, rising concerns that it will further reduce an already limited demand for formal employment (less than 20% of total employment). Without changing the Decent Work Bill, which is taken as a political constraint, the report makes proposals for reforms “at the margin” aimed at extending demand for formal employment, with specific social insurance schemes targeted at the newly formalized employees.

Abstract: We present JAS 2, a new Java platform that aims at providing a unique simulation tool for discrete-event simulations, including agent-based and microsimulation models. With the aim to develop large-scale, data-driven models, the main architectural choice of JAS 2 is to use whenever possible standard, open-source tools already available in the software development community. The main value added of the platform lies in the integration with RDBMS (relational database management) tools through ad-hoc microsimulation Java libraries. The management of input data persistence layers and simulation results is performed using standard database management tools, and the platform takes care of the automatic translation of the relational model (which is typical of a database) into the object-oriented simulation model, where each category of individuals or objects that populate the model is represented by a specific class, with its own properties and methods. JAS 2 allows to separate data representation and management, which is automatically taken care of by the simulation engine, from the implementation of processes and behavioral algorithms, which should be the primary concern of the modeler. This results in quicker, more robust and more transparent model building.  

Abstract: Fifty years have passed since the seminal contribution of Guy Orcutt [Orcutt, 1957], which gave birth to the field of Microsimulation. We survey, from a methodological perspective, the literature that followed, highlighting its relevance, its pros and cons vis-a-vis other methodologies and pointing out the main open issues.

Abstract: We study how political intermediation in the labor market interacts with search frictions. Politicians create and control (to a certain extent) business opportunities for firms, hence the creation of new vacancies. But to compete for these vacancies workers have to give their support to politicians. This leads to a fragmentation of the labor market, where politicians act as mediators between demand and supply. We show that in presence of information asymmetries (when non-affiliated workers are not able to distinguish non-affiliated firms, for which they are eligible, from affiliated ones, for which they are not eligible) the impact of political intermediation is U-shaped, and can more than double the resulting unemployment rate.