Research
Publications
Cotofan, M, Krekel, C., & Oparina, E. (2026) Exercise, Volunteering, and Mental Health: Evidence from a Nation-Wide Programme. Social Science & Medicine, forthcoming [conference talk 1] [conference talk 2]
Abstract: There is growing interest in how social prescribing, a therapeutic approach that refers individuals with unmet social or emotional needs to community activities, can improve mental health and revolutionize health care systems at cost. However, robust evidence on which interventions are effective is still scarce. We exploit a unique setting in the UK to measure the impacts of a lifestyle intervention that combines sports and volunteering on participant mental health. Specifically, we partner with a country-wide non-profit that provides infrastructure for exercising and volunteering in local communities. Using incentivised surveys linked to administrative data on participation, we compare individuals who participated with those who signed up but did not (recently or at all), including for exogenous reasons. We find that participation strongly reduces mental ill health and loneliness whilst raising feelings of belonging and connectedness to participants' local communities, as well as their trust towards others. Impacts are sizeable and persistent, returning to baseline more than six months after participation. The programme is highly cost-effective.
Rzepnicka, K., Sharland, E., Rossa, M., Dolby, T., Oparina, E., Saunders, R., Ayoubkhani, D., & Nafilyan, V. (2026). The Effect of Adult Psychological Therapies on Employment and Earnings: Evidence from England. Psychological Medicine, 56, e33. [working paper]
Media coverage: LSE Business Review, CentrePiece
Abstract: People suffering from common mental disorders (CMD) such as depression and anxiety are more likely to be economically inactive. Psychological therapies are highly effective at treating CMDs, but less is known about their impact on long-term labour market outcomes. Using national treatment programme data in England, NHS Talking Therapies (NHSTT), with unique linkage to administration data on employment and census records, we estimated the causal effects of NHSTT on employment and earnings. Overall, completing treatment led to a maximum average increase of £17 in monthly earnings (year two) and likelihood of paid employment by 1.5 percentage points (year seven). Those ‘Not working, seeking work’ saw a maximum average increase in pay of £63 per month (year seven) and likelihood of paid employment by 3.1 percentage points (year four). Our findings demonstrate the economic benefits of treating CMDs, and how investing in mental health can impact labour market participation.
Cooper, K. B., Heffetz, O., Ifcher, J., Oparina, E. & Wu, S. (2026). Teaching Happiness (Economics) in Your Dismal-Science Courses. The Journal of Economic Education, 1–15. [working paper]
Abstract: The authors of this article discuss ideas for incorporating the study of happiness and other measures of self-reported or subjective wellbeing (SWB) into undergraduate economics courses. They begin by motivating why economics students would benefit from learning about SWB, and then proceed to provide examples of ways to introduce this topic into different parts of the curriculum: macroeconomics, microeconomics, and upper-division electives.
Oparina, E., Kaiser, C., Gentile, N., Tkatchenko, A., Clark, A., De Neve, J-E., D'Ambrosio, C. (2025). Machine Learning in the Prediction of Human Wellbeing. Scientific Reports, 15, 1632. [working paper] [webinar: earlier version]
Abstract: Subjective wellbeing data are increasingly used across the social sciences. Yet, despite the widespread use of such data, the predictive power of approaches commonly used to model wellbeing is only limited. In response, we here use tree-based Machine Learning (ML) algorithms to provide a better understanding of respondents' self-reported wellbeing. We analyse representative samples of more than one million respondents from Germany, the UK, and the United States, using data from 2010 to 2018. We make three contributions. First, we show that ML algorithms can indeed yield better predictive performance than standard approaches, and establish an upper bound on the predictability of wellbeing scores with survey data. Second, we use ML to identify the key drivers of evaluative wellbeing. We show that the variables emphasised in the earlier intuition- and theory-based literature also appear in ML analyses. Third, we illustrate how ML can be used to make a judgement about functional forms, including the existence of satiation points in the effects of income and the U-shaped relationship between age and wellbeing
Oparina, E., & Srisuma, S. (2022). Analyzing Subjective Wellbeing Data with Misclassification. Journal of Business & Economic Statistics, 40(2), 730–743. [working paper]
Abstract: We use novel nonparametric techniques to test for the presence of nonclassical measurement error in reported life satisfaction (LS) and study the potential effects from ignoring it. Our dataset comes from Wave 3 of the UK Understanding Society that is surveyed from 35,000 British households. Our test finds evidence of measurement error in reported LS for the entire dataset as well as for 26 out of 32 socioeconomic subgroups in the sample. We estimate the joint distribution of reported and latent LS nonparametrically in order to understand the mis-reporting behavior. We show this distribution can then be used to estimate parametric models of latent LS. We find measurement error bias is not severe enough to distort the main drivers of LS. But there is an important difference that is policy relevant. We find women tend to over-report their latent LS relative to men. This may help explain the gender puzzle that questions why women are reportedly happier than men despite being worse off in objective outcomes such as income and employment.
Chen, L.-Y., Oparina, E., Powdthavee, N., & Srisuma, S. (2022). Robust Ranking of Happiness Outcomes: A Median Regression Perspective. Journal of Economic Behavior & Organization, 200, 672–686. [working paper]
Media coverage: VoxEU
Abstract: Ordered probit and logit models have been frequently used to estimate the mean ranking of happiness outcomes (and other ordinal data) across groups. However, it has been recently highlighted that such ranking may not be identified in most happiness applications. We suggest researchers focus on median comparison instead of the mean. This is because the median rank can be identified even if the mean rank is not. Furthermore, median ranks in probit and logit models can be readily estimated using standard statistical softwares. The median ranking, as well as ranking for other quantiles, can also be estimated semiparametrically and we provide a new constrained mixed integer optimization procedure for implementation. We apply it to estimate a happiness equation using General Social Survey data of the US.
Working papers
Oparina, E., Krekel, C., Srisuma, S. (2026). Talking Therapy: Impacts of a Nationwide Mental Health Service. [latest version]
Reject & Resubmit at American Economic Journal: Economic Policy
CEP Discussion Paper, IZA Discussion Paper
Media coverage: IZA Opinion Piece
Abstract: Mental health problems impose significant costs, yet healthcare systems often overlook them. We provide the first causal evidence on the effectiveness of a nationwide mental health service in England for treating depression and anxiety using non-experimental data and methods. We exploit oversubscription and resulting exogenous variation in waiting times across areas and time for identification, based on a novel dataset of over one million patients. We find that treatment improves mental health and reduces impairment in work and social life. We provide suggestive evidence that it enhances employment. Impacts vary across patients and services. Nevertheless, the programme is highly cost-effective.
Oparina, E., Krekel, C., Parkes, I., Srisuma, S., & D. Clark (2026). Evaluating the Effectiveness of Clinical Healthcare Services Using Population-Wide Patient Data: An Example of NHS Talking Therapies.
Abstract: We propose a novel methodology for evaluating the effectiveness of healthcare services at the population level, with the aim and ambition to complement evidence produced by RCTs. Our methodology is particularly suitable in clinical settings where services are rolled out gradually or are oversubscribed, often resulting in waitlists for access to care. We illustrate it using an example from the NHS Talking Therapies programme, a nationwide service within the English National Health Service that provides treatment for common mental health conditions such as anxiety disorders. However, our approach is applicable to many other healthcare settings, including, for example, cancer treatment programmes.
Oparina, E., Clark, A. E., & Layard, R. (2026). The Easterlin Paradox at 50.
Revise & Resubmit at The Review of Income and Wealth
Media coverage: Financial Times, Vox, Alternatives Économiques (in French), LSE Inequalities, Economics Observatory
Abstract: We use Gallup World Poll data from over 150 countries from 2009-2019 at both the individual and country levels to revisit the relationship between income and subjective wellbeing. Our inspiration is the paradox first proposed by Easterlin (1974), according to which higher incomes are associated with greater happiness in cross-sections, yet increases in a country’s GDP per head do not increase its average wellbeing. In our analysis subjective wellbeing (or happiness) is measured by the Cantril ladder on a 0-10 scale. Across individuals, other things equal, one unit of log income raises subjective wellbeing by 0.4 points. In other words, doubling income raises wellbeing by 0.3 points out of 10. Across countries, a crude regression of log income on per capita income gives a higher coefficient of 0.6. But, once social variables like health and social support are introduced, the picture changes. In rich countries, income no longer has a significant effect, either in country cross-sections or in time series: higher income only matters due to its correlation with the social variables. For low-income countries the result is also clear cut – income raises happiness in both cross-section and time series, whether the social variables are controlled for or not. For middle income countries the result is mixed.
Kaiser, C., Oparina, E., & Oswald, A. (2026). Is it Psychologically Dangerous to Live in a Rich Area? Evidence from Individual-Level Data on UK Suicides. Draft available on request.
Revise & Resubmit at The Review of Income and Wealth
Abstract: This paper uses the first-ever release of individual-level data on UK suicides to understand how humans respond to economic advantage and disadvantage and to relative deprivation in their geographical area. Under strict confidentiality requirements, we link these data to the UK Census. We report a range of new findings. Of particular importance: this paper documents evidence consistent with the view that ceteris paribus it can be psychologically dangerous to live in a rich area. Feelings of relative deprivation may matter intensely---even in decisions about whether death is preferable to life.
Bhattacharya, D., Oparina, E, & Xu, Q. (2026). Empirical Welfare Analysis with Hedonic Budget Constraints.
CEP Discussion Paper, Cambridge Working Paper
Under Review
Abstract: We analyze demand settings where heterogeneous consumers maximize utility for product attributes subject to a nonlinear budget constraint. We develop nonparametric methods for welfare-analysis of interventions that change the constraint. Two new findings are Roy's identity for smooth, nonlinear budgets, which yields a Partial Differential Equation system, and a Slutsky-like symmetry condition for demand. Under scalar unobserved heterogeneity and single-crossing preferences, the coefficient functions in the PDEs are nonparametrically identified, and under symmetry, lead to path-independent, money-metric welfare. We illustrate our methods with welfare evaluation of a hypothetical change in relationship between property rent and neighborhood school-quality using British microdata.
Kirchmaier, T., & Oparina, E. (2026). Under Pressure: Victim Withdrawal and Police Officer Workload. CEP Discussion Paper 1985. [short video]
Under Review
Media coverage: LSE British Politics and Policy, CentrePiece, CEP Policy Brief
Abstract: Domestic abuse is one of the most common forms of violence against women. To deliver justice, the police must create the conditions for victims to engage safely. Yet many victims withdraw from legal action. We ask whether these withdrawals are related to police workload. Using detailed administrative data on over 15,000 high-risk domestic abuse cases from a major English police force with quasi-random case assignment, we exploit within-officer variation in monthly caseloads. Adding ten cases to an officer's workload---the average increase over the budget cuts period---raises the probability that the victim withdraws by three percentage points, implying around 3,300 additional withdrawals each year in England and Wales. The effect is concentrated in high-risk cases, where sustained officer-victim contact matters most, and is absent for lower-risk cases and for outcomes not dependent on the victim's cooperation. Balance tests, placebo outcomes, and pre-determined workload measures support a causal interpretation. Caseload is set by staffing and budgets, making it both a hidden cost of austerity and a policy lever.
Layard, R., & Oparina, E. (2026). What is Public's Social Welfare Function? CEP Discussion Paper 2190.
Revise & Resubmit at Economica
Abstract: Optimal public policy requires a social welfare function defined over individual utilities. While there is substantial research on income-based social welfare functions, no published study has directly elicited public preferences over utility when measured by subjective wellbeing. Using a novel survey instrument with a representative UK sample (N=2,068), we estimate the public's social welfare function for life satisfaction. We find significant aversion to wellbeing inequality, with a median isoelastic parameter of 0.48. This implies a social welfare function approximately equal to the sum of square roots of individual utilities. The median respondent values improving the wellbeing of the least satisfied by one unit roughly twice as much as improving the most satisfied by one unit. Our findings provide ethically grounded distributional weights for wellbeing policy evaluation and cost-benefit analysis.
Work in progress
-
The Effects of the Nation-Wide NHS and Care Volunteer Responders Programme (with C. Krekel, A. Boler and A. Smith)
-
Evaluating a Lay-Led Interpersonal Counselling Intervention for Depression in Brazilian Primary Care: A Randomised Controlled Trial (with D. Frayman, B. Baron, and C. Krekel)
-
Comparing Effectiveness of CBT and Counselling for Depression in Routine Practice: Evidence from Population-Wide NHS Talking Therapies Data (with D. Clark, C. Krekel, I. Parkes, and S. Srisuma)
Policy reports and book chapters
Living Long and Living Well: The WELLBY Approach (with R. Layard), in: Helliwell, J., Layard, R., Sachs, J., and J-E De Neve (eds.), World Happiness Report, 2021.
Exercises (with R. Layard, J-E. De Neve, and M. Kaats), in: Layard, R., & De Neve, J. (2023). Wellbeing: Science and Policy. Cambridge: Cambridge University Press.