EuHEA Seminar Series 2023/2024

The EuHEA seminar series brings together health economists all across its member associations, and beyond, to discuss cutting-edge research

Welcome

The EuHEA Seminar Series has been established as a key activity of EuHEA to foster exchange between health economists across countries and institutions and present cutting-edge research in all areas of health economics. A Scientific Committee chaired by Céu Mateus and Bruce Hollingsworth (Lancaster University), Pedro Pita Barros (Universidade Nova de Lisboa), and Karine Lamiraud (ESSEC Paris) will coordinate the series in the academic year 2023/2024. Seminars are held online and will take place on Tuesdays, 1:30-2:30pm (CET). For details on the program, see below.

The series in fall 2023 will run from October 3 until December 19.

The series in Spring 2024 will run from March 12 until June 25.

Sessions can be accessed by clicking here.

 

Upcoming event

June 25, 2024, 1:30-2:30pm (CET)

Depends on whom you ask: discrepancies in reporting spousal care between women and men across European welfare states

Objectives: To investigate systematic differences across Europe in reporting spousal care between caregivers and cared-for persons.

Methods: Using information on care provided/received from the Survey on Health, Ageing and Retirement in Europe (SHARE), we estimate prevalence rates of spousal care and discrepancies in answers of caregivers and cared-for persons. Multinomial regressions are used to estimate systematic differences in reporting spousal care. Multivariate logistic regressions are used to estimate whether differences in reporting spousal care have an effect on the correlation of spousal care with health.

Results: Only 53.9% of dyads report care that is confirmed by both spouses. Multinomial regressions show that agreement on care being provided/received is more common when women are caregivers, while men are likely to under-report when providing or receiving personal care. Prevalence rates of spousal care across care regimes is sensitive to who reports care, as are results of estimations of the correlation of spousal care with self-reported health and depression.

Discussion: Possible explanations include social norms dictating gender roles and acceptability of men as carers. Relying on self-identification of carers in surveys or to access social benefits may underestimate their numbers.

Speaker: Ricardo Rodrigues, Lisbon School of Economics and Management
Discussant: Sean Urwin, University of Manchester
Chair: Bruce Hollingsworth, Lancaster University

 

Program Spring 2024

12 March 2024, 1:30-2:30pm (CET)

Multimarket competition and pharmaceutical firms’ exit decision

Objectives: Multimarket contacts (MMC), which correspond to firms’ overlap across markets, may facilitate tacit collusion and thus reduce the intensity of competition. While there is extensive evidence that MMC decrease within-market competition, this study aims to evaluate whether multimarket contacts also influence market participation decisions, in the Portuguese off-patent drug market.

Methods: Resorting to a simple theoretical model of multimarket competition, it is shown that the potential of MMC to enhance collusion may depend on the ability of these contacts to restore an overall symmetry between companies which may fail to be observed in individual markets. As existing empirical MMCmeasures fail to account for this, this paper puts forward a theory-founded measure which weights contacts between companies by their importance in smoothing market share asymmetries. Based on the proposed theory-founded MMC measure, the paper tests whether pharmaceutical firms with greater MMC have greater survival chances in individual markets. The monthly dataset, ranging from January 2018 until December 2021, comprised on average 199 companies and 260 markets.

Results: Based on linear probability models with fixed effects and conditional logit models, it is shown that when the unit of analysis is the firm-in-market (i.e., a firm in a given market) higher MMC with competitors do not influence pharmaceutical firms’ survival chances in individual markets. Nonetheless, when the unit of analysis is a market, higher average MMC between companies active in a market are associated with a lower probability of observing firms’ exit from that market. Indeed, if MMC in a given market increase by one unit, then the probability of observing at least one firm exit the market over the course of one year decreases by 1.14 percentage points. These distinct results suggest that the degree of mutual forbearance is essentially a feature of the individual market under consideration, and not a characteristic of inter-firm rivalry independent of the level of competition faced by other competitors in the market.

Conclusions: While cost savings in the retail prescription drug market crucially depend on competition from generic drug producers, the results from this study suggest that these savings may be lower than expected due to the mutual forbearance incentive that MMC create. The finding that greater multimarket connections between companies soften competition and, consequently, decrease firms’ exit rates gives strength to the argument that drug prices (even off-patent drug prices) need to be regulated through price caps. Indeed, direct price regulation counters the incentive that multimarket competitors have to attenuate competition.

Speaker: Carolina Santos, Nova School of Business and Economics
Discussant: Georges Siotis, Universidad Carlos III Madrid
Chair: Céu Mateus, Lancaster University

19 March 2024, 1:30-2:30pm (CET)

Building Health across Generations: Unraveling the Impact of Early Childcare on Maternal Health

Motivation. In contemporary households, women often shoulder most organisation and caregiving responsibilities leading them to play a crucial role in family dynamics. While previous research has established that public early childcare affects child outcomes and maternal employment, less attention has been given to its effects on maternal health despite its relevance within the household.

Objective. This study investigates the impact of public early childcare on maternal short- and long-term health.

Methods. Based on administrative health records covering 90% of the German population over a decade, we leverage the exogenous variation in childcare coverage rates across counties and time induced by a major German early childcare expansion in a generalised difference-in-differences setting.

Results. Our findings provide compelling evidence supporting the idea of an intra-household transmission of communicable diseases: mothers experience 4–8% more infections and 2–4% more respiratory diseases for a 10 percentage point rise in childcare coverage rates when their children are 1–2 years old. In contrast, mothers benefit from reductions in obesity and anaemia, and heterogeneity analyses show a lower prevalence of mood- and stress-related disorders for multiparous and older mothers. Our results are robust to extensive robustness checks, including the validation of the plausibility of the parallel trends and exogeneity assumptions, adjustments for multiple hypotheses testing, and sensitivity analyses for children born before the observation period.

Discussion. The policy implications of our findings extend beyond the health impacts of early childcare on mothers and shed light on the broader dynamics within families. The rise in communicable diseases and healthcare consumption, particularly during the first two years of childcare attendance, presents a notable challenge to maternal productivity. Nevertheless, our results also reveal positive effects on maternal health, specifically reductions in obesity and anaemia, along with enhanced mental health for specific subgroups. Recognising the pivotal role of maternal health in shaping the family environment, these shifts could reverberate throughout the household.

Speaker: Laia Bosque-Mercader, Universitat de Barcelona
Discussant: Francisca Vargas Lopes, OECD and Erasmus School of Health Policy & Management
Chair: Karine Lamiraud, ESSEC Business School

23 April 2024, 1:30-2:30pm (CET)

Examining pay gaps in the female-dominated (long-term) care sector: Evidence from Portugal

Objectives: This study delves into the devaluation theory, which posits that sectors and occupations predominantly occupied by women often experience subpar working conditions. Its primary objective is to elucidate the determinants of earnings and wage gaps prevalent within the care and long-term care (LTC) sector.

Methods: Employing a random intercept multilevel model using a uniquely comprehensive matched employer-employee dataset (Quadros de Pessoal) from 2021 in Portugal, the initial step of our study involves a comparative analysis between female-dominated (caring) sectors and male-dominated (non-caring) sectors. Moving beyond, an exploration is conducted to ascertain if existing theories regarding the devaluation of female-typical work further elucidate earnings disparities between female-dominated caring sectors (LTC, childcare, health care, and social work).

Results: Preliminary findings reveal persisting disparities, demonstrating that despite the consideration of individual and firm-specific factors, including occupation, female-dominated sectors such as LTC and other caring sectors experience a significant earnings penalty compared to their male-dominated counterparts. However, it is particularly the occupational segregation between caring and non-caring occupations that leads to a sizeable earnings penalty. Moreover, although a gender pay gap prevails in the highly female-dominated (long-term) care sector, it is shown to be significantly smaller, suggesting that male workers in caring sectors and occupations experience a relative pay disadvantage.

Interestingly, the study doesn’t present compelling evidence that pay differences within caring sectors result from the devaluation of female-typical work. Contrarily, despite the highest proportion of women in the labor force, our findings reveal an earnings premium among LTC  workers. Notably, this premium is more pronounced for personal carers, reaching 10% compared to similar occupations in other caring sectors like child care, social work, or health care.

Discussion: Although results confirm that low pay - compared to (non-caring) male-dominated sectors - stems from a devaluation of female-typical work, our findings also stress that there are intricate subtleties beyond gender and job characteristics that influence pay differences, particularly between care sectors. Possible other mechanisms such as compensating differentials suggest that higher earnings in LTC (compared to other care sectors) may be due to the particularly emotional nature of elderly personal care. If working conditions are particularly taxing, compensatory (financial) rewards might explain the higher remuneration in this sector.

Implications of our findings are profound, demanding a more nuanced understanding of working conditions in and differences between caring sectors.

Speaker: Viktoria Szenkurök, Vienna University of Economics and Business
Discussant: Melisa Sayli, University of Surrey
Chair: Bruce Hollingsworth, Lancaster University

30 April 2024, 1:30-2:30pm (CET)

Managerial Practices and Altruism in Health Care Delivery

We report results from a field experiment of a large and comprehensive management consulting intervention designed to improve business management and medical care delivery in the Kenyan private health sector. The intervention significantly improved management practices across management key domains including (i) operations standardization, (ii) performance monitoring, (iii) goal setting, (iv) employee incentives, and (v) financial management. We also find the intervention shifted provider preferences towards profits at the expense of altruism for patients. To elicit preferences, we use a real-stakes, modified lab-in-the-field dictator game. We also find a reduction in quality of care and increase in price consistent with quality-inelastic demand. We measure process quality using standardized patient (SP) audits; SPs are healthy individuals trained to anonymously portray standardized health scenarios during visits to health clinics. In addition, clinics increased profits by attracting more patients, charging them higher prices and providing care more efficiently. We further examine mechanisms using causal mediation analysis. We find that the effects on quality and price of services are mediated by a shift in provider preferences, while the increase in profits and reduction in unit costs are mediated by changes in management practices.

Speaker: David Contreras Loya, Tec de Monterrey
Discussant: Daniel Wiesen, University of Cologne
Chair: Bruce Hollingsworth, Lancaster University

14 May 2024, 1:30-2:30pm (CET)

Physician gender and patient health care

Objectives: In this project, we examine whether the gender of primary care providers (PCPs) affects patients’ health care utilization. We focus on PCPs as they are the first point of contact for patients in case of an illness and as they provide regular care for chronic ill patients.

Methods: We exploit a unique quasi-random procedure for filling vacant outpatient physician positions in Upper Austria to causally analyze the effect of a general practitioner’s (GP’s) gender on individual health care utilization. In this setting, patients have no influence on the substitution process. Furthermore, we show that, conditional on other observed physician characteristics, female substitutes do not self-select into certain positions. In the empirical analysis, we compare patients who have been with their male PCP for at least two years before the exit and who are now exogenously treated by either a female or a male PCP in a difference-in-differences (DID) design.

Results: Patients who are assigned a female PCP are 5.5 percentage points more likely to switch to a different PCP compared to patients who are randomly assigned a male PCP. This has implications for the PCP practices. Female physicians see significantly fewer patients and make 14 percent lower revenue compared to male physicians in a similar position. This effect seems to be driven by male physicians increasing revenue compared to their predecessor, while female physicians’ revenue remains stable. We interpret this as male physicians operating in a more revenue-maximizing manner than their female colleagues. We then move on to study the effects of being treated by a female physician on patient health care utilization. We document a significant difference in GP fees of 6.4 percent between patients with female and male successors, while  finding no effect of female physicians on other health care outcomes, such as physician fees, drug prescriptions, or hospital expenditures. This effect seems to be driven by a large significant decrease of 25 percent in health screenings by female successors compared to male successors. This gap is particularly driven by an increased propensity of male successors to carry out health screenings. As the effectiveness of health screenings is widely debated, we do not expect this to result in any long-term benefits for patients.

Discussion: Our results overall indicate moderate differences between female and male PCPs. The effects on spending for inpatient care – the most important cost component in quantitative terms – remain consistently insignificant. Thus, we do not expect a larger share of female physicians to increase health care  costs, on the contrary, this societal development will likely lead to slightly decreasing health care costs per patient.

Speaker: Katrin Zocher, Johannes Kepler University
Discussant: Torben Heien Nielsen, University of Copenhagen
Chair: Ceu Mateus, Lancaster University

21 May 2024, 1:30-2:30pm (CET)

The Effect of Centralised Procurement on Pharmaceutical Marketing and Innovation: Evidence from China

We assess how centralized procurement affects pharmaceutical marketing and innovation by evaluating the impact of a centralised procurement policy in China through which the Chinese government centralized price negotiation and purchase of pharmaceutical products. Using differences-in-differences methodology, we exploit exogenous variation of the policy across time and firms' drug portfolios to identify its impact on firm-level marketing and Research and Development (R&D) spending. We source drug portfolio data from IQVIA and expenditure data from Compustat. The findings reveal that firms exposed to the drug procurement list increase quarterly marketing and Research and Development spending by USD 409 thousand and 486 thousand (or by 11.2% and  17.6%), respectively. Further analysis suggests that the magnitude of such impacts increases with market (demand) size and competition at the drug registration (and beyond) stage. In contrast, competition at the clinical trial stage increases the magnitude of R&D spending while lowering marketing spending. R&D spending reduces drastically when firms face no competition both at the clinical and registration stages. Our results contradict the recent claim by pharmaceutical companies that enabling centralised procurement in the U.S. will hamper innovation.

Speaker: Chuanzi Yue, Imperial College
Discussant: Nahim Bin Zahur, Queen's University
Chair: Bruce Hollingsworth, Lancaster University

28 May 2024, 1:30-2:30pm (CET)

Taboo trade-off aversion in choice behaviours: a discrete choice model and application to health-related decisions

Objectives: When taboo trade-offs are involved in health policymaking, measuring preferences using discrete choice models may lead to inaccurate behavioural insights. Empirical evidence shows that individuals dislike making morally problematic or taboo trade-offs, such as human lives versus money. We introduce a recently proposed discrete choice model designed to capture Taboo Trade-off Aversion (TTOA) in the healthcare domain and empirically test the model in a typical healthcare choice context.

Methods: To capture taboo trade-off aversion, the conventional linear-additive logit model is extended by including penalties for taboo trade-offs. Using two Discrete Choice Experiments (DCEs) focussing on taboo trade-offs embodied in public health policies, we empirically compare conventional linear-additive logit models and TTOA models to explore differences in model results and Willingness-To-Pay (WTP) analyses.

Results: We find evidence for the presence of TTOA underlying choice behaviour of respondents. Discrete choice models accounting for unobserved preference heterogeneity, like latent class logit models, reveal that the magnitude of TTOA varies considerably across population groups. We also show that WTP estimates derived from TTOA models can provide richer insights, such as the compensation level individuals are willing to accept for being confronted with taboo trade-offs in policymaking.

Conclusions: Accounting for TTOA behaviour in discrete choice models can lead to more accurate behavioural information when choice behaviours are affected by taboo trade-offs. Researchers and policymakers can use the TTOA models to find mutually acceptable grounds for designing effective and morally acceptable health policies when confronted with taboo trade-offs in policymaking.

Speaker: Nicholas Smeele, Erasmus University Rotterdam
Discussant: Luís Filipe, Lancester University
Chair: Mariana Videira, Nova School of Business and Econmics

4 June 2024, 1:30-2:30pm (CET)

Opioid Epidemic: The Role of Friends

Objectives: This paper explores the role of friends in the opioid epidemic in the US.

Methods: Using data from the National Longitudinal Survey of Adolescent Health (Add Health), we focus on adults aged 25-34 and their high school best friends and adopt an instrumental variable technique to estimate peer effects in opioid misuse. We use severe injuries in the previous year as an instrument for opioid misuse and estimate the causal impact of someone misusing opioids on the probability that their best friends also misuse.

Results: The estimated peer effects are significant: Having a best friend with a reported serious injury in the previous year increases the probability of own opioid misuse by 7 percentage points in a population where 17 percent ever misuses opioids.

Discussion: The effect is driven by individuals without a college degree.

Speaker: Effrosyni Adamopoulou, ZEW and University of Mannheim
Discussant: Bill Dowd, Lancaster University
Chair: Ceu Mateus, Lancaster University

11 June 2024, 1:30-2:30pm (CET)

The gendered impact of COVID-19 on health behaviours and mental health: Evidence from the UK

Objectives: The purpose of this paper is to assess whether the COVID-19 pandemic affected the health behaviours of men and women differently, and to estimate whether the  associations between health behaviours and mental health differed by gender.

Methods: We employ nationally representative panel data from the UK Household Longitudinal Study (UKHLS). To study the differential effect of the COVID-19 pandemic on health behaviours by gender, we estimate a difference-in-differences (DID) model by regressing each outcome on gender, a pandemic indicator, and an interaction of these two variables as well as control for ethnicity, age, educational attainment, region, and month fixed effects. We also estimate DID models with an interaction of gender and time to account for gender-specific trends, an interaction of gender and age to account for changes in behaviour as respondents age, as well as individual fixed effects. The second part of our analysis investigates whether the pandemic changed the protective effect of health behaviours on mental health, differentially for women and men. We prioritise an individual fixed effects estimation strategy for this analysis due to the importance of potential confounders such as childhood circumstances, family background, and personality traits that do not change over time. Specifically, we regress mental health on an interaction of health behaviours and a andemic indicator, for women and men separately, and control for ethnicity, age, educational attainment, region, and month fixed effects. The estimation strategy we use also alleviates concerns on inaccuracies in reporting behaviour for health behaviours and mental health. As we are comparing changes across gender and/or changes before and during the pandemic, differences in reporting behaviour has limited effects, as long as these differences are constant across time.

Results: Compared to men, women were 3.2 percentage points less likely to adopt a healthy lifestyle and report 0.09 fewer healthy behaviours (corresponding to 7.0 per cent of a standard deviation) during the pandemic. Importantly, we find that the pandemic considerably weakened the protective effect of health behaviours on mental health for women but not for men. For women, adopting a healthy lifestyle was strongly correlated with mental health before the pandemic, but this relationship was no longer significant during the pandemic. This loss in significance corresponds to a 0.61-points decline in the GHQ-12 Likert score, equivalent to 10.5 per cent of a standard deviation. While the protective effect of health behaviours on mental health is stronger for women before the pandemic, it is stronger for men after the pandemic. Thus, the pandemic adversely impacted women not only in terms of changes in health behaviours but also in terms of lower psychological benefits derived from healthy behaviours.

Speaker: Karen Arulsamy, Duke-NUS Medical School
Discussant: Hans-Helmut König, University Medical Center Hamburg-Eppendorf
Chair: Pedro Pita Barros, Nova School of Business & Economics

18 June 2024, 1:30-2:30pm (CET)

Drug tenancy thresholds, consumption and newborn's health

In this paper, I study how a policy aimed at not criminalizing illicit drug consumers, to recognize their health care needs, might negatively influence newborn's health. In 2013, a policy that established maximum permissible amounts of  possession of illicit substances for personal consumption was implemented in Ecuador. This regulatory change might be one of the drivers of increased availability of drugs, which in turn, may increase consumption. I exploit variation in drug exposure at the province level in a difference-in-differences framework, which compares Ecuadorian provinces with different intensities of drug exposure before and after 2013. I find that the share of low birth weight and very low birth weight newborns raise significantly by 16.26% and 82.22% in provinces initially more exposed to drugs. I also document a significant increase in both women and men's substance use. I conclude that the increase in drug use is possibly one of the channels affecting newborn's health living in more exposed provinces.

Speaker: Grace Armijos Bravo, Universidad de Especialidades Espíritu Santo & Instituto de Economía de Barcelona
Discussant: Dolores Jiménez Rubio , Universidad de Granada
Chair: Ceu Mateus, Lancaster University

25 June 2024, 1:30-2:30pm (CET)

Depends on whom you ask: discrepancies in reporting spousal care between women and men across European welfare states

Objectives: To investigate systematic differences across Europe in reporting spousal care between caregivers and cared-for persons.

Methods: Using information on care provided/received from the Survey on Health, Ageing and Retirement in Europe (SHARE), we estimate prevalence rates of spousal care and discrepancies in answers of caregivers and cared-for persons. Multinomial regressions are used to estimate systematic differences in reporting spousal care. Multivariate logistic regressions are used to estimate whether differences in reporting spousal care have an effect on the correlation of spousal care with health.

Results: Only 53.9% of dyads report care that is confirmed by both spouses. Multinomial regressions show that agreement on care being provided/received is more common when women are caregivers, while men are likely to under-report when providing or receiving personal care. Prevalence rates of spousal care across care regimes is sensitive to who reports care, as are results of estimations of the correlation of spousal care with self-reported health and depression.

Discussion: Possible explanations include social norms dictating gender roles and acceptability of men as carers. Relying on self-identification of carers in surveys or to access social benefits may underestimate their numbers.

Speaker: Ricardo Rodrigues, Lisbon School of Economics and Management
Discussant: Sean Urwin, University of Manchester
Chair: Bruce Hollingsworth, Lancaster University

July 9 2024, 1:30-2:30pm (CET)

Behavioural Responses to Disability Insurance Generosity: Evidence from administrative records on CVD shocks

We exploit an institutional discontinuity resulting in exogenous variation of replacement rates to study behavioural responses to Disability Insurance (DI) generosity, in a novel setting of work compatibility. We use administrative data on the work and health histories of Italian private sector workers. To address
unobserved heterogeneity, we focus on individuals hit by major CVD shocks. Based on a Regression Discontinuity strategy, we find a sizeable DI response to benefit generosity, implying an elasticity of DI participation of 1.26, and a smaller employment reduction response, with an estimated elasticity of 0.15.
These findings suggest that in a setting of work compatibility DI receipt is broadly seen as a complement, rather than a substitute, to labour income, bearing relevant implications for the design of labour-inclusive DI schemes.

Speaker: Irene Simonetti, University of Amsterdam
Discussant: Silvia Balia, Università degli Studi di Cagliari
Chair: Ceu Mateus, Lancaster University

 

Program Fall 2023

3 October 2023, 1:30-2:30pm (CET)

Triage at Shift Changes and Distortions in the Perception and Treatment of Emergency Patients

In this study, we employ more than one million Emergency Department (ED) records and combine machine learning and regression discontinuity techniques to quantify previously undocumented distortions in triage nurses’ assessment of patients’ conditions and to investigate the short- and medium-run consequences for patients. We show that triage nurses tend to assign lower priority at the beginning of their shifts, and then become progressively more lenient. As a consequence, identical patients arriving at the ED just after a shift change are assigned a substantially lower priority than patients arriving just before. We employ this setting in a regression discontinuity design to show that identical patients assigned lower priority (because of this distortion) are then treated as if they were in less severe conditions and are more likely to be discharged without further care and are more likely to demand further emergency care over the following months. We interpret this as evidence that distortions in nurses’ initial assessment of patients’ urgency bias physicians’ perceptions of the conditions of the patients with long-run implications for their health and for the healthcare service.

Speaker: Simone Ferro, Università degli Studi di Milano Statale
Discussant: Igor Francetic, University of Manchester
Chair: Pedro Pita Barros, Nova School of Business of Economics

10 October 2023, 1:30-2:30pm (CET)

Violence against women: The witness dilemma

In Spain, the reform of the Violence Against Women (VAW) law (Royal Decree-Law 9/2018) made it possible to report gender violence without going to the police or court. The advantage of this kind of “soft reporting” is that victims or witnesses can report to social services and thereby gain access to legal, economic and psychological support without having to identify or denounce the aggressor. It is expected that the availability of soft reporting will help to reduce costs and overcome barriers to reporting. The goal of this paper is to evaluate the impact of soft reporting on witness involvement (the decision to report). We also analyze witness preferences and the sensitivity of potential witnesses to different costs associated with soft and hard reporting. This aspect of witness reporting has not been evaluated in Spain, and it has potentially wide relevance for other countries in Europe who are interested in adopting similar kinds of legislation. To perform this evaluation, we follow an experimental approach, embedded in an official survey conducted by the Institute of Statistics of the Government of Navarra (Spain). We find that potential witnesses are more willing to intervene when soft reporting is available. In addition, we find that the individual propensity to intervene and help the victim decreases when costs increase. We identify the switching point at which individual propensity to intervene starts to decrease, and find that it is the same for both soft and hard reporting scenarios. Finally, we find that altruistic individuals and those who belong to feminist associations are more inclined to intervene. We discuss the implications of these findings for policy makers.

Speaker: Ariadna Garcia Prado, Universidad Pública de Navarra
Discussant: Judit Vall Castelo, UB School of Economics
Chair: Céu Mateus, Lancaster University

17 October 2023, 1:30-2:30pm (CET)

Into the Far West? Investigating Health Policy-Makers' Willingness to Adopt Decrementally Cost-Effective Innovations Using a DCE Approach

Objectives
In a context of increasingly limited resources, a number of strategies, such as the adoption of decrementally cost-effective interventions (d-CEIs), which are both less clinically effective and less costly, could offer potential levers at enhancing both efficiency and equity in healthcare systems. These interventions are located in the South-West quadrant of the cost-effectiveness plane, and have yet received little attention as they are often perceived as per se “unethical” or “unacceptable”, hence the reference sometimes made to the S-W quadrant as the 'Far West'.

The purpose of our paper is to understand preferences and determinants to d-CEIs' adoption by health policy-makers. To the best of our knowledge, no previous study has inquired into health policy-makers' preferences regarding d-CEIs. Our study therefore fills a significant gap in the literature by studying health policy-makers' preferences regarding d-CEIs in a number of countries in Europe in order to identify the conditions that would allow their consideration as comparators in HTAs and their possible adoption in replacement to usual care.

Method:

We use a two-stage pairwise discrete choice experiment (DCE) survey to elicit (i) preferences for d-CEIs' attributes in forced choices and (ii) adoption preferences, i.e. the determinants of d-CEIs' adoption (unforced choices). We investigate the effect (and trade-offs) between three attributes: health loss (very small to significant), reversibility defined as the possibility to switch back to usual care (from possible to hardly possible) and cost-savings (from 5% to 15% of a fixed budget). Such trade-offs are contextualized by using two sensitivity attributes: disease severity (low and moderate) and savings uncertainty (low and high). Our final sample consists of 180 respondents with 46.7% originating from France and the remaining respondents from other EU countries.

Results:

All attributes’ levels have a significant effect in the two decision stages. The ”health loss” attribute dominates in the first stage followed by ”reversibility”: we calculate that decision-makers would require 28.3% increase of budget savings to be indifferent between a scenario of small
versus significant health losses and 14.5% to be indifferent between a scenario of possible and hardly possible reversibility. In contrast, the ”reversibility” attribute dominates in the second stage suggesting that anticipated regret may play a role in adoption decisions.

A stratified analysis of the compensation requirements according to participants’ responses of the follow-up
questions in the second stage (i.e., ”would you be ready to substitute usual care by the option you selected ” and ”would you be ready to substitute usual care by the option selected if disease severity of the patient -before treatment- changes to
moderate”), reveals the existence of preference heterogeneity across participants.

Speaker: Iván Tzintzun Valladolid, Paris School of Economics
Discussant: Petra Baji, University of Bristol
Chair: Bruce Hollingsworth, Lancaster University

24 October 2022, 1:30-2:30pm (CET)

A Healthy Start for Healthy Start Vouchers?

This paper examines the impact of a recent increase in the value of healthy start vouchers (HSV) on the purchase of healthy items, using a large and representative sample of 13 million shopping basket transactions from a major UK food retailer. We use a difference-in-differences (DiD) approach to show that a £1.15 increase in voucher values increased spending on fruits and vegetables (F&V) for single voucher users by 32p and for two voucher users by 77p. For all eligible items, the increase in the value of the vouchers increased spending by 31p for single voucher users and by 89p for two voucher users. Our analysis of the marginal propensity to consume (MPC) indicates that consumers treat vouchers similarly to cash transfers of the same value. We also find that the effects of the HSV program are greater in more deprived areas and areas with lower pre-change expenditure on eligible items, indicating potential benefits for reducing health inequalities. Our results have implications for the effectiveness of targeted benefit programs, such as the UK HSV program, in promoting healthy food choices.

Speaker: Hamideh Mohtashami Borzadaran, University of Birmingham
Discussant: Mario Martinez, Imperial College London
Chair: Pedro Pita Barros, Nova School of Business and Economics

31 October 2023, 1:30-2:30pm (CET)

Impact of Forced Sterilization on Female Labor Market Outcomes: Evidence from India

Objective: I investigate the impact of the 1975-76 forced sterilization campaign carried out by the Indira Gandhi government in India on women’s long-run labor market outcomes.

Methods: I exploit heterogeneity in the implementation of coercive sterilization at the district level using difference-in-differences (DiD) and run additional robustness and heterogeneity checks.

Results: Using large data samples from India and accounting for endogeneity concerns, I find that exposure to the forced sterilization campaign at the district level reduces long-term labor market participation by 4.5% and 1.5% in agricultural and sales occupations and increases unemployment by 4.7% and I elucidate mechanisms.

Discussion: The proposed mechanism of this is the disutility derived from having a working wife. This result is contrary to existing literature that indicates that women’s
access to contraception increases their labor market participation. My results suggest that giving women access to contraception is insufficient to improve their market outcomes.

Speaker: Niranjana Prasad, Université Catholique de Louvain
Discussant: Saurabh Singhal, Lancaster University
Chair: Bruce Hollingsworth, Lancaster University

7 November 2022, 1:30-2:30pm (CET)

Multimarket competition and pharmaceutical firms’ exit decision

Objectives: Multimarket contacts (MMC), which correspond to firms’ overlap across markets, may facilitate tacit collusion and thus reduce the intensity of competition. While there is extensive evidence that MMC decrease within-market competition, this study aims to evaluate whether multimarket contacts also influence market participation decisions, in the Portuguese off-patent drug market.

Methods: Resorting to a simple theoretical model of multimarket competition, it is shown that the potential of MMC to enhance collusion may depend on the ability of these contacts to restore an overall symmetry between companies which may fail to be observed in individual markets. As existing empirical MMCmeasures fail to account for this, this paper puts forward a theory-founded measure which weights contacts between companies by their importance in smoothing market share asymmetries. Based on the proposed theory-founded MMC measure, the paper tests whether pharmaceutical firms with greater MMC have greater survival chances in individual markets. The monthly dataset, ranging from January 2018 until December 2021, comprised on average 199 companies and 260 markets.

Results: Based on linear probability models with fixed effects and conditional logit models, it is shown that when the unit of analysis is the firm-in-market (i.e., a firm in a given market) higher MMC with competitors do not influence pharmaceutical firms’ survival chances in individual markets. Nonetheless, when the unit of analysis is a market, higher average MMC between companies active in a market are associated with a lower probability of observing firms’ exit from that market. Indeed, if MMC in a given market increase by one unit, then the probability of observing at least one firm exit the market over the course of one year decreases by 1.14 percentage points. These distinct results suggest that the degree of mutual forbearance is essentially a feature of the individual market under consideration, and not a characteristic of inter-firm rivalry independent of the level of competition faced by other competitors in the market.

Conclusions: While cost savings in the retail prescription drug market crucially depend on competition from generic drug producers, the results from this study suggest that these savings may be lower than expected due to the mutual forbearance incentive that MMC create. The finding that greater multimarket connections between companies soften competition and, consequently, decrease firms’ exit rates gives strength to the argument that drug prices (even off-patent drug prices) need to be regulated through price caps. Indeed, direct price regulation counters the incentive that multimarket competitors have to attenuate competition.

Speaker: Carolina Santos, Nova School of Business and Economics
Discussant: George Siotis, Universidad Carlos III de Madrid
Chair: Céu Mateus, Lancaster University

14 November 2023, 1:30-2:30pm (CET)

The Electoral Returns to Investing in Healthcare

In this paper, we examine whether voters respond to investments in healthcare and access to health services. We focus on the city of Rio de Janeiro, the Brazilian capital with the highest growth in primary healthcare coverage between 2009 and 2012. The newly elected government in 2008 rapidly expanded the Family Health Program (FHP) in the city, a community healthcare program introduced in Brazil at the municipal level and which is currently the largest in the world. We assess whether the FHP expansion affected the mayor's vote share in 2012 by exploiting extremely fine-grained geocoded variation in access to healthcare and in voting across time and space within the city. The identifying variation comes from changes in the share of voters at the polling booth level residing and enrolled in a FHP catchment area, within polling places fixed effects, triggered by idiosyncratic expansion of catchment areas over the 2008-2012 period. We find that FHP coverage is positively associated with the mayor's vote share. Yet, the magnitude of the FHP effect is higher for the last year of government and increases with the proximity between FHP facilities and individuals' home addresses. Our results also suggest that this impact is accompanied by relatively larger increases in vote share when the services are provided in new health facilities compared to the existing units. The evidence suggests that closer contact with services and the visible side of the service provision may be greatly rewarded by voters. Overall, this paper contributes to a better understanding of voter responsiveness of healthcare policies and its mechanisms.

Speaker: Marcela dos Santos, University of Pittsburgh
Discussant: Mariana Lopes da Fonseca, University of St. Gallen
Chair: Pedro Pita Barros, NOVA School of Business and Economics

21 November 2023, 1:30-2:30pm (CET)

Are the grandparents alright? The health consequences of grandparental childcare provision

This paper examines the causal effect of childcare provision on grandparents’ health in the US. We propose the sex ratio among older adults’ children as a novel instrument for grandparental childcare provision. Our instrument is rooted in the demographic literature on grandparenthood and exploits that parents of daughters transition to grandparenthood earlier and invest more in their grandchildren than parents of sons. We estimate 2SLS regressions using data from the Health and Retirement Study. The results suggest that childcare provision is not beneficial for grandparents’ health and may even be detrimental for physical functioning and subjective health.

Speaker: Emma Zai, Max Planck Institute for Demographic Research
Discussant: Judith Bom, Erasmus School of Health Policy and Management
Chair: Bruce Hollingsworth, Lancaster University

28 November 2023, 1:30-2:30pm (CET)

A Systematic Review of Equity Issues Experienced by Older Adults with Rare Diseases

Objectives:
To summarize the different forms of equity issues that rare disease populations face in comparison to the general population. We are also particularly interested in examining the specific equity issues that older subpopulation of the general rare disease population faces in comparison to the general rare disease population.

Methods:
We conducted a systematic literature review according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and apply a search strategy to MEDLINE via PubMed the Cochrane Library, and Embase based on search terms related to equity and rare diseases. We also considered gray literature of regulatory bodies, patient organizations, industry, and others.

Results:
We retrieved and extracted evidence from in 63 publications from which two evidence clusters and five evidence subclusters emerged, thus constituting our evidence framework. Evidence cluster 1 contained evidence retrieved from publications discussing ethical concepts (subcluster 1: n = 12) and societal preferences (subcluster 2: n= 10). Evidence cluster 2 covered regulations (subcluster 3: n= 33), access to care (subcluster 4: n=3) and health outcomes (subcluster 5: n=5).

Discussion:
Evidence that compared the older rare disease population to the general rare disease population indicated that older adults, as a subgroup of the general rare disease population could potentially encounter more equity issues. This is not surprising because it relates to a general moral and ethical debate on how specific subgroups of patients ought to be treated in societies – i.e., whether resources should be allocated to maximize the welfare of society as a whole; utilitarian , or whether everyone has an equal right to health regardless of health system efficiency; egalitarian
Reetrieved evidence shows that each society is located somewhere along the egalitarian utilitarian continuum and there is hardly any consensus in OECD countries regarding the optimal balance between the two ends of this spectrum.

However, retrieved evidence is insuffcient to enable us to conclude that older rare disease patients face more equity issues in comparison to the general rare disease population. This said, we inductively develop an evidence framework that is reflective of modern health systems and that may help identify specific research gaps. In the long run, this should lead to better understanding of the root causes of inequities, and result in improved access to care and better health outcomes for older adults with rare diseases as well as for other subgroups of the general rare disease population.

Speaker: Jean Pierre Uwitonze, University of Bern
Discussant: Joachim Marti, University of Lausanne
Chair: Céu Mateus, Lancaster University

5 December 2023, 1:30-2:30pm (CET)

Provider and patient delay: Income, utilisation, and multimorbidity predict health care accessibility in Austria, the UK and the USA during the COVID-19 pandemic

Objectives
Access to health care is known to be significantly influenced by the so-called ‘social gradient in health’. This gradient skews access to health care to disfavour socially and economically disadvantaged and deprived groups of the population. In line with this many scholars have highlighted that the COVID-19 pandemic disproportionally affects marginalised and vulnerable groups when it comes to health care. In this article socio-economic-determinants of health are analysed in their implications on health care access and health (in)equity.

Methods
This article uses different survey datasets from Austria, the USA, and the UK to illuminate the interacting effects of health, particularly of multimorbidities, and income on access to health care amidst the COVID-19 pandemic. Employed are various linear and non-linear statistical regression models to investigate the relationships between the variables, including controls, health-related behaviours and robustness checks.

Results
The main findings corroborate the expected health-income gradient effects on health care access with regards to hospitals, general practitioners, specialists, and home care. Unmet needs that are distributed asymmetrically across populations are identified in all three countries despite their different health system designs.

Discussion
Income matters because it not only mitigates having a medical condition but improves health care accessibility for (relatively) healthy persons too. Medical conditions and multimorbidity result in more barriers to health care access as well. In interaction income and health significantly reinforce inequitable access to different providers of health care. These findings have different implications for health policy.

Speaker: Thomas Resch, Medical University of Vienna
Discussant: Carolina Santos, Nova School of Business and Economics
Chair: Bruce Hollingsworth, Lancaster University

12 December 2023, 1:30-2:30pm (CET)

Can interventions help reduce diabetes incidence? Evidence from NHS DPP participation

Diabetes, a serious chronic condition, is a public health concern alongside leading to substantial economic losses. However, type 2 diabetes, is preventable through health and lifestyle changes. We study the effectiveness of behaviour change interventions on health outcomes while considering the case for the Diabetes Prevention Programme implemented in England. Using referral data from at risk (pre-diabetic) population in England, we examine factors that influence participation and whether greater programme attendance can prevent or delay type 2 diabetes diagnosis. We find that participation appears to come from less deprived communities and older age group cohorts. Our results also suggest that provider led delays can plausibly be related with an individual’s lack of incentive to participate in the programme. Since participation is non-random, we uncover causal effects using instrumental variables. We identify the effect of an individual’s level of attendance through programme design components that provide exogenous variation to individual’s participation decision. Our causal estimates indicate that attending an additional session of the programme leads to an average of 2-3% reduction in the likelihood of developing T2D, and this appears to increase with higher levels of attendance. These results appear to indicate some benefit of engaging with the programme and perhaps it may be important to consider improvements in programme design elements that can boost participation.

Speaker: Vasudha Wattal, Manchester University
Discussant: Rita Faria, University of York
Chair: Pedro Pita Barros, NOVA School of Business and Economics

19 December 2023, 1:30-2:30pm (CET)

The role of physician altruism in the physician-industry relationship: Evidence from combining experimental and observational data

Objectives: Altruism is a key professional norm that underlies the physician's role as a representative agent for patients. However, physician behavior can be influenced when private gains enter the objective function. We study the relationship between altruism and physicians' receipt of financial benefits from pharmaceutical manufacturers, as well as the extent to which altruism mitigates physicians' responsiveness to these industry payments.

Methods: We combine data on altruistic preferences for 280 physicians, identified using a revealed preference economic experiment, with information on their receipt of monetary and in-kind transfers from pharmaceutical firms along with drug prescription claims data. We first develop a stylized model of altruism in brand prescription choices and physicians' decisions to engage with drug firms. We then empirically investigate how physician altruism relates to transfers from the pharmaceutical industry to physicians, and whether altruistic preferences are associated with the relation between industry transfers and drug prescribing.

Results: Our findings reveal that physicians with less altruistic preferences obtain industry transfers that are, on average, 112.55% higher in monetary value compared to physicians with stronger altruistic preferences. This difference in payments amounts to 968.37 USD, indicating clear selection of whom pharmaceutical firms target. Furthermore, we observe that positive correlations between industry transfers and higher overall drug costs or brand prescribing rates are predominantly driven by physicians with less altruistic preferences. Our estimates suggest that, when comparing less altruistic physicians with more altruistic physicians, a 1% increase in payments is associated with an additional average increase of 0.005 percentage points (0.027%) in the share of brand drugs prescribed. We find limited evidence that patient vulnerability moderates industry influences among less altruistic physicians.

Discussion: Our results reveal that altruistic preferences are an essential determinant of the strength of physician-industry ties. We therefore provide first empirical evidence for the role of professional norms in the practice of medicine under potential conflicts of interest.

Speaker: Shan Huang, University of Zurich
Discussant: Odd Rune Straume, University of Minho
Chair: Céu Mateus, Lancaster University

 

Scientific Committee

The EuHEA Seminar Series in the academic year 2022/2023 is coordinated by:

  • Karine Lamiraud, ESSEC Business School (Chair)
  • Rossella Verzulli, University of Bologna (Co-Chair)
  • Francesca Barigozzi, University of Bologna
  • Stefan Boes, University of Lucerne
  • Fabrice Etile, Paris School of Economics
  • Geir Godager, University of Oslo
  • Mathias Kifmann, University of Hamburg
  • Shiko Maruyama, Jinan University
  • Nuria Mas, IESE Business School
  • Helen Mason, Glasgow Caledonian University
  • Paolo Pertile, University of Verona
  • Pedro Pita Barros, Universidade Nova de Lisboa
  • Katrin Zocher, University of Linz