Hans Henrik Sievertsen


Teaching & tools

Data visualisations

Public dissemination

Senior Lecturer, University of Bristol
Researcher, VIVE
Research Fellow, IZA
My CV is available here.
Contact: h.h.sievertsen[at]bristol.ac.uk

Ongoing work

Quasi-market competition in public service provision: user sorting and cream skimming
with Thorbjørn Sejr Guul & Ulrik Hvidman
Latest update: Mar 2020 (revised & resubmitted)


Although quasi-markets that introduce choice and competition between public service provid-ers are intended to improve efficiency, researchers and policy makers have been concerned about adverse effects. This paper examines the consequences of quasi-market competition on segregation of users. First, we argue theoretically why strategic responses from the providers as well as self-selection by the users in quasi-markets may lead to segregation. Second, we exam-ine empirically the impacts of a nationwide quasi-market policy that introduced choice and activity-based budgeting into Danish high schools. We exploit the substantial variation in the degree of competition that schools were exposed to because of the concentration of providers within a geographical area. Using a differences-in-differences design—and register data con-taining the full population of students over a 9-year period (N=209,517)—results show that polarization by ability increased sharply in response to the reform in high-competition areas relative to low-competition areas. These responses in high-competition regions appear to be driven both by changes in user sorting and by strategic behavior among public providers.

On Using Pareto Distributions for Measuring Top-Income Gender Disparities
with Niels-Jakob Harbo Hansen, Karl Harmenberg, & Erik Öberg
Latest update: May 2020 (revise and resubmit)

[Working Paper]

Atkinson et al. (2018) propose a measure of the glass ceiling exploiting that top incomes are approximately Pareto distributed. We clarify how this glass-ceiling coefficient describes the increasing scarcity of women further up in the income distribution and show how it relates to the top-income gender gap. If interpreting top income gender differences as caused by a female-specific income tax, the gender gap and glass-ceiling coefficient measure its level and progressivity, respectively. Using Danish data on earnings, we show that the top gender gap and the glass-ceiling coefficient evolves differently across time, the life cycle, and educational groups.

Missing a Nurse Visit
with Jonas Lau-Jensen Hirani & Miriam Wüst
Latest update: Aug 2020 (submitted)

[Working Paper]

This paper shows that the timing of universal nurse home visiting during the first year of life impacts child and maternal health. Exploiting variation from a large nurse strike in Denmark in 2008, we show that strike exposure increases child (and mother) general practitioner contacts in the first four years only for early-exposed individuals. Moreover, mothers who forego an early nurse visit (rather than a later one) have a higher probability of mental health specialist contacts in the first two years after birth. We highlight the importance of early screening and provision of information to new parents by showing that nurses in control years perform well in identifying maternal mental health risks during early home visits (likely preventing longer-term problems), and that firstborn children and children of parents with no educational background in health drive our results. A stylized calculation confirms that the short-run health benefits from early universal NHV outweigh its costs.

Grades and employer learning
with Anne Toft Hansen & Ulrik Hvidman
Latest update: Jul 2019

We identify the signaling value of a grade point average (GPA) on labor market outcomes and how fast employers learn about variation in GPA that is unrelated to labor market productivity. Exploiting a reform-induced variation in university graduates’ GPA, we find that a higher GPA causes higher earnings in the first two years after graduation, after which the signaling effect goes to zero. We provide suggestive evidence that the signaling effect is more relevant to majors related to larger wage dispersion and the private sector and that the earnings adjustment happens both within and across firms. We conduct a range of sensitivity tests to validate that the identified relationship represents a signaling effect. Importantly, we show that the recoding scheme that causes variation in the GPA is unrelated to labor market outcomes for non-treated cohorts. Our findings are consistent with a learning model whereby employers initially screen graduates based on observable signals but update their beliefs about worker productivity.

Measuring Vacancies: Firm-level Evidence from Two Measures
with Niels-Jakob Harbo Hansen
Latest update: Aug 2016

[Working Paper]

Using firm-level survey- and register-data for both Sweden and Denmark we show systematic mis-measurement in both vacancy measures. While the register-based measure on the aggregate constitutes a quarter of the survey-based measure, the latter is not a super-set of the former. To obtain the full set of unique vacancies in these two databases, the number of survey vacancies should be multiplied by approximately 1.2. Importantly, this adjustment factor varies over time and across firrm characteristics. Our findings have implications for both the search-matching literature and policy analysis based on vacancy measures: Observed changes in vacancies can be an outcome of changes in mis-measurement, and are not necessarily changes in the actual number of vacancies.


Maternity Ward Crowding, Procedure Use, and Child Health
with Jonas Maibom, Marianne Simonsen, & Miriam Wüst
Journal of Health Economics
, forthcoming.

[Working Paper]

This paper studies the impact of day-to-day variation in maternity ward crowding on medical procedure use and the health of infants and mothers. Exploiting data on the universe of Danish admissions to maternity wards in the years 2000-2014, we first document substantial day-to-day variation in admissions. Exploiting residual variation in crowding, we find that maternity wards change the provision of medical procedures and care on crowded days relative to less crowded days, and they do so in ways that alleviate their workload. We find very small and precisely estimated effects of crowding on child and maternal health. Thus our results suggest that, for the majority of uncomplicated births, maternity wards in Denmark can cope with the observed inside-ward variation in daily admissions without detectable health risks.

High-Stakes Grades and Student Behavior
with Ulrik Hvidman
Journal of Human Resources
, forthcoming.

[Published Paper] [Working Paper] [Code] [Media]

High-stakes exams carry important consequences for the prospects of reaching university. This study examines whether the incentives associated with exam grades affect educational investments. Exploiting a reform-induced recoding of high school students’ grade point average, we identify the effect of high-stakes grades on student behavior. The results show that students who were downgraded by the recoding performed better on subsequent assessments. The increase in academic performance in high school translated into an increased likelihood of university enrollment. As the recoding did not convey information about actual performance, these results emphasize that incentives are important in understanding students’ educational investments.

Neonatal Health of Parents and Cognitive Development of Children
Claus Thustrup Kreiner
Journal of Health Economics
, 69, 2020.

[Published Paper] [Working Paper] [Media]

High-stakes exams carry important consequences for the prospects of reaching university. This study examines whether the incentives associated with exam grades affect educational investments. Exploiting a reform-induced recoding of high school students’ grade point average, we identify the effect of high-stakes grades on student behavior. The results show that students who were downgraded by the recoding performed better on subsequent assessments. The increase in academic performance in high school translated into an increased likelihood of university enrollment. As the recoding did not convey information about actual performance, these results emphasize that incentives are important in understanding students’ educational investments.

The Strength and Difficulties Questionnaire and Standardized Tests: Reliability across Respondent Type and Age
with Maria Keilow, Janni Niclasen, and Carsten Obel
PLoS One
, 14(7), 2019.

[Published Paper]

Exploiting nation-wide data from the Danish National Birth Cohort, we show that children’s emotional and behavioral problems measured by the Strengths and Difficulties Questionnaire (SDQ) are closely related to their performance in standardized academic tests for reading and mathematics in sixth grade. The relationship is remarkably linear across the entire distribution for both the total difficulties score and subscale scores of the SDQ; higher scores on the SDQ (more problems) are related to worse performance in academic tests. We assess the similarity across respondent type; parent (child age 7 and 11), teacher (child age 11) and self-reported scores (child age 11), and find that teacher and parent reported scores have very similar slopes in the SDQ–test score relationship, while the child reported SDQ in relation to the academic test performance has a flatter slope.

The Socio-Economic Gradient in Children’s Test-Scores - A Comparison Between the U.S. and Denmark
with Christopher Jamil de Montgomery
Nationaløkonomisk Tidsskrift
, 1, 2019.

[Published Paper]

This paper contributes to the debate on intergenerational mobility in the U.S. and Denmark by linking parental resources to differentials in cognitive development as they develop through primary and lower secondary school in each country. Using U.S. survey data and Danish register data, we observe a socio-economic gradient along the entire test score distribution in both countries, but the gradient is always largest in the U.S. The test-score difference between the above and below median household income groups at school entry is about 20 percent larger in the U.S. compared with Denmark. Our findings show that a substantial socio-economic test-score gradient is present even in a Scandinavian welfare state with universal child policies. However, in light of the recent debate on similarities in intergenerational mobility between Denmark and the U.S., it is important to note that the socio-economic gradient in test-scores is smaller in Denmark compared with the U.S.

The Gift of Time? School Starting Age and Mental Health
with Thomas S. Dee
Health Economics
, 27, 2018.

[Published Paper] [Working Paper]

Using linked Danish survey and register data, we estimate the causal effect of age at kindergarten entry on mental health. Danish children are supposed to enter kindergarten in the calendar year in which they turn 6 years. In a "fuzzy" regression-discontinuity design based on this rule and exact dates of birth, we find that a 1-year delay in kindergarten entry dramatically reduces inattention/hyperactivity at age 7 (effect size = –0.73), a measure of self-regulation with strong negative links to student achievement. The effect is primarily identified for girls but persists at age 11.

Discharge on the day of birth, parental responses and health and schooling outcomes
with Miriam Wüst
Journal of Health Economics
, 55, 2017.

[Published Paper] [Working Paper]

Exploiting the Danish roll-out of same-day discharge policies after uncomplicated births, we find that treated newborns have a higher probability of hospital readmission in the first month after birth. While these short-run effects may indicate substitution of hospital stays with readmissions, we also find that—in the longer run—a same-day discharge decreases children's ninth grade GPA. This effect is driven by children and mothers, who prior to the policy change would have been least likely to experience a same-day discharge. Using administrative and survey data to assess potential mechanisms, we show that a same-day discharge impacts those parents’ health investments and their children's medium-run health. Our findings point to important negative effects of policies that expand same-day discharge policies to broad populations of mothers and children.
Cognitive fatigue influences students’ performance on standardized tests
with Francesca Gino & Marco Piovesan
Proceedings of the National Academy of Sciences
, 113(10), 2016.

[Published Paper] [Code]

Using test data for all children attending Danish public schools between school years 2009/10 and 2012/13, we examine how the time of the test affects performance. Test time is determined by the weekly class schedule and computer availability at the school. We find that, for every hour later in the day, test performance decreases by 0.9% of an SD (95% CI, 0.7–1.0%). However, a 20- to 30-minute break improves average test performance by 1.7% of an SD (95% CI, 1.2–2.2%). These findings have two important policy implications: First, cognitive fatigue should be taken into consideration when deciding on the length of the school day and the frequency and duration of breaks throughout the day. Second, school accountability systems should control for the influence of external factors on test scores.

Care around birth, infant and mother health and maternal health investments – Evidence from a nurse strike
with Hanne Kronborg & Miriam Wüst
Social Science and Medicine
, 150, 2016.

[Published Paper] [Working Paper]

Care around birth may impact child and mother health and parental health investments. We exploit the 2008 national strike among Danish nurses to identify the effects of care around birth on infant and mother health (proxied by health care usage) and maternal investments in the health of their newborns. We use administrative data from the population register on 39,810 Danish births in the years 2007–2010 and complementary survey and municipal administrative data on 8288 births in the years 2007–2009 in a differences-in-differences framework. We show that the strike reduced the number of mothers' prenatal midwife consultations, their length of hospital stay at birth, and the number of home visits by trained nurses after hospital discharge. We find that this reduction in care around birth increased the number of child and mother general practitioner (GP) contacts in the first month. As we do not find strong effects of strike exposure on infant and mother GP contacts in the longer run, this result suggests that parents substitute one type of care for another. While we lack power to identify the effects of care around birth on hospital readmissions and diagnoses, our results for maternal health investments indicate that strike-exposed mothers—especially those who lacked postnatal early home visits—are less likely to exclusively breastfeed their child at four months. Thus reduced care around birth may have persistent effects on treated children through its impact on parental investments.

Local unemployment and the timing of post-secondary schooling
Economics of Education Review
, 50, 2016.

[Published Paper] [Working Paper]

Using Danish administrative data on all high school graduates from 1984 to 1992, I show that local unemployment has both a short- and a long-run effect on school enrollment and completion. The short-run effect causes students to advance their enrollment, and consequently their completion, of additional schooling. The long-run effect causes students who would otherwise never have enrolled to enroll and complete schooling. The effects are strongest for children of parents with no higher education.

The Exploitation of Talent
with Nicolaj Christiansen
Nationaløkonomisk Tidsskrift
, 146(1), 2008.

[Published Paper]

This article is a review of the seminar paper Superstar Effect in Italian Football which was written for the seminar on sports economics at the Department of Economics, University of Copenhagen and awarded the McKinsey Award in spring 2008. Only the main findings from the seminar paper will be presented here. For details and technicalities we refer to the original seminar paper Christiansen and Sievertsen (2008), which can be downloaded at www.econ.ku.dk/nf/superstareffect.pdf. In short, superstar economics is the branch of labour economics that deals with the phenomenon of nonlinear and highly rightskewed income distributions, that is observed in certain activities. The puzzle is, that the most talented individuals in these activities can obtain extremely high salaries compared to their colleagues, even though they are only marginally more talented. Theoretical explanations of the puzzle are reviewed and the superstar phenomenon is analysed empirically on Italian football, where a significant superstar effect is found.

Reports & book contributions

Segregering i danske børnehaver 2005-2014
VIVE - Rapport. August, 2020.

[Download report]

En god start - Betydningen af alder ved skolestart for barnets udvikling
SFI - Rapport, 15:38. November, 2015.

[Download report] [Updated results]

Børn i Lavindkomstfamilier
with Christopher Jamil de Montgomery
SFI - Rapport, 15:22. Juni, 2015.

[Download report]

Det økonomiske fundament: Finansiering og faktorer med betydning for omkostningerne
with Ulrik Hvidman , Kapitel 4 i Styring, ledelse og resultater på ungdomsuddannelserne. Redigeret af Lotte Bøgh Andersen, Peter Bogetoft, Jørgen Grønnegård Christensen, Torben Tranæs. Syddansk Universitetsforlag 2014.

[Download book]

Infrastruktur og Danmarks internationale konkurrenceevne: et strategisk oversigtssrudie
with Ole Kveiborg , Technical University of Denmark, DTU Transport; Nr. 7:2010.

[Download report]