American Economic Review, Quarterly Journal of Economics, Journal of Political Economy, Econometrica and Review of Economic Studies.
Andrew Caplin, Victoria Gregory, Eungik Lee, Søren Leth-Petersen and Johan Sæverud · Review of Economic Studies, 2026 · Individual
Introduces a survey instrument for earnings risk that allows for quitting and being fired, fielded to Danish workers and linked to their register earnings histories. Job transitions drive subjective risk, which is many times smaller than traditional register-based estimates even after conditioning on demographics and job history, and helps explain why many hold few liquid assets.
- Data
- Copenhagen Life Panel (CLP) survey, January 2021 wave, linked on CPR number, Income-tax register earnings (IND), Monthly employer-employee match and earnings data (BFL/eIndkomst), Demographic, education, household composition and household wealth registers
- Key variables
- Survey: subjective probabilities of staying, quitting and being laid off
- Survey: subjective earnings distributions conditional on each job transition
- Benchmark: realized job transitions, time out of work and earnings growth in 2021 (administrative)
- Weights: population weights from a survey-participation model on 2020 administrative data
- Period
- Survey January 2021; administrative data through 2022
- Sample
- 10,942 employed respondents (gross sample 14,683 respondents aged 20–65) Copenhagen Life Panel respondents aged 20–65 who were employed at the time of the January 2021 survey. source
Labour markets, wages and sorting Household finance, saving and consumption Measurement and methods
Asger Lau Andersen, Kilian Huber, Niels Johannesen, Ludwig Straub and Emil Toft Vestergaard · Quarterly Journal of Economics, 2026 · Individual + firm
Builds a system of national accounts broken down into bilateral money flows between region-by-industry consumer cells, producer cells, the government and the rest of the world, and implements it for Denmark using bank transaction data and registers. Spending is locally and urban biased, and fiscal transfers raise GDP more when targeted at cells whose spending reaches cells with high unemployment.
- Data
- Danske Bank geo-coded transaction-level data (about 20% of Danish adult consumers), 2018–2019, Government tax and income registers (labour compensation, mixed income, interest, dividends, transfers), Employment registers linking adults to industry of main employment and region, Danish national accounts (SNA positions, input-output tables), 2018, CrediWire producer-to-producer transactions (about 5 million), Firm-level export and import data for manufacturing firms
- Key variables
- Outcome: bilateral cell-to-cell flows (consumer spending, labour and profit income, taxes, transfers, intermediates trade)
- Unit: consumer cells (industry of main employment x region of residence)
- Unit: producer cells (industry x region of establishments)
- Derived: domestic spending share, gravity, urban bias and spending intensity measures
- Period
- 2018 (consumer spending 2018–2019)
- Sample
- approx. 2,700 consumer cells and approx. 2,600 producer cells; around 43 million entries All Danish adults and production establishments grouped into region-by-industry cells, with consumer spending measured from Danske Bank customers. source
Macroeconomics and expectations Firms, ownership and innovation
Dmitry Taubinsky, Luigi Butera, Matteo Saccarola and Chen Lian · Quarterly Journal of Economics, 2026 · Individual
Links Statistics Denmark's consumer expectations survey to the registers to show that people's inflation forecasts and backcasts move strongly, and negatively, with their own recent and expected income changes, far more than actual inflation does. The pattern departs from Bayesian rational expectations and is consistent with affect-cued recall, where bad personal events cue pessimistic recollections.
- Data
- Danish Consumer Expectations Survey (Statistics Denmark Omnibus), linked on CPR, Tax registers from SKAT: income, liquid assets, net wealth, taxes (IND, formue), Civil Registration System / Population register (CPR/BEF), Education register (UDDA), National Patient Register (LPR), emergency room visits 1994–2018
- Key variables
- Outcome: inflation forecasts and backcasts (qualitative and quantitative)
- Explanatory: recent change in household log nominal income (t-1 vs t-2)
- Explanatory: future household income change and expected household financial situation
- Shock: family emergency room visit in the survey month
- Benchmark: realized inflation
- Period
- 2012–2019 (robustness from 2008; population registers from 1991)
- Sample
- 35050 usable responses (from 55171 survey respondents aged 25–60) Danish Consumer Expectations Survey respondents aged 25–60 in 2012–2019 without substantial self-employment income and with complete answers, linked to administrative data. source
Macroeconomics and expectations Household finance, saving and consumption
Katrine Jakobsen, Thomas H. Jørgensen and Hamish Low · American Economic Review (forthcoming), 2026 · Individual
Uses the 2009 Danish tax reform to show that higher net-of-tax wages for women reduce fertility while higher wages for men raise it. A life-cycle model of couples choosing the timing and number of children, estimated on register data, implies that allowing fertility to adjust raises women's labour supply responsiveness substantially and that human capital depreciation around births drives much of the long-run gender wage gap.
- Data
- Monthly payroll records (BFL/eIndkomst), Annual income tax data (IND), Population and family register (BEF), family definitions and parent links, Education register (UDDA), Labour market experience from ATP payments, Parental leave spell register (OF)
- Key variables
- Outcome: fertility (births)
- Treatment: 2009–2018 tax reform changes in marginal net-of-tax wage rate
- Outcome: labour market participation, part-time and full-time work from pay-slip hours
- Outcome: annual labour income
- Control: education (high vs less skilled), labour market experience
- Period
- 2004–2018 (tax sample 2004–2018; estimation sample 2010–2018)
- Sample
- 10,851,444 individual-year observations (estimation sample); 2,531,181 (tax sample) Danish individuals aged 25–60 with an opposite-sex partner, not students, self-employed, retired or on disability pension; tax sample restricted to cohabiting or married women aged 25–40. source
Gender and the family Family responses and insurance Taxation and public finance Labour markets, wages and sorting
Mathias Fjællegaard Jensen and Ning Zhang · American Economic Review, 2026 · Individual
Uses sudden first parental deaths and matched controls in Danish administrative data. Earnings fall by about 2 percent for men and 3 percent for women, mental health worsens for both, and women with young children lose the most, around 4 percent, consistent with the loss of informal childcare from grandparents.
- Data
- Cause of death registers (DODSAARS/DODSAARG), Population register with parent-child links (BEF), Income and labour market registers (IND, AKM, IDAS/IDAN/IDAP), ATP-based hours proxy, Health Insurance Register for GP/psychologist/psychiatrist visits (SYSI/SSSY), National Patient Register (LPR_ADM/LPR_DIAG), Psychiatric Central Research Register (PSYK_ADM/PSYK_DIAG), Prescription register (LMDB), Childcare registers (DAGI/BOERNFB)
- Key variables
- Treatment: first sudden parental death (heart disease, cerebrovascular disease, acute respiratory infection, accidents)
- Outcome: annual earnings (indexed to pre-death gender mean)
- Outcome: employment and hours proxy
- Outcome: mental health care use and prescriptions
- Control group: matched not-yet-treated individuals
- Period
- 1980–2019 (parental deaths 1985–2014)
- Sample
- Treated: 21919 daughters and 24354 sons whose mother died first; 74991 daughters and 85022 sons whose father died first (plus matched controls) Danish adults aged 25–50 with two known parents who experienced a sudden first parental death between 1985 and 2014, with matched controls. source
Family responses and insurance Labour markets, wages and sorting Health
N. Meltem Daysal, Hui Ding, Maya Rossin-Slater and Hannes Schwandt · American Economic Review, 2026 · Individual
Shows that infant younger siblings are hospitalised for respiratory conditions two to three times as often as their older siblings were, because preschool-aged children bring disease home. Combining birth order with local variation in respiratory disease prevalence, early-life exposure lowers younger siblings' later earnings and education and worsens chronic respiratory and mental health.
- Data
- National Patient Register (LPR), Medical Birth Register (MFR), Population register (BEF), Register-Based Labour Force Statistics (RAS), Income Statistics Register (IND), Education Register (UDDA), Academic Achievement Register (9th grade test scores), Psychiatric Central Research Register (PCRR), Health Insurance Register (Sygesikringsregisteret), Childcare enrollment data, National Child Database (breastfeeding)
- Key variables
- Outcome: acute respiratory hospitalizations in the first year of life
- Treatment: birth order (second-born vs first-born) interacted with municipality respiratory disease exposure index
- Outcome: earnings, labour force participation and income rank at ages 18–32
- Outcome: high school and college completion; 9th grade test scores
- Outcome: chronic respiratory and mental healthcare use
- Period
- Children born 1981–2017; outcomes 1980–2019
- Sample
- 1,230,180 children First- and second-born singleton siblings born in Denmark 1981–2017, at least 11 months apart, in 91 municipalities. source
Health Children, intergenerational mobility and skills Family responses and insurance
Steffen Altmann, Anita M. Glenny, Robert Mahlstedt and Alexander Sebald · Review of Economic Studies, 2026 · Individual
A large two-stage randomised experiment on the Danish public employment service's online platform gave job seekers vacancy information and occupational recommendations, with treatment intensity varying by region. Advice raises employment and earnings when few job seekers receive it, but the gains vanish at high intensity because treated job seekers crowd into the same occupations.
- Data
- Field experiment on jobnet.dk (Danish public employment service platform): treatment assignment and dashboard exposure, Jobnet job search profiles, vacancy information and registered job applications (with occupation codes), Statistics Denmark register data on employment, hours and earnings (12 months after start), Population registers and benefit payment records (UI recipients), Register data on the universe of unemployment-to-job occupational transitions (to build recommendations)
- Key variables
- Treatment: occupational recommendations, vacancy information, or both (individual randomization)
- Treatment intensity: municipality cluster share of treated job seekers (two-stage randomization)
- Outcome: working hours and earnings in the year after intervention
- Outcome: job applications by occupation (search strategy)
- Period
- Intervention from March 17, 2019; outcomes over the following 12 months (before March 2020)
- Sample
- 92,098 individuals All individuals registered as unemployed and receiving UI benefits in Denmark on March 17, 2019. source
Labour markets, wages and sorting
Anders Humlum, Jakob R. Munch and Pernille Plato · American Economic Review, 2025 · Individual
Uses quasi-random work accidents in Danish administrative data to show that injured workers enrol in bachelor's programmes that build on their previous training. Exploiting differences in how stackable degrees are, higher education moves injured workers from disability benefits into full-time jobs paying 25 percent more than before injury, and reskilling subsidies pay for themselves four times over.
- Data
- Industrial Injury Register (Arbejdsskaderegisteret) / workers' compensation claims from AES, National Patient Register (LPR), Health Insurance Register (Sygesikringsregisteret), Prescription Drug Database (LMDB), Education Register (UDDA), Course Participant Register (VEUV), Danish Register for Evaluation of Marginalization (DREAM), Integrated Database for Labour Market Research (IDA), linked to O*NET, Population register (BEF/POP)
- Key variables
- Treatment: work accident causing loss of earning capacity (vs. matched non-injured workers)
- Outcome: enrollment in and completion of formal degrees and courses
- Outcome: earnings, hours and occupation (task content)
- Outcome: disability and other social transfers
- Severity: assessed earning capacity loss and personal impairment
- Period
- 1995–2017 (work injury claims 1998–2017)
- Sample
- 14,481 injury events (14,369 distinct individuals); 4,568 events (4,528 individuals) for workers with vocational degrees with access to higher education Full-time employed Danish workers (excluding military) who suffered an accepted physical work accident with positive earning capacity loss, compared with matched non-injured workers. source
Labour markets, wages and sorting Education Health
Henrik Kleven, Claus Thustrup Kreiner, Kristian Larsen and Jakob Egholt Søgaard · American Economic Review, 2025 · Individual
Builds a model in which effort today raises earnings tomorrow, so responses to taxes are delayed and mediated by job switches. Danish administrative data confirm the model's life-cycle predictions for earnings and hours, and a quasi-experimental design based on job switchers finds a long-run earnings elasticity around 0.5, against a short-run elasticity of about 0.2.
- Data
- Linked administrative registers on earnings, occupation (DISCO), firm and demographics, full population, Employer-employee matched data (firm x occupation job cells), ATP pension contribution-based hours measure (1980–2018), Monthly uncapped hours measure (from 2008), Income tax records (2009 tax reform variation)
- Key variables
- Outcome: earnings (log earnings changes)
- Treatment: 2009 tax reform change in marginal net-of-tax rate
- Instrument: simulated net-of-tax rate change at baseline income
- Mediator: job switches (firm x occupation cell), incl. mass-layoff switches
- Outcome: hours worked
- Period
- 1980–2018
- Sample
- N = 2,501,126 and 2,428,411 in the main elasticity regressions (Table 1) Full Danish population; tax-reform analysis uses taxpayers with baseline income between 250,000 and 1,000,000 DKK. source
Taxation and public finance Labour markets, wages and sorting
Hanno Foerster · Review of Economic Studies, 2024 · Individual
Estimates a dynamic model of family labour supply, home production, saving and divorce on Danish register and survey data to study post-marital maintenance payments. Alimony carries larger labour supply disincentives and insures consumption less efficiently than child support, so welfare rises when child support is increased and alimony reduced.
- Data
- Danish register data on income, labour force status and hours worked, Marital histories and birth register (BEF/family links), Maintenance (child support and alimony) payment records and child residence after divorce, Danish Time Use Survey 2001 (DTUS)
- Key variables
- Outcome: labour supply (employer-recorded hours bins) and annual labour income
- Outcome: divorce
- Policy: child support and alimony (maintenance) payments
- Survey: housework hours and relative consumption in couples (DTUS)
- Period
- 1980–2013
- Sample
- 322,732 couples (645,464 individuals) and 4,312,826 couple-year observations; DTUS 2,105 households Danish couples in their first marriage, both spouses aged 25–58, at least one working, excluding couples with children from previous relationships. source
Family responses and insurance Gender and the family Labour markets, wages and sorting
Petter Lundborg, Erik Plug and Astrid Würtz Rasmussen · Review of Economic Studies, 2024 · Individual
Uses Danish children conceived through sperm and egg donation to separate nurture from genetic transmission, since donors are not selected on parental traits. Only mothers' education matters once genes are removed: fathers' schooling has no effect on children's test scores, while mothers' schooling raises reading scores as much as among non-donor children.
- Data
- Danish IVF register (Danish Health Data Authority), Danish Education Register (UDDA), National test scores in reading and math (from 2010), Administrative registers on demographics and labour market (birth year, gender, marital status, children, earnings)
- Key variables
- Outcome: child average standardized national test score (reading and math)
- Explanatory: mother's and father's years of schooling
- Variation: donor sperm / donor egg conception (genetically unrelated parent)
- Controls: month of first donor treatment and clinic fixed effects (waiting-list assignment)
- Period
- IVF treatments 1994–2007; demographics 1991–2016; labour market 1991–2012; tests from 2010
- Sample
- 19,509 non-donor IVF children, 820 sperm donor children and 157 egg donor children (97 matched to egg donors) Children conceived through IVF treatments in Denmark 1994–2007 with at least one national test score, compared with 30% random samples of non-adopted (650,930) and adopted (2,674) children. source
Children, intergenerational mobility and skills Education Family responses and insurance
Ufuk Akcigit, Jeremy Pearce and Marta Prato · Review of Economic Studies, 2024 · Individual
Builds an endogenous growth model with heterogeneous talent, financial resources and career choice, and links it to Danish micro data on who obtains a PhD and who becomes an inventor. R&D subsidies work better when combined with higher education subsidies that let talented but poor youth pursue research, and education policy matters most in unequal societies.
- Data
- Integrated Database for Labour Market Research (IDA), Firm-linked IDA (FIDA), PhD enrolment register (PHD), Military conscription IQ test (Boerge Prien Proeve), European Patent Office patents linked to IDA, R&D funding data (Innovation Danmark) and Ministry of Education PhD policy data
- Key variables
- Outcome: PhD enrolment/attainment
- Outcome: patent filing (inventor status)
- Talent proxy: military IQ test score
- Background: parental income
- Policy: 2002 expansion of PhD slots and R&D subsidies
- Period
- 2001–2013 (background data from 1980)
- Sample
- approximately 32 million employer-employee observations, 10,000 inventors and 37,000 unique PhDs; approximately 500,000 males with IQ data Danish population in IDA linked to PhD, IQ test and patent data. source
Innovation Education Inequality and fairness
Kristoffer B. Hvidberg, Claus T. Kreiner and Stefanie Stantcheva · Review of Economic Studies, 2023 · Individual
Links a large survey to income registers to compare perceived and actual positions in the income distribution and within reference groups (co-workers, neighbours, schoolmates). People misperceive their rank, fairness views depend on the perceived gap to others, and showing people their true position changes their views.
- Data
- Custom online survey with randomised information treatment (February–March 2019), sent via Digital Post and linked on CPR, Income registers / tax returns (IND) for respondents and all reference group members, Population register (BEF) for cohort, municipality, neighbours and siblings, Education register (UDDA) incl. school attended at age 15 for schoolmates, Workplace links from employer-employee data (IDA) for co-workers and sector, Parents' income positions via family links
- Key variables
- Outcome: perceived own income position (rank) in each reference group
- Outcome: perceived P50 and P95 incomes in reference groups
- Outcome: fairness views on inequality within reference groups
- Actual income rank computed from register data
- Treatment: randomised information on actual income position
- Events: unemployment, promotion and income shocks from income histories
- Period
- Survey 2019 (income for 2017); register income histories
- Sample
- 9,415 respondents (from 50,100 invited) Danish-born people born 1969–1973 (aged 45–50), randomly drawn by Statistics Denmark, who completed the survey. source
Inequality and fairness
Rasmus Lentz, Suphanit Piyapromdee and Jean-Marc Robin · Econometrica, 2023 · Individual + firm
Estimates a flexible model of worker and firm heterogeneity in wages and job mobility on Danish matched employer-employee data. Workers and firms sort positively, with sorting operating through both wages and non-wage amenities, and worker heterogeneity accounts for the bulk of wage dispersion.
- Data
- Matched employer-employee register data (IDA) with annual wages, Weekly labour market spell/mobility data built from Statistics Denmark registers
- Key variables
- Outcome: log wage rate
- Job-to-job mobility and transitions to and from non-employment (weekly)
- Latent worker and firm types (finite mixture, classification EM)
- Controls: gender, education, potential experience, tenure, calendar time
- Sorting channels: job preferences, segmentation, layoffs, job finding
- Period
- 1985–2013 (analysis panel starts 1987)
- Sample
- Not stated in the sources checked Danish workers after completing education, employment spells starting before age 50, followed weekly across employers from 1987 to 2013 (read from the 2018 predecessor working paper). source
Labour markets, wages and sorting
Kenneth Gillingham, Fedor Iskhakov, Anders Munk-Nielsen, John Rust and Bertel Schjerning · Journal of Political Economy, 2022 · Individual
Develops and estimates a dynamic equilibrium model of the market for new and used cars with heterogeneous households on the universe of Danish car ownership records, which record every vehicle's owner, age and scrappage. The model is used to evaluate Denmark's very high registration taxes and fuel taxes, showing how tax changes propagate through used-car prices and the age distribution of the fleet.
- Data
- Car ownership register covering all privately owned cars (Statistics Denmark extract), Odometer readings from mandatory vehicle inspections (driving), Demographic and income registers for all Danish households, Tax deduction for commuting distance (work distance), Fuel prices from eof.dk
- Key variables
- Outcome: car ownership transitions (keep, trade, buy new, scrap)
- Car types: four classes by weight and fuel efficiency, 25 age groups
- Household types: couple/single, high/low income, high/low work distance
- Outcome: kilometres driven
- Policy variable: new car registration tax and fuel tax
- Period
- 1996–2008 (estimation 1997–2008)
- Sample
- nearly 39 million observations on car ownership transitions All Danish households and all cars owned by private individuals, 1996–2008. source
Household finance, saving and consumption Taxation and public finance Measurement and methods
Steffen Andersen, Cristian Badarinza, Lu Liu, Julie Marx and Tarun Ramadorai · American Economic Review, 2022 · Individual
Combines Danish housing listings and transactions with register data on owners to estimate a structural model of listing and sale decisions with loss aversion and financial constraints. Households derive utility from gains over the original purchase price and weigh losses about 2.5 times more than gains, which helps explain the co-movement of house prices and turnover.
- Data
- Property ownership and transaction register from SKAT and the Building and Housing Register (BBR), 1992–2016, Tax-assessed property values (SKAT), Property listings data from RealView, 2008–2016, Loan-level mortgage data from Danmarks Nationalbank, 2009–2016, Civil Registration System (CPR/BEF) for demographics and household links, Income and wealth records from SKAT (IND/formue)
- Key variables
- Outcome: listing decision and listing premium relative to hedonic value
- Main regressor: potential gain or loss relative to original purchase price (reference point)
- Potential home equity relative to down-payment constraint
- Expected house value from hedonic pricing model
- Structural parameter: loss aversion
- Period
- 2009–2016 (transactions from 1992)
- Sample
- 214,103 listings (191,507 unique households, 178,933 unique properties); housing stock of 5,538,052 observations of 807,345 properties Listings of Danish owner-occupied housing between 2009 and 2016, both sold and retracted, matched to mortgages and household finances. source
Housing Household finance, saving and consumption
Wolfgang Keller and Hale Utar · Review of Economic Studies, 2022 · Individual + firm
Men and women hit by the same import shock from China adjust differently. Women in their late thirties who are displaced respond by having a child, especially high earners and those who would need to retrain, so a gender-neutral trade shock produces a long-run gender gap through the biological clock.
- Data
- Integrated Database for Labour Market Research (IDA), Central Population Register (CPR/BEF) for births, marriages, divorces and cohabitation, Domestic production database by 10-digit product (Industrial commodity statistics, VARS) to identify firms making quota goods, Industry-level imports from China and other advanced countries (instrument), Parental leave records
- Key variables
- Outcome: new births, marriage, divorce, parental leave
- Outcome: earnings, hours, unemployment
- Treatment: 1999 employer produces goods subject to MFA quota removal for China after WTO entry
- Treatment (economy-wide): industry import penetration from China, instrumented with other countries' imports
- Heterogeneity: gender and age (biological clock)
- Period
- 1999–2007
- Sample
- about 1,600,000 workers; about 10,000 textile and clothing workers The 1999 cohort of full-time Danish private-sector workers aged 18–56, followed through 2007. source
Trade and offshoring Gender and the family Family responses and insurance Labour markets, wages and sorting
Benjamin U. Friedrich and Martin B. Hackmann · Review of Economic Studies, 2021 · Individual + firm
A Danish parental-leave programme offering up to a year off per child aged 0 to 8 induced a nurse shortage that hospitals and nursing homes could not replace. Nurse employment fell by 15 percent in hospitals and 10 percent in nursing homes, 30-day readmissions of heart attack patients rose, and nursing home mortality increased sharply.
- Data
- Integrated Database for Labour Market Research (IDA) for health workers, plants and industries, Population register (BEF) for children's ages, residence and household characteristics, Social benefits records of parental leave spells (from 1995), National Patient Register (LPR), inpatient hospitalisations 1990–2000, Register of Causes of Death (DAR), 1991–2000
- Key variables
- Treatment: 1994 parental leave programme; exposure = share of nurses eligible times estimated take-up, by health sector and county
- Outcome: nurse employment in hospitals and nursing homes
- Outcome: 30-day hospital readmission rates and other quality-of-care measures
- Outcome: nursing home (elderly) mortality
- Eligibility: having a youngest child aged 8 or younger
- Period
- 1990–2000
- Sample
- Not stated in the sources checked Danish nurses, nursing assistants and doctors, and hospital patients and nursing home residents, aggregated to health sector by county, 1990–2000. source
Health Labour markets, wages and sorting
Christian Dustmann and Rasmus Landersø · Journal of Political Economy, 2021 · Individual
Uses the gender of a first-born child as a natural experiment for young fathers in Denmark. Having a son rather than a daughter lowers the father's subsequent criminal behaviour, and the reduction spills over to the father's peers and later to the child, which identifies both a direct parenting response and network effects in crime.
- Data
- Birth register (child's gender, parent links), Population register (BEF) for demographics and family links, Central Police Register crime data (charges and convictions), Victim register (from 2001), Income, education and occupation registers, Residential register with neighbourhood classification
- Key variables
- Outcome: young father's crime convictions and charges
- Treatment: first child is a son vs a daughter
- Spillover outcome: crime of male peers in the father's neighbourhood
- Peer group: males within ±3 years of father's age in the same neighbourhood
- Outcome: neighbourhood crime victimisation
- Period
- Children born 1991–2004; crime data from 1990
- Sample
- 2,803 first-time fathers aged 20 or younger; 101,132 male peers in 2,114 neighbourhoods Danish men who fathered their first child at age 20 or younger between 1991 and 2004, plus male peers in their neighbourhoods. source
Crime Family responses and insurance Children, intergenerational mobility and skills
Christopher Taber and Rune Vejlin · Econometrica, 2020 · Individual + firm
Combines a Roy model, search frictions, compensating differentials and on-the-job human capital accumulation in one framework, establishes what is non-parametrically identified, and estimates it on Danish matched data. Pre-market skills account for most wage variation, while preferences for non-wage job attributes and search frictions matter greatly for turnover and for which jobs workers choose.
- Data
- Weekly labour market spell data built from Statistics Denmark registers (all individuals aged 15–74), Integrated Database for Labour Market Research (IDA) with hourly wage measure, Danish Work Environment Cohort Study surveys (1995 and 2000) on reasons for job separation
- Key variables
- Outcome: log hourly wage (detrended)
- Job-to-job transitions between establishments (revealed preference)
- Employment and non-employment spell durations
- Worker skill/education group and labour market experience
- Decomposition of wage variance into skill, search, compensating differential and human capital components
- Period
- 1985–2003
- Sample
- 2,969,454 workers, 25,630,532 spells and 467,983 establishments (estimation sample) Danish workers after labour market entry and up to age 55, followed in weekly spell data across establishments, 1985–2003. source
Labour markets, wages and sorting
Katrine Jakobsen, Kristian Jakobsen, Henrik Kleven and Gabriel Zucman · Quarterly Journal of Economics, 2020 · Individual
Exploits the 1989 Danish wealth tax reform, which cut rates sharply for the wealthy, to estimate long-run effects of wealth taxes on wealth accumulation. Responses are sizable for the very wealthy (top one percent) and small for the moderately wealthy, with effects that keep growing over a decade.
- Data
- Administrative wealth registry (third-party reported assets and debts, since 1980), Income registers and socio-economic registers (occupation, family composition)
- Key variables
- Outcome: log taxable wealth and third-party reported wealth
- Treatment: 1989 wealth tax cut (unbound vs bound by tax ceiling: ceiling DD)
- Treatment: doubling of couples' exemption threshold (couples DD)
- Running variable for bunching: taxable wealth around exemption threshold
- Controls: pre-reform portfolio shares and income deciles
- Period
- Wealth data since 1980; quasi-experimental panel 1985–1996
- Sample
- 138,384 observations (ceiling DD) and 520,452 observations (couples DD, couples vs couples) Balanced panels of Danish households in the top 1% (ceiling DD) or married couples in the top 5% (couples DD) of the wealth distribution before the 1989 reform. source
Taxation and public finance Inequality and fairness Household finance, saving and consumption
Steffen Andersen, John Y. Campbell, Kasper Meisner Nielsen and Tarun Ramadorai · American Economic Review, 2020 · Individual
Many Danish households fail to refinance fixed-rate mortgages when it is clearly profitable. A model with inattention and psychological refinancing costs fits the data, and both frictions are stronger for households with less education, income and financial wealth.
- Data
- Loan-level mortgage data for all Danish mortgages from Danmarks Nationalbank (via mortgage bank associations), Civil Registration System / population register (CPR/BEF) for demographics and household links, Income and wealth data from the Danish Tax Authority (SKAT) (IND/formue), Education register (UDDA)
- Key variables
- Outcome: refinancing of a fixed-rate mortgage in a quarter
- Incentive: financial gain from refinancing given coupon and market rates
- Structural parameters: probability of inattention and psychological refinancing cost
- Household characteristics: age, education, income, housing and financial wealth
- Period
- 2009–2017
- Sample
- 2,376,815 household observations (9,351,183 household-quarters); 378,421 refinancings Danish households holding a single fixed-rate mortgage observed in consecutive years, 2009–2017. source
Household finance, saving and consumption
Thomas Epper, Ernst Fehr, Helga Fehr-Duda, Claus Thustrup Kreiner, David Dreyer Lassen, Søren Leth-Petersen and Gregers Nytoft Rasmussen · American Economic Review, 2020 · Individual
Links incentivised time-discounting experiments to the wealth positions of a large sample of middle-aged Danes in the registers. Patient individuals sit markedly higher in the wealth distribution, and the association survives controls for education, income profiles, school grades, initial and parental wealth, credit constraints and risk preferences, which points to saving behaviour as the channel.
- Data
- Incentivised online convex time budget experiment (February 2015) linked on CPR, Income and wealth registers based on third-party reports (1980–2015), Population register (BEF) for sampling and demographics, Education and family registers, Survey of time discounting for the 1952–1955 cohorts at ages 18–21
- Key variables
- Outcome: individual's percentile rank in the net wealth distribution
- Main regressor: experimentally measured patience (time discounting)
- Controls: education, income, parental wealth, risk attitudes
- Heterogeneity: credit-constrained vs unconstrained individuals
- Period
- Experiment in 2015; register data 1980–2015 (wealth ranks tracked 2001–2015)
- Sample
- 3,620 experiment participants (of 27,613 invited); 2,548 survey subjects in the 1952–1955 cohorts Danes born 1973–1983 who lived in Copenhagen municipality at age seven and completed the online experiment. source
Household finance, saving and consumption Inequality and fairness
Annette Alstadsæter, Niels Johannesen and Gabriel Zucman · American Economic Review, 2019 · Individual
Matches leaked offshore account data (HSBC Switzerland, Panama Papers) and tax amnesty records to Scandinavian wealth registers. Tax evasion rises steeply with wealth: the top 0.01 percent evade about a quarter of their taxes, which materially raises measured top wealth shares.
- Data
- Leaked HSBC Switzerland client list (Swiss Leaks, 2007) matched to tax records in Denmark, Norway and Sweden, Panama Papers / Mossack Fonseca leak matched to Norwegian and Swedish wealth data (not Denmark), Tax amnesty voluntary disclosures in Norway and Sweden, Population-wide administrative wealth and income tax records for Norway, Sweden and Denmark, Danish stratified random tax audits by SKAT (tax years 2008, 2010, 2012), National accounts household wealth totals
- Key variables
- Outcome: taxes evaded (share of taxes owed) by wealth group
- Ranking variable: household net wealth including imputed pension wealth
- Indicator: owning an undeclared HSBC account / Panama Papers shell / amnesty participation
- Outcome (event study): reported wealth and income after amnesty
- Random-audit detected evasion (domestic evasion)
- Period
- HSBC leak 2006–2007; amnesties since 2006; Danish random audits for tax years 2008, 2010, 2012
- Sample
- 520 HSBC households; 165 Panama Papers households; 1,422 amnesty taxpayers in Norway and 6,811 in Sweden; 18,985 randomly audited Danish taxpayers Scandinavian households appearing in offshore leaks or tax amnesties, ranked in the full population wealth distribution, combined with Danish random audits. source
Taxation and public finance Inequality and fairness
Christian Dustmann, Kristine Vasiljeva and Anna Piil Damm · Review of Economic Studies, 2019 · Individual
Using the quasi-random refugee dispersal policy, refugee allocation raised vote shares for anti-immigration parties in rural municipalities and lowered them in the largest urban municipalities. Effects are strongest where refugees are few relative to natives and where crime rises after allocation.
- Data
- Population register (BEF) to identify refugees and municipality of initial placement, Income, education, labour force and socio-economic registers for 1985 municipality characteristics, Statistics Denmark municipal and parliamentary election results, Danish National Election Survey (DNES) 1990–98
- Key variables
- Outcome: change in vote share of anti-immigration parties (Progress Party, Danish People's Party)
- Treatment: cumulative share of allocated refugees in municipal population (1986 base)
- Source of variation: Danish Spatial Dispersal Policy allocation of refugees 1986–98
- Heterogeneity: urban vs rural municipality
- Controls: 1985 municipality characteristics
- Period
- Refugee allocation 1986–1998; parliamentary elections 1990–98 and municipal elections 1989–97
- Sample
- 275 municipalities (550 observations for parliamentary elections); 76,673 refugees allocated 1986–98 All 275 Danish municipalities observed across election cycles during the 1986–98 refugee dispersal policy. source
Migration, neighbourhoods and refugees Political economy
Giacomo De Giorgi, Anders Frederiksen and Luigi Pistaferri · Review of Economic Studies, 2019 · Individual + firm
Measures household consumption from Danish tax records on income and assets and defines peer groups by workplace using matched employer-employee data. Instrumenting peers' consumption with shocks to their own non-overlapping networks and firms, the paper finds sizeable consumption spillovers that can generate macro multipliers, operating through both intertemporal and intratemporal channels.
- Data
- Administrative income and wealth tax records (tax registry incl. wealth tax data), Integrated Database for Labour Market Research (IDA) for plant IDs and coworker networks, Danish Expenditure Survey (DES) 1994–1996
- Key variables
- Outcome: household consumption imputed from income and change in assets
- Endogenous regressor: average consumption of coworkers' households (peer consumption)
- Instruments: firm-level shocks to distance-3 (non-overlapping) network members
- Network definition: coworkers in the same plant
- Mechanisms: budget shares of visible goods (DES)
- Period
- 1980–1996
- Sample
- 757,439 households Danish households with head aged 18–65 where both spouses are working employees, linked to spouses' coworkers. source
Household finance, saving and consumption Labour markets, wages and sorting
Itzik Fadlon and Torben Heien Nielsen · American Economic Review, 2019 · Individual
A severe health shock to one family member, such as a heart attack or stroke, makes spouses and adult children raise their own preventive behaviour: more check-ups, more preventive medication and lifestyle changes. Responses are sharp, immediate and persistent, and they reflect updating about one's own risk.
- Data
- Prescription Drug Database (LMDB), 1995–2011, Health Insurance Registry (SSR, primary-care and specialist contacts), 1985–2011, National Patient Registry (LPR), inpatient 1980–2011 and outpatient 1994–2011, Cause of Death Registry (DAR), Integrated Database for Labour Market Research (IDA) for demographics, education and coworker links, Family links (spouses, parents–children) from population registers
- Key variables
- Treatment: family member's first heart attack or stroke (non-fatal), or death (fatal shock)
- Outcome: statin and other cardiovascular drug take-up
- Outcome: preventive care such as GP visits and cholesterol testing
- Control group: households experiencing the same shock five years later
- Exposure groups: spouses, adult biological children, siblings, in-laws, coworkers
- Period
- 1980–2011 (shocks between 1985 and 2011)
- Sample
- Non-fatal shocks, spouses: 20,381 treated and 28,699 control (prime age 25–55), 37,828 treated and 36,392 control (older 55–85); biological children: 63,323 treated and 68,437 control (25–40), 39,783 treated and 32,926 control (40–65) Family members and close coworkers of all Danish individuals who had a first heart attack or stroke (or died) between 1985 and 2011. source
Health Family responses and insurance
Jesper Bagger and Rasmus Lentz · Review of Economic Studies, 2019 · Individual + firm
Estimates an equilibrium search model with endogenous search intensity and two-sided heterogeneity on Danish matched data. Finds positive assortative matching between worker and firm productivity, and quantifies how much of wage dispersion is due to sorting versus search frictions.
- Data
- Weekly labour market spell data from administrative registers (jobs reported to tax authorities, public transfers, mandatory pension contributions), Integrated Database for Labour Market Research (IDA): education, hourly wages and firm sector
- Key variables
- Outcome: hourly wage (last week of November)
- Key variable: job-to-job transitions and unemployment spells
- Estimated: worker skill and firm productivity types (sorting)
- Decomposition: log-wage variance into worker heterogeneity, firm heterogeneity, frictions and sorting
- Period
- 1994–2003 (spell data available 1985–2003)
- Sample
- 6,815,884 observations; 782,951 individuals; 1,698,990 job spells; 117,847 firms (NBER 2014 version) Danish workers up to age 55, after labour market entry, never observed in public-sector employment, self-employment, retirement or agriculture. source
Labour markets, wages and sorting
Sharon Traiberman · American Economic Review, 2019 · Individual + firm
Estimates a dynamic occupational choice model on Danish worker histories in which workers accumulate occupation-specific human capital that is lost when they switch. Occupations explain 57 percent of the dispersion in worker outcomes from trade, sectors only 16 percent, and rising import competition from 1995 to 2005 lowered lifetime earnings for about 5 percent of workers.
- Data
- Integrated Database for Labour Market Research (IDA/IDAS) with ISCO occupation codes, Firm and establishment statistics (FIRE, FIRM) for industry codes, Customs trade data (UHDI), UN Comtrade (CEPII BACI), Danish national accounts, Danish CPI, LIBOR
- Key variables
- Outcome: annual salary income from November employer
- Choice variable: two-digit occupation by sector
- Shock: import prices / import competition by industry
- Worker skill groups: low, medium, high education
- Occupational tenure
- Period
- 1996–2007 worker panel (tenure built from 1990–1996); import competition 1995–2005
- Sample
- 18,513,301 observations in sample frame (33,502,593 in raw data) Danish individuals aged 23–59 in IDA, 1996–2007, used to estimate an occupational choice model. source
Trade and offshoring Labour markets, wages and sorting
Petter Lundborg, Erik Plug and Astrid Würtz Rasmussen · American Economic Review, 2017 · Individual
Uses success in in vitro fertilisation treatment among childless Danish women as an instrument for having children, since IVF success is unrelated to labour market history. Children lower women's earnings substantially and persistently, mostly through hourly earnings rather than hours, and the effect is much larger for the first child than for later ones.
- Data
- IVF register held by the Danish National Board of Health (all IVF treatments in public and private clinics, 1994–2005), Linked administrative registers on education, age, marital status and number of children, Linked labour market registers: labour force status, hours worked, hourly wages, annual earnings, occupation, sector and firm
- Key variables
- Instrument: success of the first IVF treatment
- Endogenous treatment: having children (fertility at the extensive margin)
- Outcome: annual earnings
- Outcome: hourly wages and hours worked / employment
- Mechanisms: occupation, sector and firm changes
- Period
- 1991–2009 (IVF treatments 1995–2005)
- Sample
- 18,538 treated women Childless women entering their first IVF treatment with embryo implants between 1995 and 2005. source
Gender and the family Family responses and insurance Labour markets, wages and sorting
Thais Lærkholm Jensen and Niels Johannesen · American Economic Review, 2017 · Individual
Customers of banks that were hit hard by the crisis cut consumption more than customers of healthier banks. The channel is reduced credit availability: exposed households borrowed less and spent less, which shows how bank distress transmits to the real economy through households.
- Data
- Tax authority third-party reports of all individuals' loan and deposit account balances, linked to bank identity via branch numbers (Nets lists), Bank balance sheet data from the Danish Central Bank (MFI statistics), Administrative registers on demographics, education, labour market and income, Income and wealth registers (capital income, social transfers, stock portfolios, pension accounts), Motor vehicle register and real estate register
- Key variables
- Treatment: primary bank's 2007 loan-to-deposit ratio (above vs below median exposure)
- Outcome: individual borrowing (loan balances by account)
- Outcome: imputed spending = disposable income minus change in net wealth
- Outcome: car ownership (motor vehicle register)
- Separate analysis: customers of banks that failed 2008–2011
- Period
- 2003–2011
- Sample
- around 440,000 individuals, almost 3.5 million individual-years and more than 5.7 million individual-account-years 25% random sample of non-self-employed Danish residents aged 20–50 in 2007 whose primary bank did not fail during 2008–2011. source
Household finance, saving and consumption
Anna Piil Damm and Christian Dustmann · American Economic Review, 2014 · Individual
Refugees arriving between 1986 and 1998 were allocated to municipalities quasi-randomly. Boys placed in neighbourhoods with a higher share of youth criminals were later more likely to be convicted, especially for violent crime, which identifies a causal neighbourhood effect on criminal behaviour.
- Data
- Central Police Register: crime charges and convictions for the full population, Administrative registers: age, residence, parents' IDs, country of origin, immigrant status and date of immigration, Educational Institution Register and surveys, Municipality-level crime and socio-economic measures built from the registers
- Key variables
- Treatment: youth crime conviction rate (share of 15–25-year-olds convicted) in the municipality of assignment
- Identification: quasi-random assignment of refugee families under the 1986–1998 spatial dispersal policy
- Outcome: criminal conviction and charge between ages 15 and 21 (by crime type)
- Longer-term outcomes: education and employment
- Controls: family background, country of origin, municipality characteristics at assignment
- Period
- 1980–2006
- Sample
- 4,425 children Refugee children from eight source countries who arrived in Denmark before age 15 with at least one refugee parent between 1986 and 1998 and are observed to age 21. source
Migration, neighbourhoods and refugees Crime Children, intergenerational mobility and skills
David Hummels, Rasmus Jørgensen, Jakob Munch and Chong Xiang · American Economic Review, 2014 · Individual + firm
Links product-level firm imports and exports to worker wages. Offshoring raises wages of high-skilled workers and lowers wages of low-skilled workers within job spells, while exporting raises wages for all. Tasks that are routine or easily offshored lose the most.
- Data
- Firm Statistics Register (FirmStat/FIRM), Integrated Database for Labour Market Research (IDA), Firm-worker link (FIDA), Foreign Trade Statistics Register, Extrastat and Intrastat (UHDI), UN COMTRADE world trade data for instruments, Firm accounts and other firm registers (Data Appendix)
- Key variables
- Outcome: log hourly wage of workers within job spells
- Treatment: firm offshoring (imports of inputs)
- Treatment: firm exports
- Instruments: world export supply and world import demand shocks by product and country (COMTRADE) and transport costs
- Heterogeneity: worker skill type (high vs low skilled)
- Outcome: earnings losses of displaced workers
- Period
- 1995–2006 (trade data 1990–2006)
- Sample
- about 1.95 million worker-firm-year and 9,800 firm-year observations Manufacturing workers in Danish firms that both import and export, excluding single-year job spells, workers changing skill level, and firms with fewer than 50 employees and less than 0.6 million DKK in imports. source
Trade and offshoring Labour markets, wages and sorting Firms, ownership and innovation
Henrik Jacobsen Kleven, Camille Landais, Emmanuel Saez and Esben Schultz · Quarterly Journal of Economics, 2014 · Individual
The 1991 preferential tax scheme for high-earning foreigners roughly doubled the number of highly paid foreigners in Denmark. The implied elasticity of migration with respect to the net-of-tax rate is above one, far larger than domestic labour supply elasticities, and the scheme generated modest wage spillovers.
- Data
- Administrative tax and payroll records for all residents (income tax register), Civil registration (CPR) citizenship and migration records with dates of entry and exit, Foreigners' tax scheme beneficiary records (contract start and end dates), Employer links at firm and establishment level with industry
- Key variables
- Treatment: eligibility for the preferential foreigners' tax scheme (earnings above threshold)
- Outcome: number of foreigners above vs below the earnings threshold (migration response)
- Outcome: earnings of scheme beneficiaries (wage effects / bunching at the threshold)
- Outcome: spillovers on native top earners' employment and earnings
- Key variable: citizenship and years of presence in Denmark
- Period
- 1980–2010 (tax data since 1980; scheme introduced 1992)
- Sample
- Not stated in the sources checked All foreigners and natives resident in Denmark, focusing on foreigners with earnings around the scheme eligibility threshold (researchers excluded). source
Taxation and public finance Migration, neighbourhoods and refugees
Jesper Bagger, François Fontaine, Fabien Postel-Vinay and Jean-Marc Robin · American Economic Review, 2014 · Individual + firm
Builds and estimates an equilibrium search model with on-the-job search and human capital accumulation using Danish career histories. Human capital accumulation drives most early-career wage growth, while job-to-job mobility and employer competition explain the later part of wage profiles.
- Data
- Weekly labour market spell data built from administrative registers (employment, unemployment, non-participation, self-employment, retirement), Integrated Database for Labour Market Research (IDA): hourly wages, education, experience, ownership, ATP pension contributions used to measure actual experience since 1964, Firm accounts survey (Accounts Statistics) 1999–2003 with value added and hours
- Key variables
- Outcome: hourly wage of jobs active in the last week of November
- Key variable: job tenure and actual labour market experience
- Key variable: firm productivity measured by value added (accounting data)
- Key variable: job-to-job, employment-to-unemployment transitions (spell data)
- Stratification: three schooling levels (7–11, 12–14, 15–20 years)
- Period
- 1985–2003
- Sample
- 168,649, 320,638 and 66,155 workers (2,534,203, 4,344,288 and 663,362 observations) in the three education groups Danish private-sector male workers born in 1948 or later, at least two years past graduation, with histories truncated at entry into retirement, self-employment, the public sector or industries lacking value-added data. source
Labour markets, wages and sorting
Raj Chetty, John N. Friedman, Søren Leth-Petersen, Torben Heien Nielsen and Tore Olsen · Quarterly Journal of Economics, 2014 · Individual
Uses the universe of Danish savings records to show that tax subsidies for retirement accounts mostly shift saving between accounts: each krone of subsidy raises total saving by about one øre. Automatic employer contributions, by contrast, raise total saving almost one for one, because roughly 85 percent of individuals are passive savers who do not re-optimise.
- Data
- Income tax register with third-party reported earnings and pension contributions (IND), Third-party reported wealth: end-of-year assets and liabilities in taxable accounts, Population register (BEF), Integrated Database for Labour Market Research (IDA)
- Key variables
- Outcome: contributions to employer and individual retirement accounts
- Outcome: taxable saving (change in taxable assets), gross and net
- Treatment: employer pension contribution changes at job switches (automatic contributions)
- Treatment: 1999 reduction in tax subsidy for capital pensions for top-bracket taxpayers (price subsidy)
- Treatment: government mandatory savings plan (SP) contributions
- Period
- 1995–2009
- Sample
- approximately 41 million observations for 4 million individuals Danish population aged 18–60, excluding observations with self-employment income (unbalanced panel). source
Household finance, saving and consumption Taxation and public finance
Henrik Jacobsen Kleven, Martin B. Knudsen, Claus Thustrup Kreiner, Søren Pedersen and Emmanuel Saez · Econometrica, 2011 · Individual
A randomised field experiment with the Danish tax authority on about 40,000 taxpayers. Evasion is close to zero for third-party reported income but substantial for self-reported income, and audit threat letters raise reported self-reported income. Information, not morality, is what makes modern tax systems work.
- Data
- Randomized tax audit experiment implemented by the Danish tax authority (SKAT), SKAT Business Object Database: third-party information reports, pre-populated returns, filed returns and after-audit returns, item by item, SKAT computer-generated audit flags, Basic socio-economic variables and employer industry/size from SKAT records
- Key variables
- Treatment: random assignment to 100% vs 0% unannounced audit group (2007 filing season)
- Treatment: random threat-of-audit letters (100%, 50%, none) before the 2008 filing deadline
- Outcome: detected tax evasion (audit adjustments to net income and tax)
- Outcome: change in reported income in the following year
- Heterogeneity: self-reported vs third-party reported income
- Period
- 2006–2007 income years (2007 and 2008 filing seasons)
- Sample
- 25,020 employees and 17,764 self-employed Stratified random sample of Danish taxpayers (employees, including transfer recipients, and self-employed) selected by SKAT for the audit experiment. source
Taxation and public finance
Raj Chetty, John N. Friedman, Tore Olsen and Luigi Pistaferri · Quarterly Journal of Economics, 2011 · Individual + firm
Uses bunching at kinks in the Danish tax schedule to show that observed earnings responses are shaped by search costs and by firms coordinating hours around tax thresholds. Once adjustment frictions are accounted for, the structural labour supply elasticity is around 0.3, far above the small elasticities from individual-level micro studies, which helps reconcile micro and macro estimates.
- Data
- Income tax register / Income Statistics Register (IND), Population register (BEF), Integrated Database for Labour Market Research (IDA) with firm ID and occupation
- Key variables
- Outcome: bunching (excess mass) of taxable income at the top-tax kink
- Treatment: marginal tax rate changes at bracket cutoffs (tax simulator)
- Key variable: firm and occupation-level bunching (hours constraints / firm responses)
- Key variable: size of tax change across kinks
- Comparison: wage earners vs self-employed
- Period
- 1994–2001
- Sample
- 17.9 million observations of wage earners (6.8 million within DKr 50,000 of the top tax cutoff; 1.8 million self-employed observations) Danish wage earners aged 15–70, excluding the self-employed and pensioners, with self-employed studied separately. source
Taxation and public finance Labour markets, wages and sorting
Søren Leth-Petersen · American Economic Review, 2010 · Individual
The 1992 reform that let Danish homeowners borrow against housing equity provides an exogenous credit shock. Total expenditure responded only modestly on average, with the response concentrated among young households with little liquid wealth, which indicates that credit constraints bind for a minority.
- Data
- Income tax register: taxable income, transfers, final taxes, Wealth tax register: third-party reported assets and liabilities (housing, shares, mortgage deeds, cash, bonds, debt), Administrative registers on family composition and dwelling characteristics, Danish Family Expenditure Survey (DES) 1994–1996, used to validate the expenditure imputation
- Key variables
- Outcome: total household expenditure imputed from income minus change in wealth
- Treatment: 1992 credit market reform allowing home equity to be used as collateral for consumption loans
- Treatment group: homeowners (renters as comparison/consistency check)
- Credit-constraint proxy: low liquid (non-housing) assets in 1991 relative to monthly disposable income
- Period
- 1987–1996
- Sample
- 38,486 households, of which 25,275 live in owner-occupied housing (working-paper version) Households from a 10% random sample of the Danish population, aged 18–75, not self-employed, non-movers observed in all years 1987–1996. source
Household finance, saving and consumption
Rasmus Lentz and Dale T. Mortensen · Econometrica, 2008 · Firm
Estimates a Schumpeterian model of firm growth through product innovation on a Danish firm panel. Reallocation of labour from less to more innovative firms accounts for roughly half of aggregate productivity growth, with the remainder coming from innovation within firms.
- Data
- Danish Business Statistics Register: annual panel of privately owned firms (value added, total wage bill, full-time equivalent employment)
- Key variables
- Firm value added (Y)
- Total wage bill (W) used to build quality-adjusted employment
- Full-time equivalent employment (N)
- Labor productivity: value added per worker and per quality-adjusted worker
- Moments: firm growth rates including exit, and productivity growth decomposition (within, between, cross, exit)
- Period
- 1992–1997
- Sample
- unbalanced panel of 4,872 firms All privately owned Danish firms existing in 1992 with more than 20 workers, followed through 1997, excluding later entrants. source
Firms, ownership and innovation Innovation
Morten Bennedsen, Kasper Meisner Nielsen, Francisco Pérez-González and Daniel Wolfenzon · Quarterly Journal of Economics, 2007 · Individual + firm
Uses the gender of a departing CEO's first-born child as an instrument for family succession in Danish firms. Family CEOs hurt performance: operating profitability falls by at least four percentage points relative to firms that hire outside CEOs, with the largest costs in fast-growing industries and large firms.
- Data
- Købmandsstandens Oplysningsbureau (KOB) firm financial and management data (private, from annual reports), Danish Civil Registration System (CPR) with family links, Danish Commerce and Companies Agency (Erhvervs- og Selskabsstyrelsen) management/board records with CPR numbers, Industry-level data from Statistics Denmark and OECD STAN
- Key variables
- Outcome: operating return on assets (OROA) around CEO transitions
- Treatment: family CEO succession (incoming CEO related by blood or marriage)
- Instrument: gender of the departing CEO's firstborn child
- Heterogeneity: industry growth, skill intensity and firm size
- Period
- 1994–2002 (management data); financial data 1991–2003
- Sample
- 5,334 successions CEO successions in Danish limited liability (public and private) firms between 1994 and 2002, first transition per firm. source
Firms, ownership and innovation Labour markets, wages and sorting