Norwegian version of this page

Our latest publications

Scientific articles and book chapters
Other

Asheim, Geir B. & Cato, Susumu (2026). The dilemma between repugnancy and sadism extends beyond critical-level utilitarianism. Economics and Philosophy. ISSN 0266-2671. p. 1–25. doi: 10.1017/S0266267126100807. Full text in Research Archive Show summary
We consider axioms under which population principles face an unavoidable choice between accepting the Repugnant Conclusion or the Very Sadistic Conclusion. Our analysis reveals that this dilemma persists even without imposing the axiom of Anonymity. This significantly broadens the ethical implications when considering populations that differ in temporal, spatial or probabilistic dimensions or when including non-human populations. Imposing Existence Independence, combined with an Archimedean axiom that we term the Strict Limit Property, is sufficient for the dilemma to arise. These findings substantially generalize previous impossibility results in population ethics and highlight the inescapable trade-offs faced by any coherent population principle.
Nyborg, Karine (2026). Adam in the Perfectly Competitive Market. Revista de Economía Institucional. ISSN 0124-5996. 28(54), p. 179–196. doi: 10.18601/01245996.V28N54.08. Full text in Research Archive
Ayenew, Workineh; Kotsadam, Andreas; Ringdal, Charlotte & Villanger, Espen (2026). War comes home: Conflict exposure and intimate partner violence in Ethiopia. Social Science and Medicine. ISSN 0277-9536. 409. doi: 10.1016/j.socscimed.2026.119866. Full text in Research Archive
Beck, Kathryn Christine; Kravdal, Øystein; Magnus, Maria Christine; Wörn, Jonathan & Flatø, Martin (2026). From Menarche to Motherhood: Fertility Patterns by Age at Menarche Across 40 Years in Norway. SocArXiv. doi: 10.31235/osf.io/w4j79_v1. Full text in Research Archive Show summary
Age at menarche marks the onset of reproductive capacity, but its association with fertility has likely changed over time amid secular shifts in both menarcheal timing and fertility behavior. In this study, we examine how menarcheal timing is associated with first-birth timing, childlessness, and completed fertility across 40-years of birth cohorts in Norway using regional health survey data linked to administrative registers for over 60,000 women. Earlier age at menarche was associated with earlier entry into motherhood, although this association weakened in younger cohorts. By contrast, late menarche became increasingly associated with delayed first births and, in younger cohorts, with slightly lower completed fertility. These patterns were not explained by shared family background between siblings or by differences in educational trajectories between menarche groups. The findings suggest that developmental timing may play a larger role in explaining divergence in fertility outcomes, particularly with regards to the timing of childbearing, as family formation has shifted to later ages and life courses have become more individualized. These findings underscore how biological characteristics and social context jointly shape reproductive life courses and the importance of incorporating biological and developmental factors alongside social factors into demographic research on fertility change.
Boar, Corina; Knowles, Matthew; Storesletten, Kjetil & Wang, Yicheng (2026). Quantifying the Macroeconomic Impact of Credit Expansions. International Economic Review. ISSN 0020-6598. 66(5), p. 1928–1928. doi: 10.1111/iere.70024. Full text in Research Archive Show summary
ABSTRACT Credit expansions stimulate the economy. We quantify the contribution of households versus firms for this stimulus. Using causal evidence from the quasi‐natural experiment of bank deregulation across US states, we estimate a small open‐economy heterogeneous‐agent New Keynesian model. Deregulation generated lower borrowing costs. Firms' responses to such shocks account for most of the long‐run rise in output and employment. For short‐run dynamics, firms and households are equally important. Firms' large role is identified by an empirical observation: output and employment expanded gradually following deregulation. In the model, lower interest rates generate increasing capital stocks which in turn increases economic activity gradually.
Storesletten, Kjetil; Zhao, Bo & Zilibotti, Fabrizio (2026). Business Cycle During Structural Change: Arthur Lewis’ Theory from a Neoclassical Perspective. The Economic Journal. ISSN 0013-0133. doi: 10.1093/ej/ueag008. Full text in Research Archive Show summary
Abstract We document that business-cycle dynamics change systematically over the course of development. In countries with large but shrinking agricultural sectors, aggregate employment is uncorrelated with GDP, and agricultural employment falls during booms, even as agricultural labour productivity rises. We develop a unified theory of business cycles and structural change that captures these patterns. The theory emphasises the simultaneous decline and modernisation of agriculture, driven by capital accumulation. As agriculture becomes increasingly capital-intensive, traditional practices are crowded out. We estimate the model and show that it accounts for both structural transformation and business-cycle fluctuations in China.
Andresen, Martin Eckhoff & Thorvaldsen, Lars (2026). The role of escape clauses: Firm response to thin capitalization rules. Journal of Public Economics. ISSN 0047-2727. 262, p. 1–29. doi: 10.1016/j.jpubeco.2026.105757. Full text in Research Archive
Kravdal, Øystein; Grundy, Emily & Hart, Rannveig Kaldager (2026). Health Influences on Second‐ and Third‐Birth Probabilities in Norway. Population and Development Review. ISSN 0098-7921. 52(2), p. 650–650. doi: 10.1111/padr.70066. Full text in Research Archive Show summary
Abstract We examined how several health variables were associated with mothers’ probability of having a second or third child and the timing of the second birth. Data from the Norwegian Mother, Father and Child Cohort Study (MoBa) and linked registers were used. Parents with certain potentially chronic health problems before or in pregnancy were less likely than others to have an additional (especially a second) child, net of differences in relationship status, education, and economic resources. Some pregnancy‐related and largely temporary health problems during the most recent previous pregnancy and adverse birth experiences were associated with reduced probabilities of a second or third birth, or both, as was generally poor physical health six months after birth and postnatal depression. The latter health indicators may reflect health implications of pregnancy, delivery, and having a relatively “demanding” young child, as well as underlying health problems. Only a few estimates suggested associations between indicators of potentially chronic health problems of the most recently born child and subsequent fertility, but fertility was reduced among one‐child mothers whose first child had colic and among one‐ and two‐child mothers whose most recently born child slept little. Possible reasons for and implications of these health–fertility associations are discussed.
Holm, Martin B.; Jamilov, Rustam; Jasinski, Marek Aleksandr & Nenov, Plamen (2026). Tax News and Intertemporal Substitution. CEPR Discussion Paper Series. doi: https:/cepr.org/publications/dp21523. Full text in Research Archive
Datsenko, Ruslana & Holm, Martin B. (2026). Monetary Transmission to Firm-level R&D. CEPR Discussion Paper Series. doi: https:/cepr.org/publications/dp21852. Full text in Research Archive
Holm, Martin B. & Holtsmark, Katinka Kristine (2026). The Power of Substitution: Short-run Electricity Demand Flexibility. CEPR Discussion Paper Series. doi: https:/cepr.org/publications/dp21859. Full text in Research Archive
Schreiner, Ragnhild Camilla (2026). Uføretrygd i arv: korrelasjon, smitte og politikk. In Ekelund, Ragnhild; Kalstø, Åshild Male & Nicolaisen, Heidi (Ed.), Velferdsstaten og 20 år med Nav. Scandinavian University Press (SCUP). ISSN 9788294167265. p. 10–10. doi: 10.18261/9788294167272-26-16. Full text in Research Archive
Harstad, Bård & Holtsmark, Katinka Kristine (2026). The Gas Trap: Outcompeting Coal versus Renewables. Journal of Political Economy. ISSN 0022-3808. 134(7), p. 2214–2214. doi: 10.1086/740172. Full text in Research Archive Show summary
We analyze a fundamental dilemma and time-inconsistency problem facing a climate coalition producing natural gas. In the short term, it is tempting to produce more to outcompete coal. When this policy is anticipated, however, investments in renewables fall and emissions may ultimately increase. When the coalition cannot pre-commit, its policies can be counterproductive. A calibrated version applied to Europe verifies that the gas trap is quantitatively important. We discuss the robustness of this result and possible solutions, ranging from direct investments in renewables to taxes on the search and exploration of gas.
Kotsadam, Andreas; Rasmussen, Magnus Bergli; Moene, Karl Ove; Kjelsrud, Anders & Gjerløw, Haakon (2026). Strikes and Machines: Investments in the Norwegian Inter-War-Period. Journal of Economic History. ISSN 0022-0507. 86(2), p. 427–459. doi: 10.1017/S0022050725101095. Full text in Research Archive

View all works in NVA

Bhuller, Manudeep; Strøm, Marte & Wentzel, Mirjam (2026). Cushioning a Downturn: Labor Unions and the External Labor Margin. Full text in Research Archive
Wentzel, Mirjam; Bhuller, Manudeep & Strøm, Marte (2026). Cushioning a Downturn: Labor Unions and the External Labor Margin. Full text in Research Archive
Wallossek, Luisa (2026). Lost in Deduction: Taxpayers' Mistakes when Itemizing. Full text in Research Archive
Wallossek, Luisa (2026). Lost in Deduction: Taxpayers' Mistakes when Itemizing. Full text in Research Archive
Wallossek, Luisa (2026). Lost in Deduction: Taxpayers' Mistakes when Itemizing. Full text in Research Archive
Kravdal, Øystein (2026). Barn har blitt en klassemarkør. [Journal]. Dagsavisen. Full text in Research Archive
Nymoen, Ragnar & Castle, Jennifer (2026). The Nominal Exchange Rate Regression Reconsidered. Full text in Research Archive
Kravdal, Øystein & Hart, Rannveig Kaldager (2026). Health around the time of first birth matters for future fertility. Full text in Research Archive
Kravdal, Øystein (2026). Fødselstallene har lenge gått ned. Nå har forskere sett på hva som skjedde i 2007. [Journal]. Aftenposten. Full text in Research Archive Show summary
To nye forskningsartikler spør om smarttelefonen har fungert som prevensjonsmiddel.
Berg, Kristoffer (2026). Capital Gains Tax Deferral: Norwegian Evidence and Reform Simulations. Full text in Research Archive
Berg, Kristoffer (2026). Taxing Deferred and Mobile Capital Gains. Full text in Research Archive
Harmenberg, Karl; Olsson, Maria Christina & Öberg, Erik (2026). Firm Market Power, Wage Rigidity and Demand-Determined Business Cycles. Full text in Research Archive
Andresen, Martin Eckhoff (2026). Monthly Earnings Volatility and Household Pooling. Full text in Research Archive Show summary
This paper examines monthly earnings volatility and its transmission to household earnings volatility using Norwegian data on the universe of monthly pay histories. We document substantial month-to-month earnings changes: within a job, while over one-quarter of months have no earnings changes, another quarter have at least a 23% change. Accounting for multiple jobs and non-employment increases volatility, while aggregating to households reduces volatility by 9-38%. Analyses around job loss and couple formation, along with decomposition and bounding exercises, show that most of this decline reflects pooling effects rather than sorting or labor supply responses.
Andresen, Martin Eckhoff (2026). Monthly Earnings Volatility and Household Pooling. Full text in Research Archive
Kotsadam, Andreas (2026). Fear of crime. Full text in Research Archive
Kotsadam, Andreas (2026). Authoritarianism. Full text in Research Archive
Kotsadam, Andreas (2026). Penalty tax Skatteforum. Full text in Research Archive
Kotsadam, Andreas (2026). Strikes and welfare state development. Full text in Research Archive
Kotsadam, Andreas (2026). Refugee integration and voter turnout. Full text in Research Archive
Kotsadam, Andreas (2026). Refugee integration and jobs. Full text in Research Archive
Finseraas, Henning; Kotsadam, Andreas & Hauge, Karen Evelyn (2026). Nonresponse Bias and Turnout Misreporting in a Refugee Survey: Evidence from Randomized Participation Triggers. Full text in Research Archive
Nyborg, Karine (2026). Why morally motivated public good provision can lead to polarization and minimal contributions. Full text in Research Archive
Nyborg, Karine (2026). Balancing research with teaching & administrative tasks. Full text in Research Archive
Nyborg, Karine (2026). Economic Behavioral Economics of Climate Change. Full text in Research Archive
Nyborg, Karine (2026). Unwillingly Informed: the Prosocial Impact of Third-Party Informers. Full text in Research Archive
Holm, Martin B. (2026). Brennhett om renta. [TV]. NRK. Full text in Research Archive
Holm, Martin B. & Jasinski, Marek Aleksandr (2026). Lønnsvekst på lånt tid. Full text in Research Archive
Holm, Martin B. & Jasinski, Marek Aleksandr (2026). På tide med åpenhet om frontfaget. Full text in Research Archive
Holm, Martin B. & Jasinski, Marek Aleksandr (2026). Høy rente, lønnsvekst og prisvekst kunne vært unngått. Full text in Research Archive
Holm, Martin B. & Jasinski, Marek Aleksandr (2026). Når skal LO akseptere at Norge faktisk har et inflasjonsmål? | DN. Full text in Research Archive
Holm, Martin B. & Storesletten, Kjetil (2026). Vil høyere rente egentlig gi lavere inflasjon? Full text in Research Archive
Holm, Martin B. (2026). Hvorfor er offentlig sektor så stor og dyr i Norge? | DN. Full text in Research Archive
Nymoen, Ragnar (2026). {Hovedkurs, produktivitet og lønn i historisk perspektiv. Full text in Research Archive
Harmenberg, Karl; Öberg, Erik & Olsson, Maria Christina (2026). Wages and Hours Worked: Micro-Level Dynamics and Macroeconomic Implications. Full text in Research Archive
Sparrman, Eva Victoria (2026). Frontfagsmodellen: Fast anker, fleksibel praksis. Full text in Research Archive
Lind, Jo Thori (2026). KI og arbeidsmarkedet. Full text in Research Archive
Lind, Jo Thori (2026). En god KI-modell er ingen gullgruve | DN. Full text in Research Archive Show summary
De smarteste modellene kan vise seg vanskelige å tjene penger på. Det bør bekymre amerikanske teknologiselskaper og endre synet vårt på KI-kappløpet.
Prydz, Espen Beer; Lind, Jo Thori & Moene, Karl Ove (Kalle) (2026). Poverty Amidst Affluence. Measuring Miserliness in Rich and Poor Countries. Full text in Research Archive
Schreiner, Ragnhild Camilla (2026). Who Bears the Burden of Payroll Taxes at the Top? Full text in Research Archive
Schreiner, Ragnhild Camilla (2026). Who Bears the Burden of Payroll Taxes at the Top? Full text in Research Archive
Schreiner, Ragnhild Camilla (2026). Slik sløser Norge | DN. Full text in Research Archive Show summary
Handlekraft uten driftsevne gir hverken fremkommelighet på vei og bane eller kampkraft i forsvaret.
Schreiner, Ragnhild Camilla & Moum, Knut (2026). Å få flere i jobb krever mer målrettet styring | DN. Full text in Research Archive Show summary
Uten klar ansvarsplassering og systematisk måling av tiltak risikerer vi at arbeidslinjen forblir et ønske, ikke et faktisk spor.
Schreiner, Ragnhild Camilla (2026). Tall uten tillit blir bare støy | DN. Full text in Research Archive Show summary
Norge styres på tall. Dersom tilliten til tallene svekkes, kan kostnadene bli store.
Schreiner, Ragnhild Camilla (2026). Staten kan bruke mindre, men likevel bruke pengene feil | DN. Full text in Research Archive Show summary
Ansvarlig økonomisk styring handler ikke bare om årets bruk av oljepenger. Det handler også om hvilken rolle staten gradvis påtar seg.
Schreiner, Ragnhild Camilla (2026). Utformingen av skattesystemet er ikke uten betydning | DN. Full text in Research Archive Show summary
DN har rett i at Norge trenger en bred debatt om offentlig pengebruk og størrelsen på offentlig sektor. Men det gjør ikke spørsmålet om hvordan skattesystemet bør utformes mindre viktig.
Schreiner, Ragnhild Camilla (2026). Pensjonsrettigheter bør neppe skattlegges som formue | DN. Full text in Research Archive Show summary
Pensjonsrettigheter har økonomisk verdi, men bør ikke uten videre behandles som annen formue i skattesystemet. Og i alle fall ikke folketrygden, som utgjør størstedelen.
Nymoen, Ragnar (2026). Samspillet mellom lønn og priser. Full text in Research Archive
Kordt, Amund H.; Nymoen, Ragnar & Holm, Martin B. (2026). Essays on Fiscal Policy. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis examines how fiscal policy shapes labour supply, local economic activity and aggregate fluctuations in Norway. It brings together three empirical papers that analyse the effects of income taxation, intergovernmental transfers and pandemic containment policies using rich administrative and national accounts data and dynamic econometric methods. The first paper studies long-run responses of labour income and taxable income to permanent changes in the income tax schedule. Exploiting a series of legislated tax reforms between 1994 and 2019 and using a local projection IV framework, it estimates dynamic elasticities and decomposes them into income and substitution effects. The results point to a sizeable negative long-run response of taxable income to an increase in net-of-tax income, driven by strong income effects and modest substitution effects, with substantial heterogeneity across the consumption distribution. The findings are consistent with non-homothetic preferences of the Boppart–Krusell-type and have direct implications for optimal income tax design. The second paper investigates local permanent fiscal multipliers. Using variation in permanent central-government transfers to municipalities, combined with municipal accounts and firm-level balance-sheet data, it estimates the long-run impact of transfers on local value added and employment. On average, one additional krone of transfers raises local value added by more than one krone, but multipliers are much larger in municipalities where households spend a high share of their consumption locally, highlighting the importance of import leakages and local market structure. The third paper develops a daily econometric model of the joint evolution of public health and economic activity during the COVID-19 pandemic. By combining “dailylised” GDP with high-frequency health indicators, policy measures and an awareness proxy, it shows that both containment policies and voluntary behavioural responses had large but time-varying effects on economic activity, with the economic costs declining as agents adapted and vaccination reduced disease severity. Together, the three papers provide new evidence on how taxes, transfers and health policies propagate through the economy in normal times and crises.
Gaarder, Ingvil Benterud; Grindaker, Morten Haabeth; Meling, Tom G.; Mogstad, Magne & Mogstad, Ingvil Gaarder, Morten Grindaker, Tom G. Meling, Magne (2026). Green Waste. National Bureau of Economic Research. doi: 10.3386/w34649. Full text in Research Archive
Deshpande, Manasi; Kellogg, Maxwell Dylan; Mogstad, Magne & Tseng, Kuan-Ju (2026). Explaining the Historical Rise and Recent Decline in Social Security Disability Insurance Enrollment. National Bureau of Economic Research. doi: 10.3386/w35300. Full text in Research Archive
Harildstad, Anders; Nymoen, Ragnar & Slettebø, Olav Haldorsson (2026). From Manufacturing Productivity to National Wages: The Main-Course Variable and Norwegian Wage Formation, 1900–2025. University of Oslo Department of Economics. doi: https:/www.sv.uio.no/econ/english/research/publications/memoranda/2026/from-manufacturing-productivity-to-national-wages. Full text in Research Archive
Bucher-Koenen, Tabea; Wallossek, Luisa & Winter, Joachim (2026). Opt-in or Opt-out? The Power of Defaults in Pension Enrollment Choices. CESifo network. doi: 10.65864/lhwgnu4zpv. Full text in Research Archive Show summary
Default settings strongly increase pension enrollment, especially when savings incentives are high and choices are complex. We show that the effect is weaker when incentives are low, options are simple, and opting out is easy. We study the nationwide introduction of auto-enrollment for low-income employees in Germany's public pay-as-you-go pension system. We find that automatic enrollment raises participation by 23 percentage points, though most individuals actively opt out. Linking administrative and survey data shows that the default effect is stronger when enrollment incentives are higher and among individuals who lack knowledge of their enrollment status.
Markussen, Simen; Røed, Knut & Holden, Steinar (2026). Are Senior Workers Overpriced? Evidence from an Age-Differentiated Payroll Tax in Norway. IZA Institute of Labor Economics, IZA DP no 18727. doi: https:/www.iza.org/publications/dp/18727/are-senior-workers-overpriced-evidence-from-an-age-differentiated-payroll-tax-in-norway. Full text in Research Archive
Erraia, Jonas & Sevaldsen, Even André Zaborowski (2026). Temperature Dynamics and Respiratory Hospital Admissions. A Time Series Analysis of Oslo and Bergen. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis examines the association between temperature and respiratory hospital admissions in Oslo and Bergen, using weekly data from 2010 to 2024. Using an ADL model to estimate short-term temperature-admission relationships while controlling for persistence, pollution, and seasonality, the analysis finds that moderate cold temperatures (-5°C to 0°C) are associated with increased admissions for younger populations. Hot days above 20°C are associated with increased admissions for the 65+ group in Bergen, but with decreased admissions among younger groups, particularly in Oslo. A pollution index is constructed using NO₂ and PM₁₀, and an increase in the index is associated with increased admissions for children and young adults. The results are compared with a fixed-effects Poisson model, which provides a complementary specification and shows similar patterns for the main heat responses, while cold estimates are more sensitive to model specification. The findings suggest that temperature and air quality are associated with respiratory hospital admissions, with heterogeneous patterns across age groups and municipalities that may partly reflect differences in temperature distributions across cities.
Thoresen, Thor Olav; Vattø, Trine Engh & Zachrisson, Henrik Daae (2026). Reducing child poverty through cash or care. Statistisk sentralbyrå (SSB). doi: https:/www.ssb.no/inntekt-og-forbruk/inntekt-og-formue/artikler/skal-en-redusere-barnefattigdom-med-okt-barnetrygd-eller-gratis-barnehage. Full text in Research Archive Show summary
De siste tiårenes økning i barnefattigdom har fått betydelig oppmerksomhet. I denne analysen retter vi søkelyset mot hvilke tiltak som er mest effektive i fattigdomsbekjempelsen for familier med barn i alderen 1–5 år – økte kontantytelser, som høyere barnetrygd, eller tiltak som gratis barnehage.
Graber, Michael; Håvarstein, Morten; Mogstad, Magne; Torsvik, Gaute & Vestad, Ola L. (2026). Substitution and Income Effects of Labor Income Taxation. National Bureau of Economic Research. doi: 10.3386/w34987. Full text in Research Archive
Kellogg, Maxwell & Bjørneboe, Dagny Krøger (2026). Strategic changes in labour income before birth. Analysing the effects of the parental benefit scheme in Norway on labour income. Universitetet i Oslo. Full text in Research Archive Show summary
In Norway, parental benefits depend on previous labour income. I have studied the effect of the incentives in the parental benefit scheme on labour income. In theory, the policy should allow for an estimation of the Frisch elasticity. However, due to data limitations and the implementation of the policy, this could not be done in practice. The effect of the incentives could also be an alternative explanation for the child penalty, but this was not confirmed. With administrative data, I have used descriptive and empirical methods to analyse the effect of the incentives for mothers. The results are ambiguous. The descriptive analysis shows that mothers who work increased their labour income and work hours right before birth. In contrast, both analyses show that in general mothers worked less close to birth. Nonetheless, the identifying assumption of parallel trends does not hold which imply that the empirical results are biased. In general, all results should be interpreted with caution due to the lack of information on the data.
Drange, Nina & Munkvold, Rikke (2026). Effekter av statlig overtakelse på institusjonsplasseringer i barnevernet: En Difference-in-Differences-analyse av forvaltningsreformen fra 2004. Universitetet i Oslo. Full text in Research Archive Show summary
I denne oppgaven undersøker jeg om forvaltningsreformen i barnevernet fra 2004 hadde effekt på sannsynligheten for institusjonsplassering for barn i kommuner som ble om fattet av reformen, sammenlignet med kommuner som sto utenfor. Effekten estimeres ved hjelp av en Difference-in-Differences-modell (DiD) som utnytter den geografiske variasjonen i reformens implementering. Analysene viser en statistisk signifikant reduksjon i sannsynligheten for institusjonsplassering i behandlingsgruppen etter reformen, relativt til kontrollgruppen. For å teste mekanismene bak dette funnet undersøker jeg videre om reduksjonen drives av en generell nedgang i sannsynligheten for å motta tiltak fra barnevernet, eller om reformen endret tiltakspraksis for barn som allerede var i kontakt med barnevernet. Resultatene indikerer at nedgangen primært drives av at færre barn kommer i kontakt med barnevernet etter reformen, mens det ikke finnes statistisk signifikante holdepunkter for at reformen endret sannsynligheten for institusjonsplassering blant barn som allerede mottok tiltak.
Traeger, Christian & Karlsen, Jørgen Næss (2026). Exhausted - In this climate-economy?. Oil market Shocks, Climate Policy Uncertainty, and Stranded Assets. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis studies the intertemporal dynamics of oil extraction and their implications for climate policy. Using asset-level cost data, I construct Hotelling rent series that separate scarcity rents from the return on invested capital. I find no empirical support for the Hotelling rule: rents are well described by a random walk, driven by oil price shocks rather than growing at the rate of interest, and forward-looking breakeven rents are frequently negative for higher-cost producers. I then examine how producers respond to shocks that affect expectations about future market conditions. While OPEC supply shocks raise drilling and investment uniformly, climate policy uncertainty induces a reallocation from long-cycle offshore assets toward short-cycle shale, consistent with a shift toward shorter investment horizons rather than reduced extraction. Finally, I show that commercially viable reserves exceed oil-allocated carbon budgets for 1.5°C and 2°C by a large multiple under all price scenarios. Despite this, there is no evidence that markets price differential stranding risk across supply segments. These findings suggest that producer behaviour is driven by short-term considerations, implying that credible, time-bound policy is required to align market outcomes with limits on global warming.
Cao, Jin & Rønvåg, Lukas Helland (2026). The A–H Share Premium and Global Uncertainty in the Stock Connect Era. Evidence from Dual-Listed Chinese Firms. Universitetet i Oslo. Full text in Research Archive Show summary
This paper studies the sources of time-series variation in the A–H share premium during the Stock Connect era. Using a monthly panel of 152 dual-listed Chinese firms from December 2014 to January 2026, the premium is regressed on a set of global and China-specific uncertainty measures in firm fixed effects specifications. Two dimensions of global risk show the most robust associations: greater uncertainty in world trade is associated with a lower premium, while greater stress in the global financial market (VIX) is associated with a higher premium. Market-side decomposition suggests stronger H-share responses in the point estimates, though A–H differences are not statistically significant. The premium’s association with these measures changes over time at the monthly frequency studied here, with financial market stress more prominent in the early Stock Connect years and trade uncertainty more prominent in the later part of the sample. The trade uncertainty channel operates broadly across industries and is amplified for firms with less liquid and more volatile H-share markets, while the financial stress channel is concentrated in Financials and Energy. The findings add a time-varying external dimension to the literature on the A–H premium and show that these relationships have changed significantly over the course of the Stock Connect era.
Holden, Steinar & Rygh, Lavrans Stene (2026). Generative AI Exposure and Labour-Market Adjustment A Difference-in-Differences Analysis of Wage Income and Employment Using Norwegian Register Data. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis examines whether Norwegian workers initially employed in occupations with high complementarity-adjusted AI occupational exposure experienced different labour-market trajectories after the public release of ChatGPT in November 2022. The analysis focuses on annual wage income, exit from employment and occupational switching. I use Norwegian administrative register data accessed through microdata.no and compare workers initially employed in the top and bottom deciles of the complementarity-adjusted occupational AI exposure index, C-AIOE, developed by Pizzinelli et al. (2023). Treatment status is assigned using each worker’s four-digit STYRK-08 occupation in November 2021. The main wage specification is an individual fixed-effects DiD model covering 2021-2024, with 2023 and 2024 defined as the post period. Standard errors are clustered at the 2021 four digit occupation level. I also estimate an event-study specification covering 2019-2024, using 2021 as the reference year. Workers initially employed in the top decile of the C-AIOE distribution recorded stronger annual wage-income growth after 2022 than workers initially employed in the bottom decile. The estimated differential ranges from approximately 1.9 to 3.4 log points in the stayer sub sample and from 5.1 to 6.1 log points in the full sample. Event-study and placebo results indicate, however, that the wage differential was already widening before the main post treatment years, weakening the parallel-trends assumption. Exit from employment differed only modestly between the two groups. Occupational switching was substantially higher among workers initially employed in the top decile in the fixed-2021 transition sample, and descriptive occupation-year evidence confirms that this difference was already visible before 2022. The results should therefore not be interpreted as causal evidence that generative AI increased wage growth or occupational mobility. The estimates are instead interpreted as suggestive reduced-form differences between workers initially employed in occupations with different levels of complementarity-adjusted AI exposure.
Stipanicic, Fernando & Svendsrud, Erlend (2026). Public R&D Support, Business R&D Spending and Patenting: Evidence from OECD Country Panels. Universitetet i Oslo. Full text in Research Archive
Lind, Jo Thori; Mathiassen, Astrid; Nesheim, Ida & Haakestad-Bjaaland, Pernille (2026). Staples in Unstable Times. A study of staple food security in relation to shocks in the multi-shock context of Somalia. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis examines how exposure to increasingly common shocks in the Somali food system is associated with the share of staple foods in household diets, and whether these relationships vary across household types and regions. Staple foods as cereals, roots and tubers often constitute the main source of calories in low-income societies due to their relative cheapness and high calorie density (Bennett, 1941; Grigg, 1996). Moving beyond this economic interpretation of the importance of staples in food security, this thesis highlights how social and physical access to food also shape staple consumption patterns. Despite decades of recurring shocks and food crises, empirical evidence on household consumption behavior in Somalia remains limited (Hussein et al., 2021). By utilizing the first detailed household consumption data for Somali households since 1985, collected as cross-section data by the Somalia National Bureau of Statistics in 2022 (SNBS, 2022), we seek to investigate this research gap. We estimate a set of Ordinary Least Squares (OLS) models to examine the relationship between the two most reported shocks among Somali households in 2022, which were “Large rise in price for food” and “Drought or severe water shortage”, and the share of staples in their diet. Additionally, we support the analysis with a set of Linear Probability Models (LPMs) to asses whether shock exposure varies systematically across household types and regions. The results show that both shocks are associated with a small but highly significant increase in staple shares. However, these effects are not robust across specifications. Instead, staple consumption patterns appears to be largely shaped by long-term regional conditions rather than short-term shocks. The LPMs confirm that the shocks are not randomly distributed across regions, and are positively related. Taken together, the findings suggest that the two relatively short term shocks lead to only limited adjustments in an already staple-heavy diet. Instead, staple consumption patterns in Somalia appear to reflect the context of recurring shocks and instability over a long period of time. Overall, this underscores that the effects of shocks in the food system in such complex contexts must be understood within a broader set of economic, social and physical constraints on food access.
Akbari, Danial Ali & Heimvik, Dag (2026). Financing Norway’s Rearmament. Tax policy and Increased Public Defence Spending: Comparing Tax Instruments in Financing Norway’s Rearmament. Universitetet i Oslo. Full text in Research Archive Show summary
The backdrop for this this study is the political decisions from the NATO summit in the Hague July 2025 for all allies to step up their defence investments to finance the critical needs for developing more military capabilities and increase the stockpiles of munitions and spare parts. The new security political situation, the Russian war on Ukraine and the threats posed by development of new technology are among the factors fuelling the new requirements. This study assesses the different policy implications of funding the Hague defence investments pledge of 5% of GDP to core defence spending (3,5% of GDP) and defence and security related spending (1,5% of GDP) through taxation focusing on labour income, capital income and a mix between the two. This thesis focuses on how these tax instruments can be leveraged as a means to implement the defence spending target and the impact on Norwegian households. To address this question, I have developed a heterogenous agent model for Norwegian households with a two-stage income process. The model is set in continuous time and captures behavioural responses such as consumption, savings and labour supply decisions in general equilibrium. As part of the analysis, I also assess what impact the different tax instruments can have on distributional and aggregate outcomes in the general Norwegian economy. The results from this analysis show that the different tax instruments are associated with opposing benefits. The strengths of the labour tax instruments include increased aggregate outputs and stimulates to a higher labour supply, but at the same time increases inequality and aggregate welfare. The capital income tax instrument reduces inequality and can fund important parts of the welfare state. On the other hand, it can hinder the stimulation of economic activity in society. The mixed tax instrument generally provides intermediate outcomes, combining positive and negative implications from both tax instruments. The capital income tax instrument reduces inequality and improves welfare but lacks in terms of stimulating to economic activity. The mixed tax instrument generally provides intermediate outcomes, combining positive and negative implications from both tax instruments. Overall, the results from this study call attention to the trade-offs between the overall goals of a tax policy, such as economic efficiency, distortions, behavioural incentives and redistribution. A policy makers choice in tax policy consequently depends on what aspects of the respective tax instruments they want to emphasise.
Mehlum, Halvor & Yemane, Joel (2026). Norway’s Exposure to EU Safeguard Measures. An Analysis of Trade Diversion and Sectoral Vulnerability in Norwegian Exports. Universitetet i Oslo. Full text in Research Archive
Hammersmark, Eivind Moe; Casti, Carola & Brastad, Maja Kirkeby (2026). Aid volatility and education outcomes. Assessing the long-term consequences of declining trends in Official Development Assistance. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis assesses the relationship between aid volatility on the quantity and quality of education in Uganda. OLS regressions were conducted using data from The Demographic Health Survey and AidData. The thesis is a two-part analysis, first assessing the correlation between aid volatility and education outcomes, followed by the associated relationship between mother’s education and child mortality. Aid volatility is made up of total aid disbursements and education sector disbursements for locally geocoded/district-assignable aid. Correlations between both aid volatility and education, and education and mortality were found in some model specifications. However, the results are not robust and may therefore not reflect the true relationship. Further research should therefor focus on assessing different datasets to develop greater understanding of the actual relationship between aid volatility and education outcomes.
Kellogg, Maxwell & Lee, Woohyun (2026). Job loss and fertility decisions. Evidence on age heterogeneity from Norway. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis examines whether the fertility response to job loss differs by the age at the job loss. The analysis is based on Norwegian administrative register data covering 505,740 mothers over the period 1997 to 2015. I apply the dynamic two-way fixed effects event study model and the modern imputation estimator to address biases from staggered treat-ment timing. The results reveal substantial heterogeneity by age. Mothers who lose their first job at earlier life cycle stage, experience a positive and monotonically increasing fertility response. In contrast, mothers who lose their first job at later life cycle stage experience an initially small positive response that turns negative from approximately four years after the job loss. These results are consistent with the theoretical prediction that the substitution ef-fect dominates for mothers in an earlier life cycle stage, while the income effect dominates for mothers in a later life cycle stage. At an earlier stage, the income shock from job loss is modest relative to total income, so the reduction in the opportunity cost of childbearing dominates and fertility increases. At a later stage, higher accumulated wages make the in-come shock more severe, so the income loss dominates and fertility decreases. The findings suggest that mixed results in the existing literature on job loss and fertility can arise, and highlight the importance of age heterogeneity and use of imputation estimator in settings with staggered treatment timing.
Hansen, Jonas Slaathaug & Sjøblom, Ellen Elisabeth Nordqvist (2026). Beyond Aggregate Trade Effects: A Sectoral Gravity Analysis of European Integration. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis examines how European integration has affected bilateral trade flows, focusing on EUmembership and the adoption of the euro as a common currency. Using industry-level trade data from the International Trade and Production Database for Estimation, covering up to 240 countries over the period 1991–2019, the analysis applies a structural gravity framework estimated with Poisson Pseudo Maximum Likelihood and a comprehensive set of fixed effects. Moving beyond aggregate trade effects, the thesis examines whether the effects of European integration differ across broad sectors. The results suggest that joint EU membership is associated with higher bilateral trade over all, while the aggregate estimates are sensitive to how broader European integration trends are modeled. In the preferred aggregate specification, the EU level coefficient implies ap proximately 15 percent higher bilateral trade, while the inclusion of an EU-specific time trend makes the implied effect vary over time. The additional effect of joint euro adoption is less robust. Although the aggregate specification yields a negative euro coefficient, this effect becomes small and statistically insignificant when the model is estimated across broad sectors. The results therefore provide little evidence that euro adoption generated additional trade gains beyond those associated with EU membership and broader integration dynam ics. The sector-level results reveal substantial heterogeneity, with the estimated EU effect being positive and statistically significant in manufacturing, agriculture, and services, but weak and insignificant in mining and energy. This suggests that aggregate estimates conceal important sectoral differences in how European integration affects trade. Overall, the findings highlight the empirical difficulty of identifying a separate euro effect, as adoption of the common currency is closely intertwined with EU membership and broader processes of integration
Carnap, Tillmann von & Amundsen, Nora (2026). Shedding light on intimate partner violence: Evidence from solar panel adoption in Kenya. Is there an association between solar panel adoption and intimate partner violence among Kenyan households? Universitetet i Oslo. Full text in Research Archive Show summary
This thesis examines the relationship between solar energy adoption and intimate partner violence in Kenya using data from the 2022 Kenya Demographic and Health Survey. To address the non-random adoption of solar panels, exogenous variation in solar radiation is constructed using Downward Shortwave Radiation from NASA satellite data. The results show that solar adoption alone is not systematically associated with reduced intimate partner violence. However, solar-households exposed to above average solar radiation experience a lower probability of violence, with the most robust effect observed for sexual violence. This suggests that energy access may reduce economic stress or influence household dynamics. The effects are stronger in rural areas and among lower socioeconomic households. Overall, the findings indicate that the benefits of solar energy may extend beyond economic outcomes to include improvements in women's safety and household welfare.
Stipanicic., Fernando; Vivetharan., Vithushan & Mandla., Eesha Kaur (2026). Trade, Carbon Pricing & Welfare. How do trade flows, gains from trade and welfare change when pollution externalities are priced, compared to a baseline where they are ignored? Universitetet i Oslo. Full text in Research Archive Show summary
International trade models typically leave out environmental externalities, potentially overstating the gains from trade. This thesis extends the (Eaton & Kortum, 2002) Ricardian trade model to include carbon externalities, calibrated across 141 countries and 11 sectors using 2009 as the base year. The model structure follows (Artuc & Sommer, 2024). We compare two counterfactual equilibria: one where pollution is unpriced (CF1) and one where observed 2009 environmental tax rates are imposed (CF2), and analyze how trade flows, welfare and emissions change when pollution externalities are priced. The 2009 taxes are concentrated among European economies and mainly fall on pollution from production. Most countries in the sample face no tax at all. All calculations are performed in R (R Core Team, 2025). We find a positive productivity-emission correlation of r = 0,722 across 2,018 country-sector observations, meaning the cheapest exporters in 2009 were on average also the most emissions intensive. When environmental tax rates are imposed, 103 of 141 countries gain in welfare terms and 36 lose, but the losses are concentrated and large. The median welfare change is +0,28%, while the mean is –1,68%, driven by a small number of heavily trade exposed European economies facing much higher import prices. The price channel dominates over lost export income. Global emissions fall by 6,79%, with a leakage rate of 0,24%, treated as a lower bound, as the model only captures the goods trade channel. When pollution enters the utility function directly, the welfare ranking is nearly identical to the standard real income measure at (Nordhaus, 2017a) weights, with a mean gap of only 0,04 percentage points. The results suggest that part of the gains from trade in 2009 came from pollution costs that producers were not paying, and that pricing these externalities unilaterally shifts costs rather than eliminating them.
Nymoen, Ragnar & Arian, Farhad (2026). Forecasting Norwegian Inflation Using Traditional Econometric and Machine Learning Models. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis evaluates 17 classical econometric and machine learning models for forecasting Norwegian core inflation (CPI-ATE) at horizons of one, three, six, and twelve months, using a pseudo-out-of-sample framework with an evaluation period from June 2013 through December 2025. The central finding is a horizon-dependent pattern in relative forecast performance. At the one-month horizon, the naive persistence benchmark achieves the lowest RMSE and significantly outperforms all alternatives. At three months, traditional econometric and machine learning models are statistically indistinguishable. From six months onwards, the Gradient Boosting Machine dominates, reducing RMSE by 23 percent at h = 6 and 38 percent at h = 12 relative to the best traditional model at each horizon. These advantages are statistically significant under HLN-corrected Diebold–Mariano tests and robust across loss functions, sub-periods split at the January 2022 structural break, and an alternative information set using lagged predictors. A feature importance decomposition reveals that the source of forecastable variation shifts from autoregressive persistence at short horizons to structural macroeconomic fundamentals, particularly the unemployment gap and oil prices, at longer horizons, providing a coherent economic explanation for the horizon-dependent pattern. The results suggest that machine learning methods, specifically gradient boosting, can meaningfully improve inflation forecasts at the policy-relevant horizons over which monetary transmission operates.
Fehr., Nils-Henrik Mørch von der & Giskås., Ole (2026). Signalisering i det norske bankmarkedet. En markedsanalyse om signalisering kan bidra til svekket konkurranse gjennom koordinert atferd i det norske bankmarkedet. Universitetet i Oslo. Full text in Research Archive Show summary
Signalisering av boliglånsrenten kan svekke konkurranseforholdene i det norske bankmarkedet gjennom fasilitering av koordinert atferd. Signalisering av boliglånsrenten, i for eksempel media, kan svekke konkurransen da usikkerheten om fremtidige prisbeslutninger reduseres, i ytterste konsekvens kan det medføre dårligere og dyrere banktilbud (Aandal-Nilsen, 2021). En grunnleggende forutsetning for fungerende konkurranse i et marked er at aktørene ikke vet hva deres konkurrenter gjør i fremtiden, og på et selvstendig grunnlag setter priser på sine varer og tjenester (Aandal-Nilsen, 2021). Dersom aktører i et marked har tilgang på informasjon om hvordan konkurrenter setter sine priser i fremtiden, vil den grunnleggende mekanismen for effektiv konkurranse falle bort. I oppgaven introduseres insentiv- og koordineringsproblemene som vurderingsgrunnlag for koordinert atferd. Teorien sees i sammenheng med markedsforholdene i det norske bankmarkedet. Signalisering vurderes ytterligere som en mekanisme for å fasilitere insentiv- og koordineringsproblemene. Avslutningsvis gis det en vurdering på om konkurransen kan bli svekket som følge av signaliseringen. Analysen finner at bankene trolig kan ha insentiver til å koordinere seg, men at det kan foreligge betydelige koordineringsproblemer. Signalisering om fremtidig boliglånsrente kan være med på å bedre noen av koordineringsproblemene, men ikke eliminere de helt. Dermed kan det ikke med sikkerhet fastslås at signalisering svekker konkurransen i det norske bankmarkedet.
Gehrcke, Robert; Torsvik, Gaute & Skorpen, Kristian (2026). Distributed Battery Storage in Nordic Power Markets: Private Revenue, Social Value, and the Gap Between Them. Universitetet i Oslo. Full text in Research Archive Show summary
Battery energy storage systems (BESS) deliver services that the power system values: they help stabilize grid frequency, defer transmission and distribution upgrades, support local resilience, and absorb fluctuations from variable renewables. The BESS operator earns revenue only from services the market explicitly prices. In the Nordic context, this means arbitrage on the day-ahead spot market, capacity (and, for some products, activation) payments in the frequency-reserve markets, and bilateral contracts with distribution system operators. This thesis analyses the gap between the social value distributed by BESS and the private revenue an operator can capture. The qualitative analysis groups BESS services by the compensation mechanism that rewards each one. The quantitative analysis uses a linear-programming model that solves the operator's bidding problem hour by hour using realized 2023 Nordic market data and reports the resulting annual revenue stack. Norway is the country of interest; the numerical case study uses Swedish bidding zone SE3 because Sweden's reserve-capacity markets are more established than Norway's. For a 10 MW/50 MWh BESS in SE3, the model, under the model assumptions, estimates annual revenue and operational costs. The quantitative result should be read as a realized-2023 SE3 capacity price and arbitrage backtest: activation settlement is excluded because product-specific delivered activation-energy data is not available in the same public format. Additionally, the optimization model's parameters have dummy values. Capacity reservation in FCR-D Up and FCR-D Down dominates the revenue stack at the realized 2023 price levels, with aFRR, mFRR, FCR-N, and arbitrage filling the rest. A design-parameter sensitivity sweep at constant CapEx shows that rated power and energy capacity behave as complements within the modeled revenue stack: when rated power is low, the inverter rating bounds the revenue, and when energy capacity is low, the energy-reservation constraints bind. In the case-study design, the system is power-bound, so an extra inverter rating produces a much larger revenue response than extra battery capacity; whether that translates into a higher-value investment depends on the relative cost of adding power versus energy in practice, which is outside the constant-CapEx setup. The qualitative and the quantitative threads agree on the policy implication. The present Nordic market design rewards distributed BESS for a narrow set of power-bound frequency services. Most of the long-run system value (capacity deferral, congestion relief, infrastructure flexibility, resilience) sits outside the priced revenue stack. Without a combination of better tariff design, broader frequency-product markets, and targeted distribution-system-operator contracts, the private market will underinvest in distributed storage relative to the social optimum.
Cao, Jin; Wangen, Erling & Øia, Håvard Myren (2026). Government Loan Guarantees and SME Resilience. Evidence from the Norwegian COVID-19 guarantee scheme. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis evaluates the Norwegian COVID-19 guarantee scheme, which provided liquidity support to small and medium-sized enterprises through government-guaranteed bank loans. We apply a matched difference-in-differences design using administrative data on Norwegian SMEs from 2016 to 2024, combined with loan-level records from Eksfin. The matched sample consists of 2,117 treated firms and 9,395 never-treated controls. Our results show that treated firms had higher survival rates than controls, with the gap peaking at 5.3 percentage points during the most acute phase of the pandemic and remaining at 2.9 percentage points in 2024. They initially became more leveraged, which put temporary pressure on profitability and debt servicing capacity. Treated firms also retained approximately 6% more employees than controls. The estimated effects were concentrated among startup firms, while sector heterogeneity was modest. These results are consistent with the scheme having contributed to firm stabilization during the crisis.
Nilssen, Tore & Hveem, Gabriel Leite (2026). From Fuel Station to Energy Station. A Study of the Exit of Retail Petroleum from Norwegian Fuel Stations in Light of Exit Theory. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis applies the exit theory of Ghemawat and Nalebuff (1985, 1990) to the Norwegian retail petroleum market as the analytical framework. The theory predicts that larger firms face exit pressure before smaller ones as demand declines structurally, because larger capacity commitments generate a higher ongoing cost burden that becomes unsustainable sooner. Cost differences between firms can moderate this pressure, but a cost advantage large enough to reverse the ordering is rare. The capacity cost mechanism is consistent with the cost structure of Norwegian fuel stations, where fixed infrastructure costs persist largely independent of how much fuel is sold. Stations respond to declining demand in a predictable sequence: first by reducing output, then by reducing physical capacity, and finally by exiting the market entirely. The central finding is a paradox. The stations facing the greatest pressure to reduce capacity are also the stations facing the highest barriers to doing so, because the infrastructure driving the cost burden is the same infrastructure that is costly to remove. Beyond this paradox, the analysis identifies four features of the market that the framework cannot accommodate. Complementary revenue from food and convenience retail subsidizes fuel operations for multiproduct stations, buffering the capacity pressure the theory identifies. The option value of station sites as broader energy retail locations is an adaptation the theory cannot capture. Chain affiliation insulates individual stations from the direct financial pressure the theory identifies, though the degree of that insulation varies substantially between chain-operated, franchised, and dealer owned stations. The separability of market exit from station closure means that exiting retail petroleum and closing the physical station are not necessarily the same outcome. These features reveal which stations are best positioned to survive the transition. Stations that cut costs most aggressively are not necessarily those most likely to survive. Instead, those most likely to continue operating are the stations with the business model, the ownership structure, and the local market position to pursue transformation into broader energy retail before the window for doing so closes.
Mehlum., Halvor & Ackar., Dino (2026). Homeownership and Labor Market Outcomes in Norway. Evidence from Three Unemployment Margins. Universitetet i Oslo. Full text in Research Archive Show summary
The main objective of this thesis is to examine the relationship between homeownership and labor market outcomes, with a focus on unemployment risk, duration, and exit probabilities. The analysis is motivated by the Oswald hypothesis, which claims that lower mobility among homeowners could lead to worse labor market outcomes. Using individual level register data spanning from 2015 to 2023, this study estimates linear regression models for three unemployment margins; i) the probability of being unemployed, ii) the duration of unemployment, and iii) the likelihood of exiting unemployment. The results show that, across all margins, homeowners perform better than renters. These findings do not support the Oswald prediction. Instead, they fit better with explanations grounded in selection into homeownership, stronger employment stability, and financial incentives to maintain or regain employment.
Markussen, Simen; Schreiner, Ragnhild Camilla & Lund, Johan Stang (2026). Inntektsskifting under den norske aksjonærmodellen. Aktive eieres valg mellom lønn og utbytte. Universitetet i Oslo. Full text in Research Archive Show summary
Denne oppgaven undersøker om aktive eiere i norske aksjeselskaper tilpasser uttaket av lønn og utbytte som respons på skattemessige insentiver under aksjonærmodellen i perioden 2007–2022. Analysen bygger på registerdata og utnytter geografisk variasjon i arbeidsgiveravgift samt ulik eksponering for regelendringer over tid til å estimere hvordan skattedifferansen mellom lønn og utbytte påvirker eiernes uttaksvalg. Resultatene indikerer skattemotivert inntektsskifting blant aktive eiere. En økning i den marginale skattedifferansen mellom lønn og utbytte på ett prosentpoeng er forbundet med en økning i sannsynligheten for å ta ut utbytte på om lag 1,3--1,8 prosentpoeng. Blant eiere som tar ut utbytte, antyder estimatene at uttaket øker med rundt 3,0--3,8 prosent. Samlet innebærer dette en økning i utbytteandelen på litt over 0,5 prosentpoeng. Samtidig viser analysen at eierne gjennomgående tar ut mer lønn enn det som er skattemessig optimalt. Dette kan skyldes opptjening av rettigheter i folketrygden eller andre friksjoner i eiernes beslutningsprosess. Med utgangspunkt i de empiriske estimatene presenterer jeg et modellbasert anslag på omfanget av intratemporal inntektsskifting i populasjonen. Anslaget er på om lag 700 millioner kroner årlig, tilsvarende drøyt 10 prosent av de totale utbyttene blant eiere med insentiv til å skifte inntekt fra lønn til utbytte. Resultatene tyder dermed på at selv moderate forskjeller i beskatningen av arbeid og kapital kan gi skatteminimerende responser.
Gundersen., Thomas Størdal; Lorentzen., Linnea; Hvidsten., Henrik & Ryberg., Oscar F. H. (2026). Time-Varying Transmission Mechanisms of Monetary Policy and Exchange Rate Shocks. Evidence from a Rolling-Window Bayesian Structural Vector Autoregression Model for Norway. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis investigates whether the transmission of monetary policy shocks and real exchange rate shocks in Norway has changed over time, with particular attention to the 2014–2016 oil price collapse. Using quarterly data for 1990Q3–2019Q4, we estimate a Bayesian Structural Vector Autoregression model identified with sign- and zero-restrictions, and extends the framework through rolling-window estimation to trace time variation in impulse responses. The full-sample results are broadly consistent with earlier Norwegian evidence for monetary policy shocks, but they show more pronounced and persistent domestic effects of exchange rate shocks. The rolling-window analysis indicates stronger time variation after 2014–2016, especially for exchange rate shocks: exchange rate shock persistence rises, the domestic interest rate response becomes markedly weaker, and the pass-through to inflation and output becomes more visible. For monetary policy shocks, the results suggest amplified and more persistent responses in interest rates, inflation, output, and the exchange rate in the later sub-samples, although robustness tests imply that some of this variation may reflect varying shock size rather than changes to the transmission mechanism itself. The results emphasize the need for updated empirical evidence on shock transmission and suggest that Norges Bank has placed greater weight on stabilizing the real economy since the mid-2010s.
Lind, Jo Thori; Mathiassen, Astrid; Nesheim, Ida & Haakestad-Bjaaland, Pernille (2026). Staples in Unstable Times. A study of staple food security in relation to shocks in the multi-shock context of Somalia. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis examines how exposure to increasingly common shocks in the Somali food system is associated with the share of staple foods in household diets, and whether these relationships vary across household types and regions. Staple foods as cereals, roots and tubers often constitute the main source of calories in low-income societies due to their relative cheapness and high calorie density (Bennett, 1941; Grigg, 1996). Moving beyond this economic interpretation of the importance of staples in food security, this thesis highlights how social and physical access to food also shape staple consumption patterns. Despite decades of recurring shocks and food crises, empirical evidence on household consumption behavior in Somalia remains limited (Hussein et al., 2021). By utilizing the first detailed household consumption data for Somali households since 1985, collected as cross-section data by the Somalia National Bureau of Statistics in 2022 (SNBS, 2022), we seek to investigate this research gap. We estimate a set of Ordinary Least Squares (OLS) models to examine the relationship between the two most reported shocks among Somali households in 2022, which were “Large rise in price for food” and “Drought or severe water shortage”, and the share of staples in their diet. Additionally, we support the analysis with a set of Linear Probability Models (LPMs) to asses whether shock exposure varies systematically across household types and regions. The results show that both shocks are associated with a small but highly significant increase in staple shares. However, these effects are not robust across specifications. Instead, staple consumption patterns appears to be largely shaped by long-term regional conditions rather than short-term shocks. The LPMs confirm that the shocks are not randomly distributed across regions, and are positively related. Taken together, the findings suggest that the two relatively short term shocks lead to only limited adjustments in an already staple-heavy diet. Instead, staple consumption patterns in Somalia appear to reflect the context of recurring shocks and instability over a long period of time. Overall, this underscores that the effects of shocks in the food system in such complex contexts must be understood within a broader set of economic, social and physical constraints on food access.
Holm., Martin Blomhoff; Quaghebeur., Ewoud; Gjelsnes., Christopher Hammer & Skinnarmo., Lukas Trømborg (2026). The Macroeconomic Effects from Public Investments. A DSGE-based Analysis for Norway. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis examines the macroeconomic effects of government investment in Norway and how they differ from those of government consumption by using a progressive sequence of Dynamic Stochastic General Equilibrium models that culminate in a Two-Agent New Keynesian Small Open Economy (TANK SOE) framework calibrated to the Norwegian economy. Each successive specification introduces one mechanism at a time, rule-of-thumb households and distortionary taxation, Calvo pricing, investment adjustment costs, Edgeworth complementarity, an open-economy block, and implementation delays, isolating the channels through which the composition of public expenditure shapes fiscal transmission. The central finding is that public investment and public consumption deliver broadly comparable present-value output multipliers on impact, but public investment produces stronger medium- and long-run effects through the gradual accumulation of productive public capital. For both shocks, nominal rigidities reshape the labor-market response, generating co-movement of hours worked and real wages, raising the impact multiplier relative to the flexible-price RBC benchmark. Investment adjustment costs reduce private-investment crowding-out under both shocks, while Edgeworth complementarity is decisive for limiting consumption crowding-out under the consumption shock. The investment-shock multiplier is particularly sensitive to the output elasticity of public capital, implementation delays, shock persistence, the steady-state investment-to-output share, and the financing regime. Lump-sum financing yields the largest multipliers, labor-tax financing is the most distortive and can turn the long-run multiplier negative, capital taxation generates the largest steady-state deviations, and debt-financing supports short-run multipliers but only postpones the long-run cost. Finally, openness systematically dampens both multipliers through import leakage and UIP-induced exchange-rate appreciation. The thesis contributes a structured, mechanism-by- mechanism decomposition for a Norwegian-calibrated TANK SOE, complementing NORA and recent small open-economy work on fiscal transmission.
Torsvik, Gaute & Klingenberg, Pia Cathrine (2026). Konkurranseutsetting av offentlig finansierte velferdstjenester. Når er privat produksjon mer hensiktsmessig enn offentlig egenproduksjon? Universitetet i Oslo. Full text in Research Archive Show summary
Denne oppgaven undersøker et sentralt spørsmål i utformingen av velferdsstaten: Når bør offentlig finansierte velferdstjenester produseres av det offentlige selv, og når kan det være hensiktsmessig å overlate produksjonen til private aktører? Problemstillingen analyseres ved hjelp av økonomisk teori og empirisk litteratur, med hovedvekt på norske forhold, men også med komparative perspektiver fra OECD og Norden. Oppgaven begynner med en sammenlikning mellom OECD-land og Norge for å illustrere at det er ulike måter å organisere produksjonen av velferdstjenester, selv i land som har en omfattende offentlig finansiering. Den empiriske gjennomgangen viser at Norge i stor grad produserer velferdstjenester i offentlig regi, sammenliknet med flere land hvor konkurranseutsetting spiller en større rolle. Samtidig har bruken av private leverandører økt over tid, særlig knyttet til kjøp av innsatsfaktorer og på enkelte tjenesteområder. Den teoretiske delen av oppgaven bygger på en grunnleggende avveining mellom insentiver og kontraktsproblemer. Private aktører kan ha sterkere insentiver til kostnadsreduksjon og innovasjon, særlig i tjenester som er standardiserte og der kvaliteten er lett å måle. Samtidig viser teorien at slike insentiver kan føre til kvalitetsforringelse når viktige sider ved tjenestene er vanskelige å kontraktfeste. Dette gjelder særlig komplekse og relasjonelle velferdstjenester. I tillegg bruker oppgaven også transaksjonsteori og «public service motivation»-teori for å vise at det er flere hensyn som spiller inn. Den empiriske litteraturen gir ikke støtte til en generell konklusjon om at privat eller offentlig produksjon er best. Derimot viser den et tydelig mønster: private leverandører presterer ofte bedre på målbare indikatorer, men ikke nødvendigvis på helhetlig kvalitet. Effektene varierer også mellom sektorer og brukergrupper. Oppgavens hovedfunn er at det avgjørende ikke er eierskapsform i seg selv, men hvor godt tjenestene egner seg for kontraktsstyring. Kommersiell drift fremstår som mest hensiktsmessig i standardiserte tjenester med målbar kvalitet og reell konkurranse, mens offentlig egenproduksjon eller ideell drift er bedre egnet i komplekse tjenester der kvalitet er vanskelig å observere og kontrollere. Oppgaven konkluderer med at organiseringen av velferdstjenester må vurderes sektor for sektor, basert på konkrete egenskaper ved tjenestene, snarere enn på generelle eller ideologiske prinsipper.
Christensen, Amalie Halle; Warne, Sofia Corbishley & Engen, Oda Østerhagen (2026). The effect of immigration following the 2004 European Union enlargement on firms’ labor productivity in Norway. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis examines how immigration following the 2004 enlargement of the European Union affected firm-level labor productivity in Norway. The enlargement led to a substantial increase in labor migration. While immigration may enhance productivity through mechanisms such as task specialization, it may also reduce productivity due to language barriers and limited transferability of human capital. To identify the effect, we apply a first-difference difference-in-differences (DiD) framework combined with a Two-Stage Least Squares (2SLS) instrumental variable approach. The instrument exploits variation in occupation-specific language requirements, where occupations with lower Norwegian language requirements are more exposed to immigrant inflows. The 2SLS estimates indicate that increased exposure to migrant inflows leads to a small but negative effect on firm-level labor productivity. These results vary across firm sizes and industries. In particular, firms that employed fewer than ten workers experienced a greater decline in labor productivity than larger firms. Similarly, the effect was greater for labor intensive industries.
Holtsmark, Katinka Kristine; Kellogg, Maxwell Dylan & Kuraityte, Mirga (2026). Strategy First, Vulnerability Second? Evidence on How Donor Interests Shape the Likelihood and Scale of Bilateral Environmental Assistance in the 21st Century. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis examines how Development Assistance Committee (DAC) donors trade off recipient environmental vulnerability against their own strategic interests when allocating bilateral environmental aid between 2002 and 2023. The analysis focuses on bilateral environmental aid for which environmental protection is identified as the principal objective under the OECD Rio marker framework, while excluding convention-specific climate finance. This approach helps reduce the influence of donor incentives associated with financing “green” global public goods and provides a cleaner test of whether aid allocation responds to localized environmental vulnerability. The research is guided by a two-stage microeconomic model of donor behavior. In this framework, donors face fixed costs when initiating a new aid relationship, meaning that entry (the extensive margin) is determined by whether total expected benefits, often influenced by strategic factors, outweigh these fixed costs. Conditional on entry, the volume of aid (the intensive margin) is determined where marginal benefits equal marginal costs, with greater environmental vulnerability expected to increase the marginal benefits of assistance through larger developmental and environmental impacts. The empirical findings support the two-stage allocation logic proposed by the theoretical model. At the extensive margin, environmental vulnerability shows no clear association with the probability of receiving aid. Instead, this entry stage is strongly associated with strategic factors, specifically historical colonial ties and bilateral trade intensity. However, conditional on a relationship being established, the pattern shifts. At the intensive margin, higher environmental vulnerability is associated with receiving larger aid volumes. Furthermore, time-varying specifications indicate that donor responsiveness to environmental need might have strengthened since the 2010s. The findings imply that if entry into environmental aid relationships is shaped by geopolitical path dependency and economic interests, there is a risk that global assistance may not reach the populations most vulnerable to environmental challenges. Improving both the effectiveness and equity of environmental aid may therefore require greater attention to the financial and institutional barriers facing highly environmentally vulnerable countries.
Eggen, Fernanda Winger & Leonthin, Mads (2026). Bidraget fra informasjons- og kommunikasjonsteknologi (IKT) til norsk produktivitetsvekst. Universitetet i Oslo. Full text in Research Archive Show summary
Gjennomsnittlig produktivitetsvekst i markedsrettet fastlands-Norge falt fra 3,1 prosent per år i 1996–2005 til 1,5 i 2006–2013 og 0,7 i 2014–2021. Denne nedgangen er særlig problematisk i lys av skiftene Norge skal gjennom når det gjelder demografi og næringsstrukturer. Perspektivmeldingen 2024 (Meld. St. 31 (2023–2024)) anslår at antall personer over 67 år vil øke med 700 000 frem mot 2060, mens andelen i yrkesaktiv alder faller. I samme tidsrom skal vi avvikle vår største næring, oljeproduksjon. Lavere andel i arbeidsfør alder medfører at færre arbeidstimer må finansiere mer velferd. For å kunne lukke dette gapet er vi avhengige av produktivitetsvekst. Forskjellen mellom 1 og 2 prosent årlig vekst kan ved første øyekast virke minimal, men utgjør over en generasjon rundt 30 prosent i BNP per innbygger. Informasjons- og kommunikasjonsteknologi (IKT) har vært identifisert som en sentral driver av produktivitetsvekst siden 1990-tallet. Min oppgave utforsker hvor mye av den norske veksten som kan tilskrives IKT-investeringer, og om nedgangen etter 2005 reflekterer avtakende avkastning på IKT-investeringer, eller andre forhold. I oppgaven kvantifiserer jeg IKT-bidraget til produktivitetsveksten gjennom et vekstregnskap. Jeg dekomponerer arbeidskraftproduktivitetsveksten i markedsrettet fastlands-Norge i tre komponenter: IKT-næringens eget bidrag (K1), bidraget fra IKT-investeringer i øvrige næringer (K2), og en residual som omfatter øvrig kapitalfordypning og totalfaktorproduktivitet (K3). Metoden følger Eggen, Mark og Røtnes (2015), med tre oppdateringer: analyseperioden forlenges med åtte år til 2021, IKT-investeringsandelene hentes fra EU KLEMS 2024, og resultatene sammenlignes med Finland og Sverige. Norske IKT-kapitalandeler estimeres med danske investeringsandeler som proxy, justert for forskjeller i kapitalintensitet mellom norske og danske næringer. Over perioden 1996–2021 var veksten i arbeidsproduktivitet 1,88 prosent per år. Fem hovedfunn kan oppsummere resultatene fra vekstregnskapet. IKT forklarer om lag en fjerdedel av produktivitetsveksten. K1+K2 bidro med 0,51 prosentpoeng per år, tilsvarende 27 prosent av samlet vekst. Bidraget falt fra 0,80 prosentpoeng i 1996–2005 til 0,20 i 2014–2021, men målt som andel av totalveksten var IKT-bidraget stabilt mellom 26 og 30 prosent i alle tre perioder. Stabiliteten innebærer at IKT-bidraget har falt i takt med den samlede produktivitetsveksten, ikke raskere. Produktivitetsnedgangen tilskrives K3, ikke IKT. K3 falt fra 2,28 til 0,49 prosentpoeng over de tre periodene og utgjorde 75 prosent av det totale fallet. Nedgangen i norsk produktivitetsvekst etter 2005 tilskrives i dekomponeringen avtagende TFP og svakere øvrig kapitalfordypning, ikke avtagende avkastning på IKT-investeringer. IKT modnes fra en næring i høy vekst til generell innsatsfaktor på tvers av næringer. Arbeidsproduktivitetsveksten i IKT-næringen falt fra 6,1 til 1,0 prosent per år. I siste periode oversteg K2 for første gang K1: IKT-kapitalens bidrag i øvrige næringer er nå viktigere enn IKT-næringens eget produktivitetsbidrag. Denne dreiningen gjenspeiler at IKT i økende grad er en allmenn innsatsfaktor på tvers av næringer snarere enn en isolert vekstnæring. K2 er en nedre grense for IKT-bidraget. Vekstregnskapet fanger kun den direkte kapitalfordypningseffekten. Indirekte effekter som reorganisering, komplementære investeringer og nye forretningsmodeller inngår i TFP-residualen og tilskrives K3. Det kan derfor argumenteres for at det reelle IKT-bidraget er høyere enn 27 prosent.
Delgado-Prieto, Lukas & Bhatti, Ahmad (2026). The Effects of Credit Registry Reforms on Firms’ Access to Credit and Firm Performance. Universitetet i Oslo. Full text in Research Archive Show summary
I examine the effects of credit registry reforms on firms’ access to credit and firm performance. I use firm-level data from the World Bank Enterprise Surveys from 14 countries over the period 2006-2014. I apply a difference-in-differences approach to compare firms in countries that introduced or expanded credit registries with firms in countries that did not implement such reforms. The results show that credit registry reforms did not lead to an overall improvement in firms’ access to credit. The heterogeneity analysis reveals that the aggregate results masked the opposing effects across firm types. Loan access declined significantly for small and young firms, while large firms experienced an increase. The findings show a credit reallocation mechanism rather than an overall credit expansion, where better screening allows banks to shift credit towards large and old firms relative to small and young firms. Sales, employment and wages all declined in treatment countries relative to the control countries, consistent with credit being reallocated away from small and young firms that make up the majority of the sample, which experienced the largest declines in loan access. The results challenge the standard prediction in the literature that improving credit information systems expands credit access broadly, and suggest that credit registry reforms may have distributional consequences that aggregate estimates alone do not capture.
Gundersen., Thomas Størdal; Lorentzen., Linnea; Hvidsten., Henrik & Ryberg., Oscar F. H. (2026). Time-Varying Transmission Mechanisms of Monetary Policy and Exchange Rate Shocks. Evidence from a Rolling-Window Bayesian Structural Vector Autoregression Model for Norway. Universitetet i Oslo. Full text in Research Archive Show summary
This thesis investigates whether the transmission of monetary policy shocks and real exchange rate shocks in Norway has changed over time, with particular attention to the 2014–2016 oil price collapse. Using quarterly data for 1990Q3–2019Q4, we estimate a Bayesian Structural Vector Autoregression model identified with sign- and zero-restrictions, and extends the framework through rolling-window estimation to trace time variation in impulse responses. The full-sample results are broadly consistent with earlier Norwegian evidence for monetary policy shocks, but they show more pronounced and persistent domestic effects of exchange rate shocks. The rolling-window analysis indicates stronger time variation after 2014–2016, especially for exchange rate shocks: exchange rate shock persistence rises, the domestic interest rate response becomes markedly weaker, and the pass-through to inflation and output becomes more visible. For monetary policy shocks, the results suggest amplified and more persistent responses in interest rates, inflation, output, and the exchange rate in the later sub-samples, although robustness tests imply that some of this variation may reflect varying shock size rather than changes to the transmission mechanism itself. The results emphasize the need for updated empirical evidence on shock transmission and suggest that Norges Bank has placed greater weight on stabilizing the real economy since the mid-2010s.

View all works in NVA