Current research

Taxing the Rich or Insuring the Poor: Motives for supporting the welfare state
with Erling Barth and Kalle Moene
Key controversies in the debate over social insurance and redistribution revolve around the issue of the income elasticity of the demand for social insurance and redistribution. Why does not the welfare state become obsolete when national income rises? Why is there a negative rather than a positive association between pre-transfer income inequality and the generosity of the welfare state? This paper provides micro evidence on the relationship between voters’ demand for public transfers and their income, using the US’ General Social Survey 1984-2010, Norwegian election studies 1977 to 2009, as well as Norwegian individual panel data. We find that the demands for social insurance transfers and for government redistribution are driven by different mechanisms, in particular because of the role absolute income plays for the determination of the two. First, we replicate a well-known significantly negative correlation between income and support for public transfers in cross sectional data. Next we show that for social insurance transfers, this negative correlation turns significantly positive once we control for the individual’s relative position in the income distribution. We do not find this pattern for typical redistribution programs.


Trust, Ethnic Diversity, and Personal Contact: Experimental Field Evidence
with Torbjørn Hanson, Åshild Johnsen, Andreas Kotsadam and Gaute Torsvik
We combine a lab and a field experiment in the Norwegian Armed Forces to study how close personal contact with minorities affect in-group and out-group trust. We randomly assign majority soldiers to rooms with or without ethnic minorities and use an incentivized trust game to measure trust. First, we show that close personal contact with minorities increases trust. Second, we replicate the result that individuals coming from areas with a high share of immigrants trust minorities less. Finally, the negative relationship between the share of minorities and out-group trust is reversed for soldiers who are randomly assigned to interact closely with minority soldiers. Hence, our study show that social integration involving personal contact can reduce negative effects of ethnic diversity on trust.


Care for career
with Øyvind Skorge
Women remain underrepresented in leadership positions and high powered jobs. In this paper, we argue that work family-policies can reduce the underrepresentation. In particular, we contend that subsidized child daycare services increase mothers’ willingness to and opportunities for entering leadership positions. To provide empirical evidence, we make use of a daycare reform in Norway, which led to a staggered and extensive expansion of daycare services across the about 400 municipalities. With administrative register data on the whole population, our instrumental-variable approach reveals that mothers with two-year-olds in daycare are more likely to work in occupations requiring longer hours and in managerial positions. Moreover, using survey data we find that mothers with access to daycare are more inclined to view their career as equally important as men’s. Our results document how public policies can be used to address gender inequalities in access to powerful positions


Immigration, Policy Bundles, and Welfare State Generosity
with Magnus Rasmussen
In this paper we make two contributions to the literature on the relationship between immigration and the generosity of the welfare state. First, we construct an instrumental variable to address the endogeneity problem in previous estimates of this relationship. Second, we study the support for different policy bundles in order to assess a political mechanism which links immigration to welfare state generosity. When using our empirical approach, we find negative, but insignificant relationships between changes in immigration on changes in unemployment replacement rates. Our estimates of the electoral consequences of immigration show that immigration is associated with increasing support for parties that oppose multiculturalism. However, several of these parties run on pro-welfare state platforms, and we find that the negative effect on the aggregate vote share of pro-welfare state parties is small. Our results highlight the role of policy bundles to understand the short-run influence of immigration on the welfare state.


Education and Immigration Attitudes: Evidence from an Education Reform
with Øyvind Skorge and Marte Strøm
Empirical research consistently finds that high educated people have more liberal immigration attitudes, but to what degree this relationship reflects a causal effect of education is largely unknown. In this research note we rely on the staggered introduction of a Norwegian education reform to get exogenous variation in respondents’ level of education. We find no difference in immigration attitudes between those who were educated in the old and the new education system.


Social Democratic Representation and Welfare Spending: A Quantitative Case Study
The welfare state literature argues that Social Democratic party representation is of key importance for welfare state outcomes. However, few papers are able to separate the influence of party representation from the influence of voter preferences, which implies that the effects of Social Democratic representation are likely to be overstated. In this paper I study a natural experiment which makes it possible to identify a partisan representation effect. In 1995, the Labour Party (Ap) in the Norwegian municipality of Flå filed their candidate list for the local election too late, which implied that they could not run for the upcoming election. Ap had dominated local politics in Flå since the second World War, but have not managed to regain their hegemonic position after the 1995 election. I use a difference-in-difference approach and the synthetic control method to study the short and long run consequences of Ap’s decline on public welfare spending.


The Effects of Small group Instruction in Mathematics for Pupils in Lower Elementary School
with a team of researchers from NIFU, ISF, and NTNU
We conduct a randomized controlled trial where schools in the treatment group receive resources to hire an additional teacher to teach mathematics in small groups of pupils in lower elementary school. The project includes 160 schools from ten municipalities, randomized into equally sized treatment and control groups using stratified randomization. The project runs for four school years, with cross-cohort variation in the intensity and timing of treatment. The main outcome variable is test scores in the national tests in fifth grade, but we also collect our own pre-treatment and end-of-the-year test scores to measure short term effects. The analysis of test scores in fifth grade will be pre-registered in a pre-analysis plan.