Sebastian Ellingsen

I am a PhD candidate at Universitat Pompeu Fabra and the Barcelona GSE. In 2018-2019, I was a visiting student at the University of Chicago. C.V.

Google Scholar

Theme by orderedlist


I am mainly interested in questions in Economic Geography, Political Economy, and Empirical Trade. In my research, I use tools from Applied Econometrics with a particular emphasis on spatial data.

I am on the 2020-21 job market and available for interviews at the EEA and ASSA.

Working Papers

  • Free and Protected: Trade and Breaks in Long-Term Persistence, Job market paper.
    The spatial distribution of economic activity depends largely on market access and history, but countries differ greatly in the extent to which their geographies reflect these two determinants. What explains these differences? This paper explores this question using a staggered lifting of restrictions on direct trade with Europe across the Spanish Empire. I combine a difference-in-differences approach with a dynamic spatial equilibrium framework and detailed georeferenced data on maritime travel from historical logbooks to examine this issue. I show that the increase in market access induced by the reform led to a substantial reconfiguration of the economic geography in places that were initially less densely settled. Moreover, I show that modern-day settlement patterns depend less on pre-colonial population density and more on coastal access in areas subjected to the reform. Taken together, the findings show that a key determinant of persistence in economic geography is the level of development of a country as it opens up to trade.

  • Does High-speed Internet Erode Democracy? Evidence on Turnout and Polarization from a Norwegian Broadband Reform,
    with Øyvind Skorge, Øystein Hernæs.
    To what extent the surge of high-speed internet has contributed to democratic erosion is contested. One the one hand, it may crowd out voters’ consumption of traditional media with higher and more unbiased knowledge about politics, which is expected to lessen turnout and polarize those who nonetheless vote. On the other hand, it may add to individuals’ existing news consumption and increase political knowledge, which is expected to increase turnout and leave polarization unaffected. To evaluate these competing hypotheses, we exploit a large-scale broadband reform that was rolled out in a staggered fashion across Norwegian municipalities during the 2000-2008 period. Our instrumental variable analysis reveals a positive effect of high-speed internet usage on the turnout rate in municipal elections and a small (and insignificant) negative effect on polarization in vote choice. We also show that the arrival of high-speed internet increased time online without replacing the consumption of newspapers, radio, and TV. Our analysis suggests that high-speed internet have more nuanced effects on electoral participation than what is commonly asserted.


  • The impact of commercial television on turnout and public policy: Evidence from Norwegian Local Politics, Journal of Public Economics, Vol. 159, March 2018, Pages 1-15,
    with Øystein Hernæs,
    Media coverage: The New York Times - The Upshot
    We investigate the impact of commercial television on political participation and local policy outcomes. Exploiting a geographically staggered expansion of cable television after the liberalization of Norwegian broadcasting in 1981, we show that higher cable television penetration significantly reduced turnout in municipal elections. Using individual-level data, we find that cable television coverage had a negative effect on the extent to which respondents were exposed to political information through mass media. The effect is more pronounced for individuals that on average watch more cable television; namely individuals with fewer years of schooling. Consistent with an increased difference in political participation and exposure to information between more and less educated groups, we find that commercial television led to reduced public spending and increased the share spent on education. The results are evidence that commercial mass media can influence electoral politics by reducing political participation and exposure to information of its target audiences.

Work in Progress

  • Language Reform and Market Integration, with Erqi Ge.
  • Trade, Income, and Factor-Based Models of Regime Change.
  • Trade, Epidemics, and Growth in Early Modern Spain, with Gregori Galofré-Vilà.


I TA the following classes at Universitat Pompeu Fabra and Barcelona GSE.