Current projects

Here you can read about my ongoing research agenda of improving public trust in bureaucrats. Below, you can also read abstracts of other ongoing projects - all work in progress. You are welcome to reach out for more info. For more developed work see the "Under review" section.

Countering the Reputation Crisis of Bureaucrats - Field Experimental Evidence From a Nation-Wide Union Campaign

with Asmus Olsen

“It is ironic that bureaucracy is still primarily a term of scorn”
Anthony Downs, Theory of the Bureaucracy (1965, p. 439)

“It is easy to inveigh against the bureaucrat because he is a faceless enemy”
New York Times (1973)

Bureaucrats have a bad reputation. In the West, more than 40% of citizens report distrust in higher level civil servants to do what is right for the country (ISSP, 2018). But the reputation crisis is not restricted to the West and globally just short of 50% have no confidence in the civil service (World Values Survey, 2020). Although recent trends may have fueled the “myth of the useless bureaucrat” (New York Times, 2011), the skepticism has been around for ages. So much so that the term has been almost “universally regarded as an insult” (Downs, 1965, 440). As their job output is obscure to the public, bureaucrats risk becoming a straw man for citizen frustration with the state. Eventually, anti-bureaucrat sentiments may “hurt the public’s faith in government, the morale of employees who remain, and the prospects that a new generation will enter public service” (New York Times, 2021). We randomly distribute information on bureaucrat job tasks in collaboration with the largest union for bureaucrats in Denmark. Denmark is ranked in the top-2% of countries in the World Bank’s Government Effectiveness ranking and the least corrupt bureaucracy in the world according to Transparency International’s Corruption Perception Index and where bureaucrats are drawn from some of the most selective higher-education programs in the country (Barfort et al., 2019). Despite this, a growing anti-bureaucrat sentiment has taken hold throughout the last decades. One-in-three Danes do not trust that civil servants to do what is best for the country (ISSP, 2018). While this places Danes in the top-20% most trusting citizens (it is every second American), distrust increased 60% between 2006 and 2016. Hauled by the media, the trend has come to be known as “djoefication”[1]: a reference to the largest Danish union for bureaucrats “Djøf”, representing 13% of all public employees in Denmark. The anti-bureaucrat sentiment is generally referred to as “djoef-bashing”.

[1] The word has even been incorporated in the official dictionary of the Danish language as a “change of a public institution’s administration such that there is a pronounced or even one-sided focus on effectiveness, profitability, and abstract, administrative regulations”.

Broadcasting Pro-Regime Support: The Effect of Local Radio Coverage on 2021 Electoral Results in Uganda

with Camila Blanes, Corbin Duncan and Horacio Larreguy

One way authoritarian regimes win elections is by controlling the media. How does partisan media influence electoral support? And how effectively can autocratic regimes control coverage? We study these questions in Uganda where the importance of media, namely radio, for voter mobilization was amplified by the Covid19-pandemic and the not uniformly enforced ban on ground campaigning. We estimate the local coverage of radio stations ($N=252$) relying on transmitter characteristics and accounting for local topography. We measure individual radio stations' partisanship by 1) manually coding ownership data, and 2) training a machine learning model on social media posts. We geo-code polling station administrative data and merge it with local radio coverage ($N=34,684$). Results suggest that 1) opposition media positively affects opposition, but hurts regime, electoral prospects 2) but only in the 2021, not the 2016, elections 3) and incumbents respond to this by strategically (and successfully) distributing broadcasting licenses to silence critical coverage.

Distributive Politics in Competitive Vote Markets -- Evidence from Uganda

with Horacio Larreguy

Does political competition deter clientelism by encouraging politicians to rely on public goods provision? While local public goods, in theory, is a more cost-effective strategy to generate political support among large groups, studies have found that political competition intensifies clientelism. This paper studies the question in Uganda where clientelism is widespread and political competition between candidates is increasing. We rely on the universe of public expenses, down to village-level, generated by national and sub-national government between 2012 and 2018 (N=380,000). The official voter registers allow us to link villages to their respective polling station electoral results. The extent of clientelism is measured using a high density survey (N=30,000) collected during the 2016 general elections campaign. We identify effects of local political competition using a shift-share instrument leveraging how positive shifts in NRM’s nationwide popularity increase local political competition in opposition districts, but decrease political competition in regime strongholds.