David Dreyer Lassen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
David Dreyer Lassen has a total of 44 co-authors. Here are the top co-authors by number of shared publications:
James E. Alt
5 shared publications- Sune Lehmann
Department of Applied Mathematics and Computer Science, Technical University of Denmark
4 shared publications - Valentin Kassarnig
Institute of Software Technology, Graz University of Technology, Graz, Austria
4 shared publications Andreas Bjerre-Nielsen
Department of Economics, University of Copenhagen, Copenhagen, Denmark
3 shared publications- Andreas Bjerre-Nielsen
University of Copenhagen
3 shared publications - Asger Lau Andersen
Department of Economics, University of Copenhagen, Copenhagen K 1353, Denmark
3 shared publications
Latest Publications
- Open Access
- 17 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 246-250
Fairness Over Time: A Nationwide Study of Evolving Bias in Dropout Prediction
- 17 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 246-250
Open AccessFairness Over Time: A Nationwide Study of Evolving Bias in Dropout Prediction
- Open Access
- 26 June 2023
- Published by Springer Nature in Nature Climate Change
- Vol. 13 (7) , 661-670
Experience exceeds awareness of anthropogenic climate change in Greenland
Although Greenland is a hub for climate science, the climate perceptions of Greenland’s predominantly Indigenous population have remained largely unstudied. Here we present two nationally representative surveys and show that Greenlanders are more likely than residents of top oil-producing Arctic countries to perceive that climate change is happening and about twice as likely to have personally experienced its effects. However, half are unaware that climate change is human-caused and those who are most affected appear to be least aware. Personal experience and awareness of human-induced climate change diverge along an Inuit cultural dimension. Indigenous identity positively predicts climate change experience, whereas subsistence occupation and no post-primary education negatively predict attribution beliefs. Despite Greenland’s centrality to climate research, we uncover a gap between the scientific consensus and Kalaallit views of climate change, particularly among youth. This science–society gulf has implications for local climate adaptation, science communication and knowledge exchange between generations, institutions and communities.
- 26 June 2023
- Published by Springer Nature in Nature Climate Change
- Vol. 13 (7) , 661-670
Open AccessExperience exceeds awareness of anthropogenic climate change in Greenland
Although Greenland is a hub for climate science, the climate perceptions of Greenland’s predominantly Indigenous population have remained largely unstudied. Here we present two nationally representative surveys and show that Greenlanders are more likely than residents of top oil-producing Arctic countries to perceive that climate change is happening and about twice as likely to have personally experienced its effects. However, half are unaware that climate change is human-caused and those who are most affected appear to be least aware. Personal experience and awareness of human-induced climate change diverge along an Inuit cultural dimension. Indigenous identity positively predicts climate change experience, whereas subsistence occupation and no post-primary education negatively predict attribution beliefs. Despite Greenland’s centrality to climate research, we uncover a gap between the scientific consensus and Kalaallit views of climate change, particularly among youth. This science–society gulf has implications for local climate adaptation, science communication and knowledge exchange between generations, institutions and communities.
- 14 December 2022
- Published by Cambridge University Press (CUP) in Behavioural Public Policy
- Vol. 7 (3) , 721-743
Free to choose or free to lose? Understanding individual attitudes toward paternalism
In the past decades, behavioral economics has credibly identified numerous decision-making biases leading people to make choices they would not have made if better informed about the long-term consequences of their actions. This has given rise to a new reason for government interventions: internalities. In contrast to traditional reasons for government intervention, such as redistribution and externalities, overcoming internalities often involves the use of paternalistic policies. We investigate theoretically and empirically the formation of attitudes toward paternalistic policies. Theoretically, we focus on the role of self-interest and distinguish between self-interest as construed for the rational decision-maker, self-interest when self-control problems are present, and self-interest when procedural or expressive elements, such as autonomy, matter. Empirically, we employ two novel data sets: a Danish survey on political opinion combined with administrative data on actual behavior and a large-scale cross-country survey to analyze attitudes toward paternalistic policies in the health and financial domains. We show that targets of paternalism are more opposed to paternalism than non-targets both in Denmark and across nine Western democracies and rely on our theoretical priors to explore mechanisms that can explain these attitudes.
- 14 December 2022
- Published by Cambridge University Press (CUP) in Behavioural Public Policy
- Vol. 7 (3) , 721-743
Free to choose or free to lose? Understanding individual attitudes toward paternalism
In the past decades, behavioral economics has credibly identified numerous decision-making biases leading people to make choices they would not have made if better informed about the long-term consequences of their actions. This has given rise to a new reason for government interventions: internalities. In contrast to traditional reasons for government intervention, such as redistribution and externalities, overcoming internalities often involves the use of paternalistic policies. We investigate theoretically and empirically the formation of attitudes toward paternalistic policies. Theoretically, we focus on the role of self-interest and distinguish between self-interest as construed for the rational decision-maker, self-interest when self-control problems are present, and self-interest when procedural or expressive elements, such as autonomy, matter. Empirically, we employ two novel data sets: a Danish survey on political opinion combined with administrative data on actual behavior and a large-scale cross-country survey to analyze attitudes toward paternalistic policies in the health and financial domains. We show that targets of paternalism are more opposed to paternalism than non-targets both in Denmark and across nine Western democracies and rely on our theoretical priors to explore mechanisms that can explain these attitudes.
- Open Access
- 31 March 2021
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 118 (14)
Task-specific information outperforms surveillance-style big data in predictive analytics
Increasingly, human behavior can be monitored through the collection of data from digital devices revealing information on behaviors and locations. In the context of higher education, a growing number of schools and universities collect data on their students with the purpose of assessing or predicting behaviors and academic performance, and the COVID-19–induced move to online education dramatically increases what can be accumulated in this way, raising concerns about students’ privacy. We focus on academic performance and ask whether predictive performance for a given dataset can be achieved with less privacy-invasive, but more task-specific, data. We draw on a unique dataset on a large student population containing both highly detailed measures of behavior and personality and high-quality third-party reported individual-level administrative data. We find that models estimated using the big behavioral data are indeed able to accurately predict academic performance out of sample. However, models using only low-dimensional and arguably less privacy-invasive administrative data perform considerably better and, importantly, do not improve when we add the high-resolution, privacy-invasive behavioral data. We argue that combining big behavioral data with “ground truth” administrative registry data can ideally allow the identification of privacy-preserving task-specific features that can be employed instead of current indiscriminate troves of behavioral data, with better privacy and better prediction resulting.
- 31 March 2021
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 118 (14)
Open AccessTask-specific information outperforms surveillance-style big data in predictive analytics
Increasingly, human behavior can be monitored through the collection of data from digital devices revealing information on behaviors and locations. In the context of higher education, a growing number of schools and universities collect data on their students with the purpose of assessing or predicting behaviors and academic performance, and the COVID-19–induced move to online education dramatically increases what can be accumulated in this way, raising concerns about students’ privacy. We focus on academic performance and ask whether predictive performance for a given dataset can be achieved with less privacy-invasive, but more task-specific, data. We draw on a unique dataset on a large student population containing both highly detailed measures of behavior and personality and high-quality third-party reported individual-level administrative data. We find that models estimated using the big behavioral data are indeed able to accurately predict academic performance out of sample. However, models using only low-dimensional and arguably less privacy-invasive administrative data perform considerably better and, importantly, do not improve when we add the high-resolution, privacy-invasive behavioral data. We argue that combining big behavioral data with “ground truth” administrative registry data can ideally allow the identification of privacy-preserving task-specific features that can be employed instead of current indiscriminate troves of behavioral data, with better privacy and better prediction resulting.
- Open Access
- 2 July 2020
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 15 (7) , e0234003
Inferring transportation mode from smartphone sensors: Evaluating the potential of Wi-Fi and Bluetooth
Understanding which transportation modes people use is critical for smart cities and planners to better serve their citizens. We show that using information from pervasive Wi-Fi access points and Bluetooth devices can enhance GPS and geographic information to improve transportation detection on smartphones. Wi-Fi information also improves the identification of transportation mode and helps conserve battery since it is already collected by most mobile phones. Our approach uses a machine learning approach to determine the mode from pre-prepocessed data. This approach yields an overall accuracy of 89% and average F1 score of 83% for inferring the three grouped modes of self-powered, car-based, and public transportation. When broken out by individual modes, Wi-Fi features improve detection accuracy of bus trips, train travel, and driving compared to GPS features alone and can substitute for GIS features without decreasing performance. Our results suggest that Wi-Fi and Bluetooth can be useful in urban transportation research, for example by improving mobile travel surveys and urban sensing applications.
- 2 July 2020
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 15 (7) , e0234003
Open AccessInferring transportation mode from smartphone sensors: Evaluating the potential of Wi-Fi and Bluetooth
Understanding which transportation modes people use is critical for smart cities and planners to better serve their citizens. We show that using information from pervasive Wi-Fi access points and Bluetooth devices can enhance GPS and geographic information to improve transportation detection on smartphones. Wi-Fi information also improves the identification of transportation mode and helps conserve battery since it is already collected by most mobile phones. Our approach uses a machine learning approach to determine the mode from pre-prepocessed data. This approach yields an overall accuracy of 89% and average F1 score of 83% for inferring the three grouped modes of self-powered, car-based, and public transportation. When broken out by individual modes, Wi-Fi features improve detection accuracy of bus trips, train travel, and driving compared to GPS features alone and can substitute for GIS features without decreasing performance. Our results suggest that Wi-Fi and Bluetooth can be useful in urban transportation research, for example by improving mobile travel surveys and urban sensing applications.
- 1 April 2020
- Published by American Economic Association in American Economic Review
- Vol. 110 (4) , 1177-1205
Time Discounting and Wealth Inequality
This paper documents a large association between individuals’ time discounting in incentivized experiments and their positions in the real-life wealth distribution derived from Danish high-quality administrative data for a large sample of middle-aged individuals. The association is stable over time, exists through the wealth distribution and remains large after controlling for education, income profile, school grades, initial wealth, parental wealth, credit constraints, demographics, risk preferences, and additional behavioral parameters. Our results suggest that savings behavior is a driver of the observed association between patience and wealth inequality as predicted by standard savings theory. (JEL C91, D15, D31, E21)
- 1 April 2020
- Published by American Economic Association in American Economic Review
- Vol. 110 (4) , 1177-1205
Time Discounting and Wealth Inequality
This paper documents a large association between individuals’ time discounting in incentivized experiments and their positions in the real-life wealth distribution derived from Danish high-quality administrative data for a large sample of middle-aged individuals. The association is stable over time, exists through the wealth distribution and remains large after controlling for education, income profile, school grades, initial wealth, parental wealth, credit constraints, demographics, risk preferences, and additional behavioral parameters. Our results suggest that savings behavior is a driver of the observed association between patience and wealth inequality as predicted by standard savings theory. (JEL C91, D15, D31, E21)
- Open Access
- 2 March 2020
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 15 (3) , e0229789
Irresponsible parties, responsible voters? Legislative gridlock and collective accountability
Legislative gridlock is a failure of one of the key functions of government: to pass legislation. Can voters counter such political dysfunction? This paper examines whether and how voters hold politicians accountable for gridlock. We focus on the passage of the government budget, the central task of any legislature, and define a legislature to experience budgetary gridlock if it fails to pass the budget on time. We argue, based on evidence from twenty years of budget enactment data, that voters hold state legislators accountable for budget gridlock in US state governments, with gridlocked incumbents losing their seat more often than incumbents passing budgets on time. Based on established theories of party organization in American politics, we develop three competing theoretical hypotheses to guide our understanding of the observed patterns of retrospective voting. We find strong support for collective electoral accountability with voters punishing incumbent members of state legislature majority parties.
- 2 March 2020
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 15 (3) , e0229789
Open AccessIrresponsible parties, responsible voters? Legislative gridlock and collective accountability
Legislative gridlock is a failure of one of the key functions of government: to pass legislation. Can voters counter such political dysfunction? This paper examines whether and how voters hold politicians accountable for gridlock. We focus on the passage of the government budget, the central task of any legislature, and define a legislature to experience budgetary gridlock if it fails to pass the budget on time. We argue, based on evidence from twenty years of budget enactment data, that voters hold state legislators accountable for budget gridlock in US state governments, with gridlocked incumbents losing their seat more often than incumbents passing budgets on time. Based on established theories of party organization in American politics, we develop three competing theoretical hypotheses to guide our understanding of the observed patterns of retrospective voting. We find strong support for collective electoral accountability with voters punishing incumbent members of state legislature majority parties.
- Top Cited PapersOpen Access
- 11 December 2019
- Published by Springer Nature in Scientific Data
- Vol. 6 (1) , 1-10
Interaction data from the Copenhagen Networks Study
We describe the multi-layer temporal network which connects a population of more than 700 university students over a period of four weeks. The dataset was collected via smartphones as part of the Copenhagen Networks Study. We include the network of physical proximity among the participants (estimated via Bluetooth signal strength), the network of phone calls (start time, duration, no content), the network of text messages (time of message, no content), and information about Facebook friendships. Thus, we provide multiple types of communication networks expressed in a single, large population with high temporal resolution, and over a period of multiple weeks, a fact which makes the dataset shared here unique. We expect that reuse of this dataset will allow researchers to make progress on the analysis and modeling of human social networks.
- 11 December 2019
- Published by Springer Nature in Scientific Data
- Vol. 6 (1) , 1-10
Top Cited PapersOpen AccessInteraction data from the Copenhagen Networks Study
We describe the multi-layer temporal network which connects a population of more than 700 university students over a period of four weeks. The dataset was collected via smartphones as part of the Copenhagen Networks Study. We include the network of physical proximity among the participants (estimated via Bluetooth signal strength), the network of phone calls (start time, duration, no content), the network of text messages (time of message, no content), and information about Facebook friendships. Thus, we provide multiple types of communication networks expressed in a single, large population with high temporal resolution, and over a period of multiple weeks, a fact which makes the dataset shared here unique. We expect that reuse of this dataset will allow researchers to make progress on the analysis and modeling of human social networks.
- 15 October 2019
- Published by SAGE Publications in Comparative Political Studies
- Vol. 53 (6) , 892-922
Middle Class Without a Net: Savings, Financial Fragility, and Preferences Over Social Insurance
In this article, we show that it is crucial to distinguish between liquid and illiquid wealth to understand how voters form preferences toward social insurance. Many households are financially fragile despite having high incomes and wealth, because they hold little liquid savings. We hypothesize, and show empirically, that this implies that a substantial group of voters show strong support for social insurance policies despite being wealthy and having high incomes, because of their limited ability to self-insure through own savings in case of an income shock. Our empirical analysis is based on a novel dataset from Denmark, which combines administrative data with high-quality measures of individual financial assets and survey measures of political preferences. Using data for other countries from the European Social Survey, we find evidence that our results hold more generally and are not specific to the Danish context.
- 15 October 2019
- Published by SAGE Publications in Comparative Political Studies
- Vol. 53 (6) , 892-922
Middle Class Without a Net: Savings, Financial Fragility, and Preferences Over Social Insurance
In this article, we show that it is crucial to distinguish between liquid and illiquid wealth to understand how voters form preferences toward social insurance. Many households are financially fragile despite having high incomes and wealth, because they hold little liquid savings. We hypothesize, and show empirically, that this implies that a substantial group of voters show strong support for social insurance policies despite being wealthy and having high incomes, because of their limited ability to self-insure through own savings in case of an income shock. Our empirical analysis is based on a novel dataset from Denmark, which combines administrative data with high-quality measures of individual financial assets and survey measures of political preferences. Using data for other countries from the European Social Survey, we find evidence that our results hold more generally and are not specific to the Danish context.
- 1 February 2019
- Published by American Economic Association in American Economic Journal: Economic Policy
- Vol. 11 (1) , 351-379
Liquidity Constraint Tightness and Consumer Responses to Fiscal Stimulus Policy
The marginal interest rate is the price at which a household can access additional liquidity. Consumption theory posits that variation in marginal interest rates across consumers predicts differences in the propensity to spend a stimulus payment. This hypothesis is tested in the context of a Danish 2009 stimulus policy that transformed illiquid pension wealth into liquid wealth. Marginal interest rates are constructed from administrative records with account level information and merged with survey data measuring the spending response to the stimulus policy. The data reveal substantial variation in marginal interest rates across consumers, and these interest rates predict spending responses. (JEL D14, D15, E21, E43, E62)
- 1 February 2019
- Published by American Economic Association in American Economic Journal: Economic Policy
- Vol. 11 (1) , 351-379
Liquidity Constraint Tightness and Consumer Responses to Fiscal Stimulus Policy
The marginal interest rate is the price at which a household can access additional liquidity. Consumption theory posits that variation in marginal interest rates across consumers predicts differences in the propensity to spend a stimulus payment. This hypothesis is tested in the context of a Danish 2009 stimulus policy that transformed illiquid pension wealth into liquid wealth. Marginal interest rates are constructed from administrative records with account level information and merged with survey data measuring the spending response to the stimulus policy. The data reveal substantial variation in marginal interest rates across consumers, and these interest rates predict spending responses. (JEL D14, D15, E21, E43, E62)