https://sondreus.com/2024/04/05/journalism-2/https://sondreus.com/wp-content/uploads/2024/04/image-34.pngimagehttps://sondreus.com/wp-content/uploads/2024/04/image-33.pngimagehttps://sondreus.com/wp-content/uploads/2024/04/image-32.pngimagehttps://sondreus.com/wp-content/uploads/2024/04/image-31.pngimagehttps://sondreus.com/wp-content/uploads/2024/04/image-30.pngimagehttps://sondreus.com/wp-content/uploads/2024/04/image-29.pngimagehttps://sondreus.com/wp-content/uploads/2024/04/image-28.pngimagehttps://sondreus.com/wp-content/uploads/2024/04/image-27.pngimagehttps://sondreus.com/wp-content/uploads/2024/04/image-26.pngimagehttps://sondreus.com/wp-content/uploads/2024/04/image-25.pngimage2025-11-14T17:45:46+00:00monthlyhttps://sondreus.com/2019/01/01/about/https://sondreus.com/wp-content/uploads/2019/01/small.pngsmallhttps://sondreus.com/wp-content/uploads/2019/01/sondreulvundsolstad.jpgSondre Ulvund Solstad The Economist Skup 2023 FOto: Marte Vike ASondre Ulvund Solstad The Economist Skup 2023 FOto: Marte Vike Arnesenhttps://sondreus.com/wp-content/uploads/2019/01/sondre_ulvund_solstad_9-319079378-e1710329708534.jpgSondre_Ulvund_Solstad_9https://sondreus.com/wp-content/uploads/2019/01/wordpress_pic-e1562598804706.jpegwordpress_pichttps://sondreus.com/wp-content/uploads/2017/01/12118618_10206868009915422_7947049903494004213_n.jpg12118618_10206868009915422_7947049903494004213_n2025-11-14T16:25:29+00:00monthlyhttps://sondreus.com/2019/01/01/research/https://sondreus.com/wp-content/uploads/2019/01/steam_and_sail.pngsteam_and_sailSample of journeys by sail in 1750-1810 (black) versus shipping routes by modern propulsion (blue). Constructed using ship logs from Spain, France, the Netherlands and the United Kingdom (digitalized by the Climatological Database for the World’s Oceans project, some 200 000 positions), and modern propulsion ship routes from Halpern et al. (2008). I leverage the difference in port-to-port distances by sail and steam to construct an instrument for trade.https://sondreus.com/wp-content/uploads/2019/01/world_heatmap_animated.gifworld_heatmap_animatedAll countries saw more trade as a result of containerization, but the impact was not the same for all countries. The animation shows how countries' share of global trade changed in the years 1960-2015. Colors reflect trade share relative to their average for the entire era. Countries which got darker saw their share of world trade shrink, whereas countries which got lighter saw their share of trade grow. https://sondreus.com/wp-content/uploads/2019/03/5_glmnet.gif5_glmnetAI generating a theory about sport outcomes based on ~500 variables from 5000 matches.https://sondreus.com/wp-content/uploads/2019/03/4_glmnet.gif4_glmnetAI generating a theory about the whether an email is spam or not based on ~50 variables and 4600 emails.https://sondreus.com/wp-content/uploads/2019/01/2_glmnet_cropped.gif2_glmnet_croppedhttps://sondreus.com/wp-content/uploads/2019/01/2_glmnet_cap.gif2_glmnet_caphttps://sondreus.com/wp-content/uploads/2019/01/soil_calcs_pic.pngsoil_calcs_picThe plots show our method for estimating the soil power of ethnic homelands. On the left side, we plot ethnic homeland borders in Kenya in 2017 from GEO-EPR intersected with data on estimated agricultural productivity from Ramankutty et al. (2002). The Luo homeland is given a black border. In the right side plot, we zoom in on the Luo's homeland, and extract the area they control within the different grid cells, and the estimated agricultural productivity within these cells. In both plots, darker shades indicates more productive land. "Soil quality" here denotes the estimated potential agricultural productivity of the area. It is calculated using a model that incorporates data on the area's temperature, sunshine hours, rainfall, elevation, slope, and its soil's carbon density, nitrogen content, pH, water holding capacity, and more. The soil power of the Luo is the sum of the different areas times their agricultural productivity.https://sondreus.com/wp-content/uploads/2017/01/war_paths-e1551123329356.pngDisruption CostsTop 30 ocean areas for the disruption cost are highlighted. Disruption costs are much higher at naval choke-points than in open waters. The method employed here "discovers" such choke-points (such as the Panama canal) as a consequence of comparing the lengths of paths calculated with and without access to the ocean area in question.https://sondreus.com/wp-content/uploads/2017/01/both-e1542392587186.pngbothhttps://sondreus.com/wp-content/uploads/2017/01/low_accuracy.pnglow_accuracy2024-06-08T20:10:26+00:00monthlyhttps://sondreus.comdaily1.02025-11-14T17:45:46+00:00