CEPR Discussion Paper of 2025
A discussion paper titled HANKSSON written by Florin Bilbiie, Sigurd Galaasen, Refet Gürkaynak, Mathis Mæhlum and Krisztina Molnar has been titled the CEPR discussion paper of 2025 as the most downloaded paper.
Professor Øivind A. Nilsen, Head of the Macro, Risk and Sustainability Centre (MARSCe) to which Molnar is affiliated states:
"This is truly excellent news! It demonstrates that Krisztina's research is highly relevant and has attracted substantial interest within the international research community. It reflects clear and innovative thinking, as well as the crucial role of international networking. Moreover, it highlights our exceptional ability to leverage the unique Norwegian microdata to address an important macroeconomic question: does heterogeneity amplify the aggregate effects of demand policies and shocks?"
"The fact that the paper has been presented at conferences attended by the world's leading scholars," continues Nilsen, "and that it has received highly encouraging feedback from referees and editors, strongly suggests that it will be published in one of the top five journals. I am thoroughly impressed by this achievement and would like to extend my heartfelt congratulations and best regards to Krisztina."
Abstract
HANK Sufficient Statistics Out of Norway (HANKSSON) answers a core question of the heterogeneity in macroeconomics literature: does heterogeneity amplify the aggregate effects of demand policies and shocks? We provide two sufficient statistics (SS) and test them using individual-level matched data for personal characteristics, income, wealth and non-imputed, actual consumption for the Norwegian population. The first SS gauges whether heterogeneity drives a wedge between the (representative agent) average MPC and a model-consistent (heterogeneous agent) aggregate MPC. The second SS elicits whether the consumption of constrained, "hand-to-mouth," agents is more exposed to aggregate fluctuations. Our key finding is that to analyze aggregate behavior, one does not need to keep track of heterogeneity: the average and the aggregate are about the same. We show that the amplification result currently prevalent in the literature is due to using labor earnings and is overturned when using model-consistent disposable income. This is due to the strong insurance effect of taxes and transfers; when applied to our data, even the much less progressive US tax and transfer system produces no amplification due to heterogeneity. The same “close to irrelevance” conclusion arises based on the second statistic using consumption data directly. Not even during the Great Recession do we see heterogeneity contribute meaningfully to demand shock amplification.