Search

This paper studies job search behavior in the midst of a pandemic recession. We use long-running panel data from the Netherlands (LISS) and complement the core survey with our own COVID-specific module, conducted in June 2020. The survey provides data on the job search effort in terms of the number of applications of employed as well as unemployed respondents. We estimate an empirical model of job search over the business cycle over the period 2008–2019 to explore the gap between predicted and actual job search behavior in 2020. We find that job search during the pandemic recession differs strongly from previous downturns. The unemployed search significantly less than what we would normally observe during a recession of this size. For the employed, the propensity to search is even greater than what we would expect, but those who do search make significantly fewer job applications. Expectations about the duration of the pandemic seem to play a key role in explaining job search effort for the unemployed in 2020. Furthermore, employed individuals whose work situation has been affected by COVID-19 are searching more actively for a new job.

This paper investigates the impact of investment in automation- and AI-related goods on within-firm wage inequality in the French economy during the 2002–2017 period. We document that most wage inequality in France is accounted for by differences among workers belonging to the same firm rather than by differences between sectors, firms, and occupations. Using an event-study approach on a sample of firms importing automation- and AI-related goods, we find that spike events related to the adoption of automation- and AI-related capital goods are not followed by an increase in within-firm wage inequality or in gender wage inequality. Instead, wages increase by 1% three years after the events at different percentiles of the distribution. Our findings are not linked to the rent-sharing behavior of firms obtaining productivity gains from automation and AI adoption. Instead, if wage gains do not differ across workers along the wage distribution, worker heterogeneity will still be present. Indeed, in agreement with the framework in Abowd et al. (1999b), most of the overall wage increase is due to the hiring of new employees. This adds to previous findings presenting a picture of a ‘labor friendly’ effect of the latest wave of new technologies within adopting firms.

This paper examines how banks adjust their asset structure in response to changes in loan demand following natural disasters. We demonstrate how banks' asset diversification strategy helps clients smooth consumption and supports local recovery. In the empirical section, we apply the difference-in-differences method and determine that U.S. commercial banks increase real estate lending after disasters and sell government bonds to finance this disaster-driven credit surge. The theoretical section presents a novel multiple-asset dynamic credit allocation model that explains our empirical findings. We use model simulations to predict and quantify the potential impact of climate change on the asset structure and profitability of banks given different scenarios.

Fuel consumption and CO2 emission are among the main criteria to assess the environmental and economical impact of vessels on inland waterways. Both criteria, however, are directly affected by the vessels' sailing speed. In this paper, we present a mathematical programming formulation of the speed optimization problem, which aims at minimizing the aggregated fuel consumption on an inland waterway network. The network can consist of multiple river segments, connected by a set of locks, without restrictions on the configuration. To allow scalability towards realistic waterway networks, we also propose a local-search based heuristic to optimize the speed for individual vessels. We evaluate the effectiveness of the heuristic by comparing it to solving the exact mathematical programming formulation. For all computational experiments, we make use of real AIS data from a section of the Dutch river network. We observe that the heuristic is able to construct a high quality solution in realistic problem settings within reasonable amount of computation time.