This paper studies the regional economic effect of past and current tourism development in Switzerland. For this, we combine rich microdata on sectoral employment and wages with novel indicators of original tourism attractiveness based in a tailored identification strategy. By digitizing the first Swiss firm registry of 1894 and creating indicators of original tourism attractivenss from travel guidebooks via large language models (BERT) and supervised machine learning (random forest), we can disentangle the effect of initial tourism development from current tourism levels. Our results suggest strong path-dependencies. In particular, early tourism development in the 19th century is associated with higher employment and higher wages in tourism-related and non-tourism sectors, but not necessarily in tourism itself. Contrarily, higher levels of tourism today directly cause higher levels of employment today but lower wages, particularly in tourism-related sectors. We hypothesize that, while there is support for tourism-led growth, evidence points to a specific form of Dutch Disease where tourism-intense regions are locked in a sector with particularly low productivity in Switzerland.
This paper studies the effect of nearby flash floods on firms’ establishment performance in Central America and the Caribbean. To this end, I physically define flash flood occurrences from satellite rainfall data across countries and connect them with geo-located establishment survey data. I find that a flash flood significantly decreases sales and the number of employees but increases capital productivity. This decrease in output and employment is driven by establishments where financial market access is considered an obstacle to operations, whereas the increased capital productivity occurs in establishments where it is not. The construction sector is the only industry affected differently, with no negative effect on output or employment. Overall, my results suggest that flash floods negatively impact firms and that their increase in frequency and intensity due to global warming threatens economic development. Improving financial market access is an effective adaptation strategy to increase establishments’ resilience.
Natural disasters such as hurricanes cause considerable damage to the built environment and negatively affect the economy. Consequently, the exports of goods and services from the area hit are likely to fall, stimulating the now excess demand to spill over to nearby locations. Any accounting of hurricanes’ and other natural hazards’ impact on the economy that does not consider spillovers is, therefore, likely misleading. We investigate the direct effect of hurricanes on tourism and its spillovers to other destinations in the Caribbean. The Caribbean is an ideal setting to study the effects of natural disasters on tourism, as the region is both disaster-prone and economically reliant on tourism. We connect monthly airplane and cruise ship arrivals with a hurricane destruction index from a physical wind field model that considers the local economic exposure measured by the previous year’s night light activity. Evidence from local projections suggests that airplane arrivals decrease by 3% upon hurricane impact per 1% destruction according to the damage index and only recover after 14 months. In contrast, cruise ship arrivals see larger uncertainty, and their decrease is recovered fully within four months. If a destination is hit by a hurricane, the spillover effect from nearby destinations hit is positive, implying that the total negative impact of hurricanes is much lower if one considers spillovers at the regional level. However, there is never a purely positive spillover: if a destination is itself not hit, spillovers from nearby destinations are neither positive nor statistically significant for airplane or cruise ship arrivals.
Collalti, Dino, and Eric Strobl. "Economic damages due to extreme precipitation during tropical storms: evidence from Jamaica." Natural Hazards 110.3 (2022): 2059-2086.
Collalti, Dino, Nekeisha Spencer, and Eric Strobl. "Flash flood detection via copula-based intensity–duration–frequency curves: evidence from Jamaica." Natural Hazards and Earth System Sciences 24.3 (2024): 873-890.
Collalti, Dino. "The Economic Dynamics After a Flood: Evidence from Satellite Data." Environmental Resource Economics (2024).