Comment on “The Long‐Term Impacts of the 1998 Nagano Winter Olympic Games on Economic and Labor Market Outcomes”

Abstract

What is the impact of the Olympic Games on an economy? Hosting an Olympic Games potentially offers benefits for the economy of the host city or region. The benefits include infrastructure construction related to the event, improved road or rail networks, consumption by tourists, and job creation. However, there are also potential drawbacks of hosting the event, such as high costs and underutilized facilities. Economists have studied the economic effects of the Olympic Games, but the existing empirical studies show mixed results. Miyoshi and Sasaki (2016) study the long-term effects of the 1998 Nagano Winter Olympic Games on output and labor market outcomes in Nagano City and its neighboring areas. To estimate the impact of the Nagano Olympic Games on the local economy and labor markets, they compare the actual outcomes caused by the Olympic Games with a counterfactual outcome that would have been observed if the Olympic Game had not been held. The counterfactual outcome is constructed by using the synthetic control methodology developed by Abadie et al. (2010). Miyoshi and Sasaki find that the Nagano Olympic Games induced population inflows into Nagano Prefecture and had a positive impact on the local economy. They also find that the Nagano Olympic had no effect on the local labor market. Mega sports events, such as the Olympic Games and the World Cup, are expected to have not only short-run effects, but also long-run effects on the economy. Miyoshi and Sasaki (2016) place the long-run effects at center stage when they study the effect of Nagano Olympic Games on the local economy and labor market. This is a valuable paper for readers who are interested in economic effects of mega sports events. I have a few comments on the analysis presented in the paper. Miyoshi and Sasaki (2016) examine whether the Nagano Olympic Games has a positive effect on the local GDP by comparing the actual and estimated synthetic counterfactual trends in Nagano Prefecture's GDP. Their analysis shows that the Nagano Olympic Games had a positive impact on the local economy in the long run. However, their analysis did not measure the size of the impact. Since we are interested in knowing how much the event boosted the local economy, Miyoshi and Sasaki should measure the size of the impact of the Nagano Olympic Games on the local GDP. Miyoshi and Sasaki (2016) demonstrate that the Nagano Olympic Games had a significant positive impact on Nagano Prefecture's GDP and production in the real estate sector, but it did not have a significant impact on the local GDP per capita. A key to understanding these results is population dynamics. Miyoshi and Sasaki (2016) show that the Olympic Games induced population inflows into Nagano Prefecture, and argue that the increase in Nagano Prefecture's population increased production in the real estate sector, and Nagano Prefecture's GDP. The finding that the Olympic Games had a positive impact on the local population is of interest. I am wondering how the Olympic Games induced the population inflow in the long run. Miyoshi and Sasaki should provide some economic intuition for this result. To investigate the impact of the Nagano Olympic Games on the local labor market, Miyoshi and Sasaki (2016) use administrative data on job findings (Shokugyou Antei Gyoumu Tokei). As they mentioned, while this dataset comprehensively covers labor market conditions in local areas throughout Japan, it has a considerable drawback: sample selection bias. The dataset only captures job seekers and job vacancies registered at local employment service offices. Thus, unregistered job seekers and firms with vacancies are excluded from the dataset. According to the Labour Force Survey, less than half of the unemployed workers visit employment service offices as the primary way of searching for jobs. This implies that a large proportion of unemployed workers search for jobs without using the Employment Agency's services. Similarly, it is known that a large proportion of firms with vacancies search for workers without using the employment service offices (see e.g. Kodama et al., 2004). This leads to two questions: How serious is the sample selection bias? How does the sample selection bias affect the results of the paper? Sample selection bias may mitigate the positive effect of the Olympic Games on labor market outcomes. This implies that Miyoshi and Sasaki's conclusion may not be very accurate. I think that authors need to provide a better discussion of the problems caused by sample selection bias. I am also wondering how the Olympic Games affected wages, which are ignored in their analysis.