The western Pacific Ocean is not literally pacific—it is not peaceful. In the south, overlapping claims to rocky islets have led to clashes over oil and gas reserves and fishery resources between a half dozen countries. In the north, there are also conflicting claims to maritime features. Across the region low-level hostility is not uncommon.

Problems between Japan and South Korea are particularly thorny. Recently, before the current COVID-19 pandemic, tensions mounted, partly reflecting South Korean resentment over Japanese treatment of Korean women during World War II and Japanese reluctance to recognize this issue. Japanese claims to the South Korean administered island of Dokdo (Takeshima in Japanese) are a continuing touchstone inflaming the discussion.

In 2019, protests in Seoul resulted in boycotts of Japanese consumer goods and a slump in tourist travel to Japan. In the past, similar episodes appeared to have limited, transitory impact. Historically, consumer boycotts lose their steam over time, and tourists return to their customary destinations.

Can we rely on this? Can governments both cast their neighbors as desirable trading partners but also as unreasonable political adversaries? One reason to worry is that these two countries are among the most Internet-connected nations in the world. Social media reduces the costs of organizing protests, allowing people to share opinions and instantly take action. This could well mean that international incidents might now feed into long-lived social media protests.

This issue needs to be faced. Boycotts that threaten trade, travel, or transit directly endanger the region's continued prosperity. The development model in the west Pacific Ocean leverages international trade to provide for employment and growing incomes. This trade is often organized through complex and multi-country supply chains, and interruptions at any part of the global assembly line can be costly.

Looking at tourism can help us understand the problem. Tourism is seldom mentioned as a strategic industry. However, over time, events that might hinder tourist travel are also likely to discourage cross-border businesses. Hostile international relations lower the confidence of enterprises in cross-border trade and investment. A college student's reluctance to travel somewhere for a break may be mirrored in the planning of businesses, with companies looking elsewhere for markets and production sites.

In empirical work, I combine some features of gravity models and autoregressive models to explain the level of tourism between the two countries. The variable, TX, is travel from either Japan or South Korea to the other country (X=J,K) and is written as a function:

TX = F(TX_-t, Z_XY, DSeas_X, DShk_X ).

The variable TX_-t stands for lagged measures of the travel flows. The term Z includes the variables one might see in a gravity model, including the GDP for the travelers' originating country and cost variables, such as exchange rates or oil prices. The term DSeas stands for monthly seasonal dummy variables.

DShk stands for seven dummy variables allowing us to say whether incidents, including changing political relations, affected intercountry travel. These variables are:

• Terror: September 11, 2001, terrorist attacks in New York City.
• SARS/Iraq War: February 2003, the outbreak of the SARS coronavirus and the Iraq War.
• Earthquake: March 2011, an earthquake/tsunami in Japan.
• Hostilities #1: July 2012, protests in Seoul responding to tension over the Dokdo/Takeshima islands.
• MERS: May 2015, the outbreak of Middle East respiratory syndrome.
• Coopertion: December 2015, an easing of tension with the signing of an agreement regarding Japanese forces' wartime behavior.
• Hostilities #2: August 2019, Japan removes South Korea from the favored trade partner list.

The data are available from January 2000 through January 2020, before the beginning of the COVID-19 pandemic.

The model was estimated in linear form for each country. The non-dummy variables were expressed as first differences of the logarithmic values to provide for stationarity. In common with some other studies modeling tourist flows, the independent variables such as exchange rates or fuel costs were not generally significant. The lagged traveler variables and seasonal variables were, however, as were many of the dummy variables representing shocks (Table 1), especially for the Japanese model.

The coefficients cannot be simply compared to each other, as they pertain to dummy variables of different structures and time periods. However, we can compare them based on the suggested impact of each shock. Table 2 shows the model predictions if the dummy variables were set to zero—if the shocks did not occur. The table lists the change in the expected number of travelers compared to the historical record, measured over the presumed length of the shock and for several months after the end, corresponding to the lag lengths used in each model, four months in the Japanese case and five in the case of the South Korean model. 

The model suggests that Korean travelers reacted to the second period of poor relations between the two countries as they did when the earthquake and tsunami hit Japan early in the last decade. Similarly, the numbers of Japanese travelers fell off in the same manner that they did during the SARS and MERS pandemics.

This analysis suggests that political unpleasantness has the impact of a natural disaster or a pandemic. While earthquakes and diseases have some unavoidable elements, political relationships are under the control of the respective governments. The failure to resolve the long-running tensions between nations in East Asia has real costs and puts at risk future growth and development.  

_____________________

This post draws on a paper presented at the Western Economic Association International virtual meetings of March 2021 that can be downloaded at http://davidjaygreen.com. My email address is This email address is being protected from spambots. You need JavaScript enabled to view it..