If your job were suddenly in jeopardy, and you couldn’t predict next month’s income, you might put off buying that new car you had been eyeing. Similarly, an appliance company faced with sudden hardship among its customer base might delay building a new refrigerator-manufacturing plant. It turns out that uncertainty affects international trade flows, too. In a working paper published by the World Bank’s International Trade Department, we explore how that uncertainty hurts some types of trade more than others.
The research, entitled “Innocent bystanders: How foreign uncertainty shocks harm exporters,” is an attempt to address a shortcoming in scholars’ ability to anticipate the trade-related impacts of crises. In 2008, for example, trade economists failed to predict the drop in international commerce that followed the collapse of Lehman Brothers, an event that shook the global financial system. More recently, trade economists have failed to capture the full impact of the European debt crisis on exporters across the globe. While models and simulations have been getting better, the impacts of such shocks on trade are still poorly understood. Our research explores a theory that places exceptional uncertainty at the beginning of a chain-reaction that ultimately slows trade.
We find that, when the shock is big enough, uncertainty is an important factor in predicting a slump in trade flows. It finds that exporters can be hurt by uncertainty in the countries where they are sending their goods. It also finds that exporters of durable goods, such as cars, appliances, and other long-term-use commodities, are more likely than exporters of consumable goods to be affected by uncertainty in destination countries.
Our data analysis is grounded in research on producer and consumer behavior. The literature shows that both producers and consumers take a “wait-and-see” approach when the market is volatile; during uncertain periods, they freeze activity and postpone purchasing decisions. In particular, they delay the purchase of capital goods (for producers) or durable goods (for consumers) whose purchase is difficult to reverse. This resonates with every-day decision-making. If your income was uncertain, you would be likely to shop at the grocery store, but postpone your trip to the car dealership. The literature also shows that once uncertainty subsides, pent-up demand leads to a medium-term period of overshooting in production and consumption of durable goods. This also sounds intuitive: once your paycheck is assured, you decide to go buy that car and also replace your refrigerator, which had broken down in the meantime.
Data on trade flows measures consumer and producer decisions in aggregate. So for the impact of a shock to be detected in trade data, it must affect a critical number of consumers and producers. Our analysis found that if a shock is big enough, it does indeed change trade flows between countries. Specifically, we used bilateral trade data from 32 countries to look at the impact on trade of uncertainty (proxied by stock market volatility and other financial indices) in the importing country. So, for example, uncertainty in the US would affect German, Chilean, and Australian exporters. Our results show that in these moments of big-enough-shocks, uncertainty leads to an 11.5 percent reduction of aggregate exports in partner countries.
We found, as we expected, that the negative impact of uncertainty generally increases as trade relationships gain intensity in durable goods. There is a surprise, however: the countries at the highest end of the spectrum – those that are most intensive in durables-trading -- are resilient to uncertainty. For example, Japan’s trade with the US is more than 80 percent durables, and is relatively unaffected by uncertainty shocks. While this dynamic deserves more exploration, we have some hypotheses about why it might occur. One is that the prevalence of value-chain production methods might mean that some countries exporting durable goods might be removed from the final consumer. For example, it might happen that the production of a computer goes through a three-country-process: microchips made in China, assembly done in Costa Rica, and the finished product sold in the US. Each transaction involves the export of a durable good, but an uncertainty shock affecting consumers in Costa Rica would not affect demand for the product, so it would not impact the trade relationships.
Another explanation for the immunity to uncertainty in high-durables trade relations has to do with possible substitution goods. For example, an uncertainty shock in the UK might cause a decrease in demand for Sport Utility Vehicles (SUVs) from the US, which might be considered luxury goods. But consumers facing uncertainty about future income still have current income to spend. They might spend that money on laptop computers, which they consider more of an every-day staple. So while US car exports to the UK might decrease, its computer exports would increase, and both are durable goods. To understand the dynamic at that level of detail, however, we would need more nuanced information about consumer preferences in importing countries. This is an area for possible future study.
Finally, we asked whether a country’s experience in facing an uncertainty shock would help them navigate subsequent shocks. Did they learn to implement a policy, for example, that would mitigate the downturn in trade? We found that they did not. In fact, countries with prior experience in confronting a shock did not behave any differently when confronted with another. This result suggests that countries could benefit from interventions that better anticipate uncertainty shocks -- or that ensure that the shocks are as short as possible – whether through an automatic stabilizer triggered by an early indicator of crisis or some other action that could offset the trade losses brought by uncertainty. In any case, we hope that our research will add to a conversation about this issue and help countries maintain economic stability in an ever-more-connected world.
Periods of exceptional uncertainty.
Bloom, N. (2009, 05). The impact of uncertainty shocks. Econometrica 77(3), 623-685.