The 2000s saw a drastic increase in the oil price, from around $30 a barrel in 2001 to more than $100 in 2008, an unprecedented high (see Figure 1). The oil price plunged as the Global Crisis hit the world economy in late 2008, but then gradually recovered to its pre-crisis level until June 2014, when it was cut by roughly a half. It has been sliding down since then. The consequences of such oil price fluctuations on the world economy are many, but they are hard to pin down as they vary across countries (Barsky and Kilian 2004, Arezki and Blanchard 2015).
In a new paper, we aim to identify one such effect, namely the trade consequence of varying shipping costs due to varying oil prices (von Below and Vézina 2016). When the oil price reached its peak in 2008, Paul Krugman argued that “higher fuel prices are putting the brakes on globalisation – if it costs more to ship stuff, there will be less shipping”. The increase in trade costs due to pricey oil may in turn have reduced real income due to fewer gains from trade. As Chinn (2008) explained, oil prices feed (roughly) into transportation costs. As oil prices and thus transport costs rise, goods markets become more insulated. More goods become non-traded, leading to higher home bias and thus higher consumption prices. Conversely, the recent deep oil price drop could be a boon for globalisation. The trade consequences of oil prices should thus be examined carefully.
Figure 1. Shipping costs and oil prices
Note: The crude oil price is in 2014 US dollars and from the BP review. Ad-valorem costs are measured as the average US import CIF-to-FOB ratio; i.e., the ratio of US import values including freight and insurance (CIF) to that free on board (FOB). This is one of the most precise available measure of shipping costs and is computable using USITC data. Sea ad-valorem cost is taken from Hummels (2007) and is defined as expenditure/import value based on US Imports of Merchandise from the US Census Bureau. It is adjusted to changes in the mix of trade partners and products. Air $ per KG is the average air cargo freight rate (in real dollars per kilogram). It is also taken from Hummels (2007) and is originally from the International Civil Aviation Organization.
Oil prices and trade costs
The aim of our research is to estimate the effect of oil prices on trade via the change in shipping costs. We face two main challenges. The first is that oil prices influence the economy through a multitude of channels, from consumer demand through the pump price to banks' transactions via inflation forecasts. Hence we need to be careful in isolating the reduction of trade due solely to increases in shipping costs. A second hurdle is to identify the change in shipping costs which is due to oil prices. Shipping costs depend on many factors other than oil, such as port fees and insurance, and are endogenous to trade. More trade means more liners and more ports which in turn mean lower shipping costs, and this may explain why the latter have been declining continuously in ad valorem terms from 6.75% in 1990 to below 4.5% in 2011 while the oil price has been rising (see Figure 1). While concurring spikes in shipping costs and oil prices in 1990 and 2000 suggest oil prices may affect shipping costs, the effect of oil prices on shipping costs remains hard to identify. Moreover, oil prices themselves may be influenced by countries' demand for foreign goods and thus may be endogenous to trade. Additionally, a third challenge is that bilateral shipping cost data are not available for many countries, and when they are it is only for the last 20 years or so.
How can we pin down the effect of the oil price on trade?
The gravity model of trade and two identifying assumptions allow us to deal with these hurdles and identify the shipping-cost effect of oil prices on trade. The gravity equation models trade between two countries as a function of the distance between them, where distance is an exogenous measure of shipping costs. What's more, the gravity model allows us to control for any country-year shock with fixed effects, thus identifying the geography of trade patterns due solely to bilateral trade costs. To include oil prices in the model, we need to make one identifying assumption; namely, that oil prices affect long-distance trade more than short-distance trade. This can be justified by shipping costs including both fixed and variable costs and oil prices affecting only the latter, hence accounting for a large share of trade costs for long distances (see the model by Mirza and Zitouna 2010). We can thus model trade between two countries as a function of the interaction of distance and oil prices. One hurdle remains, as oil prices may increase with an increase in demand for long-distance trade, and this renders the interaction of the oil price and distance endogenous to trade. We thus instrument oil prices with the yearly number of conflicts in OPEC countries, which captures supply-side oil price shocks. This allows us to obtain an exogenous oil price-driven change in bilateral trade costs and identify the decrease in trade between any two countries due to oil prices.
Does the oil price matter for trade flows?
We estimate gravity equations using port-to-port shipping distance data from Searoutefinder, an online tool, and bilateral trade data from Comtrade for the all non-landlocked countries in the world for the period 1962-2014. We find a significant interaction of distance with the oil price. In years of high oil prices, distance matters significantly more in reducing trade. In other words, trade is less global in years of high oil prices. This pattern is illustrated in Figure 2. It shows the distance elasticity of trade as a function of the oil price. When the oil price is around $100 per barrel, a 10% increase in distance reduces trade by 15% (the shipping-distance elasticity is -1.5). The recent drop to $35 suggests a return to a distance elasticity of -1.35, a boon for globalisation.
Figure 2. Low oil prices reduce the distance elasticity of trade
Note: The solid line is the estimate of our gravity parameter, the dashed-lines show the 95% confidence interval.
What would happen to trade if the oil price rocketed to uncharted territories, hitting $200 a barrel? This scenario is not as far-fetched as it sounds. It was forecast by the US Energy Information Administration before the 2014 slump (EIA 2013). According to our estimates the distance elasticity would reach -1.9. While the oil-price rise imply an increase in trade costs due to an increase in distance elasticity, it can also be thought of as increases in shipping distance (under a fixed distance elasticity). We can thus compute the percentage increase in shipping distances that would be equivalent to an oil price rise from $100 to $200, assuming the distance elasticity would remain at -1.5 (its average level when the oil price is $100). The resulting equivalent percentage increase in shipping routes is graphed in Figure 3. Longer routes would seem even longer. Countries 10,000 km apart would feel separated by 15,500 km. This increase in trade cost is equivalent to imposing a 17% import tariff. Countries separated by 1,000 km would suffer less, facing the similar effect as a 40% increase in distance, to 1,400 km, an effect similar to a 13% import tariff.
Note: The red blue line is the mean estimate and the dashed blue lines are 95% confidence intervals. The dotted line is the density estimate of shipping distance.
We have shown that oil prices affect the geography of global trade, with trade on longer shipping routes being most affected. High oil prices in the future may indeed put the breaks on globalisation, as the distance elasticity of trade is higher in years of high oil prices. Conversely, the recent drop in oil prices could be a boon for the integration of the global economy.
Arezki, R and O Blanchard (2015), “The 2014 oil price slump: Seven key questions”, VoxEU.org, 13 January.
Barsky, R B and L Kilian (2004) “Oil and the macroeconomy since the 1970s,” Journal of Economic Perspectives, 18: 115–134.
Von Below, D and P-L Vézina (2016) “The trade consequences of pricey oil”, IMF Economic Review, forthcoming.
Chinn, M (2008) “De-globalization? Musing about oil prices and trade costs,” Econbrowser.
EIA (2013) Annual Energy Outlook 2013 Early Release Overview, US Energy Information Administration.
Hummels, D (2007) “Transportation costs and international trade in the second era of globalization”,Journal of Economic Perspectives, 21: 131–154.
Krugman, P (2008) “The world gets bigger”, The Conscience of a Liberal.
Mirza, D and H Zitouna (2010) “Oil prices, geography and endogenous regionalism: Too much ado about (almost) nothing,” CEPREMAP Working Papers (Docweb) 1009, CEPREMAP.
Authors: Pierre-Louis is Lecturer in Economics at King's College London. David von Below works as an Economist at Copenhagen Economics, with a focus on energy, climate change and macroeconomics.