Increases in agricultural commodity prices and food prices in recent years have raised concerns among policymakers about a global food shortage. For example, the director of the International Food Policy Research Institute testified in 2008 that rising prices for agricultural crops were causing food riots in many developing countries, and that, according to the Food and Agriculture Organization of the United Nations, 37 countries were facing food crises (Rosegrant 2008). By 2011, the World Bank predicted that millions more people would be driven into poverty by higher food prices in the absence of policy changes (Iman 2011). Even in the US, the press lamented the squeeze on American families facing the fastest rise in retail food prices in 20 years. In response to these concerns, numerous government agencies pondered the causes and policy implications of price volatility in food and agricultural markets.
Many academics and policymakers were quick to relate higher food prices to the persistent increase in the price of crude oil after 2003. Their reasoning was not only that oil products are required to run agricultural equipment, raising the costs of producing food commodities, and that higher oil prices may raise the price of processing, storing, and distributing food to retail customers. They also pointed to the increased reliance on biofuels in industrialised economies – notably in the US.
The US has long pursued policies favouring the production of ethanol from corn. One of the chief uses of ethanol is as an additive in producing high octane fuel for car engines. An important shift in US policy occurred in 2005, when policymakers refrained from providing liability waivers for the use of environmentally harmful gasoline additives traditionally used in producing high octane fuels in many US states. This decision effectively made ethanol produced from corn the only gasoline additive available to US refiners after May 2006, and expanded the market for ethanol. At the same time, the US introduced renewable fuel standards mandating the use of ethanol in the production of gasoline. These standards have been successively tightened since then.
The resulting increased demand for ethanol (and hence for corn) was soon followed by persistent increases in the price of corn and other food crops. A common belief among policymakers has been that the use of corn and oilseeds as a biofuel has also enhanced the transmission of oil-price shocks to agricultural commodity prices and retail food prices. One concern is that industrial uses of corn are directly competing with consumer demand for corn. A related concern is that corn is also used to feed livestock, so higher corn prices tend to be followed by higher meat prices. Finally, corn is competing with other crops such as soybeans for scarce water and for land.
Evidence on the link between oil and food prices
Our recent work examines the evidence for the view that food prices are driven by oil prices (Baumeister and Kilian 2013). It is useful to examine the data first. For example, we show that, contrary to popular views, there has been no systematic increase since 2006 in the volatility of the real prices of food commodities, and only a minor increase in the volatility of US retail food prices. In fact, the volatility of the growth rate in the real price of rice and soybeans actually fell, while that of corn increased by only 11%. The largest increase in price volatility, 29%, occurred not in corn – which presumably would be most affected by biofuel policies – but in wheat.
Retail food prices vs. food commodity prices
In addition, there is no evidence that US retail food prices closely track oil prices before or after the change in US biofuel policies in May 2006. Notwithstanding substantial variation in the real price of crude oil, indices of the real retail price of food faced by US consumers remained remarkably stable over time – even after the shift in US biofuel policies. For example, after adjusting for inflation, the retail price of food in the US since 2006 has grown at a modest average rate of 1% per year.
However, somewhat different results hold for the real prices of major food commodities produced in the US, such as corn, wheat, soybeans, and rice, which closely mirror food commodity prices in global markets. Corn and soybeans experienced average real price increases of between 12% and 15% per year after May 2006, while the real prices of wheat and rice grew at an annual rate of 7%. Most livestock prices experienced average real increases closer to 4%. Hence, the distinction between retail food prices and the prices received by farmers for grain crops and for livestock is important.
Thus, policy concerns about a looming food shortage, if they have any foundation, must be related to rising real crop prices received by farmers as opposed to rising real food prices faced by consumers. The discrepancy between the slow growth in real consumer food prices and the more rapid growth in the crop prices received by farmers is explained by the small producer share in the food prices paid by US consumers. For example, the farm value of wheat in the price of bread is only about 5%, so even substantial wheat price increases are associated with only small increases in the price of bread.
Explaining increases in food commodity prices
This evidence raises the question of how much of the increase in food commodity prices can be explained based on rising oil prices. Our study shows that after May 2006, unexpected increases in the real price of oil have indeed been followed by systematic increases in some US crop prices. For example, a 1% real oil-price shock tends to be followed by a persistent and statistically significant increase in the real price of corn that peaks at 0.5% one year later. These effects should not be interpreted as the causal effects of higher oil prices, however, because the real oil-price shocks since May 2006 mainly reflected broad-based shifts in the demand for industrial and agricultural commodities (Kilian and Murphy 2013). One way of illustrating this point is to show that even the real price of fertiliser – which does not depend on the price of crude oil but on the price of natural gas – appears to respond positively to real oil-price shocks. Given the divergence of the prices of oil and natural gas in recent years, this evidence can only be explained by increased demand for both crude oil and food products.
In fact, 27% of the variation in the growth rate of the real price of corn since May 2006 can be explained by unexpected changes in global demand. Corn-ethanol mandates are by no means the only explanation for the remaining variation. Our analysis suggests that a combination of idiosyncratic demand shocks, idiosyncratic harvest shocks, and government policies is required to explain the residual variation in the crop price data. In other words, there is no simple explanation of the evolution of the real prices of corn, wheat, and soybeans.
This conclusion also holds for the real price of rice, which is perhaps the single most important food commodity for many developing countries. There is no doubt that increases in the real price of rice were not caused by US biofuel policies, because rice production is water-intensive and restricted to a handful of areas of the US mainly along the Mississippi, whereas corn, soybeans, and wheat are primarily grown on dry land in the Midwest. In other words, rice is not competing with corn, soybeans, or wheat in production. Although both rice and oil prices surged after 2006, Figure 1 illustrates that the real price of rice received by US farmers continued to increase even as the real price of oil collapsed in the second half of 2008, and that the real price of rice declined in 2009 and 2010, while the real price of oil recovered. This evidence highlights the lack of a systematic statistical relationship.
In fact, it can be shown that even the direct effect of higher oil prices on farm prices is quite small. For example, all else equal, a 100% increase in the real price of farm fuel (which mainly consists of diesel fuel) would only raise the cost of farming by about 5%, given the small cost share of farm fuel in farm production. A case study based on the large independent variation in the real price of oil associated with the invasion of Kuwait by Saddam Hussein in 1990 reveals no evidence of this major oil-price shock having been transmitted to farm prices.
This leaves the concern that unexpected increases in the real price of oil are associated with higher retail food prices because of the increased cost of processing, transporting, and distributing food to retail customers. Focusing on different stages of production starting with the price received by the farmer and ending with the price paid by the retail customer, we document that there is no evidence of a systematic effect of oil-price shocks on any of the price spreads along the value-added chain. The lack of feedback from oil prices to food prices should not be surprising given the overall stability of real retail food prices in recent years, even as the real price of oil fluctuated considerably (see Figure 2). This finding is nevertheless remarkable, as it runs counter to the conventional wisdom in the policy debate.
Figure 1. Real prices of rice and crude oil, May 2006–May 2013
Source: Baumeister and Kilian (2013).
To summarise, we find no evidence that rising food prices have had a significant negative impact on consumers in the US. One would expect similar results for other industrialised economies (and increasingly for emerging economies), in which food consumption consists mainly of processed foods.
A different question is to what extent increases in the real price of food commodities affect the welfare of poor people in the least industrialised countries, who tend to consume less processed food and spend a larger share of their income on food. The answer is not obvious, because poor farmers in developing countries are likely to be producers as well as consumers of food commodities. Providing a reliable answer to this question requires detailed household survey data for extended periods. Recent studies show enormous variation in the effects of higher food commodity prices on poverty in developing economies. For example, there is evidence that an increase in the real price of rice raises welfare in countries such as Vietnam, given the relatively egalitarian land distribution in that country, while lowering welfare in countries such as Pakistan with a higher fraction of poor urban consumers (Ivanic and Martin 2008). Our interpretation of this evidence is that, whether a country benefits from rising food commodity prices or not, in important part appears related to the agricultural, exchange rate, and trade policies it pursues, which tend to predate recent food commodity price increases. We make the case that a more nuanced analysis is called for in developing policy proposals for dealing with these problems than is embodied in many current policy statements.
Figure 2. Price of crude oil and US Consumer Price Index for food relative to US non-food CPI, January 1974–May 2013
Source: Baumeister and Kilian (2013).
Disclaimer: The views expressed in this article are those of the authors and should not be attributed to the Bank of Canada.
Baumeister, C and L Kilian (2013), “Do Oil Price Increases Cause Higher Food Prices?”, CEPR Discussion Paper 9689.
Iman, P (2011), “Food Price Rises Pushing Millions into Extreme Poverty, World Bank Warns”, Guardian, 14 April.
Ivanic, M and W Martin (2008), “Implications of Higher Global Food Prices for Poverty in Low-Income Countries”, World Bank Policy Research Working Paper 4594.
Kilian, L and D P Murphy (2013), “The Role of Inventories and Speculative Trading in the Global Market for Crude Oil”, Journal of Applied Econometrics, forthcoming.
Rosegrant, M W (2008), “Biofuels and Grain Prices: Impacts and Policy Responses”, Testimony for the US Senate Committee on Homeland Security and Governmental Affairs, 7 May.