Narrative Global Food Commodity Market Shocks

Online appendix to “Macroeconomic Effects of Disruptions in Global Food Commodity Markets: Evidence for the United States” This appendix documents the construction of a narrative series of exogenous global food commodity market shocks. More specifically, we rely on historical documents to identify episodes of changes in food commodity prices that were unrelated to the state of the economy, i.e. movements where the proximate causes were disturbances in food commodity markets. The series is constructed by reading FAO reports, newspaper articles (e.g. the Financial Times archive), disaster databases (e.g. EMDAT) and several other online sources (e.g.Google). The task is daunting given the global level of the analysis. There are continuously, many times even conflicting, events affecting food commodity markets somewhere in the world. We therefore only select episodes that fulfill the following criteria: 1. There has to be an event that is important enough to affect food commodity markets at the global level, such as weather shocks in a major food production region, or unantici- pated news on the volume of global food production (e.g. a sizable revision of expected agricultural production by the USDA). 2. The event should have an unambiguous significant effect on global food commodity prices. A shift in commodity prices is considered to be significant if either the quarterly change in food commodity prices or accumulated change over two subsequent quarters is at least one standard deviation different from the sample mean. 3. There should be no developments in the macroeconomy, alternative events or macroe- conomic news that may also have a recognizable impact on food commodity prices. For example, we exclude admissible food market events if there is simultaneously a signif- icant shift in crude oil prices (one standard deviation different from its sample mean), or in economic activity (e.g. a global or US recession). Put differently, we eliminate or minimize possible endogenous movements in food commodity prices to current or future fluctuations in the business cycle, i.e. the event in food commodity markets has to be the proximate cause of the price shift. All ambiguous cases are not selected as an episode. By applying these criteria to the historical records, we were able to identify 13 episodes that could reasonably be interpreted as major exogenous food commodity market disturbances that are unrelated to the state of the economy. 6 of these episodes are unfavorable food market disruptions, whereas we have detected 7 favorable shocks to food commodity markets. The remainder of this appendix motivates the selection of each episode. We also provide excerpts of articles and reports on which we based our motivation. Relevant quotes are marked in bold. Unless otherwise mentioned, changes in real food commodity prices are calculated as the change in the food commodity price index from the International Monetary Fund (IMF), deflated by US CPI. The real oil price series is the refiner acquisition cost of imported crude oil, deflated by US CPI. Cereal production is the global (annual) production of corn, wheat, rice and soybeans downloaded from FAOSTAT, aggregated on a caloric-weighted basis as described in the paper.