Since the 2007-08 food crisis, many thoughtful analyses have addressed the causes and impacts of high and volatile international food prices and proposed solutions to the crisis. These studies have covered global as well as local food price dynamics and policy reactions. The food price problem is, however, far-reaching, and its impacts are wide and interrelated. The price formation mechanism has become highly complex and dynamic. Policy actions are politically and economically sensitive. This situation calls for continuous and comprehensive assessments of the problem to provide timely and evidence-based knowledge for policy makers. This paper reviews existing evidence and theories and presents new thoughts and insights from analyses to enlighten the course of actions to be taken. Our review implies that the current body of literature concentrates on high food prices. Commodity price analysis should, however, differentiate between three types of price changes: trends, volatility, and spikes. While price trends are important in the long term, volatility and spikes are more important in the short to medium terms. Descriptive statistics indicate that all three price changes are increasing over time and show strong correlations among themselves. A rising medium-term price trend has triggered extreme short-term price spikes and increased volatility. An assessment of the costs of price volatility has shown that the existing literature follows a conventional marginal-cost approach that considers only few cost components. Direct and immediate components have not been adequately analyzed, and long-term effects have been overlooked. The effect on child nutrition and health is one such long-term effect. Under-nutrition in early childhood has negative consequences for lifetime earnings capacity because of the physical and mental impairment it causes. Economy wide distortions and misallocations also threaten the long-term development of commodity-dependent economies. Measuring and estimating the cost of food price volatility should factor in ongoing processes such as economic growth and technological changes. The supply, demand, and market explanations for high and volatile global prices have been differentiated as exogenous and endogenous factors. To help further identify the drivers of food price changes, they are categorized as root causes, intermediate causes, and immediate causes. Both empirical and theoretical evaluations suggest extreme weather events from the supply side, biofuel production from the demand side, and speculation in commodity futures from the market side are the three most important root causes of observed price volatility. The theoretical and empirical effects of speculation in commodity futures are not yet well understood. However, speculative trading in commodity futures should not be viewed as a random bet that can be smoothed out through the price system. It is important to consider the market and nonmarket contexts that guide the behavioral and strategic choices of speculators. Whereas speculation caused by manipulative, disorderly behaviors and financialization are damaging, speculation caused by demand and supply in physical markets can serve as price discovery, liquidity, and risk-hedging mechanisms. Our empirical analysis to quantify the importance of these factors shows that speculation effect is stronger than demand- and supply-side shocks for short term price spikes. Overall policy interventions at global, regional, and local levels should concentrate on reducing price spikes and protecting poor people from short- and long-term crises. The viii formulation and implementation of such policies must be supported with timely information and research-based evidence. A comprehensive portfolio of policy actions is proposed here, rather than over-extended individual measures to address the root causes or over-regulation of markets to address volatility and spikes. Evaluation of policy instruments should weigh the true costs associated with both, action versus inaction. Research must focus on developing price and food security indicators and models that will guide policy implementation also in the short run. Such models are currently missing.
