FOOD PRICE SURGES, TRADE POLICY RESPONSES, AND THE WELFARE IMPLICATIONS TO TANZANIAN HOUSEHOLDS

Food price turbulence worldwide in 2007/2008 renewed interest in the role of trade policies in food security. In East/Southern African countries, simultaneous increases in the price of maize - the major staple food in many of these countries - were observed during this period. Several countries imposed discretionary trade policies, such as export bans and tariff waivers, citing food security concerns as the primary reason. The two main objectives of this dissertation are to empirically investigate the role of discretionary trade policies, particularly export bans and tariff waivers, in food security and to assess the impact of the maize price changes on household welfare. To achieve these objectives, this dissertation is composed of three essays: the first investigates the price dynamics and the role of trade policies; the second assesses the impact of food prices to consumption patterns;and the third estimates the short and long run impact of food price changes on household welfare.The first essay tests the effectiveness of export bans as a price insulation policy using panel data of local maize prices across 35 markets in 7 countries (Tanzania and its neighboring countries of Burundi, Kenya, Malawi, Mozambique, Uganda, and Zambia). While the data have a relatively high rate of missing values, I use a multiple imputation method to utilize as much of the available information as possible for the analysis. Using an extended competitive storage model, I analyze the dynamics of local maize prices and test the effect of export bans. Not only I reject the null hypothesis of no impact of export bans on local maize prices, but I find that the export ban is associated with higher maize prices. In other words, an implementation of export ban induces higher prices due to the speculative demand instead of keeping the price low through the increased availability at the local market. Moreover, the impact of the export ban is conditional on the existence and magnitude of trade costs: with a high trade cost, the trade flow would not exist prior to an export ban and the ban would not binding. I find that the upward pressure of the export ban on local prices would be underestimated, if such non-binding cases are not taken into account.The second essay estimates a large-scale demand system of foods, and assesses the impact of price changes on food consumption and nutrition consumption patterns of Tanzanian households. Malnutrition is a widespread issue in Tanzania. For securing a sufficient supply of food, particularly grains, price policies such as subsidies or protective trade measures are exercised in Tanzania and many of its neighboring countries. However, evidence on the effectiveness of price policies for improving nutritional intakes are mixed. In the literature, studies on nutrient consumption are particularly scarce, and the roles of income and socioeconomic characteristics is often investigated separately from price responsiveness of food consumption. I assess the price responsiveness of foods and their associated nutritional intakes by estimating a flexible demand system - Exact Affine Stone Index demand system to overcome the limitation of alternative demand systems - using Tanzanian living standard surveys.Using the estimated elasticities, I find that the price of maize is not very influential to the caloric intakes even among very poor households due to these households' high capability to substitute foods and secure the level of energy intake. However, in exchange, intakes of other nutrients are greatly reduced when the maize price increases, due to the major role of maize as a relatively cheap source of calories and other nutrients.Evaluated at the historical maize price changes from 2007-2008, simulations indicate a small decline in the average energy intakes. However, the decreases in average micronutrient intakes are large and the deficiency incidence rises significantly.The third essay extends the framework of the second essay by incorporating the production of foods in households and the general equilibrium effects of food price changes to the wage, assessing the overall welfare changes. Many households in Tanzania are both consumers and producers of food crops.Higher food prices increase the real income of such households, and thus have a potential to significantly reduce the loss or even realize a net gain. In addition, higher prices of agricultural goods - or, in general, traded goods - may increase the demand for labor as the production level is increased.This higher demand, in turn, would increase the return to the wage, under a set of certain assumptions. This general equilibrium effect of food prices to the wage level is evaluated empirically using the living standard survey of Tanzania.Using the estimated price-wage elasticities and the data of budget and income shares, I assess the average welfare impact of the historical maize price increase across sub-populations of households. I find that the net effect through consumption, production, and wage earnings are, on average, positive for all the household across income levels.However, the net gains are much stronger for wealthier households and almost negligibly small for poorer households because of the dominantly strong gain from the wage earning increases compared with the smaller gains from production net consumption. Moreover, I identify the rural and/or farm households experience disproportional net loss, for significantly higher reliance on maize in consumption, compared with urban and/or non-farm households.In summary, I find that the role of discretionary trade policies - export bans and temporary tariff rate changes - to be insignificant or contrary to the intended effect of price insulation. By assessing the impact of food prices on household welfare, I find potential gains from higher food prices, maize price particularly, although such gains are concentrated in certain sub-populations.I also find adverse effects of higher maize price on food consumption and nutritional intakes are not immediately visible due to a high capacity of households to substitute. These findings suggest, as a general statement, evaluations of policies regarding the food price require a careful examination of the conditions under which the intended policy may be effective in achieving its objective, and an assessment of the heterogeneous impacts of the intended policy on various sub-populations.Failing to do so, policies that are aimed at improving the welfare of people may lead to unintended consequences of failing the objective, misidentifying the target populations, or missing the potential opportunities.