Essays on spatial and vertical price transmission

This is the final outcome of a three-year joint PhD programme carried out at the Sapienza University of Rome, Italy, and at the Friedrich-Schiller University of Jena, Germany, within the framework of a co-tutelle agreement between the two universities. The first essay (“Vertical price transmission, geographical dispersion and the structure of the Ugandan coffee market”) extends the vertical price transmission analysis through a structural approach, which evaluates whether spatial oligopsony power is prevailing in Ugandan coffee market and in case how strong it is. The first paper tests whether in markets, such as Uganda, where infrastructure quality is poor and transport costs are relevant, geographic dispersion of smallholder farmers allows traders to exploit their market power against farmers with a large impact on market structure and reduction of farmers' welfare. By building upon (Sexton, 1990), the study brings an original contribution to the literature, since (Sexton, 1990) develops just a theoretical model and it is not interested to do any econometric exercise. (Sexton, 1990) employs a single spatial price gap equation instead of a system of well-founded behavioural equations in agricultural markets, which is indeed a major improvement delivered by this essay. Moreover, in this analysis the approach to spatial price gap determination is combined with the oligopsony modelling and SUR technique in order to produce empirically testable hypotheses. Without such transformations the approach by (Sexton, 1990) cannot be employed for any empirical exercise. Indeed, the idea of the role of distance is taken from (Sexton, 1990) and introduced in an original way in a more sophisticated model, which is micro-founded at three levels, i.e. demand and supply of agricultural commodities by traders and farmers as well as conditional demand of inputs by farmers. Since the wholesale-farmgate price spread is net of transport costs, results confirm that geographic dispersion of smallholder farmers plays a significant role on price margin and that there is room for local oligopsony, because traders exploit their market power and overcharge transport and transaction costs to farmers. Indeed, farmers are not able to skip traders in the value chain, because a significant information asymmetry is prevailing in the market. Traders exploit farmers' ignorance because the latter are small and dispersed as well as they lack information about current market prices because of villages remoteness and poor communications with marketplaces (Courtois and Subervie, 2015). Moreover, farmers are not aware of actual transport costs faced by traders, which carry larger quantities of coffee than single smallholder farmers and spread fixed costs over a larger amount of crop. The second essay (“Spatial price transmission and trade policies: new evidence from selected sub-Saharan African countries and crops with high frequency data”) assesses the impact of trade policies on spatial price transmission of maize, rice and wheat in Cameroon, Kenya and Tanzania. This paper improves the existing literature in the field (see, inter alia, Anderson and Nelgen, 2012a, Anderson and Nelgen, 2012b and Anderson and Nelgen, 2012c), because it estimates the impact of tariff and non-tariff trade policies on spatial price transmission in the agricultural markets using monthly data. Employment of monthly data allows assessing more precisely short-lived movements of the analysed series, which could disappear because of aggregation bias at lower yearly frequency, thus providing a better identification of insulation policies. Furthermore, this essay focuses on the impact of both tariff and non-tariff barriers on spatial price transmission by taking advantage of the combination of the FAO-GIEWS (Global Information and early warning system) database and trade policies information from the FAO-FADPA (Food and Agriculture Policy Decision Analysis) with the recent release of the World Bank World Integrated Trade Solutions (WITS) Database (UNCTAD, 2016) (FAO, 2016) (FAO, 2016b) (World Bank, 2016). This latest WITS release provides monthly ad-valorem equivalent tariff rates consist of tariff, para-tariff and non-tariff measures. In particular, non-tariff barriers comprises technical measures, such as sanitary or environmental protection measures, as well as others traditionally used as instruments of commercial policy, e.g. quotas, price control, exports restrictions, or contingent trade protective measures, and also other behind-the-border measures, such as competition, trade-related investment measures, government procurement or distribution restrictions (UNCTAD, 2015). The empirical methodologies of this study, like threshold, fractional integration and panel estimation, allow to separately estimate the confounding factors and clean the estimates of the variables of interest from them. In particular, while the confounders cannot be identified, the coefficients of the variables of interest are consistent and they can be properly identified, conditional on the estimate of the confounders. An additional value added of this work is the possibility to separately estimate the impact of trade policies within the two regimes of behaviour of the domestic price series: in the first regime the trend of domestic prices is increasing, in the second one the trend is decreasing. It highlights that trade policies play a role both in case of increasing and decreasing domestic prices, but their relevance is much larger, if prices are increasing. The policy implication is that trade policies were able to insulate the country from the price shocks on the international markets during the food price spike crisis, when it was mostly needed. By presenting high frequency analyses and techniques able to detect non-linearities in the data generating process we thus provide results which are different from the standard literature (Anderson and Nelgen, 2012b and Anderson and Nelgen, 2012c). Note however that although the impact of these instruments is proved to be relevant in the short term during the food price spike crisis, these policies could not be regarded as long term solutions. The third essay ("Shocks, price transmission and Food consumption with changes in Price risk aversion) looks seriously at the issue of time variant price risk aversion parameters. This is a fundamental question to address in the investigation of both spatial and vertical price transmission in a risky environment. To this end, the essay assesses the behaviour of farm households, which consume and produce crops at the same time, and answers the following key research questions: i) whether the occurrence of exogenous shocks induces a change of price risk aversion over time and then ii) how the time-varying risk aversion parameter affects production and consumption pattern by the farm households. This research employs the risk aversion parameter introduced by (Bellemare et al., 2013), which takes into account not just the household psychological risk attitudes, but also the market imperfections and availability of institutions which facilitate risk-bearing (Mendola, 2007) (de Janvry et al., 1991). Nevertheless, unlike (Bellemare et al., 2013) the essay develops a microfounded empirical model, where the risk aversion parameter is allowed to change over time and not just across households. This empirical model is estimated within a two-stage structural approach. The results of the empirical analysis suggest that the risk aversion parameter is not constant over time and that households can become more risk averse, if they face adverse market conditions in the previous periods. Furthermore, this paper provides evidence that peasants do not aim just at need satisfaction, but they behave in an optimal way and make sure their food security in the medium term. Indeed, they prefer to increase their income instead of directly consuming the harvested crop, because the reduction of dietary energy consumption derived from the giving up the harvest for sale is more than offset by the rise of food purchasing power due to the larger profits obtained.