An examination of large-scale land acquisitions and their possible effects on growth volatility

Abstract

Abstract Introduction Large-scale land acquisitions in developing countries became notorious in the early 2000s. Research carried out in those early years explored their effects from multiple angles (e.g. ecology, water, gender, local population, etc.) and found their outcomes mostly negative. More recently however, academics started publishing articles which viewed such deals as a form of investments, and highlighted their possible positive outcomes. Since majority of developing and least developed countries' economies rely heavily on agriculture it is surprising that so far there had been no detailed analysis of the effects of agriculture either on economic growth or growth rate volatility. At most, analyses included agriculture as one of economic sectors and, in conjunction with other sectors, explored its effects on the economy. However, those studies are not detailed enough to understand agriculture specific economic effects, and to guide relevant policies either in general or in relation to large-scale land acquisitions. Ideally, my research would explore the economic effects of large-scale land acquisitions on growth rate volatility in developing countries. However, such direct approach was not feasible due to lack of necessary data in relation to large-scale land acquisitions. Therefore, I first explored large-scale land acquisitions from various perspectives with a focus on the composition of agricultural produce coming from such acquisitions. Secondly, I explored the effects of crops' production and export structure on growth rate volatility in developing and least developed countries. Methodology Two different research methodologies are used in my thesis; one for large-scale land acquisitions chapter, and the other one for the two analytical chapters. Exploration of large-scale land acquisitions was limited by available data. Data was gathered from an open-sourced Land Matrix database. The database compiles information on land deals in developing countries which took place after the year 2000, and involve an area of at least 200 hectares (my dataset spans from 2000 to 2017). After eliminating entries where information was incomplete, I explored different attributes of investor and host countries (e.g. where investments go/come from; how diversified investments are, etc.), and then focused on detailed analysis of crops' production structure (e.g. what types of produce are grown on such lands, how diversified crops are, etc.) In the two analytical chapters of my thesis I explore influence of crops' production and export structure on growth rate volatility in developing and least developed countries. I used Food and Agricultural Organisation's data on agricultural crops' production and export values in 64 countries from 1970 until 2010. Due to the dynamic nature of my dependent variable I employed generalised method of moments regression analysis to understand their effects. Chapter exploring crops' production structure explores individual crops as well as different crop groupings (e.g. plantation/non-plantation; perishable/non-perishable; food/non-food; crop groups e.g. vegetables, fruits, etc.), and different forms of diversifications. On the other hand, richer data used in chapter exploring crops' export structure enabled me to further divide those same crop groupings into their processed and unprocessed parts. Results The chapter exploring large-scale land acquisitions provided many interesting information on investors behaviours. It seems that investors prefer to invest in domestic region when available or the closest appropriate region in the cases of Europe and Northern America. The majority of deals explored are intended for food production, preferably non-plantation, and non-perishable ones (e.g. cereals). Among non-food crops the top two are bio-fuels and rubber. The analytical chapters showed that increased share of production value of plantation crops in GDP significantly decreases growth rate volatility however their exports don't have any significant effects. Ratio between non-plantation and plantation crops is also significant with increased share of production of non-plantation crops relative to that of plantation crops increasing growth rate volatility. Plantation crops are present in almost half of all deals in the Land Matrix. Increased share of production of perishable crops in GDP again decreases growth rate volatility. Their growth rate volatility decreasing effect is strengthened with higher level of openness. None of the examined groups' exports have significant effects on growth volatility. Perishable crops represent 35 percent of crops produced on those lands whereas 65 percent are non-perishable. Increased share of production of food crops in GDP significantly decreases growth rate volatility however their exports have no significant effect. In the database 72 percent of deals are intended for food crops production. Productions of four crop groups have growth rate volatility decreasing effect (vegetables; nuts; coffee, cocoa, and tea; and rubber) whereas oil crops increase growth rate volatility. Openness significantly strengthens growth rate volatility decreasing effect of two crop groups: coffee, cocoa and tea; and tobacco (which although not significant has growth rate volatility increasing effect). Surprisingly none of the crops groups' exports significantly influences growth volatility. Vegetables account for 30 percent of food crops in the dataset, and rubber for 33 percent of non-food crops. On the production side, the only significant diversification variable is diversification between food and non-food crops. Higher diversification between food and non-food crops increases growth rate volatility which is a surprising result and could probably be explained by diversification's negative effect on domestic food supplies. On the other hand higher diversifications of exports of different crop groupings significantly decrease growth volatility.