No description available.

The purpose of this thesis is to theoretically and empirically investigate the effects of institutional inefficiencies on markets and on non-elite groups, and to better understand the factors that prevent efficient institutions from evolving through the process of institutional change. It commences by reviewing the literature relating to institutions and institutional change and presenting a theoretical framework. It then presents three empirical chapters that aim to address the key questions and hypotheses relating to how inefficient institutions affect markets and why inefficient institutions persist. The first two empirical case studies of are of institutionally driven market failures that currently exist in Tanzania’s coffee and maize markets (coffee grading and maize farm gate buying). These chapters demonstrate how these failures contribute to market inefficiency and how they lower the incomes of some of the poorest groups participating in these market chains. The findings demonstrate that there is no automatic welfare maximising process in the functioning or the evolutionary path of institutions because even though these institutions are inefficient, they remain constant and largely unchallenged in the market. In other words, inefficient market institutions do not spontaneously disappear even though they disadvantage large groups. The findings also raise questions about how these inefficient institutions evolved and why they persist. The third case study of Tanzania’s agricultural market liberalisation reforms addresses these questions. It describes shifting alliances and local level resistance and shows how competition between groups around the reform period has changed their respective abilities to influence institutional change over time. Initially, elite power was characterised by the capture of local and village governments by big agricultural cooperatives during the liberalisation reform period. Subsequent to the reforms, private sector traders and processors have become powerful and influential even though they were the market underdogs for many years. This is because they have invested in reducing their influence costs by establishing strong business associations and by building strong relationships with local and village government authorities. It is argued that groups with low influence costs are more powerful and can build the links that are necessary for influencing institutional change more easily. The analysis of Tanzania’s agricultural market reforms also shows that these relative positions of power and influence evolved through a long process of distributional conflict at the micro level. The complexities, contradictions, delays and reversals of Tanzania’s agricultural market liberalisation reforms were largely determined at the most disaggregated level. Massive institutional change was taking place, but its path was steered by a drawn-out process of distributional conflict in rural villages that is still ongoing today. The findings of the coffee and maize chapters are directly linked to this above described process of distributional conflict, relative power and institutional change since the inefficient institutions analysed in the coffee and maize markets emerged as outcomes of the liberalisation reforms. What this thesis shows is that institutional change depends, to a large extent, on the preferences and responses of the most influential interest groups. The historical perspective is also important in that it acts as a clarifying lens for what may otherwise seem to be an opaque set of groups, structures and incentives. This is what this thesis has sought to achieve. By combining quantitative institutional impact investigations with interest group-based political economy and historical analyses, this research has been able to reveal the thread that links current economic outcomes with long-standing group conflict dynamics.

One cannot study poverty in Tanzania without understanding the agricultural sector, which employs more than two-thirds of the population and accounts for nearly a quarter of national GDP. This thesis examines three themes that focus on the difficulties that rural Tanzanians face in achieving a reasonable livelihood: the adverse legacy of a failed historical policy, a difficult climate, and market failures. The first empirical chapter examines the legacy of the villagization program that attempted to transform the predominantly agricultural and rural Tanzania. Between 1971 and 1973, the majority of rural residents were moved to villages planned by the government. This essay examines if the programs e↵ects are persistent and have had a long-run legacy. It analyzes the impact of exposure to the program on various outcome measures from recent household surveys. The primary finding of this study is that households living in districts heavily exposed to the program have worse measures of various current outcomes. The second empirical chapter examines the role of reliability of rainfall, which is important in Tanzania as agriculture is predominantly rain-fed and a small fraction of plots are irrigated. This chapter investigates if households cope with this major risk to income by re-allocating their labor supply between agriculture, wage labor, and self-employment activities. This chapter combines data on labor allocation of households within and outside of agriculture from the National Panel Survey with high-resolution satellite-based rainfall data not previously used in this literature. The primary finding of this study is that households allocate more family labor to agriculture in years of good rainfall and more labor to self-employment activities in years of poor rainfall. Market failures are often cited as a rationale for policy recommendations and government interventions. The third chapter implements four tests of market failures suggested in the literature, all of which rely on the agricultural household model but di↵er in how market failures are manifested. The common finding of these tests is that market failures exist in agricultural factor markets in Tanzania, although significant heterogeneity exists. Markets are more likely to fail in rural areas, remote locations, and are more likely to affect female-headed households. Households are also more likely to face market failure when they try to supply labor to the market than when they try to hire labor from the market.