The subject of my dissertation is the study of coal procurement by electric utilities in the US over 2 decades, from 1979 to 2000. Energy markets are typically characterized by severe contracting problems. Buyers and sellers therefore employ various instruments, such as contract length or complex pricing arrangements, to restrict these problems. Relationship specific investment, wherein buyers make investments specific to their suppliers, has been advanced as a prominent explanation for contractual length.

Investment decisions are however endogenous in length or pricing, making causal identification of the role of investment specificity difficult. In my first chapter, I attempt a resolution. I use the 1990 Clean Air Act Amendment as an exogenous shifter of the extent of relationship specific investment. A key feature of the Amendment's design helps me define a difference-in-difference model arguably free of the endogeneity issues discussed above. I find that the plants forced into switching - Phase I plants located in the US Midwest - are more likely to choose fixed price contracts than those that were not. Further they also write contracts of shorter terms, with the reduction being approximately 30%.

Considerably little is known about the performance implications of contractual choices. These form the basis for Chapter 2. Here I find prices to be lower, by between 5% to 20% of the total transaction price, but the probability of renegotiation higher, under fixed price contracts than under escalator or cost-plus contracts. Contract choices appear consistent with a trade-off between establishing incentives ex-ante and lowering negotiation costs ex-post, with relationship specific investments in particular making such a trade-off compelling.

Chapter 3 considers the regulatory environment these utilities were subject to. Both incentive based regulation as well as the restructuring of electricity generation are smaller in comparison to relationship specific investment in terms of their effects on contractual decisions. Consequently, when evaluating the effect of these reforms, ignoring the contractual structure of fuel procurement - and therefore investment specificity - leads to large and significant biases in their impacts.