Identifying and measuring the value of information technology in developing economies.

Abstract

In this dissertation, I study the value of information technology in developing economies. More specifically, I focus on two topics. The first half of the dissertation examines the relative benefit of IT investments made by firms in different developing economies. The second half looks more deeply at how IT helps firms, particularly by focusing on the role of uncertainty and risk aversion. The first chapter establishes the foundation for my analysis of the relative value of IT. In this chapter, I derive both static and dynamic models describing the benefits of IT investments in developing nations. Interestingly, these models can also be used to compare the relative benefits of IT investments made in different countries. In the next chapter, I empirically validate the analytical models of the first chapter. I find that while the models in the first chapter are supported by a static analysis, a time-series analysis has conflicting results. Portions of the analytical model are not supported and the empirical results indicate that IT may have a relatively small role to play in encouraging growth. In addition, in this chapter, I identify a new metric for measuring IT at the country level. The third chapter shifts gears from the first two. Instead of looking at the relative value of IT, it begins to open up the black box of IT to ask how it helps firms enter emerging markets. In particular, it looks at uncertainty and risk aversion which can be reduced through IT investments. Therefore, a firm that invests in IT may get a competitive advantage. Perhaps more significantly, we also show that IT and other forms of physical infrastructure are complements, not substitutes. The last chapter builds on the previous chapter, adding competition into the model. Although the general research questions are similar, the added complexity of the new model requires a new solution technique. I devise a search tool based on classifiers to search for solutions. My results---though only a proof of concept at this stage---indicate that classifiers are an effective tool for finding generalizable solutions to facility placement problems.