This thesis consists of three essays examining the relationship between institutions and economic development. Essay one focuses on private participation in infrastructure. Over the past decade private involvement in the provision of infrastructure services has grown increasingly common in a large number of countries around the world. Increased activity brought along a good deal of controversy, most frequently relating to the cancellation of high profile projects. This paper analyzes this phenomenon empirically, using project level panel data from the 1990-2005 period. My first finding is that, contrary to popular belief, infrastructure project cancellations are rare. Second, contract cancellations are not randomly distributed, but seem correlated with a number of factors. I find that cancellation rates are higher for water sector projects, countries with a poor track record of protecting property rights and those with more effective local bureaucracies. Neither the level of GDP per capita nor its growth rate seem to be important factors, but larger current account deficits are correlated with more cancellations. Essay two examines the economic rationale for industrial policies aimed at supporting small firms with the intention of improving the rate of innovation and economic growth. I argue that such policies, while very common in the last few decades, frequently ignore two fundamental facts. First, a firms' size is largely determined by the economic environment surrounding it, and in particular by the uncertainty it must face. Attempts to actively micromanage the mix of small to large firms while ignoring the environment they operate within is more likely to be harmful than helpful. The second often overlooked observation is that small and large firms often play complementary roles in the process of innovation. Instead of attempting to actively pick winners with certain characteristics, policymakers' efforts are better spent on building a framework which is conducive to all innovation, wherever it may originate. In the third paper I analyze the real world impact of direct financing programs for small and medium enterprises. I base my analysis on two specific SME financing schemes implemented in Romania between 1998 and 2004, but my findings are broadly applicable. I argue that direct funding programs can suffer from two major flaws: a failure to address the financial system's binding constraints, and a difficulty in dealing with imperfect information. I find that both problems were acutely relevant in Romania, where they created programs that appeared successful at the firm level but in fact had very limited impact.