Doctor of Philosophy / Department of Economics / Lance Bachmeier / In the first chapter I study the effects of oil price shocks on economic activity at the U.S. state-level, an innovative feature of this dissertation. States which rely more heavily on manufacturing or tourism are more adversely affected by adverse oil price shocks, while states which are major energy producers either benefit or experience insignificant economic changes from historically large oil price increases. Additionally, oil price increases from 1986 to 2011 have not impacted state-level economies to the same degree as increases from 1976 to 1985. This discrepancy can be attributed to a fundamental change in the structure of the U.S. economy, for example, a declining manufacturing sector or an increase in the efficiency with which energy is used in the production process. In the second chapter I explore the effects of alternative measures of energy price shocks on economic activity and examine the relative performance of these alternative measures in forecasting macroeconomic activity. The alternative energy prices I consider are: gasoline, diesel, natural gas, heating oil and electricity. I find that alternative measures of energy price shocks produce different patterns of impulse responses than oil price shocks. The overwhelming evidence indicates that alternative energy price models, excluding a model containing gasoline prices, outperforms the baseline model containing oil prices for many states, particularly at short-to-mid forecast horizons. In the third chapter, which is coauthored with Lance Bachmeier, we determine whether accounting for oil price endogeneity is important when predicting state-level economic activity. We find that accounting for endogeneity matters for in-sample fit for most states. Specifically, in-sample fit would be improved by using a larger model which contains both regular oil price and endogenous oil price movements. However, we conclude that accounting for endogeneity is not important for out-of-sample forecast accuracy, and a simple model containing only the change in the price of oil produces equally accurate forecasts. Accounting for endogeneity is particularly important in an environment in which rising oil prices were caused by a growing global economy, such as in the years 2004-2007.