Master of Arts / Department of Economics / Lance J. Bachmeier / The traditional view of oil price movements is that they represent exogenous changes in the supply of oil. In that case, oil price increases will hurt output. Recently, some have questioned whether oil price increases are actually due to higher demand for oil, in which case higher oil prices will be followed by higher output. This thesis develops a model that allows changes in the price of oil to have different effects depending on whether the price of oil and output growth are moving in the same direction (so that the increase in the price of oil was primarily due to an increase in the demand for oil) or in the opposite direction (so that the increase in the price of oil was primarily due to an oil supply shock). The paper presents three sets of results. First, we present the model results for the 1965-2008 time period. Then we look at the 1986-2008 period separately. Finally, we construct a forecasting model for the U.S. industrial production index. The model developed does not require making identifying assumptions and can be used with the data that is available on the internet, and is well understood. Maximum likelihood estimation, which is commonly used for non-linear estimation, is used to estimate the model. We find in-sample evidence in favor of our new model for the 1986-2008 subsample. The new model is unable to provide better out-of-sample forecasts for the 1986-2008 time period.