Regional analysis of an urban and rural economy in the State of Oregon: a cointegration approach

Abstract

Throughout regional literature the debate of the relationship of diversity and stability is abundant. Various indices have been devised to measure these two economic phenomena and are analyzed with regression techniques or input-output models. This dissertation will present another method to test their relationship using Johansen's (1988) multivariate cointegration technique. This procedure is conducted on two regional economies, the Portland Metropolitan Area (PMA) and Southwest (SW) Region, within the state of Oregon. These two economies were chosen because of the urban and rural nature of the PMA and SW respectively. This defining difference will indicate each economies differing levels of diversity and resulting stability. Within each regional system a new and statistically valid aggregation scheme based upon pairwise cointegration is used. This scheme enables unique characteristics of each economy to be illustrated and analyzed, which enables the study to be more accurate. Pairwise cointegrated aggregation reveals that previous techniques have considerable biases as the aggregates fail to maintain the properties of their individual components. The combination of this technique with the analysis of the Vector Error Correction Mechanism of each cointegrated system provides a test to verify whether correct aggregation and specification have taken place. Each economy's relative diversity and stability is revealed within their respective cointegrated systems and persistent profiles. The cointegrated system illustrates diversity through the number of common trends. Persistent profiles enable an analysis of stability to be conducted by analyzing how quickly an economy returns to long run equilibrium after the introduction of a system wide shock. These results combined with impulse response functions lead to policy recommendation for both economies to improve upon their diversity and stability.