Regions going through a natural resource boom tend to have higher average incomes and employment relative to the rest of the country. For policy analysis, a question that often needs to be answered is to what extent the economic growth in the extraction region spills over to neighboring areas. This thesis develops a detailed methodology for analyzing the economic effects of geographically localized shocks within the framework of a parsimonious spatial general equilibrium model, including various methods for estimating key parameters. This model-based approach is being offered as a complementary tool for applied researchers conducting economic impact analysis. Existing empirical methods such as input-output analysis or difference-in-difference estimation techniques are often not optimal for analyzing spatially correlated data, and this model-based methodology can be used to overcome their limitations. Another important advantage of this methodology is that it is computationally tractable and has a relatively low data requirement, which can make a particularly big difference in studying developing countries where data quality and availability can often be an insurmountable challenge. Following the exposition of the methodology, this thesis presents two separate applications, one involving a developed nation and the other a developing one. In the first case, the methodology is applied to analyze the economic impact of the shale energy boom that's been occurring in and around Bakken counties in western North Dakota and eastern Montana over the past decade. In the second case, the methodology is used to analyze the economic impact of the Oyu Tolgoi copper-gold mining project in the Southern Gobi region of Mongolia. A common conclusion that is drawn from the two applications mentioned above is that economic booms fueled by natural resource extracting industries are largely local and have limited spillover effects on neighboring regions.