Advances and Declines in the Rural Working Poor: Complementing Traditional Econometric Results with Case Analysis

We employ a multi-method approach to more fully explore determinants of greater than expected rural county-level increases and decreases in the proportion of working poor in four states. An econometric model by Anderson, Goe, and Weng (2007) using 1990 and 2000 Census data in the North Central region of the U.S. supplies the error terms to identify our outlier counties. We show that counties performing better than expected may be more self-reliant than counties performing worse than expected (Michigan), that regional attributes contribute greatly to overall performance (Ohio), that the structure of local employment patterns also influences outcomes (Missouri), while devolution of federal government and long-term commuting patterns may also affect outcomes (South Dakota). Future efforts should attempt to replicate these research strategies and may serve to inform the direction of best practice in federal data collection efforts.