Approved for public release, distribution is unlimited / The Navy has several tools at its disposal to shape the force, including early separation incentives. This analysis looks at the issues surrounding the separation incentives including the discount rate for government use, the discount rate used by individuals to make a decision (personal discount rate), comparable civilian sector wages for separating personnel, promotion probabilities, and the costs to the government of maintaining personnel until retirement. Using the information provided from researching these topics, a model was created to identify the present value of retirement payments for officers given their current rank, age, and years of service. The model provided a maximum amount the government should be willing to offer to separate an officer early. This model was compared to a second model which determined the minimum amount an officer would be willing to accept to separate before retirement. Recommendations for future methods of shaping the force were provided to maximize savings to the Navy based on data generated from the model. / Captain, United States Marine Corps

Measures of American inequality offer sparse coverage of subnational units and rely on surveys of self-reported family and household incomes. This dissertation details the development of new inequality datasets at the county, state, and national levels from alternative lenses: sector wages; industry earnings; and average incomes. Sector and industry data are particularly rich, detailed, consistent, and reliable. These new metrics from underutilized data sources contribute to debates over the lived effects of inequality. American economic inequality concentrates in some places more than others, arising from different causes. This dissertation considers ecological associations between inequality, voter turnout, and election outcomes at the state and county levels and multilevel models of individual participation and candidate preference, with voters nested within their state contexts. Aggregate voter turnout has been lower in states with higher levels of income inequality for the last several presidential elections, though this relationship did not strengthen with rising inequality. Likewise, some inequalities have strong associations with state- and county-level presidential election outcomes in certain years, but the patterns are irregular. Multilevel models of voters in states do not indicate a strong relationship among inequality per se and individual behavior. / text