In order to provide enough economic growth so that the majority of individuals within the jurisdiction are satisfied with government services, state governments typically pursue two budgetary choices for economic growth: overall increased production and stabilization (Bivens, 2014). According to two budgetary choices as a path to economic growth, this research investigates the relationship between capital expenditure or savings and economic growth. It covers the years 1990 through 2013 and uses a paneled data set at the state level in the United States. The first model for this study is the structural equation model (SEM), which examines the direct and indirect effects of capital expenditure and state government savings on economic growth by including the volatility of the total expenditure as a mediating factor. Then, this dissertation investigates the relationships among capital expenditures, the total expenditure volatility and savings by using the endogenous growth or the OLS regression model. This dissertation can conclude that both of the two budgetary choices for state governments are effective for economic growth. Under controlling state characteristics, they are positively related to economic growth, which supports the allocation role of government for economic growth. However, this study finds that state governments do not find any supportive evidence on the fact that they can attain the stabilization role of government for economic growth. Even though they spend money on savings or capital expenditure from Keynesian macroeconomic theory, it does not lead to budgetary stabilization of the total expenditure. Thus, this dissertation leaves the missing links of the relationship between both fiscal policies and volatility inconclusive while it supports that volatility can negatively affect economic growth.