Essays on the government and trade deficits

Abstract

A given expenditure can be financed by raising taxes, borrowing, money creation, and/or a combination of these three. What are the differential impacts of these alternative methods of financing public expenditures? Is there any relationship between government budget deficits and trade deficits? Economists have differed on these issues without an apparent resolution. This dissertation uses an extended U.S. time series data in order to garner further evidence about the relationships between and among the series that define the government budget deficits and trade deficits. First, the two major theoretical foundations, the traditional hypothesis and the Ricardian equivalence theorem were thoroughly reviewed and their implications formally derived. In short, traditional hypothesis suggests relationships to exist between budget deficits on the one hand, and other variables such as interest rates, exchange rates, output, employment, and trade deficits, on the other hand, while the Ricardian proposition suggests otherwise. Based on these theoretical implications, a number of deficit measures were then defined. Both quarterly and annual data frequencies were also employed in defining these deficit measures. Next, the time series properties of each individual series and the series related to each other were investigated. Specifically, the results of this research indicate not only most of the series that define the budget and trade deficits follow stochastic trends but also share common trends. The presence of common trends (or cointegration) between and among the series implies that the series may balance out in the long run even though they may depart from each other in the short run. This may also be interpreted as government policies being reversed eventually and/or economic forces bringing the variables in line to each other in the long run. The apparent relationship between the relevant series also leads one to reject the Ricardian proposition in favor the alternative.