This thesis aims to study the impact of some of the credit market imperfections on various fiscal decisions. It comprises of two papers, each of which sheds light on how the established results in literature are altered when studied in different environments, with more realistic elements. In chapter 1, we set up a dynamic stochastic general equilibrium model with financial frictions affecting the cost channel and an endogenously derived probability of default. We then study the effects of an expansionary government spending shock. Our exercise highlights that a positive shock to government spending, a demand side shock, increases the cost of firms' marginal costs and hence, their loan requirements. With higher borrowing, their probability of default goes up. The commercial bank takes this into account and charges a higher finance premium, discouraging the firms from borrowing as much. This results in a lower level of economic activity. The government spending multiplier is thus smaller when risky loans are borrowed to finance working capital, instead of fixed capital. In addition, we derive the multiplier to be less than one. With a lot of start-ups borrowing to meet their day-to-day expenses, this result extends a plausible explanation to why during the Great Recession, the impact of government spending was not as large as it was expected to be. In chapter 2, we derive the optimal level of capital taxation in the presence of agents with different discount factors. We set up a real business cycle model with patient and impatient households that borrow and lend amongst themselves, as per a borrowing constraint. Our results show that if the Ramsey planner's weights on different households are such that he is indifferent between redistribution towards patient and impatient households, the borrowing constraint is not binding. Moreover, we get the classical result of zero optimal capital taxation in the distant long run. However, if the Ramsey planner chooses the borrowing constraint to be always binding, he will favour redistribution from impatient households to patient households. As time moves forward, this ultimately leads to a negative optimal tax rate on the capital returns of patient households, a contradiction to the seminal Chamley-Judd result.