In recent years, federal deficits and debt have increased to phenomenal levels. In 1993, total debt reached $4.35 trillion, and the debt burden per capita amounted to $16,848. Measured in 1987 dollars, total debt and per capita debt were $3.5 trillion and $13,565, respectively. For a family of four, the debt burden was $54,260 in 1987 dollars.(1) Legislators have made several efforts, such as the Gramm-Rudman-Hollings (GRH) Act of 1986 and the Budget Reconciliation Bill of 1993, to reduce deficits and debt. But the former attempt has failed and the impact of the latter remains to be seen. Clearly almost all countries incur debt to provide adequate public services, finance the construction of infrastructures, and pursue full employment and economic growth.
Federal debt may facilitate or deter economic growth, depending on the level of the debt. When economic growth is slow or when the private sector does not have incentives to invest, the federal government may need to pursue fiscal and/or monetary policy to stimulate the economy and may resort to debt financing. On the other hand, when economic growth is normal or above the long-term trend, an increase in federal government debt may be detrimental to real GDP growth. This is because an increase in federal debt may push interest rates upward and reduce private investment. The three views of the impact of federal deficits and debt on economic growth more clearly explain these arguments. (For a detailed analysis of these three views, see Sawhney and Di-Pietro, 1994.) The stimulus view (Eisner, 1984) argues that if deficits and debt are measured correctly, higher deficits and debt will stimulate employment, consumption, investment, and economic growth. The crowding-out view (Friedman, 1988) maintains that higher deficits and debt will reduce economic growth due to rising interest rates and lower investment and capital formation. The Ricardian view (Barro, 1989) holds that deficits and debt do not have any impacts on economic growth because the decrease in future income and consumption due to more tax burdens will offset the increase in current government spending. Because these hypotheses represent different relations, a quadratic form is an appropriate way to test the positive, horizontal, and negative relations.
This paper extends the work of Barro (1979), Eisner (1992), Joines (1991) and others to examine the federal government debt's impact on economic growth and to test if an optimal debt ratio exists that will maximize the economic growth rate.(2) Barro (1979, p. 966) argues that his "result supports the surprising proposition of the theory that the debt/income ratio does not have a 'target' value but rather moves 'randomly' in accordance with the realizations for the federal expenditure and income shock." Few studies demonstrate that this may not be the case. A number of studies (Eisner, 1992; Joines, 1991) choose different debt/GDP ratios so that the debt would not grow faster than the GDP or so that the debt is within a reasonable range. However, they do not test if these debt/GDP ratios are optimal.
II. LITERATURE SURVEY
Deficits and debt and their relationships with other macroeconomic variables have been investigated extensively. Eisner (1992) is optimistic about the deficit and debt issue. After adjusting for deposit insurance, net investment, state and local budgets, and inflation tax, he estimates that the deficit in 1991 would be a very small amount of $17 billion rather than the reported value of $269 billion. He indicates that a constant debt/GDP ratio may be maintained so that the debt does not grow faster than the GDP. He chooses a constant public debt/GDP ratio of 47.1 percent as the one that corresponds to a deficit of $188 billion. However, he does not indicate whether this 47.1 percent is the optimal debt ratio. Joines (1991) selects three public debt ratios for comparison and regards the 40 percent debt ratio as the long-run historical average. …