A public debt theory is constructed in which the Ricardian invariance theorem is valid as a first-order proposition but where the dependence excess burden on the timing of taxation implies an optimal time path of debt issue. A central proposition is that deficits are varied in order to maintain expected constancy in tax rates. This behavior implies a positive effect on debt issue of temporary increases in government spending (as in wartime), a countercyclical response of debt to temporary income movements, and a one-to-one effect of expected inflation on nominal debt growth. Debt issue would be invariant with the outstanding debt-income ratio and, except for a mirror effect, with the level of government spending. Hypotheses are tested on U.S. data since World War I. Results are basically in accord with the theory. It also turns out that a small set of explanatory variables can account for the principal movements in interest-bearing federal debt since the 1920s.