those evidenced by private ownership of public debt, may not be perceived as being on a par with other components of net worth by the private sector as a whole if higher future taxes, increased macroeconomic instability, or reduced real economic growth are expected.61. Even so, one may suspect that only weak effects can readily be swamped by changes in other potentially relevant factors that were omitted from the regressions.
This study shows that government saving habitually moves with cyclical and other factors in a well-defined pattern. Therefore, only deviations from a rule that implies government dissaving on the average but with large swings over the business cycle are treated as fiscal surprises.
In this expectational milieu the crucial distinctions are not between an automatic and a discretionary fiscal change or between a temporary and a permanent change. This allows one to dispense with labels for particular fiscal actions.62. What matters is not the labels attached to isolated policy steps but whether the resulting overall fiscal stance is consistent with the rule or off the trajectory to which it is expected to return. So long as there are no political or economic grounds for suspecting that the fiscal policy rule will change, a rational assumption is that the rule will continue to be observed and that government debt will grow indefinitely.63. Further development of the concept of the fiscal rule and its measurement may yield clearer insights than have been obtained in this study.
Additional work on other input variables may also prove rewarding. A great deal of care was lavished on the construction of the real after-tax rate-of-return variables, including accrued capital gains. Because I was dissatisfied with the common practice of picking the rate of return on a single asset, such as corporate bonds, to represent the saving incentives of households when actual household portfolios are, and must be, much____________________