Academic journal article Journal of Money, Credit & Banking

Joint Implications of Consumption and Tax Smoothing

Academic journal article Journal of Money, Credit & Banking

Joint Implications of Consumption and Tax Smoothing

Article excerpt

OVER SEVERAL decades the permanent income hypothesis progressively became a central tenet of macroeconomics. More recently, the tax smoothing hypothesis also came of age, with a detailed account of it having been included in the second edition of Romer's (2000) popular graduate textbook. Sargent (2001) has called attention to the "isomorphism" between these two theories. (1) The present paper shows that interaction between a consumption smoothing model and a tax smoothing model generates two new hypotheses, one concerning optimal forecasts of government outlays and the other, optimal forecasts of tax revenues.

Both hypotheses are in the class of exact linear relations. The exact linear relation for optimal forecasts of government outlays builds on the one for tax revenues and also on the government budget constraint given in Bohn (1991) and the tax smoothing contributions of Huang and Lin (1993) and Ghosh (1995). In turn, the exact linear relation for tax revenues builds on its counterpart in the consumption smoothing contribution of Campbell (1987).

Bohn (1991) explores the implications of the government budget constraint for subsequent movements in outlays and revenues following a shock to the budget deficit. (2) Huang and Lin (1993) combine a log-linear version of the government budget constraint with a tax smoothing assumption that allows changes in tax revenues to be replaced with changes in real output in the constraint. Thus, the budget deficit acts as a signaling variable for two kinds of shocks, one to government outlays and the other to the tax base, with the latter proxied by real output. The budget deficit potentially reflects both types of shock. By contrast, Ghosh (1995) seeks to pin down the type of shock that is signaled by the budget deficit. The theoretical development in his paper achieves this by assuming that real output grows at a constant rate. Under this assumption, together with tax smoothing, the budget deficit signals only one kind of shock, to government outlays, since, in effect, the tax base is non-stochastic, in contrast to Huang and Lin (1993) where the tax base is stochastic but unforecastable.

The novel feature of our approach is that the tax base is stochastic, yet potentially forecastable by way of the permanent income hypothesis. If the tax base is non-asset income, then the optimal predictor of changes in the tax base, under consumption and tax smoothing, is private dissaving suitably defined. Our measure of private dissaving takes account of capital gains and was first introduced by Lettau and Ludvigson (2001). The log-linear government budget constraint due to Huang and Lin (1993) is the exact public sector analogue of the measure of private dissaving due to Lettau and Ludvigson (2001). Using this constraint, we then show that a log-linear variable composed of the budget deficit and private dissaving is the optimal predictor of changes in government outlays. This variable, which we term fiscal pressure, signals only one kind of shock, namely, to government outlays, because it optimally takes account of the effect of expected changes in the tax base on the budget deficit.

It should be noted that our model, like that of Huang and Lin (1993) and Ghosh (1995), is in the class of what Sargent (2001) describes as the "Barro" (Barro 1979) model of tax smoothing, that is, an economic environment in which the government cannot issue contingent claims, with a consequence being that tax rates are stochastic and forward-looking. By contrast, the "Lucas-Stokey-Sims" model allows payment on public debt to be state contingent, and implies deterministic and backward-looking tax rates; see Lucas and Stokey (1983) and Sims (2001).

In this paper, the exact linear relation for outlays and for revenues is tested with annual post-World War II U.S. data. We find that fiscal pressure, the variable that controls for the effects of shocks to the tax base on the budget deficit, has considerable predictive power for expected future changes in government outlays. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.