This PDF is a selection from an out-of-print volume from the National Bureau of Economic Research Volume Title: Inflation, Tax Rules, and Capital Formation Volume Author/Editor: Martin Feldstein Volume Publisher: University of Chicago Press Volume ISBN: 0-226-24085-1 Volume URL: http://www.nber.org/books/feld83-1 Publication Date: 1983 Chapter Title: Fiscal Policies, Inflation, and Capital Formation Chapter Author: Martin Feldstein Chapter URL: http://www.nber.org/chapters/c11330 Chapter pages in book: (p. 61 - 80) Fiscal Policies, Inflation, and Capital Formation The large unprecedented government deficits in recent years have stimu- lated speculation about their adverse effects on inflation and private capital formation. While it is clear that deficits may have no adverse effect in an economy with sufficient unemployed resources, the effects of a deficit when there is full employment are less clear. Is a persistent increase in the government deficit necessarily inflationary? Does it ne- cessarily reduce private capital formation? Is it possible to avoid both adverse effects? A primary purpose of this paper is to answer these questions in the context of a fully employed and growing economy. A closely linked issue is the relation between private saving and capital formation when money and other government liabilities are alternatives to real capital in individual portfolios. John Maynard Keynes, Roy Har- rod, and James Tobin have all emphasized the possibility of excess saving when individuals will not hold capital unless its yield exceeds some minimum required return. When the return on capital is too low, an increase in saving only reduces aggregate demand. If prices are flexible downward, this causes deflation until the increased value of real balances causes a sufficient reduction in saving; if prices cannot fall, the excess saving results in unemployment. Three ways of averting such "excess saving" have been emphasized in both theory and practice. The thrust of the Keynesian prescription was to increase the government deficit to provide demand for the resources that Reprinted by permission from American Economic Review 70 (September 1980): 636-50. This paper is part of the NBER Program of Research on Capital Formation; a previous version was circulated as NBER working paper no. 275 (August 1978). I am grateful for financial support from the National Science Foundation and the NBER. The paper presents my views and not those of the NBER. 61 62 Inflation and Tax Rules in Macroeconomic Equilibrium would not otherwise be used for either consumption or investment. In this way, aggregate demand would be maintained by substituting public consumption for private consumption. A second alternative prescription was to reduce the private saving rate. Early Keynesians like Seymour Harris saw the new Social Security program as an effective way to reduce aggregate saving. The third type of policy, developed by Tobin, relies on increasing the rate of inflation and making money less attractive relative to real capital. In Tobin's analysis, the resulting increase in capital intensity offsets the higher saving rate and therefore maintains aggregate demand. This paper will examine ways of increasing capital intensity without raising the rate of inflation. The analysis will also show why, contrary to Tobin's conclusion, a higher rate of inflation may not succeed in increas- ing investors' willingness to hold real capital. An important feature of the analysis in this paper is a monetary growth model that distinguishes between money and interest-bearing govern- ment bonds. With this distinction, we can compare government deficits financed by borrowing with deficits financed by creating money. It is possible also to examine the effect of changes in the interest rate on government debt while maintaining the fact that money is not interest bearing. The two types of government liabilities also permit analyzing the distinction between the traditional liquidity preference and a demand for government bonds that I shall call "safety preference." In practice, this safety preference may be much more important than the traditional liquidity preference. The first section of the paper develops the three-asset monetary growth model that will be used in the remaining analysis. Section 5.2 then considers the effects of changes in the government deficit. The effects of increased saving on aggregate demand and capital intensity are de- veloped in section 5.3. 5.1 A Three-Asset Model of Monetary Growth The model developed here differs from the traditional monetary growth model (see, e.g., Tobin, 1955, and David Levhari and Don Patinkin, 1968) in two important ways. First, instead of the usual assump- tion that all taxes are lump sum levies, the current analysis recognizes taxes on capital income that lower the net rate of return.1 Second, the government deficit is financed not only by increasing the money supply 1. My 1976 paper (chap. 3 above) and my 1978 paper with Jerry Green and Eytan Sheshinski (chap. 4 above) show the importance of recognizing capital income taxes in analyzing the effects of inflation in a monetary growth model. Corporate and personal taxes were distinguished there but will not be in this paper. 63 Fiscal Policies, Inflation, and Capital Formation but also by issuing interest-bearing government debt.2 Throughout the analysis we maintain the simplifying assumption that the savings rate out of real disposable income is fixed; the tax on capital income affects the allocation of saving but not the saving rate itself. Because the analysis in the sections that follow will focus on comparative steady-state dynamics, only these steady-state properties will be discussed here.3 Note that this assumption of steady-state dynamics implies that all growth rates are constant and therefore that all quantities are correctly anticipated. The economy is characterized by an exogenously growing population nt (1) N = Noe The labor force is a constant fraction of the population and technical progress is subsumed into population growth. Production can be de- scribed by an aggregate production function with constant returns to scale. The relation between per capita output (y) and the per capita capital stock (k) is (2) y=f{k) with/' > 0 and/" < 0. For simplicity, output is measured net of deprecia- tion and real depreciation is not explicitly included in the analysis. 5.1.1 The Government Budget Constraint Government spending (G) plus the payment of interest on the govern- ment debt must be financed by either tax receipts, money creation, or borrowing. Total real tax receipts (T) are the sum of a lump sum tax (To) and the revenue that results from taxing the income from real capital at rate T.4 The money created by the government (M) is the only money in the economy and does not bear interest. The time rate of change of the stock of nominal money is DM; the real value of the extra money created in this way is DMIp. Government bonds bear interest at rate /; the 2. Green and Sheshinski (1977) examine an economy with both bonds and money but assume that such bonds are perfect substitutes for private capital in investors' portfolios. Their analysis generally focuses on quite different issues. Tobin and Willem Buiter (1978) recently developed a three-asset model in which government bonds and money are imper- fect substitutes for each other as well as for real capital. I received a copy of their unpublished paper only after this paper was submitted for publication. Because the tax structure that they assume is very different from the taxes described herein, their conclu- sions are frequently different from my own. Benjamin Friedman (1978) also recently developed a three-asset model but used it to analyze quite different questions (the short-run effects of monetary and fiscal policies in an economy with unemployment and fixed prices) from those that are the focus of my own research. 3. Section 5.3 will, however, consider the possibility of disequilibrium behavior associ- ated with excess saving. 4. The tax rate T is best thought of as a corporation tax rate. The personal tax on real capital income and on the interest on government debt is not specially recognized. 64 Inflation and Tax Rules in Macroeconomic Equilibrium nominal market value of these bonds is B and the real value of new borrowing is DB/p.s The government's budget constraint may be written Alternatively, it will be convenient to denote the real government deficit by A and write In the steady state, the ratio of real money per unit of real capital (M/pK) must remain constant. This implies that the rate of growth of M is equal to the rate of growth of pK, or with Dplp = IT,6 (5) * + n Similarly, the steady-state rate of growth of nominal government bonds equals the inflation rate plus the real growth rate of the economy: (6) 7T + n v ; B Substituting these expressions into (4) and dividing by the population gives the steady-state per capita deficit: (7) A JL £ N v ' pN v ' pN With lowercase letters representing real per capita values, (7) can be rewritten (8) 8 = (IT + n){m + b) The real per capita deficit equals the product of the economy's nominal growth rate and the real per capita government liabilities. 5.1.2 Portfolio Behavior7 The real value of household assets is the sum of the real values of government liabilities and the capital stock:8 5. These bonds may be thought of as Treasury bills although their maturity is irrelevant for steady-state analysis as long as that maturity is finite. I ignore changes in the market value that would temporarily result from changes in the interest rate if the maturity were not very short. 6. This uses the fact that in the steady state k = K/N is constant, implying DKIK = n.