2  Growth Effects of Taxes, Expenditures and Deficits

There are now numerous endogenous growth models incorporating fiscal variables and which are capable of yielding predictions of long-run or steady state growth effects arising from fiscal changes.[2] Most such models have focused on one side of the government budget or the other – usually the tax side.[3] Barro (1990) and Cashin (1995) analyse both taxes and expenditures simultaneously, though both models preclude deficit finance. For present purposes, the Barro (1990) and Barro and Sala-i-Martin (1992) models provide a useful starting point. They adopt the standard Ramsey framework in which the consumption path of a representative consumer is obtained by maximising an inter-temporal utility function over an infinite horizon. There are n producers each producing output (y) according to the production function:

(1)     y = Ak^1-αg^α

where k represents private capital and g is a publicly provided input (per capita). There are therefore constant returns to total (public plus private) ‘capital’ inputs, k+g.[4] The government also produces consumption (‘unproductive’) goods, C, which enter consumers’ utility functions but have no effect on production. The government balances its budget in each period by raising a proportional tax on output at rate τ and lump-sum taxes of L, giving the constraint:

(2)      ng + C = L + τny

Of course, lump-sum (or non-distortionary) taxes do not affect the private sector’s incentive to invest in the input good, whereas the taxes on output do. Thus, with an isoelastic inter-temporal utility function, Barro and Sala-i-Martin (1992) show that the long-run growth rate in this model (γ) can be expressed as

(3)      γ = λ(l-τ)(1-α)A^1/(1-α)(g/y)^α/(1-α) - μ

where λ and μ are constants that reflect parameters in the utility function. Alternatively, using (2), (3) can be re-written as:

(4)    

Equations (3) and (4) show that the growth rate is decreasing in the rate of distortionary taxes (τ) and increasing in government productive expenditure (g), but is unaffected by non-distortionary taxes (L) or unproductive expenditure (g_c).[5]

The growth effects of the alternative combinations of taxes and expenditures in the Barro and Sala-i-Martin model are summarised in the four cells in the north-west corner of the matrix in Table 1 below. In addition, though Barro and Sala-i-Martin (1992) exclude the possibility of deficit finance, the framework is readily extended to include fiscal deficits, and their predicted effects on growth are also shown in Table 1 (and are discussed further below).

It is immediately obvious from the Table that the predicted effects of taxes or expenditures on growth rates depend on: (i) the type of tax or expenditure considered (and hence the tax/expenditure mix); (ii) the total level of expenditures; and (iii) how this is financed (compensating tax or expenditure change). This is reinforced when budget surpluses/deficits are included.

As Table 1 shows, even where all government expenditure is productive, the use of distortionary taxes to finance this can, at sufficiently large tax/expenditure levels, generate negative growth effects.[6]

For this framework to be useful empirically, it is important to be able to distinguish productive from unproductive expenditures and distortionary from non-distortionary taxes within public budgets in practice. On spending, a typical ‘first approximation’ is to treat government consumption spending as ‘unproductive’ (i.e. it affects consumers’ welfare but not private production efficiency) and treat investment spending as ‘productive’. The latter usually includes (some or all?) education and health spending because of their effects on human capital accumulation. The growth effects of public expenditure on current transfers such as social security remains a debated issue. If these merely affect welfare they can be treated analogously to other ‘unproductive’ expenditures. However, transfers may affect savings rates, inequality,enforcement of property rights, etc. and could therefore be either growth enhancing or retarding depending on the empirical relevance of these potential growth mechanisms.

On taxation, in the Barro (1990) model ‘distortionary’ taxes are those distorting the decision to invest – essentially capital and labour income taxes. With no labour-leisure-education choices, consumption taxes are non-distorting. However, as Mendoza et al. (1997) show, human capital investment can be affected by consumption taxes when labour supply is endogenous. Clearly, in practice, almost all taxes are distortionary to some degree and the key issue in searching for long-run effects of various taxes is whether these distortions can be expected to be substantial or minor with respect to the main determinants of long-run growth, such as investment or technical progress.

In extensions to the Barro-type model allowing budget deficits, whether these affect growth depends on whether Ricardian Equivalence (RE) is assumed to hold – that is, whether the private sector anticipates future taxes and adjusts its savings to compensate fully for changes in public sector savings. Where RE does not hold, budget deficits are generally expected to be growth-retarding. This can arise because total savings are reduced (if the private sector does not fully adjust its savings or government borrowing finances consumption goods provision), hence reducing factor accumulation. Alternatively, as Tanzi and Zee (1997) argue, if deficits are perceived as unsustainable, then changes in tax/expenditure policy and/or monetary policy will be anticipated. Either is likely to retard growth via effects on investment from increases in expected inflation or uncertainties associated with possible fiscal policy changes. Even if monetary policy is designed to ‘neutralise’ the inflationary effects of a budget deficit, growth is still likely to be retarded by the associated increases in interest rates.