The government deficit equals the non-government surplus, by definition. Although the relationship is straightforward, it often meets with resistance. It might be worth reflecting on the orthodox interpretation of the identity and the problems with it.
This post is prompted by a comment by Peter D:
This point – that the govt. deficit = non-govt. surplus – seems so self-evident once you think about it that it keeps amazing me how on earth anybody could dispute it. Can you identify what exactly prevents people from appreciating this point? And it is not only laymen. I am not an economist, but the sectoral balances identity G-T=S-I+M-X is supposedly taught in any introductory macroeconomics course, is it not? If so, how could economists fail to understand this? Is there something we are missing?
Others may have different explanations for the resistance to this interpretation of the identity. If so, I’d be interested in reading them. For me, two points spring to mind. One is the way the identities are usually presented in introductory textbooks and presumably lectures. The other is the different causation usually attributed to the variables in the identity by neoclassical economists. As with any identity, the sectoral balances equation shows a relationship that must hold by definition, but says nothing in itself about causation.
In introductory textbooks, the student is typically presented with the identity in one or more of the following alternative forms at different points in the presentation:
GDP = C + I + G + (X – M) I + G + X = S + T + M I = S – (G – T) – (X – M)
The first version of the identity stresses the components of spending that contribute to GDP. The second version highlights the equality of actual injections and actual leakages. The third version breaks up the private-sector balance in a way that hints at a neoclassical understanding of causation that should not have survived the contributions of Kalecki and Keynes, nor the Cambridge Capital Controversy.
Economists using the sectoral-balances framework, including MMT proponents, instead often express the identity in a form that groups the terms into the three sectoral balances:
(T – G) + (S – I) + (M – X) = 0
I will not discuss this form further in the present post since it is the usual one dealt with here and at other MMT blogs. For a discussion of the sectoral-balances approach, see this post by Scott Fullwiler and the links provided in it.
Neoclassicals often read the third version of the identity as suggesting that private saving S encourages private investment to the extent that budget deficits (G – T) don’t crowd it out.
There is additional room for private investment, in this view, to the extent the country runs a capital account surplus (which will occur if there is a current account deficit, M > X). Strictly speaking, the current account deficit, CAD, is equal to imports minus exports plus net income payments. I’ll just assume, for simplicity, that CAD = M – X. By identity, the CAD equals the capital account surplus, KAS. This is because, if there is a net leakage out of the domestic economy to imports, there will be net capital inflow, which in this view enables more private investment.
The neoclassical perspective becomes easier to follow if we write the third version of the identity as:
I = S – BD + KAS
where BD is the budget deficit.
I = S + BS + KAS
where BS is the budget surplus. This is read by neoclassicals as suggesting that private saving, S, and public “saving”, BS – or “national saving”, S + BS – leave room for private investment.
In the long run, it is then supposed that private investment will adjust to the level of “national saving” through real interest-rate adjustments.
This view of causation has been discredited by the work of Kalecki and Keynes, and also by the capital debates.
Kalecki and Keynes
Kalecki points out the identity for aggregate profit, P:
P = Cp + I + (G – T) – Sw + (X – M)
where Cp is capitalist expenditure out of profit and Sw is worker saving out of wages.
To interpret this identity behaviorally, Kalecki observes that capitalists can choose what they spend but not what they earn. The same logic applies to other economic agents. Workers cannot control what they ultimately save, since their saving will depend on income. The government can choose what it spends but not ultimately the tax revenue it receives, since tax revenue depends on the amount of income generated in the economy. Lastly, the foreign sector does choose its expenditure (export demand) but not import spending, which depends on domestic income.
The variables in the profit identity that are completely subject to choice within the domestic economy are therefore I and G. The capitalists’ choice of I and the government’s choice of G will together help to determine the level of income, which in turn will determine the endogenous variables, including P.
The profit identity is simply a different version of the same accounting identity discussed previously. To see this, we can rearrange to get:
P – Cp + Sw = I + (G – T) + (X – M)
Note that (P – Cp) is that part of profit not consumed, i.e. capitalist saving out of profit. So the left-hand side of the identity is capitalist saving out of profit plus worker saving out of wages. In other words, private saving, S. So we have:
S = I + (G – T) + (X – M) => I = S + BS + KAS
However, from Kalecki’s analysis, the appropriate interpretation of causation differs from the neoclassical one. Autonomous private investment I will (along with government expenditure and export demand) create new income, which will at the same time create private saving, tax revenue and imports.
Keynes makes a similar argument. In accordance with the marginal propensity to consume, some of the new income created by the injections will go toward further consumption as part of a multiplier process. The sum of leakages (S + T + M) will equal the sum of injections (I + G + X) at every stage of this multiplier process. Production will keep expanding until the sum of leakages (equal to injections) reaches a proportion of income that is consistent with the propensities to tax, save and import.
The analyses of Kalecki and Keynes suggest that it does not make sense to interpret private saving or a budget surplus as encouraging private investment. Provided there is excess capacity, an increase in private investment will add to income and create an equivalent amount of leakage from the circular flow of income. Similarly, inside capacity limits, government deficits will not crowd out private investment. Rather than depleting a finite pool of saving, deficit expenditure will add to income, saving and other leakages.
Implications of Kaleckian and Keynesian Analyses
Importantly, the macroeconomic explanations of Kalecki and Keynes suggest that the reconciliation of desired leakages and injections occurs independently of wages, prices and interest rates. In a two-sector closed economy without government, this work also showed more specifically that the reconciliation of saving and investment behavior has nothing to do with interest rates (see Thinking in a Macro Way). Instead, autonomous changes in investment create equivalent changes in saving, with the subsequent multiplier effects adjusting income until this higher level of saving is made consistent with saving desires of households. A similar argument concerning injections and leakages in general applies to an open economy.
It might still be wondered whether interest rates might not play the equilibrating role traditionally supposed by neoclassicals in situations where there is a change in saving behavior (such as a change in the marginal propensity to save). Consider a spontaneous attempt by households to lift their level of saving. The insights of Kalecki and Keynes indicate that the initial impact of this will be a reduction in consumption demand. This will result in a build up of unsold inventories. Firms will tend to cut back production and employment in response. This behavioral response will result in a decline in income, largely defeating the private-sector attempt to increase private saving. In the simplest two-sector model, private saving will end up being a higher proportion of a lower income, with overall saving left unchanged. In these circumstances there will be little impetus for firms to increase investment when unsold inventories are building up as a result of weak demand. There is already excess capacity without firms adding even more through additional investment. The much likelier outcome is simply that the attempt to increase private saving will be thwarted by negative income adjustments.
Even though it is likely that the central bank will lower nominal interest rates and so reduce the nominal cost of borrowing, expected revenues are also likely to fall due to the weak demand. On top of that, deflationary pressures and expectations of deflation may prevent the nominal interest-rate reduction from translating into lower real interest rates. So there is no guarantee that real interest rates will fall, even if such a real reduction in rates could be relied upon to induce an increase in private investment to match the higher intended private saving. In neoclassical terms, there could be real interest-rate “stickiness”.
Cambridge Capital Controversy
But the problem with neoclassical causation goes deeper than this. Neoclassicals traditionally argued that provided real interest rates did adjust, this would ensure private investment moved to the level sufficient to absorb full-employment saving. Once full employment was established, it would then be impossible for investment to be increased without a corresponding reduction in consumption to allow for extra saving to make room for the investment. In the long run, it was claimed, the real rate of interest would at last come to the fore to equilibrate investment and saving, even though it was conceded that income adjustments played the role identified by Kalecki and Keynes in the short run. Although the initial impact of increased private saving might be a reduction in output and income, the neoclassicals argued that eventually real interest rates would restore full employment at a higher level of saving and investment.
The capital debates demonstrated, on logical grounds, that this is not the case. There is no monotonic inverse relation between the real rate of interest and private investment. For instance, Paul Samuelson, the leading neoclassical participant in the debate, conceded the following:
The phenomenon of switching back at a very low interest rate to a set of techniques that had seemed viable only at a very high interest rate involves more than esoteric difficulties. It shows that the simple tale told by Jevons, Böhm-Bawerk, Wicksell and other neoclassical writers — alleging that, as the interest rate falls in consequence of abstention from present consumption in favour of future, technology must become in some sense more ’roundabout,’ more ‘mechanized’ and ‘more productive’ — cannot be universally valid. (“A Summing Up”, Quarterly Journal of Economics vol. 80, 1966, p. 568. This quote is reproduced from the link provided above.)
The capital debates made clear that there is no basis for supposing real interest-rate adjustments can be relied upon to adjust investment to a higher level of saving, including in the long run.
Post Keynesian and MMT Insights
More recently, Post Keynesian Economics (PKE) and Modern Monetary Theory (MMT) have demonstrated that in a modern monetary system interest rates are not determined in the way neoclassicals have traditionally supposed. Operationally, the short-term nominal interest rate is an exogenous policy variable. Longer nominal rates could also be set by the central bank if it so desired, but otherwise tend to move in line with shorter rates on the basis of inflationary expectations and arbitrage.
Prices and inflation are also susceptible to strong policy influence through fiscal policy. As the monopoly issuer of its own currency, the government’s spending choices and level of net spending ultimately determine the price level. Exogenous changes in fiscal policy can be made if the price level is rising faster than preferred.
The interest-rate target of the central bank (nominal rate) is therefore a policy variable. The inflation rate is also largely policy determined. Moreover, the central bank can maintain whatever nominal interest-rate target is desired irrespective of fiscal policy and the inflationary implications of fiscal settings. This means that the real rate of interest is also ultimately a policy variable.
One important implication is that fiscal deficits do not in themselves cause higher interest rates, and so cannot be said financially to crowd out private investment. In fact, to the contrary, fiscal deficits in isolation would drive the short-term interest rate to zero due to an influx of reserves. If the central bank is intent on targeting a positive nominal rate, it needs to step in to maintain its target.
The reason for differing interpretations of the sectoral balances identity is ongoing disagreement over causation. However, the neoclassical view of causation has been discredited by Kalecki, Keynes and the capital debates, and also relies on arguments that do not apply to a modern monetary system.
The Kaleckian behavioral observation that economic actors can choose what they spend but not what they receive in income (or, in the case of government, extinguish in taxes) suggests that it is autonomous expenditures (private investment, government expenditure, export demand) that determine the level of saving, tax revenue and import spending.
The work of Kalecki and Keynes suggests that household saving intentions and firms’ investment plans in the two-sector model (or desired leakages and injections in the open-economy model with government) are reconciled independently of wage, price and interest-rate adjustments. Instead, the desires of the various actors are reconciled through income adjustments driven by autonomous expenditures.
Further, a spontaneous increase in private saving will result in a build up of unsold inventories and cutbacks in production and income that are likely to thwart the additional intended private saving. Falling demand and income are not conducive to a strengthening in private investment. In this context, falling nominal interest rates may not be sufficient to induce increased expenditure. If prices are also falling, it is not even clear that the real rate of interest will fall.
However, even if the real rate of interest did fall, the capital debates have made clear that there is no valid basis for supposing that this will induce an increase in investment sufficient to absorb the higher intended saving.
Finally, MMT shows that it would be operationally impossible for government deficits to cause financial crowding out of private investment. When there is deficit expenditure, the central bank has to step in to prevent the short-term rate falling to zero whenever its interest-rate target is positive.