Domestic Financial Policies under Fixed and under Floating Exchange Rates

Fleming asks how the choice between a fixed and a floating exchange rate bears on the relative effectiveness of monetary policy versus budgetary (fiscal) policy as tools for raising domestic monetary demand, a bearing he says “is not always kept in mind when such systems are compared” (p. 369). He works with “a simple Keynesian model” of a small open economy: taxation and private after-tax income vary directly with national income; private expenditure on consumption and investment varies directly with after-tax income and inversely with the interest rate; the interest rate varies directly with the income-velocity of money (the ratio of income to the money stock); the trade balance varies inversely with domestic expenditure and directly with the exchange rate; and net capital imports vary directly with the interest rate (p. 369). All magnitudes are expressed in constant domestic wage units, and no account is taken of changes in spending propensities arising from terms-of-trade effects, nor initially of exchange speculation (p. 370). A full algebraic version of the model is worked out in the Appendix (pp. 377-379).

An increase in public expenditure raises income and, if the economy was underemployed, output and employment, but it worsens the current-account balance of payments (mainly through higher imports) while raising tax revenue by less than the increase in spending (pp. 370-371). Because “unchanged” monetary policy is defined here as a constant money stock, keeping the money stock fixed while income rises requires the interest rate to rise, which both damps the increase in income somewhat and produces a favorable shift in the capital account (p. 371). Since the current account worsens and the capital account improves, the net effect of a fiscal expansion on the overall fixed-rate balance of payments is ambiguous – it is more likely to deteriorate the higher the marginal propensity to import and the responsiveness of exports to expenditure, and the less sensitive are the interest rate and capital movements (pp. 371-372). If the balance of payments does deteriorate, a shortage of reserves may eventually force the authorities to abandon the expansionary policy (p. 371).

It depends on the model’s parameters, and can go either way. If, at a fixed exchange rate, a rise in public expenditure would have worsened the balance of payments, then under a floating rate it instead causes the currency to depreciate, which partially restores the trade balance and adds a further stimulus to income beyond what occurs under a fixed rate (pp. 371-372). But if a rise in public expenditure would instead have improved the fixed-rate balance of payments – an outcome Fleming notes is not merely an “academic curiosum,” citing R. R. Rhomberg’s econometric model of Canada as an empirically relevant case – then under a floating rate it causes the currency to appreciate, worsening the trade balance and yielding a smaller net stimulus to income than under a fixed rate (pp. 372-373). In the polar case of infinitely interest-elastic capital flows, the appreciation induced by a fiscal expansion completely offsets the direct stimulus, leaving national income and the interest rate unchanged (p. 373).

A monetary expansion lowers the interest rate, stimulating private investment and consumption both directly and via the multiplier, raising income, output, and employment while worsening the current-account balance of payments (p. 373). Because the expansion itself lowers the interest rate (even after income rises), it also worsens the capital account. Unlike a fiscal expansion, a monetary expansion under a fixed rate is bound to worsen the overall balance of payments in every case, so the associated income and output gains can be sustained only for as long as reserves allow, or to the extent the balance of payments would otherwise have been in surplus (p. 373).

Because the balance-of-payments deterioration that a monetary expansion always produces forces a currency depreciation under floating rates, and the resulting improvement in the trade balance provides an additional stimulus to income – both directly and via the multiplier – that raises income above what a fixed exchange rate would have allowed (pp. 373-374). Unlike the fiscal-policy case, there is no scenario in which a monetary expansion causes the exchange rate to appreciate, since a monetary expansion’s balance-of-payments effect is unambiguously adverse; a floating rate therefore unambiguously reinforces monetary expansion’s effect on income (p. 374). The size of this extra effect grows with the interest-sensitivity of international capital flows: with zero sensitivity, income rises by the same amount as in a closed economy; with infinite sensitivity, money income rises in the same proportion as the money stock (p. 374).

Fleming sets up a monetary expansion (“Policy A”) and a budgetary expansion (“Policy B”) calibrated so that, under a fixed exchange rate, the two produce the same aggregate increase in income; he then shows that when both are switched to a floating rate, Policy A’s income effect will never be less than, and will generally exceed, Policy B’s (pp. 374-376). The reasoning is that, with income held equal across the two fixed-rate policies, Policy A involves a larger money stock and lower interest rate than Policy B, so Policy A produces a more unfavorable capital-account position under a fixed rate; switching to a floating rate therefore requires a deeper currency depreciation to restore external balance under Policy A than under Policy B, generating a larger trade-balance improvement and hence a larger extra stimulus to income (p. 375, Appendix paragraph 19). At zero interest-sensitivity of capital movements there is “nothing to choose between the two policies”; as that sensitivity becomes very large, income under Policy A exceeds income under Policy B in roughly the same proportion as the money stock under A exceeds that under B (p. 375).

Yes. Under a fixed exchange rate, monetary expansion can be sustained only as long as reserves hold out (except where the external accounts were originally in surplus), whereas budgetary expansion, if capital movements are sufficiently interest-sensitive, may be sustained indefinitely because the reserve drain can be offset by the induced capital inflow (p. 375). Under a floating exchange rate, by contrast, both types of policy can be sustained indefinitely so far as the balance-of-payments situation is concerned, since the exchange rate itself, rather than reserves, does the adjusting (pp. 375-376). Fleming adds a caveat (fn. 21) that the interest-sensitivity of capital movements reflects both a one-time relocation of existing capital and a recurring reallocation of new savings, so the sensitivity – and hence the gap between the two regimes – is likely to be greater in the short run than in the long run.

The main argument assumes the absence of exchange speculation. Fleming then notes that under a floating rate, speculation that is “equilibrating” – damping the exchange-rate movements that a change in financial policy would otherwise cause, as he says was generally true of Canada in the 1950s – will narrow the difference in effectiveness between monetary and budgetary policy, while “disequilibrating” speculation, which exaggerates those movements, will widen that difference (p. 376). Under a fixed rate, he separately notes (fn. 22) that if confidence in the peg is less than complete, fear of provoking destabilizing capital flows may itself limit the size and duration of expansionary policies, particularly monetary policy, whose balance-of-payments effect is in any case more adverse than that of budgetary policy.

Three, tightly stated: (1) a given increase in the money supply always has a greater expansionary effect on income under a floating exchange rate than under a fixed rate; (2) whether a given fiscal expansion has a greater or smaller expansionary effect under floating versus fixed rates is uncertain, depending on the model’s parameters; and (3) in all but extreme cases, the stimulus to monetary demand from a given increase in the money supply, relative to that from an equivalent-under-fixed-rates increase in budgetary expenditure, is greater under a floating exchange rate than under a fixed rate (p. 369, restated pp. 374-376). Fleming further observes that the choice of exchange-rate regime affects not only the relative effectiveness of the two policy types but also their relative sustainability: fixed rates constrain monetary expansion via the reserve position while allowing fiscal expansion more room, whereas floating rates remove the balance-of-payments constraint from both (pp. 375-376).

Fleming works throughout with a small open economy whose wages are fixed in domestic currency, so all magnitudes are measured in constant domestic wage units, and he explicitly sets aside any change in spending propensities that might arise from terms-of-trade effects (p. 370). The core comparison in the base case abstracts from exchange speculation, which is reintroduced only afterward as a qualification rather than built into the formal model (pp. 370, 376). The paper’s definition of “unchanged” monetary policy – a constant money stock – is presented as one defensible convention among others; Fleming notes that R. A. Mundell, in a contemporaneous companion paper, instead defined constant monetary policy as a constant interest rate (p. 371, fn. 6). Several of the paper’s directional results are explicitly conditional on parameter magnitudes – e.g., whether a fiscal expansion appreciates or depreciates the exchange rate, and how much more effective monetary policy becomes under floating rates – and Fleming flags the empirically relevant case of Canada (via Rhomberg’s econometric model) as an illustration that the less intuitive “appreciation” case is not merely a theoretical curiosity (pp. 372-373).