Anyone who has argued for substantial deficit expenditure to sustain demand and employment levels in the aftermath of the Global Financial Crisis has probably run up against opposition from supporters of the Austrian School. This observation applies as much to orthodox advocates of fiscal stimulus as it does to heterodox economists.
The derision of Austrians tends to be aimed at the supposed inflationary impacts of current fiscal policy settings. Three of their most frequent arguments are:
1. Deficit expenditure, if unmatched by public debt issuance, would expand the “money supply”. This, in turn, would increase the claims on real goods and services, and result – or so it is claimed – in a reduction in the real claim represented by a unit of the currency.
2. Even when deficit expenditure is matched by debt issuance – as it currently is – government spending and taxing diverts resources from “more efficient” private sector activities to “less efficient” private or public sector activities.
3. Deficit expenditure unmatched by public debt issuance is by definition inflationary because inflation for the Austrians is defined as expansion of the money supply, not a persistent rise in the general price level.
In this blog, I want to address these arguments.
“Money Supply” and Claims on Real Output
A preliminary point to note is that the current practice of issuing public debt to match deficit expenditure has no impact on the degree to which fiscal policy alters “claims on real goods and services”, even though the debt issuance does neutralize the impact of fiscal policy on the amount of reserve balances in the system.
When deficit expenditure is matched by debt issuance, as is currently the case, there is an overall increase in net financial assets equal to the amount of the deficit, but no overall change in reserves.
If, instead, deficit expenditure were not matched by debt issuance, there would be an increase in net financial assets equal to the increase in reserves, and also equal to the amount of the deficit.
In either case, the impact on net financial assets would be the same. Under the current practice, the increase in net financial assets is in the form of bonds. In the absence of debt issuance, the increase in net financial assets would be in the form of reserves. Either way, net financial assets – which are a “claim on real output” – would be altered to the same extent, whether debt is issued or not.
But there is something else going on other than just an increase in “claims on real output”. There is also an increase in the amount of real output, provided the deficit expenditure is used to activate production. So real claims and real goods both increase.
Austrians need to explain why net financial assets should be kept too scarce to sustain full employment. That is, what is their basis for determining whether the current amount of net financial assets is “correct”?
It has been pointed out many times in previous blogs that modern monetary theory (MMT) takes a position on this question. It is that fiscal policy should alter net financial assets (“claims”) by an amount just sufficient to enable full employment and non-government net saving desires (for an illustration of this point, see Parable of a Monetary Economy). This policy stance will correspond to a particular change in net financial assets.
The extent to which net financial assets need to be increased can be altered through redistributive policies, and this may well be appropriate, but the point remains that for a given distribution and non-government net saving disposition, there will be a level of deficit expenditure consistent with full employment and price stability.
Deficit Expenditure and “Efficiency”
As for the argument that private markets are efficient and government expenditure causes ‘distortions’, let’s break this down with a simple example to see what is really at stake. Imagine the government gives an extra $100 to an unemployed person and the recipient spends the money at the shops. Suppose – as is clearly the case at present – that the shops (and their suppliers) have sufficient spare capacity to respond to this extra $100 of demand by selling more goods at current prices.
Where is the supposed inefficiency? The shops have supplied output at a price determined by the cost structure and degree of competition within their respective sectors. So there is no necessary impact on cost-efficiency.
What the Austrians call inefficient in this instance is the fact that this $100 has been spent by someone who under private-market conditions would not have possessed the $100. It is really a complaint about distribution, not efficiency.
By calling this a problem of efficiency, the Austrians are really referring to what the neoclassical orthodoxy calls ‘allocative efficiency’ or ‘pareto efficiency’. The argument is that the recipient of the $100, by spending it, will activate production that would not have occurred in a purely free-market economy, and this is supposedly an ‘inefficient’ use of resources.
The argument does not stand up to orthodox analysis of allocative efficiency. It is well established in orthodox theory that once we are outside the world of the ‘first best’ (meaning once there are at least some distortions), the orthodox Theory of the Second Best shows that additional market distortions (e.g. government interventions) do not necessarily reduce allocative efficiency. Once there are at least some distortions – and no orthodox economist would claim an absence of distortions in the real world – the introduction of additional distortions does not, in general, move the economy further away from allocative efficiency, and in fact may move the economy closer toward allocative efficiency.
In other words, it is not valid, on orthodox grounds, to presuppose that in a second-best world an additional distortion will reduce allocative efficiency. Further, it is impossible not to have distortions in the real world because of the presence of what the orthodoxy terms externalities, market failures, public goods, incomplete information, etc. The presence of these problems means that a purely “free-market” economy is not pareto efficient in any case. Since there is no such thing as a first-best world, the market, left to its own devices, is not allocatively efficient, on orthodox grounds.
Now the Austrians, of course, are not orthodox, even though they believe in ‘marginal’ behavior like the orthodoxy. They object to the mathematical analysis used by the orthodoxy, and it is this mathematical work that shows the logical flaws in the notion of an allocatively efficient pure free-market economy. So we can’t really argue with the Austrians directly on these questions, since they object to mathematical logic when applied to models of the economy. Their position is really just an assertion, not an argument.
Even so, the orthodoxy has demonstrated results that shed light on whether the Austrian position on efficiency could really be tenable. Specifically, the orthodoxy has determined the set of highly restrictive conditions that would need to hold for a purely free-market economy to deliver pareto efficiency, and gone on to demonstrate in their Fundamental Theorems of Welfare Economics that if such a first-best world existed, then: (i) any free-market equilibrium outcome would be pareto efficient; and (ii) any desired pareto-efficient outcome could be attained through appropriate lump-sum transfers (redistributions) of income. (It has also been demonstrated that in a first-best world, a command economy could deliver pareto efficiency just as easily as a market economy, since the conditions underpinning a first-best world include perfect information.)
So, in a first-best world, any initial distribution of income is consistent with pareto efficiency. That means it would be possible, in a first-best world, to impose lump-sum taxes (e.g. property taxes, tolls, poll taxes) and lump-sum transfers and then let the market economy deliver pareto efficiency.
These findings make clear that the Austrian argument about the supposed efficiency benefits of private markets (which they merely assert without proof) really reduce to an assertion about income distribution. On the one hand, it is invalid to argue that non-lump-sum transfers (i.e. distortions) inevitably reduce efficiency once we are not in a first-best world (which we are not, and never can be). This is a lesson of the theory of the second best. On the other hand, it is equally invalid to argue that the status quo distribution would be more efficient than any other distribution if we were in a first-best world, since any distribution would be compatible with pareto efficiency. This is a lesson of the fundamental theorems of welfare economics.
The Austrians dislike redistribution. That’s all. Nothing in their position pertaining to efficiency holds up to logical scrutiny.
They also appear to dislike full employment. Otherwise, why would they think that a level of net financial assets too scarce to sustain full-employment output is appropriate, and that attempts to expand net financial assets to enable full-employment output constitutes inflation?
It makes no sense to assert that no expansion of net financial assets is appropriate if such an expansion enables not only more “claims” on real goods, but more “real goods”.
Here, the Austrians rely on another assertion that has been shown to be without basis on mathematical (logical) grounds. They think a pure free-market economy will automatically deliver full employment through wage and price adjustments. There is no basis for this claim, as the orthodox participants in the capital debates openly admitted. Further work by orthodox general equilibrium theorists has revealed many more problems with the claim of an automatic tendency to full employment. (For more on this topic, see Nobel-nomics and Unemployment is a Government Policy Choice.)
Again, Austrians are left in the position of having to reject mathematical analysis (including of the orthodoxy) and clinging to an assertion about market economies that they have not been able to establish on logical grounds. They have not demonstrated – and cannot demonstrate (because the mathematics shows it is impossible to demonstrate) – an automatic tendency to full employment in a purely free-market economy.
In sum, the Austrians wish to keep the level of net financial assets at a level insufficient to enable full-employment output. There reasons cannot be defended on the grounds of pareto efficiency or a supposed automatic tendency to full employment. They are left with a complaint about the redistributive effects of deficit expenditure. In that case, they should be throwing their energies into arguing why one distribution of income is more justifiable than another, even if it coincides with less output and employment and, quite possibly, lower ‘allocative efficiency’.
The Expansion of the “Money Supply” as “Inflation”
The notion that an expansion of the money supply constitutes inflation, in and of itself, is an Austrian definition. It is not anybody else’s definition. For everybody else, inflation means a persistent rise in the general price level (i.e. an increase in the weighted average of all prices).
But let’s consider the Austrian definition of inflation as an expansion of the money supply. If the concern is with the impact on “claims on real goods”, I have responded above. The definition is misleading, in that respect, because the amount of “real goods” can change along with the amount of “claims on real goods”.
If, in contrast, the ultimate concern is with the effect of changes in the money supply on the general price level (i.e. the change in the “claim on real goods” represented in one dollar), then the definition would only be useful to the extent that changes in the money supply were more or less synonymous with proportional changes in the general price level.
The former concern is what bothers the Austrians, and has already been considered in preceding sections. The latter concern is more likely to trouble non-Austrians. In the past, this concern was held by the so-called quantity theorists.
Money supply in the quantity theory referred to currency plus deposits. It excluded reserves, which were part of the so-called money base but not part of the money supply. Quantity theorists claimed that the central bank could reliably control the money supply through its control of reserves, based on the notion of a stable money multiplier.
More specifically, the quantity theorists argued that by adjusting reserves, the central bank could achieve a predictable, multiplied change in the money supply; i.e. they argued that M = mB, where M was the money supply (currency plus deposits), B was the money base (currency plus reserves), and m was a stable multiple (the money multiplier).
We know that no such stable relationship holds, especially under current circumstances, because:
a) The velocity of money (the amount of times on average a unit of money is used in transactions in a given period), and hence the so-called money multiplier, are highly unstable. For example, velocity collapsed with the onset of the Global Financial Crisis. See this link for charts. When velocity collapses, a given money supply is used less for expenditures, and upward pressure on the general price level is reduced.
b) We are nowhere near full employment and full capacity, so even if an expansion in the money supply could be engineered by the central bank through reserve manipulation (and the velocity of money did not collapse at the same time), the additional demand would impact on real output, not just prices. Again, an increase in the money supply enables an increase in real output, not just “claims” on real output.
c) The central bank cannot control the money supply exogenously, since (i) it is private banks who determine credit creation on criteria that have nothing to do with possessing prior reserves, and (ii) the so-called money multiplier is unstable. The central bank can only attempt to influence the demand for credit through its choice of interest-rate target and then supply whatever level of reserves is required to maintain that interest rate. The money supply adjusts endogenously to money demand through the credit creation of the private banks.
In terms of the quantity equation, MV = PY, the quantity theorists assumed a full-employment equilibrium situation in which M (the money supply) was exogenous, V (velocity) was stable and determined by real factors, and Y (real output) was at the full-employment level and hence at its maximum. Under those three assumptions, an exogenous change in M would cause an equiproportionate change in P (the general price level). As has already been discussed, the quantity theorists then made a further assumption that M was a predictable and stable multiple of the money base, B, and that the central bank was free to choose whatever level of B it wanted, irrespective of interest-rate targeting imperatives, supposedly ensuring that the money supply was exogenously controllable. (For more discussion of the quantity theory, see Misplaced Faith in Quantitative Easing.)
In reality, none of these assumptions hold. The central bank cannot exogenously control the money supply, and even if it could, this would not necessarily control the general price level. Further, any expansion in the money supply would not necessarily increase “claims on real output” more than proportionately to the increase in “real output” itself.