Posts Tagged The Neutrality of Money

The Myth of Neutral Money

Conventional economic theory purports that money is neutral: that is, changes in the money supply do not affect the ‘real’ economy (patterns of trade, production and consumption). Instead, the only interaction between the monetary and the real economy is thought to be through the determination of nominal quantities such as prices, wages, exchange rates and so forth. Though a change in the quantity of money may create short term disruptions, the economy eventually will settle at the same long term equilibrium as before.

An extreme interpretation of the neutrality of money would lead to absurd conclusions, such as the idea that the ‘real’ economy would operate the same whether it had a gold standard or hyperinflation. I’d therefore interpret the ‘neutral money’ view as the claim that, at ‘normal’ levels of money, a change in the money supply will not alter the long term economic equilibrium. This viewpoint was described well by Milton Friedman in his famous ‘helicopter drop’ story, which I will use as the basis for my critique.

Friedman began his story by imagining a community in economic equilibrium:

Let us suppose, then, that one day a helicopter flies over our hypothetical long-stationary community and drops additional money from the sky equal to the amount already in circulation-say, $2,000 per representative individual who earns $20,000 a year in income. The money will, of course, be hastily collected by members of the community. Let us suppose further that everyone is convinced this event is unique and will never be repeated….

…People’s attempts to spend more than they receive will be frustrated, but in the process these attempts will bid up the nominal value of goods and services. The additional pieces of paper do not alter the basic conditions of the community. They make no additional productive capacity available. They alter no tastes….the final equilibrium will be a nominal income of $40,000 per representative individual instead of $20,000, with precisely the same flow of real goods and services as before.

It is first worth noting that the ‘real’ benchmark for Friedman’s equilibrium is somewhat hard to define – after all, the real economy is an artificial construct with no real world counterpart. Economic agents must necessarily negotiate with and act on the nominal: as John Maynard Keynes pointed out, workers do not have control over the general price level, and hence can only impact their nominal wages. Clearly, nobody has control over the ‘general price level’ (itself surely a problematic concept), so Keynes’ argument also applies to prices, exchange rates and other variables (sorry, economists, no ‘as if‘ arguments allowed). Nominal variables are actually observable, while proponents of money neutrality have no moneyless baseline by which they can judge real activity, despite repeatedly appealing to the idea.

More generally, I find the idea expressed by Friedman – that the economy will tend toward a stable, long term equilibrium, perhaps oscillating in the short term – is often used by economists, but is rarely fully justified. It is merely assumed that the economy will behave this way, and any erratic behaviour – such as money illusion, and sticky wages/prices – can be dismissed as short term ‘noise’. However, seems to me that such an idea can only be sustained by sweeping potential problems under the rug. Indeed, this supposed ‘noise’ (a) could be more relevant to understanding the system than the equilibrium and (b) could have a permanent impact on the economy and therefore equilibrium itself.

It is entirely possible – common, even – for a system’s behaviour to differ markedly from its equilibrium value(s). This is true even if the system has some tendency toward the equilibrium**. In examples such as Friedman’s, a monetary disturbance will surely alter people’s perceptions (something Friedman acknowledges), and they will engage in economic activity based on these altered perceptions, continually adjusting as they overshoot or undershoot their plans. Hence, a monetary shock could push the economy out of equilibrium and into a long term trajectory that has little relation to its initial position. Furthermore, if there are constant changes in the money supply, any tendency toward equilibrium will be continually thwarted. As Irving Fisher put it, “equilibrium is seldom reached and never long maintained”.

In fact, monetary disruptions can have even more fundamental effects than this. Due to path dependence, a monetary disturbance could change not only the immediate behaviour of the system but also the long term equilibrium itself. If money is invested based on people’s altered perceptions, long term capital goods can be created that otherwise would not have been. This phenomenon is all the more pronounced if new money is not evenly distributed but injected at specific points, something known as Cantillion effects. Friedman considers this possibility, but dismisses it without must justification (“during the transition some people will consume more, others less. But the ultimate position will be the same”. Erm, why?). The fact is that a company, individual or government who finds themselves with a relatively higher income due to a monetary injection could make important investments, altering long term patterns of production and consumption.

The role of finance

All of these effects would be important even in Friedman’s imaginary world. But it only becomes clear quite how important they are when we consider the nature of the modern banking system is particularly important, something entirely absent from Friedman’s example. This is because in neoclassical theory, banks are generally assumed to be ‘intermediaries’ who take money from Peter and loan it to Paul. The result is that banks only really ‘smooth things out’ by matching borrowers and lenders, and hence can be assumed away, perhaps save for one or two ‘frictions’ (transaction costs, interest rate mark ups). Effectively, we model the economy as if Peter is loaning directly to Paul, and from there we suppose that the nominal amount of money lent & borrowed is arbitrary, having no impact on Paul’s ‘real’ activity.

However, as has been comprehensively discussed in the blogosphere, this is not how banks work in the real world. Rather than taking money from Peter and loaning it to Paul, banks simply create a loan for Paul out of nothing. The ‘other side’ of the loan is not Peter’s deposit; it is a deposit that belongs to Paul, created at the same time as his loan, at an amount equal to the loan itself. At the moment the loan is issued, the money supply expands, and when the loan is repaid, the money supply will contract. Hence, the real economic activity Paul engages in is inextricably intertwined with the change in the money supply. The goods and services Paul buys, or the business he starts, or the assets whose price he bids up are a direct consequence of the same decision that expands the money supply. We cannot say that only prices will be affected, as the loan has a clear impact on production and consumption patterns in the real economy.

Furthermore, the constant extending and repaying of credit means the money supply is always expanding and contracting, with no discernible regularity. This is in stark contrast to the idea that the quantity of money simply moves from one long-term quantity to another, or increases at a constant rate. The idea of an underlying ‘real’ equilibrium simply becomes irrelevant when the nominal economy is constantly shifting like this, as irrelevant as discussing a surf board on calm waters if we want to understand its motion when it’s riding a wave.

Lastly, I previously noted that nominal variables are the variables which are actually observed and used in the real world, and nowhere is this more important than in the financial sector*. It is clear that by doubling the quantity of money in circulation, the relative value debts and assets would halve, which would have a big impact on the economy – imagine waking up to find your savings and mortgage were now worth half as much! Plans would be thwarted; firms, households and the government would find themselves in dramatically different financial situations: better or worse depending on whether they were a debtor or creditor. Bankruptcies and spending sprees would surely ensue. Likely, it would be a highly chaotic situation.

The constant interaction between the real and nominal – whether due to people’s perceptions, the financial sector, Cantillion effects, or what have you – means that they are impossible to separate. This leads me to question how useful the idea of real variables is, and whether theories should use nominal variables instead. This is especially important when trying to understand the role of assets and the financial sector – in fact, economist’s ‘real’ benchmark, and their adherence to the neutrality of money, which allowed them to gloss over the role of money and finance, surely helped blind them to the financial crisis. Perhaps further acknowledging the importance of money and the nominal could be a positive step forward for economic theory.

*I have seen people suggest that such variables should be made real to ‘correct’ the problem. Well, this was tried in Iceland, and it didn’t work. You simply cannot force the world to behave like theories; you have to do things the other way round.

**This is easy to show using difference or differential equations. Try, for example, plugging values into y(t+1) = y(t)*(1 – a*(y(t) – y’), where 0 < y < 1, a is some constant, and y’ is the equilibrium value of y. There is a negative feedback loop, yet depending on the value of a, and the initial values, the average can be far from y’ for long periods of time.


, , ,