Throughout the history of economic thought, opposing perspectives on interest and money have created fundamental divides between the various schools. In the one camp, interest is regarded as having a real determination, with monetary policy ultimately at the mercy of the markets. In the other camp, the determination of interest is considered to be monetary, and a politically determined variable.
This post is a little broader than normal. It was prompted by a couple of questions that are not really connected, but acted as a catalyst for what follows. I will probably explore individual aspects of the subject matter here in future separate posts.
Interest on public debt
The first question was one I read recently on a discussion forum. It is a common question:
I don’t say that all public debt is bad. Only public debt that is disproportionately large.
Let me ask you a question. Suppose the public debt is so high that the interest on it is 100% of GDP. Do you recognize that that would be a problem?
MMT suggests that the rate of interest has a monetary determination, and is in fact a policy choice. In view of this, it would make little sense for the central bank to maintain interest-rate targets that caused the government’s interest obligation to be 100% of GDP. Scott Fullwiler provides a comprehensive analysis of the MMT position in this paper.
In fact, MMT economists typically argue for a zero interest-rate policy (e.g. see this post at billy blog).
Ideally, there would be no debt issuance (reserves would just be allowed to mount in the case of deficit spending) and a zero (or near zero – e.g. 0.25%) interest rate paid on reserves. There is little reason to provide a risk-free positive rate of interest. It is basically just corporate welfare.
Even so, let’s briefly reflect on the consequences of a 100% of GDP interest-rate obligation before exploring the MMT position. A key question, in this scenario, would be: what are the recipients of these interest payments going to do with them? If they are going to spend out of them beyond the capacity of the economy to absorb the extra expenditure at current prices, there will be inflation. If, instead, they are simply going to allow the interest payments to add to their existing financial wealth, it will not be inflationary until and unless they do decide to spend them at a rate too rapid to be absorbed by the economy at current prices.
But even if the situation was non-inflationary, this does not mean the policy would be a desirable one. There is still the question of why this risk-free interest income should be handed over to the recipients. Paying out an amount of interest equal to GDP represents a large handout to a section of society for riskless and unproductive activity.
It is clear that even though there is no need to match government net spending with bond issuance, some financial interests would object to the discontinuance of the practice, and would also oppose a zero interest-rate policy. For one thing, simply allowing government net expenditure to result in a build up of reserves makes patently clear that the corporate sector is playing absolutely no financing role in the process. And if they are not playing a financing role, it is hard to see why they should be entitled to an interest payment.
In a revealing post, Bill Mitchell recounts how the class-interested politics of this issue played out in Australia in the early 2000s. The Australian government ran fiscal surpluses ten years out of eleven from 1996-2006. Initially, the government stopped issuing debt. After all, they were in surplus. By conventional wisdom, there was no need to borrow when taxes exceeded spending.
But the reduced supply of government bonds caused consternation in the financial community, which felt it “needed” the risk-free interest rate as a basis for “pricing” riskier financial assets (although zero would serve just as well as any other rate as a basis for such “pricing”). The response of the government – ridiculously – was to start issuing debt even though it was running surpluses. It could hardly be made any clearer that public debt issuance is corporate welfare, not financing of deficits.
As for the 100% of GDP interest obligation, strong economic growth causes the fiscal deficit to shrink endogenously through the automatic stabilizers. Public debt, or the build-up of reserves if debt-issuance is discontinued, will automatically decline as a proportion of GDP. Fullwiler, in the paper linked to above, points out that, under most scenarios, provided the nominal rate of interest is kept below the nominal rate of GDP growth, this will be sufficient to ensure “fiscal sustainability” in the sense the term is used by neoclassical economists (that is, deficit expenditure under these circumstances will not eventually cause runaway inflation).
Consider what happened during and after WWII. The US fiscal deficit during the war reached twice its current level as a percentage of GDP. This government net expenditure and bond issuance directly preceded the only sustained period under capitalism of generally strong growth, low unemployment, and fairly stable inflation. In numerous countries, unemployment remained very low in the twenty-five year period after the war alongside modest, stable inflation.
The opponents of fiscal deficits and public debt in the 1930s and 1940s were just as vehement as now, and the arguments proved to be baseless.
Returning to the MMT perspective on monetary policy, there is little reason to think a zero interest-rate policy would be inflationary if pursued in conjunction with appropriate fiscal policy and financial regulation. Fiscal policy has the advantage of being more readily targeted at the specific sources of inflationary pressures, in contrast to the bluntness of monetary policy. (See Bill Mitchell’s Asset bubbles and the conduct of banks.) Those who think low interest rates are inevitably inflationary need to explain the monetary transmissions mechanism they have in mind.
The argument used to be based on the so-called money-multiplier mechanism. It is getting harder to find prominent central bankers who would defend that position now, and some are beginning to make their thoughts public on the matter. For example, the money-multiplier theory is explicitly dismissed as fundamentally incorrect in a recent BIS paper.
Statements by Ben Bernanke indicate that he, too, is fully aware that monetary policy does not work directly from reserves to money. That is not what he is trying to achieve through quantitative easing (QE), for instance. He is hoping for either interest-rate or wealth effects on aggregate demand and an increase in expected inflation (because this would lower the real rate of interest, which neoclassical economists hope might encourage demand for credit and private investment).
From the perspective of MMT, Bernanke’s hopes are forlorn. Interest rates don’t have strong effects on aggregate demand. More importantly (since it is not even clear that Bernanke is confident on the likelihood of significant interest-rate effects), the attempt to induce wealth effects is ill advised. If this policy succeeded, all it would do is encourage another unsustainable build up in private debt.
The appropriate policy approach under current circumstances is to enable the private sector to get its debt under control, not encourage it to go further into debt. This can be done through either orderly debt write downs or government net expenditure that sustains demand, income and private saving, enabling the private sector to repair balance sheets.
The rate of interest
There is a fundamental difference between the neoclassical and MMT perspectives on interest rates. Neoclassical theory depicts real interest rates as being determined by real factors (summed up as productivity and thrift) and reflecting scarcity. According to MMT and most other Keynes-influenced approaches to macroeconomics, interest rates have a monetary determination and are distributive variables.
I like the way Fullwiler sums it up in the paper linked to above:
In short, the orthodox concept of fiscal sustainability is flawed due to its assumption that a key variable—the interest rate paid on the national debt—is set in private financial markets as in the orthodox loanable funds framework. On the contrary, as a modern or sovereign money (Wray, 1998, 2003) system operating under flexible exchange rates, interest rates on the US national debt are a matter of political economy (Fullwiler, 2005, 2006).
The sharp difference in perspective over interest-rate determination is partly what the capital debates were about. Once it is acknowledged that capital goods are heterogeneous, interest rates cannot be assumed to play, at the macro level, the role traditionally attributed to them by neoclassical theory, nor have the real determination traditionally conceived in that approach. This is well established, but largely ignored by neoclassical economists (or avoided, by avoiding macro questions).
The capital debates established that the notion of an aggregate production function is invalid outside a single-commodity world. This means, in turn, that the supposed downward-sloping aggregate ‘demand for capital’ schedule, based on the ill-defined concept of an aggregate marginal productivity of capital, is also invalid. In neoclassical theory, this aggregate ‘demand for capital’ function was supposed to form one half of the ‘loanable funds’ market and, in combination with the preferences of savers, determine the real rate of interest.
But on top of the aggregation problems uncovered in the capital debates, there is no such thing as a ‘loanable funds’ or ‘investment-saving’ market in any case. This concept, along with its associated notions (e.g. “crowding out”), is not applicable to the current monetary system. Investment (and government expenditure) are not dependent on the prior existence of some pool of saving (or tax revenue), and in fact this would be operationally impossible in a modern monetary system.
Vertical transactions (those between the consolidated government sector and non-government) impact the level of net financial assets. Horizontal transactions (those within the non-government, such as private bank lending) essentially leverage off the level of net financial assets provided through vertical transactions. (These concepts are explained very clearly in this paper by Warren Mosler and Mathew Forstater.)
Vertically, government spending creates tax revenue. That is, ex nihilo creation of net financial assets through spending makes possible the destruction of net financial assets through taxing.
Horizontally, private loans create deposits. That is, non-government assets create corresponding non-government liabilities.
In the real economy, investment creates (desired) saving through income adjustments. More generally, there can be no leakages until there have been injections. Government spending, private investment and export demand create private saving, import spending and tax revenue. There can be no private saving until there is some income to save out of, and there can be no income until production is initiated through spending.
MMT on the fringe?
The other question that motivated this post is the following:
MMT is just a fringe theory, and also relatively new. Why should I take it seriously?
The extent to which MMT is fringe and new depends on which aspect is considered.
Understanding of the chartalist nature of money has a long history. Mosler and Forstater (in the paper just linked to) note:
Interestingly, the Chartalist view of a tax driven currency can be found in the writings of Keynes (not to mention Adam Smith!), the Post Keynesians, and the Circulation theorists, yet it is almost always presented as an aside, with the implications remaining unexplored (see Wray, 1998, on Smith, Keynes, and Post Keynesians such as Minsky; for the Circulationists, see Graziani, 1988).
In general equilibrium theory, economists now make use of the notion of a tax basis for the currency (a fundamental chartalist position) because otherwise it is hard for them to work money into their theory in a way that makes sense; i.e. that can explain a motive for holding money rather than simply engaging in barter. None of the commonly heard post-rationalizations of money can explain this point, but instead fall into an infinite regress, such as if A is willing to hold money because B is, why was B willing to hold money? Why was the first person who held money willing to do so? Once there is a satisfying answer to this question, a lot is learned about what drives a particular monetary system.
Endogenous-money theory has been strongly established in Post Keynesian theory for some time now. Endogenous-money theory is entirely consistent with MMT. Basically, MMT provides insights into the interactions between the vertical and horizontal components, whereas earlier endogenous-money work in Post Keynesianism focused on the horizontal aspects.
Further, aspects of endogenous-money theory are increasingly finding their way into neoclassical monetary economics. It does not seem out of the question that this will increasingly occur in the case of MMT as well, because close analysis of monetary operations and the endogeneity of money leads to these insights.
It may not be apparent to non-economists how confounding the events of the last few years have been for neoclassical economists. On a personal level, I became aware of MMT while searching for different explanations of the crisis. Neoclassical theory, as it existed prior to the crisis, had no answer for how and why the crisis arose, and certainly didn’t see it coming. They had no idea how to respond to the crisis other than revert to what they had rejected as “Keynesian” (i.e. fiscal measures) in the past. They could not conceive, within their theory, that a massive expansion of reserves by the Fed could possibly have no impact on the broader money supply and prices.
In contrast, if it is asked who anticipated the crisis, the answer is either economists who always expect crisis (e.g. Austrians) or economists who had a relevant reason for expecting crisis (debt deflationists understood much of the story, MMT almost the whole story). Wynne Godley and others working within the sectoral-balances approach, including Randall Wray (a leading MMT economist, and Minsky’s most well known student), have a trail of papers dating from the late 1990s pointing out that the fiscal austerity of the period corresponded, as a matter of accounting, to non-government running deficits (spending more than its income), resulting in an unsustainable build up of private debt. This point was emphasized, for example, in a paper by Godley and McCarthy in 1998 entitled, ‘Fiscal Policy Will Matter‘, and in another by Godley and Wray in 1999, ‘Can Goldilocks Survive?‘. This argument, though undeniably prescient in retrospect, had no place at all in neoclassical theory of the time.
Now, in contrast, things have shifted quite a lot in neoclassical economics. Although there are extremists (e.g. the Chicago School) who simply ignore reality and persist with their fantasies no matter what, this is not how the moderate majority of the profession operates. Their modus operandi, when they see valid ideas from outside of neoclassical theory that help reconcile empirical contradictions within it, is to find ways to smuggle the new ideas into neoclassical theory without doing so to such an extent that it dismantles their entire theory. They did this with Keynes, and bastardized his theory in the process. What many people know as Keynesian economics is not Keynesian economics at all, but the neoclassical synthesis.
Currently, some neoclassical economists are doing the same with endogenous money (e.g. rejecting the money multiplier) and Minsky. There has been a significant increase in neoclassical interest in Minsky, which would have been unthinkable until recently.
So, yes, MMT can be regarded as fringe and relatively new in some respects, but it has emerged out of long theoretical traditions (e.g. chartalism, the principle of effective demand, functional finance) or well established monetary theory (e.g. money endogeneity) in response to the changed institutional realities of the post-Bretton Woods era, and I think there are already signs that the newer aspects are becoming less fringe, and it seems likely the trend will continue.