Optimum Taxation Policy and the Impact of Public Debt Under Modern Monetary Theory
We have been sympathetic critics of MMT on this blog, seeking to integrate some of its key insights with circuitism. Its basic insight, that a currency issuing country can spend before it taxes, is broadly correct – as state money is spent into existence. Also taxes net of state spending can equally destroy the purchasing power of money (we do not hold that taxes per-se destroy money – as Parguez, Seccarrecia and the ‘State in the Circuit’ theorists for example argue as – rather more subtlely that demand is altered by the net change in debt. So if a government spends less than the debts it writes down through tax raising then it reduces net effective demand and contra increases it – credit extinguished is purchasing power destroyed and vice versa.) This is an implication of the Steve Keen modification of Walras’s Law to incorporate changes in aggregate supply and demand through credit financed investment as first introduced at the MMT/Circuitist meeting last year. Though we accept that demand for taxes creates demand for currency we consider that this operates through the framework of credit based money – that is it is demand for credit that creates demand for money per-se as a store of value whereas taxes create demand for a particular currency as a unit of account. This is but a nuancing of the chartalist position and reflects historical cases where money has existed outside state monopoly currencies. Here we don’t really disagree with Randall Wray
we are not trying to claim that, um, taxes drive money, that that is necessary. It is merely sufficient. Taxes will drive a money. That’s a sufficient condition. Um, it’s perfectly conceivable – and people have many stories about this – of private entities creating a unit of account, denominating liabilities in that unit of account, and then exchanging those liabilities and using those in payment. It works pretty well in theory. It’s just that in practice we don’t find these. [rather I would say they are rare for example money in prison camps and the 18C Scottish banking system]
A final issue is that sectoral identities express accounting constraints but are not substitutes for a full monetary theory of production. So we would argue for example that ‘classical’ position (put most ably by Torrens) that demand itself will not lead to growth unless that demand is made effective through increased production of goods and services achieved through capital formation and accumulation. The purpose of state spending is to maintain that growth not per-se to increase money, indeed if the nominal purchasing power of money is decreased by credit creation by more than the growth in production then state money creation can harm price stability.
Within this framework it is difficult to reconcile neo-classical ideas such as a ‘state budget constraint’, ‘ricardian equivalence’ or ‘debt overhang’ (the latter might apply to the private sector but not to a sovereign currency issuer), but, in the post-Reinhart-Rogoff fallout (despite their ‘we are not really austerians ‘ post-hoc protestations in the NYT today) , does this leave any role at all for debt levels, particularly high public debt levels, in affecting economic growth?
So where does that leave the role of taxes? The MMT position is that taxes are used to control aggregate demand. (see Warren Mostler) This is correct though formally we would argue that there are dual channels. Firstly the writing down or creation of state debt based money is balanced by creation of private sector financial assets, taxes then can be used to net increase or decrease the money supply. For none-debt based money there is no accompanying financial asset so taxes in this case are necessary to directly rather than indirectly offset state money creation. But in the latter cases taxes would not be operating at the margin, so would have to be higher and this would suppress capital formation.
What then should the optimal rate of tax be? The optimal rate will need to be very different between debt based and none debt based state money regimes, according to the financial obligations the state sector owes to the private sector and between different real interest rates. If the real rate of interest is low then there is no effective difference, future debt obligations will be near zero. In these cases if there is a demand problem governments could create none debt based money to reflate demand and to take advantage to drive investment. The great disadvantage with this is the lack of a contractual obligation by the state to contract money later once the initial impulse from credit is spent, there is the political risk that taxes will not rise to deflate aggregate demand. The risk of debt free money is that the private sector could price in an inflation in the value of money and shift capital towards regimes with contractual obligations for debt repayment. Even in cases where interest rates remain low for a considerable period (as in Japan) there will be the considerable risks to the net inflow of state funds from an future upturn in interest rates which has seen an export of capital.
In cases where there are high real interest rates then future public debt servicing requirements will be high. To avoid the deflationary effects of paying down debt the state will need either to raise taxes – to maintain fiscal/demand neutrality, or further raise public debt, with implications for implied future deflation of demand. So we can see even in a world without a state budget constraint debt can be a problem if the change in debt impacts on future demand. The basic functional Lerner functional finance position which MMT inherits though is correct, growth in debt is manageable if outpaced by growth in the economy. In Keyne’s fundamental insight the ‘debt pays for itself’ if a boost to demand leads to a multiplier effect leading to increased tax revenues. Of course infinite growth in debt is impossible so there must be fiscal limits on the scope for such interventions without some form of debt write down.
As well as the impact of changes in debt there may also be impacts from the relative deficit levels. Public spending has a high local multiplier it is likely to be spent locally, whereas private investment in the local economy will see short term imports of capital goods and longer term benefits from exports. So austerity induced by high levels of public debt is likely to see capital outflows from those who receive interest payments to higher yielding markets without austerity induced stagnation. Indeed that is just what we have seen in Japan, where as Aziznomics correctly points out the private sector has now largely delevered leaving on high levels of public debt. Over the next few months graduate economometric students are likely to have a field day charting the relationship no longer only between public debt and growth but overall debt, private debt and changes to each under different initial conditions of growth and inflation. This is likely to be interesting but simply exposes the theoretical gaps we have in terms of the relationship between finance and debt.
Does old fashioned IS/LM help use here? No as the IS and LM curves only intersect at equilibrium and with high unemployment the labour market must be out of equilibrium and so must one or another of the other markets. as Hick’s stated in his late mea culpa it must be replaced by a disequilibriium understanding of the three way relationship of the money, labour and goods markets.
What then of the MMT solution to force the labour market back into equilibrium through an employment guarantee? Given the multiplier effects of public spending a rise in confidence could see capital repatriation and a rise in growth that would more than offset rises in future interest payments from the rise in debt. Indeed this is just what we have seen in the early months of Abenomics. Though prices are still declining now 3/4 of japanese consumers expect prices to rise. More radical measures could be tried to back an MMT inspired programme, for example a Central Bank could use its balance sheet expansions to invest in equity of firms that invest, hire workers and raise wages.
If then employment is forced back to a full employment equilibrium point the policy test is whether equilibrium can be restored between the goods and money markets. If there is still excess demand for money (hoarding) and ‘credit deadlock’ then the issue is how the holders of that money can be induced to spend it. Some ideas:
-A tax on idle balances used either to avoid austerity to fund interest on public debts;
-Cutting payroll taxes to zero and/or introducing a negative payroll tax (subsidy) to those at the bottom of the labour market;
-Shifting taxes to those things that will not deter capital formation and accumulation, such as land, &
-Directly monetising deficits/and or the central bank burning bonds it buys from state money creation. This would be inflationary but hey if you want to meet a minimum inflation target hey this is the way to do it.
This post was prompted by Tim Wortsall, of all peoples, post on whether progressive taxation makes sense a MMT world. In a sense he was right, and perhaps Randall Wray in response was not, the principles of taxation radically change in an MMT world as the overiding princple is how taxation can create inducements to restore full employment not how taxation can fairly fund public spending.
Hicks, J. (1982), Money, Interest and Wages, Collected Essays on Economic Theory, Oxford, Basil Blackwell.
Brett Fiebiger Modern Money Theory and the ‘Real-World’ Accounting of 1-1, The U.S. Treasury Does Not Spend as per a Bank
Scott Fullwiler, Stephanie Kelton & L. Randall WrayModern Money Theory: A Response To Critics
Brett Fiebiger A Rejoinder to “Modern Money Theory: A Response to Critics”