3.8  Is government creditor or debtor?

/73/ In MMT, relations between government, central bank, banks, and companies and citizens are interpreted in a way which assumes that government, in cooperation with the central bank, issues its own currency, and that banks are just 'true intermediaries' between the two as well as between government and taxpayers.[1] In MMT's interpretation, government is not debtor but creditor.

Even if it now may be redundant, I want to make clear again the NCT view in this respect. Banks are no intermediaries but are the pivotal actors who decide on primary credit creation and thus the money supply. Central banks react to the banks' proactive monetary initiative and willingly refinance the banks. In western countries, central banks do not finance government. /74/ The reality today is a fractional reserve banking regime backed by central banks and government – in fact, a situation of state-backed banking rule. Government is in no way a primary creditor. Government belongs in the category of nonbanks. To the banks it is debtor, and the debt mountains which almost all advanced industrial governments have accumulated are truly 'majestic'. Governments at high levels of expenditure tend to be chronically short of tax revenue and thus depend on banks to fill the deficit with additional debt. If levels of financial investment and debt become too high and can no longer be served out of current flows of income, or if sovereign debtors default for other reasons, banks (and nonbank financial intermediaries) threaten to fail and then depend on the central banks and the dependent governments to bail them out.

MMT's divergent interpretation of the situation rests on two storylines with no clear interconnection. The one has banks as intermediaries between government and central bank. The other has banks as intermediaries between government and taxpayers, and again dates from Mitchell-Innes:

'The government, the greatest buyer of commodities and services in the land, issues in payment of its purchases vast quantities of small tokens which are called coins or notes, and which are redeemable by the mechanism of taxation, and these credits on the government we can use in the payment of small purchases in preference to giving credits on ourselves of transferring those on our bankers.'[2]

This does not correspond to reality, especially not in a basically non-cash monetary system. Treasury coins (with a residual monetary importance next to nothing) are sold to the central banks, but for the rest coins and notes as well as reserves and deposits are loaned, not spent into circulation. This is not different even if one follows the over-simplification of subsuming government and central bank into one category, called 'government' or 'public sector', as distinct from the 'private sector'. Government and central bank finances cannot sensibly be consolidated into one balance sheet. If the central bank credits banks, it is not the government that credits. If banks credit the government, the reserves involved do not flow back to the central bank but from the banks' account at the central bank on to the government account there, whence it flows back to the banks, or the private sector respectively.

In Mitchell-Innes' quote there is no primary or secondary government 'credit' involved, nor a taxpayer 'credit' to the government – simply payments of available money out and in. Nor does government spend coin into circulation by purchasing something. /75/ Government sells the unimportant amount of coin it still mints to the central bank, depending on the demand for coin as it results from everyday payment habits. The demand for banknotes results from the same habits. The monopoly on banknotes rests with the central bank, and notes are not spent into circulation either, but rather are loaned to the banks or sold to them in exchange for reserves loaned to them. The domineeri

ng monopoly of credit and deposit creation is with the banks, and what they create determines the fractional demand for reserves. Mosler, however, writes:

'business and households in the private sector are limited in how much they may borrow by the market's willingness to extend credit. ... They must borrow to fund expenditures. The federal government, on the other hand, is able to spend a virtually unlimited amount first, adding reserves to the banking system, and then borrow, if it wishes to conduct reserve drain'.[3]

Well, the US government could spend its own banknotes, also its own currency on account, if Congress decided to update its constitutional prerogative of 'coinage' and take back the monetary prerogative from the banking sector. As long as this is not the case, the government cannot freely spend any amount of money. The central bank does not, either. It does not spend money other than for its own office, but lends reserves to the banks according to banks' demand. As an action of quantitative easing in times of crisis, the central bank may lavishly offer reserves to the banks nearly for free, but it is up to the banks how far they make use of this and what they do with the reserves.

Besides, in economically stronger nation-states – in particular the US, with the dollar as the dominant global reserve currency – the government can of course rely on being promptly served by the banks when it decides to borrow. But this is another aspect and not – at least not expressly – part of the MMT storyline. It is implicit in MMT's assumption that the US would be able to incur debts and foreign-account deficits 'forever' (4.2–3). But any government must be careful not to stretch things too much. Sooner or later banks and bond markets will start to think twice, especially if an ever growing part of creditors is non-domestic.

/76/ As opposed to Mosler's thesis, a government in the present banking system cannot spend money without having taken in the money before, just like companies and households. What the government issues is nothing but sovereign bills and bonds, underwritten in the first instance by an exclusive group of large banks who have the privilege of participating in this business. Taxes are not normally used to redeem such sovereign debt to the banks. Tax receipts are immediately spent back into circulation.[4] The same is true of reserves the banks may have had to transfer to the treasury. The banks get the reserves back immediately, since government immediately spends what it receives. Bonds falling due are not normally redeemed either, but are revolved, i.e. payments on interest and principal are made by taking up new debt to the same or a greater extent.

MMT sometimes blurs the distinct meaning of 'unit of account' and 'means of payment'. In some passages one can get the impression that government IOUs are identified with the currency which is created by the banks as a consequence of selling government IOUs to them.[5] However, government IOUs do not circulate as money. There is nothing to quibble about: the state goes into debt with the banks. Far from being the originator or creditor of the money, government is the debtor – actually the biggest debtor of all. The taxpayer’s role in this game is to pay for state institutions and public infrastructure, for government transfers and for interest to banks – and to nonbanks, to the degree that banks on-sell government bonds to investment funds and individuals.

In another source, the MMT thesis of government freely spending whatever it decides is summarised in this way:

'A modern monetary system can best be thought of as a system of debits and credits where government deficit spending credits the private sector and payment of taxes debits the private sector. One might think of deficits as 'printing money' and taxes as 'unprinting money''.[6]

The first sentence is true, but of no relevance here in that it just regards payments back and forth. The second sentence is right in that deficit spending to a degree implies 'printing money' through the banking sector. /77/ But the statement is wrong on the 'unprinting', and wrong in that it withholds the pivotal role of the banks in the process. MMT describes banks' role in this as if it were unremarkable: 'Private banks intermediate between taxpayers and government, making payments in currency and reserves on behalf of the taxpayers'.[7] This is either trivial or misleading. It is trivial in that banks and central banks technically manage all payments among all groups of actors. No doubt they do a good and very useful job in that respect. The important monetary and financial question, however, is who primarily issues the money by spending or loaning it into circulation. In this respect, MMT's analysis is misleading. Far from being 'inter­media­ry', it is the banks who are the determining originators of the money supply.

For sure, banks also react. They react to the demand for money from financial markets, from other banks, from government, from businesses small and large, from private customers. But it is always up to the banks' individual discretion to what extent they want to meet demand from these actors. It is the banks, and the bond markets beyond, who hold the reins.

What about the banks' role as intermediaries between government and central bank? It might look as if there were a mechanism by which central banks fully monetise government debt, something Friedman has previously mused on.[8] Central banks are interdicted to directly take up new bonds from the government. In the emergency since 2008, though, they have heavily absorbed government bonds from the secondary market in a continual attempt to prevent a meltdown of bond markets. As regards the initial placement of government bonds, however, central banks are not allowed to buy these; nevertheless the central bank contributes to financing government debt in that, firstly, the government sells new bonds to banks and the banks, secondly, can sell or lend them on to the central bank in exchange for the reserves and cash the banks have paid to the government.

The snag with this story is that it only applies partially. The banks do not need to have refunded the reserves they had to use for making bond-related payments to the government, because they get the reserves immediately back upon government expenditure anyway. /78/ To the degree that banks extend the overall money supply, they of course need to be refinanced, but only fractionally, at about 2.5–10%, not to the full amount of credit and deposits they made out. The reserves needed for making out payments to a government central-bank account are more or less part of the existing stock of reserves built up over time. Placement of government debt has after all developed into a continual large-volume activity.

If the MMT story were right, central banks would hold most of government debentures. They do not. Central banks hold government bonds only to a minor extent. In the US as much as in Europe, the biggest part of government debt is not passed on to the central bank but
a) sold on to other creditors who pay by on-lending already existing deposits (secondary credit), and
b) kept in the banks' proprietary portfolio.

/78/ In the euro area, on average about 55% of government debt is held by domestic and foreign banks, 33% by funds and insurance companies and the remaining 12% by households.[9] Holdings of public debt by European central banks (in pre-crisis times) have been about 0–4%, since central banks come to hold government debt only when extending the fractional money base of notes and reserves or fine-tuning repo operations; and even in those operations, there are not just government bonds involved, but other kinds of securities too. In the US, the Fed system's share of government bonds is at about 10–15%, and thus not too important either. Domestic and foreign banks hold significantly less sovereign debt than in Europe, while the share of public funds such as Medicare or the Social Security Trust Fund and private investment funds within the country and abroad is correspondingly higher.[10]

MMT however, as discussed in 2.5, believes in treasuries and central banks jointly exerting control over the banks' money supply via base-rate policies. To Wray, it appears that 'the treasury cooperates with the central bank, providing new bond issues to drain excess reserves, or with the central bank buying treasuries when banks are short of reserves. ... for this reason, bond sales are not a borrowing operation (in the usual sense of the term) used by the sovereign government, instead they are a tool that helps the central bank to hit interest rate targets'.[12]

/79/ New bond issues, however, do not drain excess reserves as these flow from the banks’ operational balances to the treasury account, and from there immediately back to the banks' operational balances. Reserves are drained when the central bank re-sells or gives back bonds to the banks which it had absorbed before. In view of governments' creditor breakdown, the above mechanism represents quite a far-fetched interpretation. Government does not issue the currency on bank account, nor do government or central banks proactively issue coin, notes and reserves. Government borrows. In the present system of fractional reserve banking, government belongs in the group of nonbanks, and the central bank is just the lender of least reserves and last resort for the banks, rather than being the first and sovereign issuer of the money supply to the government. It could be that way. But the predominance of banking doctrine and official political thought will want it to be different.

Against this background, a conceptual pair introduced by MMT, i.e. 'horizontal' and 'vertical' money, is ill-conceived from the outset. According to Mosler, government spending results in the creation of 'vertical' money for it would increase banks' reserves without a corresponding liability in the banking system. In contrast to this, when banks extend credit this is considered to be 'horizontal' for it occurs completely within the banking system and creates a corresponding liability (bank deposits).[13] The 'vertical' part, though, as just explained, does not correspond to operational facts. Treasuries have to take in central-bank deposits (reserves) or bank demand deposits before they are able to spend them. Furthermore, and to the benefit of the private monopoly of bank money, central banks in advanced countries are not allowed today, to directly credit treasury accounts; and what they re-finance by purchasing sovereign bonds is much less than what government needs to get.

'Vertical vs horizontal' money is an unrealistic construction. It 'consolidates' two institutions – treasury and central bank – into one category while in actual fact /80/ the two are different and fulfill different functions separately from each other (4.1–2). This obscures and re-interprets the fact that government has to borrow from the banks, becoming ever more dependent on these the more indebted government gets; and that central banks primarily serve private banking interests rather than those of the government.


- - - - 

[1] Wray 2012 280.

[2] Mitchell-Innes 1914 153ô52.

[3] Mosler 1995 8–9.

[4] Also see Walsh/Zarlenga 2012 4.

[5] Wray 2012 xv, 39–40, 259.

[6] Pragmatic Capitalism. Blog by Cullen Roche, here: Understanding the Modern Monetary System, http://pragcap.com/resources/understanding-modern-…etc.

[7] Wray 2012 111, 276.

[8] Friedman 1969b.

[9] ECB, Monthly Bulletins, Table 6.2.1.

[10] Arslanalp/Tsuda 2012, 12, 52-53. - U.S. Treasury Department, Monthly Treasury Bulletin, June 2012.

[12] Wray 2012 112.

[13] Cf. a summary of vertical and horizontal money by Mosler, published in an article by Roche 2010.A