Originally published October 9, 2011 on the New Economic Perspectives blog.
Last week we began to analyse fiscal and monetary policy formation by a government that issues its own currency. We went through a list of false statements about sovereign government spending, and offered a list of general statements that do apply. Let us now begin to examine in more detail the government’s budget and impacts on the nongovernment sector. This week we will look at the relation between budget deficits and saving, and the effects of budget deficits on bank reserves and interest rates.
Budget deficits and saving. Recall from earlier discussions in the Primer that it is the deficit spending of one sector that generates the surplus (or saving) of the other; this is because the entities of the deficit sector can in some sense decide to spend more than their incomes, while the surplus entities can decide to spend less than their incomes only if those incomes are actually generated. In Keynesian terms this is simply another version of the twin statements that “spending generates income” and “investment generates saving”. Here, however, the statement is that the government sector’s deficit spending generates the nongovernment sector’s surplus (or saving).
Obviously, this reverses the orthodox causal sequence because the government’s deficit “finances” the nongovernment’s saving in the sense that the deficit spending by government provides the income that allows the nongovernment sector to run a surplus. Looking to the stocks, it is the government’s issue of IOUs that allows the nongovernment to accumulate financial claims on government.
While this seems mysterious, the financial processes are not hard to understand. Government spends (purchasing goods and services or making “transfer” payments such as social security and welfare) by crediting bank accounts of recipients; this also leads to a credit to their bank’s reserves at the central bank. Government taxes by debiting taxpayer accounts (and the central bank debits reserves of their banks).
Deficits over a period (say, a year) mean that more bank accounts have been credited than debited. The nongovernment sector realizes its surplus initially in the form of these net credits to bank accounts.
All of this analysis is reversed in the case of a government surplus: the government surplus means the nongovernment sector runs a deficit, with net debits of bank accounts (and of reserves). The destruction (net debiting) of nongovernment sector net financial assets of course equals the government’s budget surplus.
Effects of budget deficits on reserves and interest rates. Budget deficits initially increase bank reserves by the same amount. This is because treasury spending leads to a simultaneous credit to the bank deposit account of the recipient and to that bank’s reserve account at the central bank.
Let us first examine a system like the one that existed in the US until recently, in which the central bank does not pay interest on reserves. Deficit spending that creates bank reserves will (eventually) lead to excess reserves—banks will hold more reserves than desired. Their immediate response will be to offer to lend reserves in the overnight interbank lending market (called the fed funds market in the US).
If the banking system as a whole has excess reserves, the offers to lend reserves will not be met at the going overnight interbank lending rate (often called the bank rate, but in the US this is called the fed funds rate). Hence the banks with excess reserve positions will offer to lend at ever-lower interest rates. This drives the actual “market” rate below the central bank’s target rate for overnight funds.
Once the rate has fallen sufficiently far away from the target, the central bank will intervene to remove the excess reserves. Since the demand for reserves is fairly interest inelastic, lowering the offered lending rate will not increase the quantity of reserves demand by very much. In other words, it is difficult to eliminate a position of system-wide excess reserves by lowering the overnight rate. Instead, the central bank must remove them.
The way that it does this is by selling from its stock of treasury bonds. That is called an open market sale (OMS). An OMS leads to a substitution of bonds for excess reserves: the central bank’s liabilities (reserves) are debited, and the purchasing bank’s reserves are also debited. At the same time, the central bank’s holding of treasuries is debited and the bank’s assets are increased by the amount of treasuries purchased.
Since the bank’s reserves decline by the same amount that its holdings of treasuries are increased, this is effectively just a substitution of assets. However, it now holds a claim on the treasury (bonds) instead of a claim on the central bank (reserves); and the central bank holds fewer assets (bonds) but owes fewer liabilities (reserves). The bank is happy because it now receives interest on the bonds.
It is easy to see that the same process would be triggered even if the central bank paid interest on reserves—as is now done in the US. Once banks have accumulated all the reserves they want, they will try to substitute for higher-earning treasuries. They will not push the overnight rate below the central bank’s “support rate” (what it pays on reserves)—since no bank would lend to another at a rate below what it can receive from the central bank. Instead banks with undesired reserves will immediately go into the treasuries market to seek a higher return.
The impact, then, will be to push rates on treasuries down. In this second case, the central bank need not do anything—it does not need to sell bonds since it maintains its overnight interest rate by paying interest on reserves.
In practice, a central bank that adopts this second procedure usually pays a slightly lower rate on reserves than it charges to lend reserves. As discussed earlier, in the US the central bank lends “at the discount window” and at the “discount rate”. It might charge 25 basis points (0.25 percentage points) more on its lending than it pays on reserves. For example, it might charge 2% on loans and pay 1.75% on reserves. The “market” interest rate on interbank lending will remain approximately within that band since a bank needing reserves has the option of borrowing at the central bank at 2%, while a bank having extra reserves can earn 1.75% simply by holding them at the central bank.
That’s enough for today—just about over my 1000 word target! Send your comments and questions.