A normal gap between supply and demand for some subset of currently produced commodities is not a serious problem, because it is balanced by excess demand for other currently produced commodities. As industries suffering from insufficient demand shed workers, industries benefiting from surplus demand hire them. The economy rapidly rebalances itself and thus returns to full employment – and does so with a configuration of employment and production that is better adapted to current consumer preferences.
DeLong then explains the problem:
By contrast, a gap between supply and demand when the corresponding excess demand is for financial assets is a recipe for economic meltdown. There is, after all, no easy way that unemployed workers can start producing the assets – money and bonds that not only are rated investment-grade, but really are – that financial markets are not adequately supplying. The flow of workers out of employment exceeds the flow back into employment. And, as employment and incomes drop, spending on currently produced commodities drops further, and the economy spirals down into depression.
Leaving aside that the problem is a bit narrower--money rather than "financial assets," this is also a good description of the problem. And then DeLong explains the solution:
Thus, the first principle of macroeconomic policy is that because only the government can create the investment-grade financial assets that are in short supply in a depression, it is the government’s task to do so. The government must ensure that the money supply matches the full-employment level of money demand, and that the supply of safe savings vehicles in which investors can park their wealth also meets demand.
I strongly agree that somehow allowing that "the money supply matches the full employment level of money demand" is essential. Given the monopoly power of the Fed over base money, I would agree this is a "government responsibility" under current conditions. However, I do not favor having the government guarantee that "the supply of safe savings vehicles in which investors can park their wealth also meets demand."
In North America, governments appear to have muddled through. They have not provided enough bank guarantees, forced enough mortgage renegotiations, increased spending enough, or financed enough employment to rebalance financial markets, return asset prices to normal configurations, and facilitate a rapid return to full employment. But unemployment has not climbed far above 10%, either.
I do not favor having the government provide any safe assets for investors to purchase. In an ideal world, the national debt would be zero, and so there would be no government bonds, short or long term, to provide "safe" investments. As for "bank guarantees," I oppose those as well. Admittedly, getting the national debt to zero isn't too likely in the foreseeable future. And I have never favored going "cold turkey" on deposit insurance. Still, expanding government guarantees during recessions is something to be avoided.
So, what do we make of the claim that recessions involve excess demands for financial assets matched by excess supplies of currently produced goods and services. Has there ever been a recession where people were calling their brokers asking to get into the stock market, and the reply was, "sorry, just can't find any stock for you to buy?" Of course not. Many stocks are traded on organized exchanges and any "shortage" results in higher prices, with prices rising minute by minute. Firms can't sell products and workers can find jobs all because of a booming stock market? Absurd.
But, of course, DeLong really isn't worried about all financial assets, just those that that are "investment grade" that only government can create during a recession. Because of its power to tax, the government's debt or guarantees are always investment grade. Leaving aside the long record of government default on debt, cataloged in Reinhart and Rogoff's This Time Is Different, is it really true that a government that had no debt and refused to guarantee private debt would be subject to recession? Would it be perpetual recession?
Nick Rowe's discussion of the disastrous impact of unobtainium seems to apply. If people suddenly think of a nonexistent and desirable good, does that result in surpluses of existing goods as people sell them in order to fund purchases of "unobtanium?" Rowe argues that this is implausible and that the problem with surpluses of currently produced output, along with the labor and other resources used to produce it, isn't shortages of any old kind of asset (financial or unreproducible real,) but rather with money--assets that are used as the medium of exchange.
So, what does DeLong's argument amount to? As Yeager pointed out in 1956, an excess demand for bonds can push down their yields to very low levels. If the yields on those assets become so low that holding money is a better alternative, then an excess demand for them "spills over" into an excess demand for money. Quasi-monetarists have never claimed that recessions are only caused by increases in the demand to hold money that appear as a bolt from the blue. Generally, an excess demand for money would be caused somehow, and a spillover from an excess demand for some other sort of asset is possible.
It is true, however, that if an excess demand for government-guaranteed assets is spilling over to create an excess demand for money, then an increase in the supply of such assets would solve the problem. Budget deficits and government bailouts would correct the excess demand for money. Of course, those policies might cause other problems as well. Further, the sensible rule that a central bank should never make open market purchases of assets with zero yields can involve a type of government guarantee. For example, quantitative easing with long term to maturity government bonds implies, at the very least, the central bank is risking its own capital to protect those holding reserve balances or currency from interest-rate risk.
To emphasize again the monetary nature of the problem, consider the following alternative monetary order. Hand-to-hand currency is solely issued by unregulated, uninsured "free banks." Think of them as shadow banks. Conventional banks can offer insured transactions deposits and other savings vehicles. Hand-to-hand currency, checks, and electronic payments are cleared through the central bank. The reserve balances at the central bank pay interest. The Treasury funds some of its debt by Treasury bills. There are various short and long term government bonds.
As DeLong suggests, there is a decrease in the demand for currently-produced output and an increase in the demand for government guaranteed financial assets. Hand-to-hand currency is not one of those assets, and worries about the financial conditions results in a reduction in the demand for hand-to-hand currency.
Some of the financial assets in demand are traded continuously on dealer markets, and so, their prices rise and yields fall to clear their markets. Suppose that at the "zero-bound," there remains a shortage. What happens? The most likely scenario would be that their prices rise and their yields fall below zero.
But what about the zero nominal bound? Who would lend at a negative interest rate? Won't they just hold money? No, because "money" is not a government guaranteed financial asset. Or, at least, the zero-nominal-interest, hand-to-hand currency, isn't a government guaranteed, low risk, financial asset.
Of course, in the scenario described above, banks have a variety of interest bearing, government-insured deposit accounts. Households and firms may well choose to hold those rather than government bonds. The transactions deposits issued by conventional banks even serve as money.
However, like everyone else, these banks demand more of the government guaranteed assets, and are facing a progressively lower yield on their asset portfolios. This reduces their demand to raise funds by issuing deposits of all sorts, and so the interest rates they are willing to pay on those deposits. As the yields on their low risk earning assets fall, they lower the yields on the low risk liabilities they issue. If those yields are reduced and go to zero, there is nothing to prevent the yields from continuing to fall below zero.
But won't banks that try to "pay" negative interest on insured deposits suffer a currency drain? Not if the hand-to-hand currency is issued by risky, uninsured "free banks."
Finally, banks keep interest bearing accounts at the central bank. Won't banks faced with progressively lower yields on government bonds and other "investment grade" assets just hold reserves? Won't excess reserves just build?
That depends on what happens to the interest rate the central bank pays on reserve balances. If the central bank lowers those rates along with other interest rates, particularly with the interest rates it earns on the short term to maturity government bonds it holds, then the interest rate banks can earn on reserve balances will fall with other interest rates. Again, to less than zero if necessary.
But won't the central bank face increased currency demands? Won't banks simply hold vault cash? Won't the banking system face a currency drain? Again, no. In this alternative monetary institution, currency is solely issued by free banks. There is no reason for anyone to accumulate currency as a store of wealth.
Suppose that there was a surplus of currently produced output and a shortage of government guaranteed financial assets. The result would be negative yields on those assets, so that people would have to pay for the privilege of having their savings guaranteed by the taxpayer. Those who wanted to earn a yield on their financial holdings would have to bear risk. They could buy stocks, BAA corporate bonds, or maybe even long term to maturity government bonds. With the market for "safe" assets clearing, along with the markets for other financial assets, there is not a surplus of currently produced output. While there might be a shift in the composition of demand, from capital goods to consumer goods, or even a reduction in the supply of labor, nothing like a general glut of goods could appear.
This alternative monetary institution has much to recommend it as a reform. (Think of it as a step along the path to a currencyless payments system.) However, exploring such systems also help us understand the real world by way of contrast. What is the aspect of the existing monetary arrangements that makes surpluses of currently produced output possible? The central bank stands ready to issue zero-interest bearing currency on demand. This limits the possible decrease in the interest rate paid on balances at the central bank. And that is what creates the lower bound on the interest rate and upper bound on the price of "investment grade" financial assets.