Policy Responses to the Depression--February 2-May 1, 2009--Part I

By February 2, when I finished my book, the government had already adopted, or signaled the imminent adoption of, a number of policy responses to the depression, and I discussed these in the book. In the two months since, there have been a variety of new developments on the policy front, many quite dramatic and some important. But none has changed the fundamental character of the anti-depression program that could be discerned in the earliest days of the Obama presidency and was discussed in the book.

The anti-depression program as it has emerged--which incidentally exhibits considerable continuity with the Bush Administration's program--is divisble as follows: (1) reducing interest rates and expanding the money supply (these are not quite the same thing), which can in turn be divided into (a) conventional Federal Reserve money-supply expansion and (b) "quantitative easing"; (2) bailing out (that is, recapitalizing with Treasury money) the banks; (3) mortgage relief; (4) keeping the Detroit automakers from liquidating; and (5) a "stimulus" package (that is, Keynesian deficit spending), which can in turn be divided into (a) tax relief, (b) expanded benefits (unemployment benefits, health benefits, food stamps, etc.), and (c) public works, such as construction of transportation infrastructure.

All these are discussed in my book, so as in my previous blog entry I will just be updating. But there is plenty of updating to do, and so I have decided to split my discussion into three separate entries. This first entry examines the first three measures above, all related to banking. The second entry will focus on the other three measures, and the third will be an assessment of the six measures.

(1)(a) One way to stimulate demand in a depressed economy is to reduce interest rates, as that encourages borrowing, and most borrowing is for purposes of spending, whether for consumption or for investment; either form of spending stimulates output and therefore employment. Traditionally, a central bank, such as the Federal Reserve, that wants to reduce interest rates purchases short-term government bonds. (The procedure the Fed currently uses, involving repurchase agreements, is slightly different, but the older procedure is more intuitive so I will pretend it is still the one used.) The cash that the sellers receive in these sales is deposited in bank accounts and thus increases the amount of the money that banks can lend. Moreover, the central bank's purchases, by increasing the demand for bonds, drives up their prices, which in turn reduces the interest rate that buyers receive, because the dollar amount of interest is fixed when the bond is sold and so becomes a diminishing percentage as the price of the bond rises. Thus the central bank's purchases exert an additional downward pressure on interest rates.

The Fed uses as the target for its traditional money-creating activity the rate at which banks lend money to each other for repayment the following day (the overnight or federal-funds rate). The lower that rate, the more fully the banks' reserves (i.e., their cash) are employed in making loans. Although the overnight rate is obviously a short-term rate, it influences long-term rates, such as interest rates on mortgages. It does this both because of substitutability of short-run for long-run loans and because the lower the costs of borrowing to banks, the lower the rates which they are forced by competition to lend their borrowed capital.

Beginning last fall, the Federal Reserve pushed the overnight rate down virtually to zero, and it has kept it there. But as I say in my book, you can lead a bank to money but you can't make it lend. The banks prefer to hoard any additional cash they receive (or invest it in very safe securities or use it to buy other banks) than to lend it, because lending is risky in a depression. The demand for loans is reduced, and many banks are undercapitalized as a result of the collapse of the home-mortgage market, in which the banks were and are deeply invested. (Oddly the Federal Reserve is paying interest on bank reserves. On the one hand, this increases the banks' income, but on the other hand it reduces their incentive to lend their "excess" reserves, that is, the cash they are not required by the regulatory authorities to retain but are free to lend or otherwise invest.) The more hoarding there is, the less money is available for investment, so interest rates rise.

This is not to say that the injection of added cash into bank accounts has been pointless. It has doubtless reduced the decline in lending that set in last September when the banking industry teetered on the brink of outright insolvency. But bank lending remains far below normal levels.

(1)(b) The Federal Reserve has supplemented its downward pressure on the overnight bank rate with what is called "quantitative easing." This means that it is buying other forms of debt besides the short-term Treasury bills that it buys when it wants to reduce the overnight rate. The other debt it is buying includes credit-card debt and mortgage-backed securities. With the overnight rate effectively zero, manipulating that rate cannot have any further effect on the cash that banks have available for lending. Moreover, the only direct effect of the overnight rate is on bank balances. Banks are not the only lenders, and buying the debt of other lenders is a way of getting cash to firms that may be less wary about lending than most banks at present are. The quantitative-easing program was under way on February 2, but is being expanded.

Through both forms of "easy money" policy--reducing the overnight rate and buying debt other than short-term government bonds--the Federal Reserve has already increased the money supply by a trillion dollars and is planning a further increases of more than another trillion dollars. The major increase in the money supply that the Fed is engineering is fraught with dangers that I will discuss in a subsequent entry.

(2) The Treasury is lending the banks some $700 billion in an effort, supplementing the Fed's efforts, to increase bank solvency and by doing so encourage the banks to do more lending. With most of the money now spent and Congress unwilling to appropriate more for this purpose because of the intense unpopularity of "Wall Street," the Treasury has created a complicated plan (the "Geithner Plan") for increasing banks' capital without a further direct and immediate infusion of federal money. (The Treasury requires appropriations from Congress in order to be able to feed money to banks; the Fed does not. The reason for the difference in authority and the policy and economic implications of the difference are issues that I will discuss in a future entry.)

The Geithner Plan involves indirect but sizable subsidies to hedge funds and other private investors, including banks, to induce them to buy the overvalued assets (mortgage-backed securities and other forms of securitized debt) owned by banks. It is believed that these assets are carried on the banks' books at inflated values in order to fend off demands that the banks increase their capital. No one is fooled, but the concern is that as long as these assets are on the banks' books the difficulty of valuing them will deter private investment in banks and thus leave them in a weakened state.

I am dubious. Most corporate balance sheets contain assets that are difficult to value; that doesn't prevent the corporations from raising money. In any event, it is unclear what exactly is gained by the government's paying hedge funds and other investors (including strong banks) to induce them to buy the overvalued assets, rather than paying the banks directly and taking preferred stock in return, as in the original bank bailouts.

True, the investors will be putting up much of their own money, and not just operating as a conduit for a government subsidy, but that will leave the investors with less of their own money to invest. So what is gained? The aim seems to be just to avoid having to go to Congress for a further appropriation for "Wall Street." (The subsidy will be disguised by taking the form of generous guaranties by the Federal Deposit Insurance Corporation.) The problem is that helping banks in this way substitutes a complex three-party transaction (government, private investor, bank) for a simpler two-way transaction (government, bank), and that the private investors may be loath to go into partnership with government, lest the government under pressure from an angry public and Congress try to "claw back" any "exorbitant" profits that the investors make.

Impatience with the plan leads some economists to advocate the government's "nationalizing" the weak banks, but that would be a mistake. This is not only because of the manifest inability of the government to manage banks competently, but also because the vexing problem of valuing the overvalued assets cannot be avoided in this way. The banks are not broke; if the government takes them over, it will have to compensate the owners for the net value of the assets that the government takes, including any overvalued assets that, despite being overvalued, have some value. Perhaps what the government could do would be to take (with compensation) all the good assets of the bank, leaving the overvalued ones with the shareholders; then the bank's balance sheet would be "clean." But then what would it do with the bank? Run it? Sell it? The practical complications would be immense.

(3) My book discusses several forms of mortgage relief. One, the proposal to allow bankruptcy judges to "cram down" first mortgages on primary residences--that is, to reduce the mortgage to the current market value of the residence, with the difference between the debt and the crammed-down mortgage thus becoming an unsecured debt of little value--is dead. It never promised much relief because it would have required the borrower (the mortgagor) to declare bankruptcy, whereas in the usual case a mortgagor whose house is worth less than his mortgage can, if he considers the house to have become a worthless investment, simply abandon it. Unless he is rich, he won't be sued for the unpaid portion of the debt. Moreover, while cramdown would have benefited some homeowners, it would have hurt lenders and thus have undermined the bank bailouts.

Other forms of mortgage relief, however, such as subsidizing first-time home buyers in order to stimulate the purchase of houses whose current owners cannot afford them and thus to avoid defaults, pushing down mortgage interest rates, and facilitating modification of mortgages (to reduce foreclosures) by immunizing mortgage servicers from legal liability should the modification hurt a junior lender, are going into effect, or will go into effect.

The goals of mortgage relief are twofold. First, anything that reduces homeowners' liabilities or increases the net value of their homes encourages them to spend money. They are less indebted and their savings, which include the net value of their home, are worth more. So mortgage relief is a form of stimulus. Second, anything that invcreases the value of mortgages increases the capital of banks, which remain heavily invested in mortgages and mortgage-backed securities.

The congressional appropropriation sought for the program of mortgage relief--$75 billion--is modest in relation to the total amount of mortgage debt ($12 trillion). But the part that involves simply granting legal immunity to services does not require an appropriation to make it effective.