In a new paper, ‘A Public Option for Bank Accounts (or Central Banking for All),’ we propose that the Federal Reserve (the ‘Fed’) offer bank account services to the general public and not just to banks and other select financial institutions. Under our proposal, U.S. residents would be permitted to open bank accounts with the Fed, which we call ‘FedAccounts.’ These accounts would provide all the same functionality as ordinary retail bank accounts (except for overdraft coverage) as well as all the special features that banks currently enjoy on their central bank accounts. For example, FedAccounts would pay a high rate of interest (known as the interest-on-reserves or IOR rate), transfers between FedAccounts would clear instantly, and balances would be nondefaultable no matter how large. We believe that FedAccounts would, among other things, significantly reduce the number of unbanked and underbanked U.S. households and significantly increase the efficacy of monetary policy by improving ‘pass through,’ the transmission mechanism through which the central bank’s target interest rates affect rates in the broader financial markets. We also expect that FedAccounts would generate revenue for the federal government, mitigate regulatory complexity, improve financial stability, and put downward pressure on levies like interchange fees that function as tolls on commerce.
In our paper we address numerous potential obstacles and objections to FedAccounts. In this post we look more carefully at four concerns we have heard as we have discussed and presented the project over the past few months. These are: (1) whether FedAccounts would distort the allocation of credit in the economy; (2) whether Congress might ‘raid’ FedAccounts to fund government spending; (3) whether, perversely, in an adverse market environment the existence of FedAccounts would worsen financial panics by encouraging runs on private money products (eg, repos, commercial paper, and money market funds); and (4) whether FedAccount balances would be more vulnerable to levy or garnishment than traditional bank account balances.
I. Credit Allocation
Some commenters have raised concerns about the effect of FedAccounts on credit allocation in the economy. The first thing to note is that, under our proposal, the Fed would not be authorized to extend credit to individuals or nonbank businesses (outside its existing emergency authority). Accordingly, the only way for FedAccounts to affect credit allocation is indirectly due to the movement of balances by individuals and businesses from commercial banks to the public ledger. We do not think, however, that such migration would be distortive. First, under our proposal, the Fed would allow banks to maintain their existing balance sheet size by borrowing from the discount window to replace lost deposit funding. Banks currently hold about 35% of U.S. loans and 9% of U.S. securities; these figures need not change at all. Second, even if as a result of FedAccounts banks’ average cost of funds were to rise to match that of nonbank lenders (which we think would be a good thing), banks’ lending rates depend on not their average but rather their marginal cost of funds (ie, the federal funds rate). FedAccounts would not change the federal funds rate and therefore would leave banks’ lending rates unchanged.
Moreover, even if the Fed chose to gradually shift its portfolio away from discount window loans to Treasury securities or agency debt, we do not expect the effects on credit allocation to be deleterious. Credit portfolios need not be and often are not funded with money claims. If there are profitable lending opportunities, we expect the market to ferret them out. To the extent that Congress seeks to subsidize certain types of lending (such as small business loans, residential mortgages, or student loans), such subsidies can be extended in more targeted and efficient ways through specific legislation. Adam Levitin, whose important work in this area has influenced our own thinking, makes a similar point in his paper ‘Safe Banking.’[i]
Others worry that FedAccount balances could be used improperly by Congress to serve various short-term political ends, in much the same way that Congress has raided the Fed’s capital surplus account twice in recent years. This concern is misplaced. FedAccounts would not increase the Fed’s capital surplus account or lead to the accumulation of ‘raidable’ funds. FedAccount balances would be liabilities on the Fed’s balance sheet, and the corresponding assets would consist of discount window loans and government securities. FedAccount balances would not constitute surplus or equity or end up in a vault. In a world with FedAccounts, Congress would have no greater ability to fund its spending with the Fed’s funds than it already has; after all, the Fed’s balance sheet is presently $4 trillion, more than quadruple the size of its pre-crisis peak, and the Fed could always create additional money if it decided to do so, allowing it in theory to monetize all government spending. Accordingly, central bank independence, not balance sheet size, is what prevents Congress from monetizing appropriations.
Another concern is that FedAccounts might exacerbate panics by incentivizing households and businesses to move their funds to the Fed. As Benoît Coeuré puts it (in the context of debates over central bank digital currencies), although runs ‘could also happen now between deposits and cash, a digital currency would make it cheaper and faster, making ‘digital bank runs’ more frequent and more severe.’[ii] But money claimants can already ‘digitally run’ to safer alternatives in a crisis. In 2008, for example, billions of dollars poured into Treasury money market funds and healthy banks like JP Morgan Chase that were widely perceived to be too big to fail. While these flows might be directed to FedAccounts in a future crisis, this presents no special cause for concern. Indeed, the flow of funds to the Fed could give the Fed greater visibility and improve its ability to fashion better liquidity support. Besides, as we discuss in the paper, we expect the FedAccount program to reduce the attractiveness of shadow money products overall, thereby dramatically reducing the likelihood of panics in the first place.
The concern over panics is really an indictment of shadow money itself and would be better addressed by eliminating unauthorized money creation. Indeed, it would be odd to conclude that we should refrain from offering digital money instruments because their existence would exacerbate runs on shadow money products. It would be logically equivalent to contending that we should eliminate FDIC deposit insurance because the program might raise the likelihood of shadow money panics.
IV. Levy or Garnishment
One question we have been asked is whether FedAccounts, because they would be operated by a federal government agency, would be more susceptible to levy or garnishment than existing accounts. They would not. In this regard, FedAccounts would be no different than traditional bank accounts. Private creditors would need to obtain a money judgment in court before levying or garnishing a FedAccount. Failing to observe these legal process requirements would leave both the private creditor and the Fed subject to legal liability. Further, federal government creditors such as the Treasury Department (I.R.S.) or the Department of Education would have no better access to FedAccount balances than to ordinary bank account balances, because Federal Reserve Banks are separately chartered legal entities operated under the aegis of an independent agency. To further ensure the privacy of FedAccount information, we also propose that Congress codify stringent statutory protections for FedAccounts mirroring those that currently exist for tax return information.
We recognize that, as with any major initiative, the Fed will have to carefully consider the effect of the FedAccount program on various components of the financial system. But, as far as major initiatives go, FedAccount is a remarkably simple and straightforward extension of existing policy that promises to achieve numerous, heretofore hard-to-reach policymaking goals. In 1913, Congress created the Federal Reserve System. The following year, the Fed printed the first Federal Reserve Notes—those pieces of paper that, today, most Americans carry around in their wallets. Over a century later, following the rise of the internet and fundamental changes in how people pay for goods and services, we think it is time for the Fed to offer these same notes in electronic form. That is what FedAccount would do.
[i] Adam Levitin, ‘Safe Banking: Finance and Democracy’ (2016) 83 U CHI L REV 357. (‘One concern about Pure Reserve Banking would be that it would result in a contraction in credit. It is not clear that this is the case. Much depends on how much consumers and businesses really want to assume credit risk. To the extent there is a contraction of credit, however, it is a right-sizing, because the level of credit would reflect risk internalized pricing, rather than subsidization.’)
[ii] Benoît Coeuré, ‘The Future of Central Bank Money’ (Remarks to the International Center for Monetary and Banking Studies, Geneva, Switzerland, 14 May 2018, 6) <https://www.ecb.europa.eu/press/key/date/2018/html/ecb.sp180514_4.en.html>.
Lev Menand is a former Treasury senior adviser now with the U.S. District Court for the Southern District of New York, John Crawford is a Professor of Law at UC Hastings, and Morgan Ricks is a Professor of Law at Vanderbilt University.
This paper was first presented at the 5th Annual Conference of the Journal of Financial Regulation.