By Dan Kervick
Warren Mosler has made an interesting proposal concerning how we should think about Treasury securities:
… with today’s floating fx/non convertible currency tsy secs (held outside of govt) are logically additions to ‘base money’, as the notion of a reduction of govt reserves (again, gold, fx, etc) is inapplicable to non convertible currency.
That is, with today’s floating fx, I define base money as currency in circulation + $ balances in Fed accounts. And $ balances in Fed accounts include both member bank ‘reserve accounts’ and ‘securities accounts’ (tsy secs). And to me, it’s also not wrong to include any other govt guaranteed debt as well, including agency paper, etc.
That is, with floating fx, ‘base money’ can logically be defined as the total net financial assets of the non govt sectors.
And deficit reduction is the reduction in the addition of base money to the economy, with the predictable slowing effects as observed.
The point of this post is to ‘reframe’ govt deficit spending away from ‘going into debt’ as it would be with fixed fx, to ‘adding to base money’ as is the case with floating fx where net govt spending increase the economy’s holdings of govt liabilities, aka ‘tax credits’.
So Warren is proposing that we extend the definition of the monetary base to include not only holdings of currency and reserve deposit balances, but also include Treasury securities and government-sponsored agency debt. Let’s use the term “MM” to stand for the types of financial assets just described – short for “Mosler Money.”
There is certainly much to be said in favor of the proposal. For one thing, if we think about financial assets in terms of their associated risks, we can note that the new capital rule approved by government regulators and scheduled to go into effect in January, 2015 – described here in the Community Banking Guide jointly issued by the Fed and FDIC – assigns a zero risk weight to both cash and “direct and unconditional claims on the U.S. government, its agencies, and the Federal Reserve.” So the regulators regard dollars and treasuries as equally risk-free.
On the other hand, one might raise questions on the score of liquidity and maturity. Dollars are non-interest bearing, non-maturing liabilities of the government, while treasuries mature and earn interest. They don’t function in exactly the same way. Also, its not clear that all treasuries are quite as liquid as dollars. Treasury bonds and notes are usually thought to be less liquid than T-bills.
But leaving these issues aside, I have one minor quibble about Mosler’s proposal, and then a few follow-on questions to pose.
The quibble is not about Warren’s proposal to count MM as the monetary base, but the further claim that we can go further and identify MM with the net financial assets of the non-government sector, as Mosler suggests in the third paragraph above. Whether one agrees or disagrees that Treasury debt is sufficiently “money-like” to classify it as part of the monetary base, the conceptual motive behind the Mosler’s monetary base proposal seems clear enough. But the non-government sector holdings of US government liabilities are only the asset side of the non-government economy’s consolidated balance sheet. The net financial assets of the non-government sector should be equal to the non-government’s total holdings of government-issued financial assets minus the non-government sector’s total financial liabilities to the government. Federal Reserve notes, reserve balances and treasuries are only the asset side of that equation. On the other side would be such things as outstanding tax obligations to the US Treasury, all principle and interest owed to the Fed and all other accounts payable owed to the Treasury or other federal agencies: in other words the net financial assets of the non-government sector are equal to MM minus the net debt of the non-government sector to the federal government.
Notice that it is possible for the Fed to increase MM without increasing the private sector’s net financial assets. If the Fed buys a mortgage backed security from a bank, the bank’s financial asset holdings will be relatively unaffected, as it trades the security for an amount of cash of roughly equivalent value. But MM will be increased, since the mortgage backed security is a financial asset, but not part of MM.
Now for my follow-on questions:
If we adopt Mosler’s extended view of the monetary base as MM, then it becomes an even more perplexing question as to why we maintain the current division of responsibilities for issuing different classes of government liabilities. Why is the central bank the only part of the government currently permitted to issue non-maturing and non-interest-bearing liabilities (i.e. dollars), while the part of the government that is responsible for conducting fiscal policy is responsible for issuing most of the interest-bearing and maturing liabilities (i.e. treasuries)? If these are just two different forms of money, then why does one agency emit one form and the other emit the second? And why do we let that division lead to so much political conflict and wrangling?
To take a stylized example, suppose the government buys a fleet of 500 cars from auto manufacturer Universal Motors. Suppose the negotiated price of the fleet is $10 million, and that the automobiles cost Universal Motors $9.5 million to produce, store and deliver to the government. The Treasury issues T-bills with a face value of $10.1 million and sells them to the primary dealers for $10 million (where many will then be re-sold in secondary markets). So in other words, the bills yield 1% interest. The Treasury then purchases the cars from Universal Motors by cutting a check for $10 million. What is the balance sheet outcome for the two private sector parties involved here?
Dealers: + $100,000
The dealers now have $10 million fewer dollars, but T-bills worth $10.1 million.
Universal Motors +$500,000
They have sunk $9.5 million of costs into production and delivery of automobiles, but receive $10 million from the government.
But what in the world do these dealers have to do with this transaction, and what public policy purpose was served by the government making a $100,000 addition to the financial net worth of dealers in the financial sector? It seems like an arbitrary and irrelevant side-deal. Suppose instead the Treasury were permitted to emit dollars the same way that the Fed is allowed to emit dollars. Suppose the treasurer then says: “We need to buy some cars. So we’re going to create $10.1 million dollars, and give $10 million to Universal Motors in exchange for some cars and $100,000 to some banks for free” What sense could be made of that statement?
Now here is part of a plausible answer: the government gives free money to the financial sector to accommodate the growth in financial sector liabilities to the non-banking sector of households and firms. Just as you and I can earn interest on our deposits if we switch them from checking to savings accounts or CDs, so the banks earn interest on their money by shifting some of it from from reserve accounts at the Fed to government securities. And this continual injection of interest income from the government into the banks is needed to accommodate the continual injection of interest income from the banks into the private sector. The Fed, in coordination with the Treasury which actually issues the debt, is just playing its legally appointed role of providing a “flexible currency” and allowing the money supply to grow in response to growth in the broader economy.
Fair enough. But shouldn’t those operations by the central bank within the banking system be kept separate from the operations of the Treasury? The Fed could offer its own risk free savings vehicles and securities, generating the required interest flows for its banks, while the Treasury attends entirely to fiscal policy. Why do we have a system in which the nation’s central banking system is entangled with the government’s other spending operations? Why not permit both the Fed and the US Treasury to issue either non-interest bearing dollars or interest-bearing securities, as their policy needs dictate?
At this point, the defender of the current arrangement is likely to appeal to the requirements of monetary policy and monetary stability, and the desirability of keeping those responsible for monetary policy at some remove from the daily business of the legislature’s taxing and spending decisions. One might be convinced that unless the government has an independent monetary authority of this kind, the temptations to conduct a reckless monetary policy will be too great. Because its constituents love spending and hate taxes, Congress will frequently expand its deficits too rapidly, or in unpredictable ways, and destabilize the currency. So the system that has been set up is that the government voluntarily subjects its budget to monetary constraints that an independent monetary authority then adjusts and manages. In other words, the Treasury is required to operate on a budget, and can’t just issue dollars. Only the monetary authority is permitted to do that. The Treasury must issue securities, and swap them in the financial market for dollars that have already been issued. And the monetary authority – i.e. the Fed – determines the prices in that market by deciding how aggressively it, itself will participate in the market to re-purchase the securities.
That the government actually needs an independent monetary authority of this kind is a highly debatable point. But for now, let’s accept that point for the sake of argument. The question still remains as to why that monetary authority has been vested in the Fed, which is the central bank of the United States. Monetary policy is not identical to central bank policy. Rather, central bank policy and fiscal policy should be seen as two different aspects of government financial operations, both of which are inseparable from the government’s monetary policy. The Fed has a banking system to govern; the Treasury has authorized spending to carry out. But the fact that monetary policy has been assigned to the central bank is a historically contingent arrangement which is by no means a necessary one. An alternative arrangement would be to set up something like a Monetary Policy Board, on which sit representatives from the central bank, the Treasury and the Congress. The board’s role would be to establish a comprehensive monetary policy and coordinate fiscal policies and central bank policies with an eye to their combined impact on the nation’s monetary policy.
Under this arrangement, both the Treasury and the central bank would have the authority to issue non-interest-bearing dollars; and both would have the authority to issue interest-bearing securities. (We could even say that both have the authority to collect taxes, to the extent that we view the assessment of fees by the Fed on its member banks as taxes.) In some cases, the board might decide that the nation’s needs are best met by an increase in the amount of non-interest-bearing dollars emitted by the Treasury, with Fed emissions of dollars into the banking sector limited. In other cases the nations’s policy needs might recommend that the Treasury reduce its emissions of dollars, and increase its debt issuance, while the Fed should increase its emissions of dollars. The board’s joint recommendations could be submitted to Congress as part of its annual budget process, and enacted as legislation, with further recommendations offered as needed during the course of the year.
The government would establish a single consolidated balance sheet and Master Account, with the Fed’s operating accounts and securities accounts, as well as the Treasury’s operating account and securities accounts, existing as sub-accounts on that single balance sheet. Dollars emitted by either the Treasury or the Fed would then be entered as liabilities on that balance sheet and both would be part of currency in circulation. Similarly, securities issued by either the Treasury or the Fed would both count as part of the debt to the public. The Treasury general fund currently exists as an account at the Fed, but under the new arrangement, it would be a sub account on the Master Account. If, for example, the Treasury has been authorized to emit $20 billion in new, non-interest bearing dollars during the new year, then the Fed would have no role or discretion in that operation.
The economist Abba Lerner discussed some of these issues in the context of what he called functional finance. Governments can issue bonds, but they can also issue money. The Second Law of Functional Finance, according to Lerner, is that the government should issue bonds and exchange them for money only if it is desirable that the public should have less money and more bonds. But these decisions need to be made with respect to the policy needs of the whole government. There is no reason that the decision over the proportion of bonds to currency in the government’s issue of MM should be a de facto responsibility of the central bank, and determined by that bank in the process of handling its own special responsibilities for the banking or credit sector.
Some argue that the provision of safe assets to the private sector should be seen as part of the government’s responsibility, and that is one of the reasons for the existence of government bonds. Again, fine. But that seems to be a job for the bankers, not the treasurer. If people want a safe savings vehicle, let them deposit the funds with the central bank (or at member banks that belong to the central banking system) at various term lengths and rates. Why should the fiscal authorities, whose job it is to carry out spending operations, be worrying about whether they should also issue debt instruments for the purpose of providing savings vehicles?
An economy grows, develops and progresses as a result of its investments in the creation of new value. But in the private sector, there are two main forms in which that investment can occur: savings-driven investment and credit-driven investment. In savings-driven investment the firm, household or entrepreneur invests out of equity already accumulated, paying for goods, labor and other services out of savings and building new value out of the present value that already exists. If the investment fails to produce as much value as anticipated, the investors bears the full loss. With credit-driven investment, the household, firm or entrepreneur pays for goods, labor and other services by issuing promises for future payment, effectively financing the development out of a share of the value that will come into existence in the future. The existence of credit-driven finance permits more rapid growth, but also involves more risk on the part of more people, since it builds networks of obligations extending into an inherently uncertain future, and some losses will be born by the creditors.
But a government that issues the nation’s currency can, in principle, engage in a type of financing different from both of these: what we could call, “anticipatory money-financing”. As the economy grows, government monetary policy is used to regulate growth in the supply of money along with the economy, continually monetizing the new value in the economy while maintaining stable prices. Money is generally conceptualized for accounting purposes as a liability of the government, but its a peculiar type of liability, since it’s not an obligation by the government to make any further payment. Since the money can be used to pay taxes, issuing it might reduce future tax payments. But a government that can issue money clearly doesn’t need the tax payments. So the risk involved in money-issuance is only a monetary policy risk: that the issuance might result in unintended fluctuations in prices, loss of asset values, etc. In a country in which there are massive unmet needs and unemployed people and resources, however, and abundant opportunities to create new value, new growth and new markets driven by strategic and transformative government investment, the private sector should easily be able to absorb new money via direct spending without destabilizing prices.
Returning now to Mosler’s MM, we see that we can issue the MM in non-interest-bearing and interest-bearing forms. And the Treasury could continue to rely entirely on the current system of swapping the interest-bearing MM with the financial sector for non-interest-bearing MM. And if, as Mosler suggested recently, the Fed adopts a more formal policy for targeting Treasury yields, things will function reasonably well without requiring any difficult new legislation. On the other hand, we will still run into the endless political wrangling and melodrama over the growth in treasury debt. Is this really the best system we could have?
In an economy in which the most pressing needs are in the sector of households, small businesses and other non-banking companies, the present system seems like a wasteful contraction of fiscal space. The system imposes pointless extra monetary policy constraints on the government resulting from wholly unneeded side-injections into the financial sector, and muddies up the process of targeting new issues of money at the places in the economy where it is most needed. It gives the central bank too much discretion in determining the yield curve for MM, and thus would permit an aggressively anti-government central banker to gobble up fiscal space by allowing yields to rise. If our concern is Universal Motors and similar enterprises, along with all the people who work for these enterprises or could be working for them, there is no need to be injecting cash into Goldman Sachs as an operating requirement for carrying out deficit spending.
Cross-posted from Rugged Egalitarianism