Originally posted on August 15, 2009 at the New Economic Perspectives blog.
What is a federal government surplus?
When the federal government’s revenue exceeds its spending over the course of a year, it is running a budget surplus and outstanding Treasury securities will be reduced by the same amount over the year. In 1999, the federal government’s surplus was $99 billion and it is projected to grow to $142 billion for fiscal year 2000. This means that US taxpayers will pay $142 billion more in taxes this year than the government spends. More concretely, taxpayers will write checks to the Internal Revenue Service in the amount of $1.914 trillion, while the US Treasury will write checks received by Americans in the amount of only $1.772 trillion—a difference of $142 billion. The only way that taxpayers can write checks to the IRS that exceed the amount of checks received from the Treasury is to surrender $142 billion of Treasury securities to the government. In other words, running a surplus necessarily means that the Treasury is reducing nominal wealth of the non-government sector. This is why federal budget surpluses reduce outstanding Treasury debt and non-government sector net nominal worth.
What is the long-term effect of running perpetual government surpluses?
On current projections, the federal government will run surpluses over this decade that will total more than $2.9 trillion, leading to an equivalent reduction of non-government sector net nominal wealth—of $2.9 trillion. This wipes out almost 80% all of the publicly-held US Treasury debt, including that now held by foreigners. No one can accurately predict how the economy will react to such an unprecedented reduction of its nominal wealth—especially when the most liquid assets will be removed from private portfolios. However, throughout our history, the US has experienced exactly six periods of substantial reduction of federal government debt, achieved through persistent budget surpluses, and each of those periods ended in one of our nation’s six depressions. Our last period of substantial surpluses occurred between 1920 and 1930, when Treasury debt was reduced by 36%; the Great Depression began in 1929. For a more recent example, Japan began to run government surpluses in 1987, which reduced non-governmental nominal wealth and generated a deep recession that has already lasted a decade. Note, however, that neither the US in the 1920s nor Japan in the late 1980s came close to draining $2.9 trillion worth of wealth from the economy, even after adjusting for higher prices today.
Doesn’t a budget surplus allow us to save for the future?
Those who believe that a surplus can be “saved” for the future, or “used” to finance tax cuts or spending increases simply do not understand the nature of a surplus. Does anyone really believe that we can “save for the future” by burning $3 trillion worth of private sector wealth? During any period, the government can always choose to spend more (or less), in which case the surplus over the period may be lower (or higher); similarly, it can increase (decrease) taxes and thereby may increase (decrease) the surplus. But, as Gertrude Stein said, “there is no there there”-a surplus exists only as a deduction from private sector income. The negative household saving that some commentators are finally noticing is merely the accountant’s flip-side to the budget surplus. A government surplus necessarily reduces private sector savings and cannot be “saved for the future”.
How do budget surpluses impact non-government sector financial balances?
There is another, less transparent, impact of government surpluses on the non-government sector. At the level of the economy as a whole, when one sector spends more than its income, another necessarily spends less for the simple reason that in the aggregate, total spending equals total income. Let us, then, disaggregate the economy into three sectors to determine the implications of government surpluses for the other sectors. First, we can consolidate all levels of government into a public (or, government) sector, and likewise consolidate households and firms into a domestic, non-government (or, private) sector. For completion, we must add a foreign (“rest-of-the-world”) sector. At the aggregate level, the spending of all three sectors combined must equal the income received by the three sectors. It is clear that if the public sector is spending less than its income—that is, it is running a surplus—this must imply that at least one other sector is spending more than its income (in other words, is running a deficit). Mathematically, the sum of the balances of the three sectors must equal zero. It is convenient for our purposes to write this as:
{Public Sector Surplus} + {Foreign Sector Surplus} = {Private Sector Deficit},
which merely moves the private sector balance to the right-hand-side and reverses the sign (in other words, writes the balance as a deficit rather than a surplus, since a negative surplus is the same thing as a deficit).
Because the US has been running a balance of payments deficit in recent years, this means that the foreign sector is in surplus (the rest-of-the-world receives more US dollars than it spends). A few years ago, our public sector ran a sufficiently large deficit to more than offset the foreign sector surplus, so that our domestic non-government sector was able to run surpluses. However, in the past two years, the US public sector’s balance has turned toward surplus. When combined with our balance of payments deficit (or foreign sector surplus), this means that the domestic private sector’s balance (that is, its savings) has turned sharply negative—toward large and growing deficits. The non-government sector deficit is now approximately equal to 5.5 percent of GDP—far and away the largest private sector deficit the US has seen in the post-war period.
Will the federal government really run surpluses for the next decade?
It is very difficult to take seriously any analyses that begin with the projection that our government will run surpluses for the next decade. Part of our skepticism comes from the inherent difficulty in making projections. More importantly, it is difficult to believe that our economy can continue to grow robustly as the government sucks disposable income and wealth from the private sector by running surpluses. When the economy slows, the surplus will eventually disappear—automatically—as unemployment compensation rises and tax revenue falls due to the slowdown. As the government spends more and taxes less, the surplus will vanish.
6 RESPONSES TO “A PRIMER ON GOVERNMENT SURPLUSES”
- deanx | August 17, 2009 at 2:38 am |A personal Thank You for this bit of enlightenmet. I have always felt that government surpluses are mythical and can be only thought of in a Current Account perspective. This dovetails well with your other thoughts on Social Security … it is not possible to ‘escrow’ public funds intergenerationally. Every Fiscal Year must stand on its own. That is also why privatizing the Trust Fund is a fig leaf for ‘private taking’ of the revenue stream.
- JKH | August 17, 2009 at 6:04 am |While the policy case against cumulative government surpluses is clear, there are operational aspects of the argument presented here with which I strongly disagree. This pertains to the interpretation of government surpluses as saving or not. Given a theory that otherwise puts substantial emphasis on disproving the myth of government financial constraints, it is ironic that there is such a rigidly held position denying the potential for government saving.First, saving is a flow rather than a stock concept. An economic unit’s management of its balance sheet subsequent to a flow has nothing directly to do with characterizing the nature of the preceding flow itself.A government surplus (i.e. excess of tax over expenditure) typically shows up in the first instance in the form of (cumulative period) government deposit balances that are first held with the commercial banking system. This reflects transfers of funds from non-government accounts to government accounts as a result of tax payments.If the government subsequently transfers the cumulative amount of the surplus from its commercial bank accounts to its central bank account, that decision reflects a “voluntary constraint”. There is no operational reason why the government can’t let its surplus accumulate in the commercial banking system. That would then become a financial asset representation of the surplus. (In fact, the same holds even if the government transfers balances to the central bank and holds them there.)This financial asset is available to the government as the means by which future government expenditures can be paid for in settlement through the banking system. Such settlement is an alternative to issuing net new bank reserves, currency, or government bonds. (Settlement will occur by transferring such deposit balances to the government’s account at the central bank, which covers any overdraft temporarily created by the expenditure and has the effect of withdrawing any net new reserves that would otherwise have been created as a result of the expenditure.) The difference is that settlement is achieved by reducing a government asset rather than increasing a liability.The fact that the government has the fiat currency means to spend without relying on such asset sources has nothing whatsoever to do with confirming whether or not such a surplus should be categorized as saving. The fact that the government doesn’t need to accumulate such surpluses and financial assets prior to spending doesn’t mean such amounts if accumulated aren’t available as an alternative means for settlement of future expenditures….. Continued following…
- JKH | August 17, 2009 at 6:05 am |Continued from previous….There is a very simple way to demonstrate that such a surplus amount (a financial asset in the form of deposit balances with the commercial banking system) must represent saving. The sector financial balances model explained on this blog is the result of rearrangement of standard accounting identities, wherein investment is netted out from the mix. It is easy to construct a hypothetical example referencing this model that proves that government deposits as described above must represent saving.Suppose the level of investment in the economy is equal to I. Suppose also that the government surplus is equal to I. And suppose businesses borrow from the commercial banking system in order to invest in the amount I. (Assume a balanced current account to simplify the example.)Clearly, banking system loans and deposits will have expanded by the amount I. Loans have created deposits. Investment in the amount of I has created income in the same amount. That amount of income must also be equivalent to total saving S in the economy.Suppose the government engages in absolutely no expenditure or tax transactions except for the following: suppose the government creates a surplus financial position by taxing the full amount of income created from investment I. (There is no theoretical or operational constraint preventing such a tax.) Then those income earners will transfer the bank deposits they accumulate as a result of that income to the government as payment for the tax. The aggregate size of these deposits will be equivalent to the amount I.The government will then have deposits in the commercial banking system equal to I = S. The original non-government earners of income resulting from expenditure on investment will have saved nothing, because their entire income has been taxed, leaving disposable income of zero. This leaves the government as the only possible source of aggregate saving S = I, because the non government sector has saved nothing. (Again, we assumed the current account is in balance.)This proves the government has saved….. Continued following…
- JKH | August 17, 2009 at 6:06 am |Continued from previous….The government’s net financial asset accumulation in this example equals the gross saving S in the economy. Generalizing beyond this example, any government surplus represents, along with net financial saving, a direct share of the gross saving in the economy. (I’m assuming away the typical classification problems involved in government expenditure, which is an entirely separate methodological measurement issue, and beside the point being made here.) This can also be seen by noticing that any government surplus is equal to the difference between (non-government) investment and non-government gross saving. This relationship is derived back from the parent accounting identities in a similar way as are the various SFB model equations.The example assumes that the government maintains its deposit balances with the commercial banking system. But that preliminary assumption is not necessary in the final analysis. It is only important to demonstrate that this financial asset is a legitimate form of saving. From that point on, whatever the government does in terms of managing its balance sheet is completely irrelevant to the issuing of identifying the incidence of saving that has already occurred.For example, as typically represented in this blog, the government can simply buy back or mature its debt with its surplus. This is equivalent to an exchange of government balances at the commercial banks for non government held government bonds. At this stage, the central bank happens to become operationally involved in further intermediary operations for this type of transaction. But this is completely irrelevant to the issue of identifying whether or not saving has occurred. The fact that the government may transfer its deposit balances to the central bank (with obvious bank reserve effect) and then transact bond purchases with the non government sector from that point (with obvious reversing reserve effect) is irrelevant to the saving issue. Such balance sheet transactions, with or without central bank intermediation, have no effect on the fact that saving has already occurred at the flow level. Once we acknowledge correctly that these balance sheet transactions are irrelevant to the identification of saving, it becomes very obvious again that we should be logically indifferent to the two alternative operational modes of the government’s deployment of its initial surplus deposit balances – asset accumulation or liability retirement. In short, the idea that the government appears to make tax money “disappear” by retiring bond liabilities has absolutely nothing to do with the issue of whether or not saving has occurred.A corollary observation to this discussion is that it is not necessary for the non government sector to sell bonds to the government in order to pay taxes. This is the government’s choice, rather than a non government balance sheet requirement. In the example above, the non government deposits required to pay taxes already exist as a result of income generated from expenditure on investment. Taxes are paid from these deposits. The government then decides if it wants to bid for bonds or retire maturing bonds in order to return the broad money and reserve balances that it has taxed away back to the non government sector. This is a separate balance sheet transaction that follows the actual event of government saving. It falls into the category of “voluntary constraint” in the way in which the government manages its balance sheet….. Continued following…
- JKH | August 17, 2009 at 6:09 am |Continued from previous….The only slight complication to this story is the case in which non-government entities use currency to pay taxes. (This case in which banks pay taxes is simple; they simply credit government deposit accounts and debit their own equity positions. Central bank reserves are not immediately affected by such a transaction.) Using currency to pay taxes would be unusual. In this case, the medium of exchange is both the means for paying taxes and a government liability form. It is reportedly the case that the government will destroy the currency so submitted. Essentially, the payment of taxes using currency results in the extinguishment of a liability for both the taxpayer and the government. In this case, there seems on first look to be no initial asset resulting from the payment of taxes. But that appearance again is misleading. The standard process may well be the destruction of such redeemed currency, which reduces the net liability position of the consolidated government entity. But this practice is another one of those “voluntary constraints”. At an operational level, there is nothing preventing the government from re-depositing currency in the banking system, if that is the form in which taxes have been collected. But before going further on this currency point, we should ask why on earth would the taxpaying public, who choose to hold a certain level of currency for reasons of normal convenience, choose to inconvenience themselves in an abnormal way by paying a material amount of tax due in currency? It makes little sense for a taxpayer to accumulate disproportionate amounts of currency in order to settle a transaction that is regularly settled through bank deposits. That doesn’t mean it can’t be done, but is it a material issue in the analysis here? Notwithstanding that practical point of relevance, it would nevertheless be operationally feasible for the government to present currency at its commercial banking facilities in exchange for a deposit liability, thereby accumulating a financial asset analogous to the example above, even with the original payment of taxes having been made using its own liability as the medium of exchange. The banks could may well subsequently re-circulate this redeemed currency in order to replenish the public demand that would likely result from the bizarre choice of having used a disproportionate part of its existing stock of currency just for the purpose of settling tax liabilities. In any event, the government has created a financial asset from its tax surplus, just as in the first example above. So in the final analysis, the case of taxes paid with currency also shows that the issue of whether or not the government has saved in accumulating it’s surplus has nothing whatsoever to do with the choices it makes, voluntary constraints or otherwise, in managing its balance sheet….. Continued following…
- JKH | August 17, 2009 at 6:09 am |Continued from previous….Beyond the analysis above, I don’t see why discrediting the idea of government saving should be an essential element of the modern monetary theory. Nothing described above contradicts key points of the theory, such as the fact that banks are not reserved constrained or that the money multiplier theory is a terribly flawed concept. Nor does it contradict any desired policy orientation toward cumulative deficits rather than surpluses. But the sector financial balances model is derived from accounting identities that depend on consistent mathematical summation. These original equations should be re-constructible from the derivative net financial asset formulation, but this can’t be the case if the idea of government saving is simply denied. The error being made, in my view, is an attempt to force a particular government balance sheet construction as the basis for the interpretation of a corresponding saving question, when balance sheet management subsequent to an identified or not identified saving flow has absolutely nothing to do with whether or not saving has occurred in the case of any economic unit. As noted above, the government balance sheet is not operationally constrained in such a way as to prohibit the accumulation of financial assets as the result of surpluses. Moreover, the capacity for operationally unconstrained government expenditure in itself doesn’t disprove the capacity for government saving. Similarly, the capacity for net reserve, currency, or bond issuance doesn’t disprove the capacity for saving. So the fact that government saving is operationally unnecessary doesn’t mean that it cannot occur. And the fact that government saving isn’t necessary doesn’t mean that when it does occur it isn’t sufficient on its own as an offset to an equal amount of government expenditure, and as an alternative to what otherwise would be net reserve, currency, or bond issuance.End
