Does seigniorage represent an unlimited source of government revenue? Discuss theoretically as well as with reference to empirical evidence that is available.
Government can always print money to meet its budgetary deficits and the revenue generated by mere printing of money is termed as seigniorage. Neumann (1996) explained it as gross flow of resources to the government sector, in real terms, associated with money creation. Seigniorage can also be termed as "raising of revenue by money creation and can be used by the government to finance expenditure when taxes cannot be raised from other sources" (Mohammad Ali Moradi 2001).
This essay will start with a theoretical model of seigniorage followed by an empirical study with respect to limitations of seigniorage as a revenue source. The main objective through out this essay would remain to explore whether seigniorage is an unlimited source of revenue for the government?
Theoretical Model of Seigniorage.
The0 essay focuses over seigniorage in real terms. The approach presented below, for theoretical understanding of seigniorage and its limitations as a source of government's real revenue, has been inspired from Olivier Blanchard (2003).
M = Nominal Money Stock.
Change in nominal money stock.
P = Price Level.
Then real revenue (in terms of goods) of the government, which it can generate by printing additional money,, will be called seigniorage.
(1) can be rewritten as follows.
is rate of nominal money growth and is real money balance.
(2) explains a relationship between seigniorage, real money balances and the rate of nominal money growth. Now we need to determine how much balance of real money are people willing to hold and how it is related to nominal money growth. We know that, under equilibrium in financial markets, LM relationship can be written as follows.
LHS in (3) is equal to real money supply (supply equals demand in equilibrium) and RHS is equal to real money demand (Y = real income and = a function of interest rate ). (3) implies two important characteristics which are as follows.
Higher Y (real income) will induce people to hold larger balances of real money ().
Higher rate of interest will increase the opportunity cost of holding money.
Both conditions 1 and 2, stated above, hold in the time of stable economics conditions and during hyperinflation, but in the case of latter, we need to modify (3) further.
We know that nominal interest rate is
i = Nominal Interest rate
r = Real Interest rate
= Expected inflation.
Put (4) in (3)
During hyperinflation, it would not be a bad approximation that both Y and r remains constant whereas fluctuates too much. Therefore Y and r in (5) can be assumed constant.
and are assumed to be constant in (6)
(6) tells us that real money balances depend upon expected inflation and as expected inflation increases, opportunity cost of holding money also increases and people will reduce their real money balances.
Now put (6) in (2)
Seigniorage = (7)
Now if government selects a constant rate of nominal money growth during hyperinflation and this rate is maintained forever (this is one of the assumptions) then how much seigniorage can be generated by this constant nominal money growth rate. If nominal money growth rate is constant forever, inflation and expected inflation would both equal nominal money growth (again we assume that output growth is zero). This implies that
Put (8) in (7)
Seigniorage = (9)
According to (9), net effect of nominal money growth on seigniorage is ambiguous. It has a dual relationship.
Given real money balances, seigniorage increases by printing more money. (first expression of in (9)).
When more money is printed, inflation increases and in turn real money balances decrease (second expression of in (9))1.
It also means that while government prints money, the nominal revenue collected by the seigniorage increases. But the advantage in real terms is off set by increased inflation caused by printing money and eventually real seigniorage revenue declines after a certain increase in nominal money supply. Real seigniorage will start declining after a particular point where the cost of printing money (inflation) will be greater than nominal revenue generated through it. The above discussion can be represented by the following diagram (adopted from Romer 1996).
The Seigniorage (S) Laffer curve
(Source: Romer 2006, p.422)
In the above figure
gm = rate of money growth, S = Seigniorage revenue and Smax = maximum seigniorage revenue corresponding to gmax. and gmax is the maximum money growth after which real seigniorage starts to decline.
Is Seigniorage an un-limited source of revenue?
The figure above is quite similar to "Laffer Curve" which represents the relationship between tax revenue and tax rate in the same manner, as has been described above between seigniorage and money growth.
The above figure and the empirical evidence (described later in this essay) suggest that seigniorage is not an unlimited source of revenue. When rate of growth of nominal money is low, there is small reduction in the real money balance which leads to an increase in seigniorage. But when growth in nominal money is higher, the reductions in real money
1. When Seigniorage
balances are also larger (due to increasing inflation). There comes a point, gmax (as described in the above diagram), after which any further increase in the money growth decreases real seigniorage.
Seigniorage and Inflation Tax.
Inflation is a sort of tax on real money balances. So inflation tax can be represented by the following equation.
Inflation Tax =
= rate of inflation
= real money balances.
Now, we know that = , therefore above equation becomes
Inflation Tax = = Seigniorage
Empirical Evidence of Seigniorage:
We already know that printing of money causes the holders of cash and cash equivalent in an economy to pay the inflation tax, which regress the real benefit from the seigniorage, as it was explained by Kiguel and Neumeyer (1989). The same phenomenon can be observed in Pakistan during the last few decades.
The empirical evidence about seigniorage and inflation endorses that seigniorage and inflation tax are correlated notions as when seigniorage exceeds a certain limit the inflation tax becomes harder to avoid and devaluation of the real money takes place as elucidated by Keynes (1923). The empirical evidence described in this essay has been inspired from Arby (2006) who discussed in his paper that during 1970s the seigniorage remained handicapped in creating the real revenue when inflation boosted until 20% on average. After 1970 (when inflation remained below 20%), a direct relation between inflation and seigniorage was being observed in Pakistan.
Increasing Inflation and decreasing Seigniorage in Pakistan.
An inverse relationship, at very high and low inflation rates, between inflation and seigniorage was observed in Pakistan during early 1970s and early 2000s. The inverse relationship was observed due to following reasons.
During early 1970s, inflation was around 20% and government could not generate more real resources by simply printing money. Seigniorage as percentage of GDP declined from 7.4% in 1972-73 to 5.3% in 1973-74. Moreover during early 1970s government was relying more on monetary supplies and the commercial banks were also nationalised. Therefore, the large portion of seigniorage distributed by commercial banks went into the pockets of government. This phenomenon is also supported by findings of Baltensperger & Jordan (1997), who investigated the earnings of seigniorage between the central and commercial banks.
In 1999-2000, inflation was at its ever lowest value. This situation was coupled by inflow of forex reserves along with freezing foreign currency accounts by government of Pakistan. All these factors combined together and gave the government an incentive to increase its seigniorage revenue. Seigniorage as percentage of GDP increased from 2.5% in 1998-99 to 3.8% in 1999-00. As expected, this increase in seigniorage also increased the inflation.
Figure 1 below gives the graphical representation of the relationship between seigniorage and inflation in Pakistan.
(Source: Arby 2006, p.7)
If we start this graph from late 60s, then the inverse relationship between increasing inflation and decreasing seigniorage (as percentage of GDP), around 1971, can also be observed.
The table below shows the contribution of seigniorage to GDP of Pakistan from 1972-73 to 2004-05.
(Source: Arby 2006, p.9)
Behaviour of Seigniorage and Inflation in Pakistan.
The Decade of 1980:
Generally a direct relationship is seen during this era between inflation and seigniorage (Arby 2006). The average seigniorage (as %age of GDP) during this time was 4.8% as compared to 6% in 1970s. Seigniorage and inflation remained low in 1980s due to large inflow of external resources, aids and remittances to Pakistan. Inflation and seigniorage followed a downward trend till such favouring macro economic factors were removed.
The Decade of 1990:
During early 1990s, seigniorage was once again employed for revenue purposes which caused an increase in inflation as well. During the later half of the decade, seigniorage as a percentage of GDP started to decline, along with inflation due to the following reasons.
Privatization of commercial banks, thus reducing the government seigniorage from nationalized banks.
Introduction of resident foreign currency accounts in Pakistan increased forex reserves thus increasing dollarization and declining tax base.
Autonomy given to State Bank of Pakistan in late 1990s which reduced the reliance of government on Central Bank and seigniorage. As Berument (2002) mentioned, that more independent a central bank is, lesser is the seigniorage and inflation thus resulting into more value of real money.
During mid 1990s, we again see an inverse relationship between increasing inflation and decreasing seigniorage (as percentage of GDP) which substantiates the fact that seigniorage is not an unlimited source of government revenue in Pakistan.
Late 90s and Early 2000:
Inverse trend between seigniorage and inflation, starting from 1998-99 till 1999-2000 has already been explained above.
Inflation Tax Trailed Laffer Curve:
As high revenue can only be achieved until a specific level in the period of high inflation, as explained by. After that the inflation causes the real monetary base to decline and cuts down the nominal revenue - Wanniski (1978). If inflation and inflation tax (seigniorage) are integrated in a graphical form, the figure exhibits the characteristics of a Lafer Curve. Seigniorage remained a limited source of real revenue in Pakistan.
The Figure below shows a decline in the inflation tax (seigniorage) as the inflation increases in Pakistan. The trend is the same as that of Laffer Curve.
(Source: Arby 2006, p.10)
A brief cross country analysis, presented in the table below, tells us the following.
Governments are relying less on central banks' for seigniorage because of the increases autonomy given to different central banks in different countries.
Governments try to seize the opportunity of creating revenue (financing purchases / budget deficits) when ever inflation is low, which in turns increases inflation in subsequent periods. But the maximum limit of money supply at which real seigniorage starts declining is not breached by most of the countries.
Seigniorage in Different Countries (% of GDP)
(Source: Arby 2006, p.11)
Through a theoretical model and an empirical investigation, we have shown that seigniorage is not an unlimited revenue source for governments. Different governments generate resources through printing money, when inflation rates are low. But as inflation rate reaches a particular level, government revenue starts to decline along with their reliance on seigniorage.