In my previous post I showed that Krugman’s recent piece on Argentina completely glossed over the data in its assertions that the inflation in that country was due to fiscal deficits**. I also, somewhat offhandedly, referred to his argument as being ‘monetarist’. This caused some degree of confusion so I thought I should probably clear it up.
Okay, so let’s first try to get a grip on the monetarist position on deficits. This was most clearly brought out in the debates during the late-70s and early-80s under Thatcher in Britain. Note that in this period the fiscal deficit was referred to as the PSBR (Public Sector Borrowing Requirement).
In his seminal account of the monetarist era in Britain, The Scourge of Monetarism, Nicholas Kaldor lays out clearly the monetarist position on the relationship between fiscal deficits, inflation and the money supply.
In the Green Paper on Monetary Control of March 1980 it is asserted that “it is sometimes helpful to examine how a particular control will affect items on the asset side of the banking system”. The Paper then proceeds to state an accounting identity which shows the change in the money stock (£M3) in a given period as the sum of five separately identified items, of which the PSBR is one… The main monetarist thesis is that the net dissaving of the public sector in ‘inflationary’ in so far as it is ‘financed’ by the banking system and not by the sale of debt (bonds or gilts) to the public. (pp48-49)
Kaldor’s account is a rather nice and clear view of the monetarist theory of the fiscal deficit. If the deficit is financed by selling debt to the public it ‘crowds out’ private investment by driving up interest rates. But if it is funded by the banking system — i.e. the central bank — it is inflationary. Sadowski, for example, completely missed this distinction when he wrote in response to my piece,
Pilkington is evidently drawing from a peculiar rewriting of the history of the Thatcher years, when the UK’s public sector borrowing requirement (PSBR) was targeted for reduction as a means holding down interest rates. No form of monetarism, not even an imaginary special Thatcherian variation, believes that inflation is a fiscal phenomenon.
Both parts of Sadowski’s arguments are completely incorrect. First of all, in the Thatcher years the monetarists closely watched to what extent the fiscal deficits were being funded by the central bank and chalked these up as a major cause of inflation. Second of all, Really Existing Monetarism under Thatcher did indeed believe that the (monetised) fiscal deficits added — by identity — to the M3 money supply and thus to inflation. It is Sadowski that is rewriting history by asserting otherwise.
Now, how does this monetarist argument regarding the relationship between (monetised) fiscal deficits, the money supply and inflation square with Krugman’s? Well, let’s turn to his piece on Argentina to see.
Running deficits and printing lots of money are inflationary and bad in economies that are constrained by limited supply; they are good things when the problem is persistently inadequate demand. (My Emphasis)
Do you see that? That is identical to the monetarist argument made in the late-70s and early-80s. And that is why I called Krugman’s argument ‘monetarist’… because it is monetarist!
So, what was the Keynesian position in those years? I don’t want to run through this in too much detail as it is enormously complex but Kaldor gives us a taste.
[The monetarist view] ignores the fact that the net saving, or net acquisition of financial assets of the private sector will be the same irrespective of whether it is held in the form of bank deposits or bonds. The part of the current borrowing of the public sector which is directly financed by net purchases of public debt by the banking system — and which has its counterpart in a corresponding increase in bank deposits held by the non-banking sector — is just as much part of the net saving of the private sector as the part which is financed by the sale of gilts to the private sector. (p49)
Now, I don’t want to get into a debate regarding the truth of the above statement, that is not my point here. My point is simply that Krugman’s argument is far closer to the monetarist position than to the Keynesian position. Anyone who states otherwise is either ignorant of these debates or does not understand the implications of endogenous money theory sufficiently well. And that is the end of the story.
Update: Thanks to Nick Edmonds who pointed to the following document in the comments section. The document reads:
It is not the intention to achieve this reduction in monetary growth by excessive reliance on interest rates. The Government is therefore planning for a substantial reduction over the medium-term in the Public Sector Borrowing Requirement (PSBR) as a percentage of Gross Domestic Product (GDP). The relationship between the PSBR and the growth of money supply is important but is not a simple one; it is affected by the economic cycle, the rate of inflation and the structure of the tax and public expenditure flows generating the borrowing requirement. But although the relationship between the PSBR and £M3 is erratic from year to year, there is no doubt that public sector borrowing has made a major contribution to the excessive growth of the money supply in recent years. (p16 — My Emphasis)
________
** Note that Mark Sadowski questioned the data I provided and supplied IMF data that was somewhat different. Although I am suspect of the IMF data, as Argentina basically told the institution to shove it in 2001-2002 they have every incentive to exaggerate the fiscal deficit, even if we take it at face value it makes the same point that I originally made: namely, that inflation and fiscal deficits are not correlated. You can read the exchange between Sadowski and I here.
Fixing the Economists 
Would a monetarist say that deficits (and money printing) are ONLY inflationary when there’s an agg supply constraint?
This is like taking Keynes’ comment, paraphrasing, “when the economy is at full employment we are back in the world of the classical theory” and calling Keynes a classical economist.
I think it’s more accurate to view PK’s position as how we used to discuss the old AS curve–there’s a Keynesian region (deficits have no impact on prices); an intermediate region (anything that increases AD impacts both Q and P); and the classical region (at full employment any increase in AD will only impact P).
The latter is the world of the monetarist. Deficits are always inflationary unless the government somehow “pre-finances” its spending by issuing bonds to the public.
Thanks, now I don’t have to go back and read Kaldor. He was an MMTer before there was MMT.
(1) If there is an inflation in monetarist-land then there must be aggregate supply constraint — even if there is 10% unemployment. They would just claim that the NAIRU has risen. So, for them — as for Krugman — the very existence of inflation indicates that there is an aggregate supply constraint.
(2) Yes, it is well-known that Keynes’ comments in that regard are at odds with contemporary Post-Keynesian theory. The GT was written “on the terms of” the mainstream. Later Keynes rejected the microeconomics of the mainstream and hence that comment becomes irrelevant.
(3) Yes. AS-AD curves are basically in line with actual monetarist thinking. They are not very Keynesian at all. That’s why you don’t find them in Kaldor or Robinson or whatever.
As agreed, PK’s focus on demand side in Arg ignores other important influences on inflation. But you are still stretching when you call him a monetarist for focusing only on AD (in that original post you link to).
I don’t know many Keynesians–of all stripes–who would deny that excess demand won’t impact prices “at some point.” Minsky states, inflation is mainly the result of financing too many claims (excess demand) on the supply of consumer goods. MMTers agree that deficits can be inflationary “at some point.”
You focus ONLY on the first part of PK’s statement and say “AHA! See, I told you, he’s a monetarist!” When, in fact, he says “it depends” on the conditions in the economy.
It appears that you, on the other hand, believe demand can never have an impact on inflation?
Post-Keynesians — and early Keynesians — view inflation as primarily a wage issue. Excess demand does not cause inflation directly. Typically it goes through the wage channel as the labour market tightens up. Sometimes it hits due to bottlenecks in certain sectors.
In an open economy it often hits primarily through the current account and via a devaluation of the currency. All of these dynamics are at play in Argentina.
In some circumstances government deficits — usually in wartime (WW2, Korea, Vietnam etc.) — can generate the above dynamics. But most of the time they come from elsewhere.
Monetarists like Krugman have a conception, however, that the channel through which the inflation is created is simply by generating too much money (exogenously) which then chases too few goods. That is the monetarist conception and it is wrong a priori. Once we accept endogenous money it cannot be correct because “money” is no longer seen as something that is actively “created” via policy. Rather it becomes a residual of the dynamics mentioned above.
“Running deficits and printing lots of money are inflationary and bad in economies that are constrained by limited supply”
You can change that statement to:
“Running deficits and printing lots of money are inflationary and bad in economies that are constrained by limited supply and insufficient private net-saving.”
Then it works.
Which is why I suggested, almost flippantly, that we should stop using ‘net-saving’ and start using ‘voluntary taxation’ instead.
You also have to take the wider context. Krugman, like the monetarists, simply assumes that there is a supply constraint any time he sees inflation. That is why his analysis of Argentina was so crap and misleading.
Interesting post.
This accounting identity for £M3 is:
I was taught this is high school economics in the 80s as the central plank of monetarist policy and it seems funny to me now that it’s hard to find even a mention of it (although perhaps not, given how absurd it is). There is no doubt, though, that the PSBR in this period was seen principally as one of the elements in controlling monetary growth, albeit also with an eye to interest rates, see e.g. clause 4 of the MTFS.
Click to access CA1DCDBC51264429B4A5E242596C5A09.pdf
Yep, that’s right Nick. Anyone who learned macro in the 1980s was taught that the PSBR/fiscal deficit was a key component of monetary control. This had a massive impact on Krugman’s generation. That is one of the reasons why Mankiw was perfectly correct to say that New Keynesianism should more properly be called ‘New Monetarism’.
From a monetarists perspective inflation will occur when growth in NGDP is greater than growth in RGDP.
A govt deficit financed by selling bonds to the private sector will be inflationary to the extent that it affects NGDP and RGDP differently. At less than full employment the deficit will boost NGDP (as the deficit increases people’s income) and will boost RGDP to the extent that it simulates increased output (private or public). You can invent scenarios where the deficit will be inflationary of deflationary depending upon which parameters you use.
If the CB buys some of these bonds from the public with newly created money then unless it is doing so to counteract a fall in NGDP (caused by the deficit or something else) this will likely be inflationary. since the new money will increase spending more than it will increase output (even at less than full employment).
I agree that Krugman thinks it is the second scenario that is happening in Argentoina and this argument is consistent with monetarist theory.
Very odd statement given that NGDP/RGDP – 1 = CPI.
Agree with most of the rest though.
I think it is true to say “inflation will occur when growth in NGDP is greater than growth in RGDP”. The formulate NGDP/RGDP – 1 will give you the rate of this inflation.
Philip,
“But if it is funded by the banking system — i.e. the central bank — it is inflationary”.
“Financed by banking system” doesn’t only refer to the central bank.
If the Treasury sells bonds to the banking system (and the banks hold on to them), then when the Treasury spends this results in a net increase in bank deposits (i.e. M3 money). I think this is what Kaldor means in the quote above.
by the ‘banking system’ I mean the commercial banks.
The Treasury doesn’t spend the result. The process result is the issuing of the bonds (which are a form of new money). Spending first, once those transactions are resolved, decide form of saving.
Very important that you linearise the circular process from the correct starting point – otherwise you end up risking making the same mental mistakes as monetarists.
If commercial banks hold bonds then they are a direct replacement for bank reserves. They offset the increased holdings of bank deposits in the rest of the private sector.
If commercial banks don’t hold the bonds, then both bank reserves *and* deposits drop and the banks’ balance sheets shrink – because other entities are holding the bonds directly. It’s precisely the same effect you get when people draw cash.
When the Treasury sells bonds to banks, bank reserves go down but there is no reduction in bank deposits. When the Treasury spends, bank reserves and bank deposits increase. So in the case of a fiscal deficit involving only bond sales to banks (i.e. “financed” by banks), there is an overall net increase in bank deposits (i.e. M3 money) as a result of the deficit spending.
If instead the Treasury sells bonds to non-banks, bank reserves and bank deposits both go down by the same amount, and increase by the same amount when the Treasury spends. So in this case there is no overall increase in bank deposits (i.e. M3 money). This is what Kaldor is referring to in the quote above.
This is what Kaldor is referring to in the quote above, when he says “The main monetarist thesis is that the net dissaving of the public sector in ‘inflationary’ in so far as it is ‘financed’ by the banking system and not by the sale of debt (bonds or gilts) to the public”.
Because deficit spending ‘financed’ by bond sales to (commercial) banks leads to a net increase in bank deposits (M3 money) overall, whereas deficits ‘financed’ by bond sales to the non-bank public do not.
And of course that is simply an artefact based upon an incorrect view of the accounting and the idea that banks use deposits to expand lending.
If you sit any intermediate entity between a source and a target the balance sheet will expand.
‘Inflation’ seems to be used here in an almost Austrian sense – bigger balance sheets.
“There is no necessary relationship between the size of the PSBR and monetary growth.”
Milton Friedman in evidence to the Treasury and Civil Service Committee of the House of Commons.
Cited in “Public Expenditure: Its Defense and Reform” by
David Heald, p.51.
But the Thatcher government certainly thought there was. From the Medium Term Financial Strategy (link in my post above):
“… although the relationship between the PSBR and £M3 is erratic from year to year, there is no doubt that public sector borrowing has made a major contribution to the excessive growth of the money supply in recent years.”
Yes, only monetised PSBR has a relationship to the M3. The Treasury under Thatcher was clear about this.
The ambiguity is in the word ‘monetarism’. In some countries during the 1980s, ‘monetarism’ either meant ‘Thatcherism’ or at least ‘Thatcherite macroeconomic policy in the years from 1979-1985 approx.’ On the other hand, there are some people (not insiginifcant in number) who take ‘monetarism’ to mean roughly ‘the macroeconomic ideas of Milton Friedman and similar macroeconomists either insofar as they followed Friedman or influenced him’.
The March 1980 Green Paper is not a good source if you’re in a context (as we almost always are now) where ‘monetarism’ means the latter. For example, Friedman said that “only a Rip van Winkle”, who was ignorant of macroeconomics from about the early 1950s onwards, would have written the sections in the Green Paper that listed the methods for controlling the money supply. And it takes only a superficial familiarity with Friedman’s theories to know that what matters for the money supply is how deficits are funded, not their existence. Friedman continually dismissed the notion that the 1980s US deficits were inflationary. And you would search long and hard and fruitlessly for Friedman to say favourable things about the credit-counterparts identity, which is a very different way of thinking about money supply determination than the money-multiplier approach.
Kaldor’s book was written in a time and context in which the former use of ‘monetarism’ was predominant. That time has passed.
For a good paper that discusses (at some points) some of the interrelations between American monetarist ideas and British macroeconomic policy in this period, I recommend Edward Nelson’s “The U.K.’s rocky road to stability”. For a British perspective on the differences, I recommend Tim Congdon’s “British and American Monetarism Compared” in his book “Keynes, the Keynesians and Monetarism”.
I’ll clear this up for you: I’m not interested in cultist arguments about what is and is not “true” monetarism. For me “monetarism” encompasses the policies that were enacted in the 1970s and 1980s by governments and central banks that aimed at controlling the money supply (and failed miserably).
That’s a flawed definition from the start, since there was a great deal of variance between different countries in method and doctrine.
Anyway, if you’re going to exercise your liberty to use words however you want, it would save a lot of your time and others if you define them clearly when you deviate from standard usage. If your definition of monetarism excludes Milton Friedman, then it’s deviating from standard usage.
http://en.wikipedia.org/wiki/Monetarism
Looks like a standard definition to me. It’s you who are insisting that only what Friedman says counts as True Monetarism. Anyway, I’m not interested in having such a ridiculous and obviously unproductive semantic debate.
“The main monetarist thesis is that the net dissaving of the public sector in ‘inflationary’ in so far as it is ‘financed’ by the banking system and not by the sale of debt (bonds or gilts) to the public”
It’s worth pointing out that this would contribute towards an F if a student of Friedman had written it. Friedman et al did not believe that government borrowing from commercial banks is equivalent to borrowing from the central bank.
I think Kaldor is talking about the central bank here.
It’s not clear from the phrase ‘banking system’. In ordinary contexts, undistorted by the idea that government borrowing that does not affect base money can increase the money supply, I’d assume that it does, but it really depends on who he is criticising at the time.
I think it’s clear from this part that Kaldor is talking about the commercial banks rather than the central bank:
“the net saving, or net acquisition of financial assets of the private sector will be the same irrespective of whether it is held in the form of bank deposits or bonds. The part of the current borrowing of the public sector which is directly financed by net purchases of public debt by the banking system — and which has its counterpart in a corresponding increase in bank deposits held by the non-banking sector — is just as much part of the net saving of the private sector “
I don’t see it. Why would government debt purchases by commercial banks increase deposits in the non-banking sector?
Say for example bank A has £100 in reserves and £100 in customer deposits.
The Treasury then sells £100 bonds to bank A. Bank A’s reserves decrease by £100 as it pays the Treasury for the bonds (and the Treasury’s balance at the central bank increases by £100).
The quantity of bank A’s customer deposits is unaffected by the bond purchase.
So bank A now has £100 in bonds and £100 in customer deposits.
The Treasury then spends £100 to pay Mr Brown, who has an account at bank A. This results in an increase in bank A’s reserves AND bank A’s customer deposits. Both increase by £100.
So now bank A has £100 in reserves, £100 in bonds, and £200 in customer deposits.
So overall there has been a net increase in the quantity of bank deposits held by the non-bank public as a result of the deficit spending.
Yeah, I see where you’re coming from. But you’re missing one key element. Namely, the fractional reserve aspect.
Let’s stick with your example and assume 100% fractional reserves. So, the starting point when bank A has £100 in reserves and £100 in deposits is a sort of “equilibrium”. At the end of your example there is a disequilibrium insofar as the bank has £100 in reserves and £200 in deposits. If the central bank did not accommodate by providing more reserves the interest rate would start to climb.
The point is effectively the same if we assume reserve requirements of 10% but the numbers change.
Yes you’re right that the central bank would have to provide some extra reserves to the commercial banks in that case, presumably by buying some bonds from the banks (assuming there is a required reserve ratio and the central bank wants to maintain a certain interbank rate of interest).
But say instead that the government bonds are initially bought by non-banks.
Sticking with my simple example, say bank A has £100 in reserves and £100 in deposits. A customer of bank A (Mr Brown) buys £100 bonds from the Treasury. Bank A’s reserves and deposits both go down by £100. The Treasury’s balance at the CB increases by £100.
So now, bank A has £0 reserves and £0 deposits, and Mr Brown has £100 in bonds.
The Treasury spends £100 to pay another customer at bank A (Ms Jones).
Bank A now has £100 in reserves and £100 in deposits.
Overall, in this case, there is no net increase in the quantity of bank deposits held by the non-bank public as the result of deficit spending. Nor is there a need for the central bank to provide additional reserves.
Sure. In this case there is no net increase. But Kaldor was talking about a situation where there was a net increase.
So far as I can see this can only happen if the CB accommodates the reserves. I think this is basically the case that Kaldor is making in the book and is the Post-Keynesian endogenous money argument.
You guys seem to be over-complicating this. Assuming no external sector, Kaldor is simply saying that the public sector deficit increases the private sector’s net surplus. In the end, if the bond that backs the deposits created by government spending are held by the banks, there’s simply a shift from reserves to bonds on the banks’ BS without any decrease in deposits (money supply is increased); if the non-bank sector buys the bonds, there’s a shift from deposits to bonds held by that sector, so no net increase in money supply. These are the results regardless of CB action.
If the CB monetizes, it’s buying the bond back from either the bank, increasing bank reserves again, or the non-bank, increasing their deposits.
I should’ve added, that last action also ends up increasing the money supply.
Pingback: Philip Pilkington: Inflation is NOT Always and Everywhere a Monetary Phenomenon | naked capitalism