Genuinely New Ways of Understanding IS-LM Watch: Karl Smith Is Really Smart and Thoughtful: Wednesday Hoisted from the Archives Weblogging
Let me start with this:
Hoisted from the Archives: October 8, 2011: IS-LM Watch: Three Ways of Looking at a (Closed-Economy) IS Curve: Karl Smith Raises My intelligence Department:
1) The way I was taught: the national income identity:
Y = C(Y-T) + I(r) + G
Spending (the right hand side) must equal production (the left-hand side, which is also income). If spending is greater than production, inventories fall and businesses hire more workers, raising production and income.
2) The way I found in Hicks's (1937), "Mr. Keynes and the Classics": the Wicksellian flow-of-funds through financial markets:
S(Y - T) = I(r) + [G - T]
Planned borrowing (the right-hand side) to finance investment by businesses I plus borrowing to finance the government deficit G-T must equal saving (the left-hand side). If planned borrowing is greater than saving, firms must cut the prices of the bonds they are issuing and so the interest rate rises. As the interest rate rises, the velocity of money increases.
3) Now Karl Smith comes along with a third way: bank lending: the BL curve:
S(Y - T) = BL(i; π, ρ)
where "BL" stands for bank lending. It is no longer the case, as it was in (2), that governments and businesses issue as many bonds as they wish and that businesses decide how much they want to borrow at the current interest rate r. Instead, we have a Stiglitzian credit-rationing situation: banks take a look at their cost of funds and at the riskiness spectrum of borrowers, and decide how much their borrowers--businesses and governments--are going to be allowed to borrow.
Thus planned savings (the left-hand side) must be equal to bank lending BL (the right-hand side). And bank lending is a function of:
- the short-term safe nominal interest rate i (that is the cost of funds to well-capitalized banks,
- the expected inflation rate π, and
- a risk factor ρ that depends on (a) the average quality of the loans that banks are making and (b) on banks' risk tolerance.
If planned savings are greater than bank lending, businesses find that they cannot go forward with investment projects and so spending falls below planned levels, pushing income and savings below planned levels as well.
As Karl writes, this third approach brings to the foreground things that are hidden in (1) and (2). In (1) and (2) of course an increase in government spending holding taxes constant shifts the IS curve: it raises the right-hand side while doing nothing to the left-hand side. But in (3), the reason that an increase in government spending holding taxes constant increases bank lending is that it raises the quality of the average borrower and so reduces the required risk premium ρ:
Government borrowing changes the game... because the government is... always a good credit risk. Indeed, in a world where reserves are swapped for government bonds [because the government can always print reserves] the government can’t not be a good credit risk. Thus a rise in government borrowing suddenly makes overall lending safer and the BL curve moves out. Governments which may directly default (rather than inflate) lose traction.... It is not at all clear that Greece can move the BL curve...
This is, I think, a genuine insight. It is not something that I had ever thought of before, and I am a smarter person this afternoon than I was this morning.
Now let me pick up with four additional observations:
First, the government is not always a good credit risk--as Karl knows in his comments about Greece. Expansionary government spending is stimulative only to the extent that it reduces the riskiness of those top tranches of potential bonds that banks are thinking about funding. If expansionary government spending has no effect because government bonds are as risky as any other bonds, or if increased deficits increase risk in the system as a whole, then austerity is, indeed, expansionary.
Second, government-provided loan guarantees can be a very effective form of economic stimulus: they reduce the riskiness of what would otherwise be very risky bonds, and make banks willing to fund more of them. Government-provided loan guarantees aren't monetary policy. You can call them announcements of fiscal policy--they do involve the government's taxing and spending in those future states of the world when the guarantees are invoked. But it is probably better to call them something like "banking policy" or "credit policy".
Third, Karl's "BL" curve is a way of thinking about the IS-LM framework that I had never seen before--or at least that I do not remember ever seeing before, and it is I think a genuine insight because it highlights aspects of the situation that approaches (1) and (2) conceal. But you could get to where Karl is going using approach (1) or (2)--you would just have to think hard about where all of the spreads between the short-term nominal safe interest rate i in the money market and the long-term real risky interest rate r that determines business investment come from.
Fourth, in all three approaches you can interpret them in one of two ways: as that what is going on in financial markets--supply and demand for bonds--determines equilibrium interest rates and that the interest rate than determines the velocity of money which determines spending, of that supply and demand for cash determines equilibrium interest rates and that given that interest rate the requirement that the ex post flow of funds balance determines spending. As Hicks pointed out back in 1937, you don't have to choose--both channels are operating, and everything is simultaneously determined.
I return to this because Karl appears to be very frustrated this morning:
BL-MP: Critique This View: My description of how the private banking sector interacts with the Central Bank and the Fiscal Authorities seems uncontroversial to me. It seem to arise simply from the widely accepted facts about how these things operate. Yet, I have chatted with several well known economists offline and read much commentary online. The way they talk about finance dynamics implicitly assumes that my model of the world is incorrect. What I’d want to know is where I am going wrong here…. Let me lay it out then….
10) The credit worthiness of bond can be imagined this way: The banking system will dole out funds in an exchange for bonds. There is a ranking of bonds in terms of which will get funding ahead of others. The lower you are on the ranking the worse your credit. At the same time, the less willing the banking system is to walk down the list the worse your credit.
11) The extent to which the banking system is willing to walk down the list depends in large part on the cost of funds which is determined by the central bank….
13) When the Central Bank stands as Lender of Last Resort then that automatically moves bonds up to the top of the list….
15) Because of all of this any institution – but most interesting from our point of view Central Governments – can permanently extract enormous amounts of resources from the economy without raising taxing or printing money….
17) This is key because creditwise it is irrelevant what anyone, anywhere might think of the government’s eventual willingness to pay out of taxes or printing money. They cannot be moved from the front of the list and so they will not be denied funds.
18) A central government which strains its banking system through large amounts of bank extraction will see the private savings rate rise and economic growth slow as both private borrowing for consumption and investment are pushed further down the list.
19) The Central Bank could choose to offset this by allowing higher inflation or it could simply allow the private sector to contract.
20) In any case all bond dynamics are completely anchored by who gets money first versus who gets money last and that is almost completely under the control of the Central Bank, should it choose to exercise such control.
From my perspective, (15) needs to be changed to read "any institution perceived as rock-solid and credit-worthy can extract…" And then you need to worry about what makes an institution rock-solid and credit-worthy in a depression and panic--and what makes it rock solid is its ability to get somebody to raise taxes or print money and give it the proceeds if necessary.
Thus (17) also needs to be changed. It is not "creditwise… irrelevant" what people think of the government's eventual willingness to tax or print. What people think of the government's eventual willingness to tax or print if necessary is of the essence of credit worthiness.
Otherwise, what Karl says appears to be to be correct although in places infelicitously expressed--references to money being created or destroyed when he means not "money" in the M2 sense but rather what we economists call "purchasing power" or "flow of spending". It is, as I said above, a Stiglitzian credit-rationing view of the flow-of-funds through financial markets. But that is a fine view to hold. It definitely highlights important aspects of the situation that are hard to see in the standard formulations of the IS-LM framework.
And its bottom line--that a central bank that can guarantee loans on a large enough scale either explicitly or implicitly by taking private risk onto its balance sheet can push the economy pretty much anywhere it wants to--seems to me to be correct.