The changing game between monetary and fiscal authorities

Who's in charge of aggregate demand?

Monetary policy affects aggregate demand; fiscal policy affects aggregate demand. How the monetary authority acts may depend on how it expects the fiscal authority to act, and vice versa. What happens depends on the rules of the game they are playing. The rules of the game have changed in the last month in many countries. In Canada the new rules are very unclear.

It is usually a very bad idea to have fiscal and monetary authorities both trying to target aggregate demand independently. If the Bank of Canada wants (say) slightly lower aggregate demand than the Department of Finance, then if the Bank sets monetary policy to get aggregate demand where the Bank wants it to be, Finance will respond by loosening fiscal policy, to which the Bank responds by tightening monetary policy, to which Finance responds...etc. The eventual equilibrium (which we only get to if the players also care about something other than aggregate demand) is one where fiscal policy is very loose and monetary policy very tight, which neither player wants.

In the 1960's it was common for fiscal authorities to play the lead role in targeting aggregate demand. Budget statements made reference to "creating jobs" (for looser fiscal policy) or to "restraining inflation" (for tighter fiscal policy). Central banks were more subordinate players, cast in a supporting role of adjusting interest rates (and sometimes exchange rates) to influence not total demand, but the composition of demand between consumption and investment (or between domestic absorption and net exports).

Monetarism changed all that. Milton Friedman lost the battle over targeting monetary aggregates, but he won the war nevertheless.The 1970's and 1980's were a muddled period, where it wasn't clear who was in charge of what, but by the 1990's the outcome was clear: the roles had reversed. Central banks now played the lead role and were responsible for aggregate demand (and targeting inflation). It was fiscal authorities who now played the subordinate role of deciding on the composition of demand: between government and private sector, between consumption and investment, and between domestic absorption and net exports (via the effects of fiscal policy on interest rates and exchange rates).

During the 1990's, and until very recently, the government could ignore the effect of fiscal policy on aggregate demand, because it knew the central bank would take offsetting action. The government could concentrate on the national debt, and on microeconomic reasons for changing fiscal policy. If fiscal policy were loosened, the central bank would tighten; if fiscal policy were tightened, the central bank would loosen. So fiscal policy didn't have any effect on aggregate demand, because the central bank made sure it didn't.

The rules of the game have now changed again. In the US, the new rules are clear. Since the Fed wants to but cannot lower interest rates further, fiscal policy can now play a role in determining aggregate demand. In Canada the new rules are not yet clear. Unclear rules are dangerous (who has right of way at this intersection?).

The Bank of Canada has not yet cut interest rates to zero (I think it should, but that's another question). Yet the next budget will probably see a looser fiscal policy deliberately chosen to increase aggregate demand. Should the Bank of Canada take expected looser fiscal policy into account when deciding on interest rates? Should the government take expected future interest rate cuts into account when deciding on fiscal policy?

If both players reason in this way, then both players try to take the lead role, and the play is a mess. If the Bank targets a (say) lower level of aggregate demand than the government, then the Bank will raise interest rates (or not cut them as much as it would have done) in response to fiscal loosening. And in response, the government will loosen fiscal policy further, and in response the Bank of Canada will...etc. We may end up with an undesirable mix of relative fiscal looseness and monetary tightness.

A second possibility is that the Bank of Canada will cede the lead role. Realising that fiscal policy will respond to its actions, it cuts interest rates further than it wants to, purely in order to forestall further fiscal loosening. The Bank lets the government choose aggregate demand. Back to the 1960's.

A third possibility is that the government will once again cede the lead role. Realising that monetary policy will respond to its actions, it loosens fiscal policy less than it wants to, purely in order to forestall monetary tightening (or less monetary easing). The government lets the Bank once again choose aggregate demand. Back to the 1990's.

(Game theorists will note that the first possibility is the Nash Equilibrium; the second has the Bank playing Stackelberg leader, where the Bank picks a point on the government's reaction function; the third is the Bank playing Stackelberg follower, where the government picks a point on the Bank's reaction function. Also note the paradox: it's the Stackelberg follower who plays the lead role in choosing aggregate demand.)

Regardless of who you think ought to be playing the lead role (I think it should remain the Bank, but that's another question), this uncertainty over who is in fact playing the lead role is not a good thing. I am fairly confident that the Bank of Canada and the Department of Finance are talking behind the scenes, and know what roles they are playing; but the audience doesn't. And given a minority government, the audience isn't even sure if the understudies aren't playing the lead role. Back to the muddled 1970's and 1980's.

Comments

[Gabriel]

Since commitment is possible, why not look at cooperative solutions?

Another point here... there's only one (consolidated) government budget constraint. Seigniorage, taxes, public debt, it's all in the same pot.

[Nick Rowe]

Gabriel:

Your second point first: all the "profits" go in the same pot, so we can't push the analogy to duopolistic firms too far. But if the central bank and government disagree on the right level of aggregate demand, or disagree on the inflation or output/employment target, or disagree on the numbers in the model of the transmission mechanism, they can still have diffeent objective functions defined over {monetary policy, fiscal policy} space, and that's all we need to get a fun game going.

Your first point: I'm still trying to think up a good answer. For some reason, governments want an independent central bank (they seem to do better). So the government has in effect deliberately ruled out a cooperative solution to the game. Something to do with the government's inability to pre-commit would be the standard answer, I think.

[kthomas]

Nick, good stuff. The only part I do succintly agree with is that fact that monetary policy is now, more or less, useless. I sincerrly hope Friedman and his ilk are turning over in thier graves.

[Winslow R.]

" If the Bank of Canada wants (say) slightly lower aggregate demand than the Department of Finance, then if the Bank sets monetary policy to get aggregate demand where the Bank wants it to be, Finance will respond by loosening fiscal policy, to which the Bank responds by tightening monetary policy, to which Finance responds...etc. "

Have you thought this through? Great!

Could you explain how the Fed lowering the Fed funds rate changes aggregate demand? Before you answer, please consider that the lowering of interest rates would lower interest payments to lenders from borrowers which would lower aggregate demand provided by lenders while raising aggregate demand created by borrowers. A wash?

The Fed raising interest rates and inverting the yield curve can force fixed rate loans, made by lenders, upside-down thereby destroying the lender's capital base (as in the present case and previous cases in history). It therefore seems the Fed can destroy aggregate demand quite easily but with a very blunt instrument.

Is there any direct evidence where monetary policy can be isolated as causing an increase in aggregate demand? No new technological boom, no fiscal deficit spending, no relaxation in lending standards, just a simple lowering of the Fed Funds rate causing an increase in GDP.

Logically monetary policy doesn't seem to have any ability to alter demand unless lenders are less likely to spend than borrowers or we want to destroy the financial sector.

[Nick Rowe]

Winslow: I am glad you made that comment. A lot of people (including expert economists) totally miss the perfectly valid point you have just made. But then you have missed another valid point that they make. So I'm going to take the opportunity to kill two birds with one stone (not an environmentally correct metaphor).

If the price of apples falls, it creates "income effects" and "substitution effects". (ECON1000, Micro)

Income effects: A fall in the price of apples makes buyers of apples better off (and so increases their demand for stuff), but it also makes sellers of apples worse off (and so decreases their demand for stuff). (Where "stuff" can include apples). But in macroeconomics, at least in a closed economy where we neither export or import apples, the value of apples bought equals the value of apples sold. So the gain in real income to the buyers is equal to the loss in real income to the sellers. So, unless the buyers and sellers of apples have different marginal propensities to consume, it's a wash. Exactly as you say.

This is something few macroeconomists seem to understand, so I will say it explicitly: price changes do NOT have aggregate income effects in macroeconomics; they only have distribution (of income) effects, if any. (This is not true if the country is a net importer or exporter of the good in question).

That was the bit you got right. Here's the bit you forgot:

Substitution effects: A fall in the price of apples will cause buyers of apples to substitute away from consuming pears into consuming more apples, now that apples are relatively cheaper compared to pears. And it will cause sellers of apples to substitute away from producing apples into producing more pears. So it creates an excess demand for apples both from the supply and the demand sides.

In macro, to a first approximation, we should ignore income effects of price changes and look only at substitution effects. Only look at income effects if we either export or import a relatively large amount of the good in question. Only look at distribution effects if you are interested in them in their own right, or have some idea about which direction they go, because otherwise they probably wash out.

For "apples" read "loans".

A fall in the rate of interest causes a substitution effect, towards consuming more goods today, and fewer goods tomorrow. Forget income effects.

[Winslow R.]

For "apples" read "loans".

Substitution effects: A fall in the price of 'loans' will cause buyers of 'loans' to substitute away from consuming pears into consuming more 'loans', now that 'loans' are relatively cheaper compared to pears. And it will cause sellers of 'loans' to substitute away from producing 'loans' into producing more pears. So it creates an excess demand for 'loans' both from the supply and the demand sides.

Huh? :)

Thanks. I am inclined to agree 'at best' aggregate demand may be shifted forward or backwards by raising and lowering interest rates and therefore the current transmission mechanism is a very weak policy tool.

Why not strengthen the monetary policy and actually alter aggregate demand through a channel similar to fiscal policy?

A 'monetary tax' can be implemented by redesigning the broken transmission mechanism where the Fed loans directly to the states on a per capita basis at the current Fed funds rate in large quantities ($6 trillion). This will allow the monetary authority to remove ('tax') funds from circulation in large quantities, subtracting from aggregate demand. In slow times, by lowering the interest rate, the monetary authority could slow or stop the removal ('tax') of funds from circulation. These loans would greatly strengthen the Fed and the effect of its monetary policy tool.

Any ideas why this wouldn't work?

[Andrew F]

Seems like state governments would be slower to respond to a change in the rate.

[Winslow R.]

"Seems like state governments would be slower to respond to a change in the rate."

Good point, and one of reasons to post such ideas is to improve them.

Perhaps state governments would tie their tax rates to the current fed funds rate allowing for monetary policy to be counter-cyclical and immediate.

To get even closer to the foundation of the economy, perhaps the Fed could make per capita loans direct to those citizens that desired them.

Thanks.

[Nick Rowe]

Winslow: Let me re-phrase. The (real) rate of interest is the relative price of present vs future goods. A 10% interest rate (adjusted for inflation) means that if you want to consume 100 more goods today, you must cut consumption by 110 next year (or 121 the year after that, etc.)

So a fall in the real rate of interest reduces the relative price of buying goods today (for consumption or investment) vs. buying goods in the future. That's the substitution effect.

[Nick Rowe]

Andrew: "state" as in "provincial"? (US - Canada translation).

I don't know about slower. But it is interesting to add provincial/state governments to the game. Provincial fiscal policy has a smaller multiplier (because of greater marginal propensities to import). (Though maybe the fixed vs flexible exchange rate question would change that, since each province has fixed exchange rates relative to other provinces). Also, provinces are small relative to the Bank of Canada, or the Federal government, so couldn't really play Stackelberg leader. Would be an interesting exercise to throw provinces into the game, as a third set of players.