Monetary and fiscal policy ought normally move together

We normally think of monetary and fiscal policy as alternative methods of stabilising fluctuations in aggregate demand. It is only in abnormal times, like the present, when central banks' interest rate instruments are at or near the zero lower bound, that we might want to use both monetary and fiscal policy together. Only if we are nervous, and afraid our belt might fail, might we wear both monetary belt and fiscal braces.

But this isn't right, even if you accept the consensus of the past couple of decades: that monetary policy should play the lead role in macroeconomic policy by stabilising aggregate demand to keep inflation on target; and that fiscal policy should be decided on purely microeconomic grounds. Government spending decisions, and deficits, should respond to monetary policy in the same way as private spending decisions.

Suppose you take an absolutely orthodox microeconomic perspective on government investment decisions. All government investment projects should be evaluated on standard cost-benefit grounds. You calculate the costs of the investment, and the benefits, carefully including all externalities and distributional effects, and calculate the Net Present Value of the investment. If the NPV is positive you spend the money; if the NPV is negative, you don't.

In other words, government investment decisions should be evaluated in the same way as private investment decisions, except that the government also takes account of externalities and the distribution of well-being.

But a change in the rate of interest will affect the NPV of government investments just as it affects the NPV of private investments. A lower real interest rate should increase the NPV of both private and government investment projects, make more private and government investment projects "profitable", and lead to an increase in both private and government investment spending.

Suppose the economy is in macroeconomic equilibrium, with inflation forecast to stay at the target, and then (say) consumption demand falls. At a given interest rate, AD will fall too, and inflation would be forecast to drop below target. So the central bank loosens monetary policy, allowing interest rates to fall, to stimulate investment demand. That is an orthodox macroeconomic analysis of monetary policy. But under an equally orthodox microeconomic analysis of fiscal policy, that increase in investment demand should be seen equally in government as well as private investment. It would be an inefficient allocation of resources, between private and government sectors, if private investment decisions were evaluated using a changing interest rate while government investment decisions were evaluated using a constant interest rate. There would be too much fluctuation in private investment and too little fluctuation in government investment.

So if the exogenous shocks are in private consumption demand, we should expect to see monetary and fiscal policy moving together, even if we don't have a deliberate active fiscal stabilisation policy. The central bank uses monetary policy to lead government microeconomists to act like macroeconomists, even if they don't think like macroeconomists. We should expect to see increased fiscal deficits when interest rates fall. (And some puzzled macro-econometricians, expecting to see deficits causing higher interest rates).

If the exogenous shocks were in private investment demand, we should still expect to see monetary and fiscal policy moving together. But now, private and government investment would move in opposite directions.

It is only if the exogenous shocks were in fiscal policy itself that we should expect to see monetary and fiscal policy moving in opposite directions. If governments ran out of "profitable" investment opportunities, we should expect to see private investment, encouraged by lower interest rates, taking up the slack.

And it is not just government investment spending that should respond to interest rates. We accept that private consumption should respond to interest rates; government consumption should respond to interest rates too, for exactly the same reason. A fall in the rate of interest lowers the relative price of present vs. future consumption, so you consume more now and less next year, whether "you" refers to a private or government decision-maker.

Taxes are a bit trickier. Looking at the optimal time-structure of taxation from an orthodox microeconomic public finance perspective, the government should choose the time-path of taxation to minimise the present value of the excess burden of taxation (subject to satisfying the long-run government budget constraint). A fall in the rate of interest lowers the incentive to suffer the excess burden of taxation today rather than next year, and so should lead the government to lower tax rates today relative to next year. There is no exact analogy between private and government income, but the answer is nevertheless the same. Just as private agents should work less today and more next year when the rate of interest falls, so the government should tax less today and tax more tomorrow.

To sum up, even if government fiscal policy is decided from an optimal orthodox microeconomic perspective, with no thought of playing a role in macroeconomic stabilisation, we should still normally expect to see fiscal policy responding as if it were run by macroeconomists. We should expect to see the full gamut of active counter-cyclical fiscal policy -- increased government investment and consumption, lowered tax rates, and increased structural deficits -- whenever monetary policy is loosened by lowering real interest rates.

And we should see reduced government spending, increased taxes, and structural surpluses, when monetary tightening raises real interest rates.

Comments

[Adam P]

Nick, this is a really good point!

[Nick Rowe]

Thanks Adam!

Did I get the tax shifting effect the right way round? I think I did, but I'm not 100% sure. Lower interest rates mean you lower taxes today? Couldn't quite do a mental Euler.

[bill woolsey]

Lower interest rates so lower taxes? Of course.

Households should borrow more through the tax system when the interest rate is lower.

Deficits should be higher when they are easier to fund.

By the way, I didn't realize that Canada was ruled by a benevolent despot. In the U.S., we have these guys called "politicians" in charge. The government microeconomists all work for them, and
from what I understand, hear whatever they want, regardless of what the government microeconomists
tell them, and are inclined to especially listen to the ones that tell them what they want to hear.
Economist says, "Yes, I suppose it is possible that you could have your cake and eat it to, under certain conditions" Politician hears, "wah, wah, wah, wah... you can have your cake and eat it too.. wah, wah, wah.."

You are so luck to have a benevolent despot up there.

[Nick Rowe]

bill: sure. But then if we believed that what we say ought to happen makes no difference ever about what actually happens, why are we bothering to make any normative statements about economic policy? And influencing public opinion is perhaps as important as influencing government opinion.

[Scott Sumner]

Nick, Sorry I got to this late. I think you are right about public investment, wrong about public consumption, and I am not sure about taxes. If interest rates fall because of increasing thrift, then the public is signaling that they want to consume less now and consume more later. In that case (I think) the government should accommodate their wishes by having less public consumption and more public investment.

I believe you could get your result if there was some fundamental difference between public consumption and private consumption. But is there?

I am not sure about your reasoning process for public consumption. You start from the lower interest rates, and reason forward from that point. But in general don't you have to start the analysis from the factor that caused the price change, not the price change itself?

Analogy. Suppose you had private and public oil companies. The private companies all suddenly realized that a oil substitute is about to be developed and so they start pumping oil like crazy. The price falls and the sleepy public company says to itself "oops, prices are falling, better cut back on production." And yet we know that the public company should behave just like the private companies.

Don't be put off by all this quibbling. I think public investment is more important than public consumption, so I think you have a really interesting point. It's not my field, so I can't give you an answer on taxes, it might depend on whether the real long term interest rate is a random walk, or trend reverting.

[Nick Rowe]

Scott: Damn! I think you might be right about consumption. I thought I had such a simple and clearly valid point. Curses! It will depend on the underlying reasons, and whether it's a change in average time preference, related (say) to demographics, or to something else.

On taxes, there is an Euler equation that links current marginal deadweight cost, expected future marginal deadweight cost, and the one period rate of interest. It's just a question of getting the sign right when you differentiate that First Order Condition. It's exactly analagous to the Irving Fisher condition that the MRS between present and future consumption equal (1+r). Just has more negatives, since the govt is minimising the PV of deadweight costs rather than maximising lifetime utility.

I don't like the idea that real interest rates could be a random walk, since a random walk has infinite long run variance, and real interest rates have been around for a few millenia, and they are currently nowhere near + or - infinity.

[Scott Sumner]

Nick, You are right about my random walk comment, it doesn't make sense. I was actually thinking of something slightly different, a sort of EMH that says it is not obvious, ex ante, when it is a good time to buy long term bonds (or consols) At any given point the price of a long term bond is as likely to rise as to fall, isn't it? I have no idea whether this relates to the tax issue, but I thought it might as you were basing it on the assumption that low interest rates are a "good time" to push more taxes into the future. Maybe so, but aren't long term interest rates almost as likely to be higher and lower next year? And if not, doesn't that mean the EMH is wrong? I have a hunch my argument is wrong (as I am not good at finance) but that's the sort of thing I was implying.

Perhaps the counterargument is that the optimal path of taxes depends on the slope of the yield curve. That would probably still get you the results you want regarding fiscal and monetary policy, but on a slightly different basis. I say that because yield curves usually slope upward in recessions.

[Nick Rowe]

Scott: The effect of interest rates on the timing of taxes doesn't depend on any violation of EMH. It's really just like the reasoning that says investment should depend on real interest rates.

Suppose we have higher tax rates today. That increases the deadweight loss today. But it lets us have lower tax rates and lower deadweight losses today. If the one-year rate of interest increased to 10%, you would be able to cut next year's tax revenue by $110 for every extra $100 this year, so there would be a bigger incentive to have higher taxes this year relative to next year.