Yates on Lags and a Balanced Budget

Well they are all a twitter.

Yates reports that Wiesenthal of Bloomberg asked what MaMos would think of a balanced budget rule.

He then added how Noah Smith twitted a link to a Scott Sumner post claiming that long and variable lags are a myth.

Yates then writes a blog post discussing how Sumner must be wrong--referring to characteristics of new Keynesian models.

Under current conditions, I believe the U.S government should run a surplus under "normal" conditions.   Government spending should be kept to a minimum and tax rates set to generate slightly more revenue.   It should then gradually pay down the national debt.

If the economy does poorly, then the government will collect less tax revenue.  I would strongly oppose having the government raise tax rates to balance the budget.   The budget surplus would, and should, decrease.   Now, if the problem is so severe that the result is a deficit, I do think the government should borrow and add to the national debt.   However, I favor a constitutional debt limit, and one reason for running a surplus under normal conditions is to reduce the national debt below the debt limit so that temporary deficits and borrowing would be consistent with the constitutional debt limit.

I am not too worried about the national debt being paid off, but if that did happen a balanced budget is appropriate and in that scenario, a temporary reduction in tax revenue should allow for a budget deficit and a temporary national debt.   Then a return to the rule of thumb that there should be a surplus when the economy returns to normal.

A permanent contraction in real output, on the other hand, should result in less government spending and a reduction in the provision of public goods.

I have no interest in using discretionary fiscal policy, including tax cuts, to return spending on output to a target growth path.   Even less do I favor introducing special government spending projects for the purpose of creating jobs.

Note that I don't worry much about the impact of taxes and spending on spending on output.

Yates apparently hasn't heard, but Market Monetarists favor targeting nominal GDP in the future.   We do not favor adjusting a target interest rate according to the past deviation of nominal GDP from a target level.

Now, for long and variable lags.

For Milton Friedman, the thought experiment was always a change in the growth rate of the money supply.   We are in equilibrium at one growth rate of the money supply   Suppose we shift to a higher growth rate?   How long will it be before the economy returns to long run equilibrium?   That is, a higher inflation rate, higher nominal nominal interest rate, with real interest rates, real output, and employment all back to their natural, long run equilibrium values?   Well, the lag between the new growth rate of the money supply and that final long run equilibrium is long.   And further, it appears to take longer sometimes than other times.

However, one problem with the use of empirical evidence to evaluate Friedman's claim is that there has never been a monetary regime that targets the growth rate of the money supply, much less one that has tried out various growth rates and allow adjustments to the new equilibrium.   Friedman tested this thought experiment in worlds with quite different monetary regimes.

Most importantly, with a gold standard, changes in the quantity of paper money and deposits (both demand and time) are unlikely to be new, persistent increases in the growth rate of the quantity of money.    Frequently they will be accommodating a temporary or even permanent change in the demand to hold money, and even if they are truly an excess demand for or supply of money given the existing trajectory of nominal income, there is a very good reason to expect they will be temporary.

This same reasoning applies to price level target.

Market Monetarists don't favor changing the growth rate of the money supply and letting the economy settle down to a new equilibrium,  How long it would take for a change in the quantity of money to have its full effect on inflation is of little concern.

What is of more concern is the lag between current monetary conditions and spending on output.   How changes in spending on output would be split between changes in prices and output is of less concern.   We are not concerned about inflation per se, but neither is the goal to goose demand to create output and employment.  Nominal GDP level targeting means that the goal is to get spending on output on target.

Years ago, Sumner advocated targeting the price level one month in the future.   But for the last decade or so he has advocated targeting nominal GDP one or even two years in the future.  The stabilization of current spending on output would largely be generated by the expectation that spending on output will be on target in the future.    And further, the reason that nominal GDP will be expected to be on target in the future will mostly be the expectation that future monetary conditions will be consistent with nominal GDP being on target in the future.

Milton Friedman and the old monetarists typically assumed that all changes in the quantity of money are permanent.   How long will it take for that permanent change to have its full impact on the price level?   How long will the transitory changes in real interest rates and real output and employment last?

But that tells us little about what happens if the quantity of money were to rise and then fall.   Should the analysis of a permanent increase in the quantity of money be transitioned over to a temporary increase in the quantity of money?   Suppose the quantity of money rises by $50 billion this quarter and then falls by $100 billion the following quarter.  Do we really expect that a year later there will be a quarter of higher real output growth followed by a quarter of lower real output growth?   And then a year after that a quarter of a higher price level followed by a quarter with a lower price level?    What possible market process would allow for such a result?

What firm would raise prices because of a boom in demand a year ago, especially when it was followed by a recession nine months ago?

Under a gold standard, the inflation rate depends on the supply and demand for gold.   Maybe projecting past experience into the future was a good enough rule of thumb    Current changes in the price level are likely to be reversed with long run stability.   Looking at what happens to various measures of the quantity of money says little.

Looking at data from a period where the central bank uses a smoothed interest rate target to keep inflation and unemployment from rising too high will provide little information about a regime that targets a growth path for spending on output.

Market Monetarists believe that the central bank should make no commitment regarding the future path of money market interest rates or the quantity of any measure of the quantity of money.   We certainly anticipate that nominal GDP will deviate from target.   The only commitment is that the quantity of money and interest rates will be at the level expected to keep future spending on output--future nominal GDP--on target.