61 Comments
User's avatar
Benjamin Cole's avatar

Seems right to me, NGDP should be targeted.

I do wonder about the size of the Fed's balance sheet, and how much IOER should be paid.

There are practical realities, such as the Congress will never balance the budget.

Some nations build central bank balance sheets, and seem to prosper, such as Japan and more recently Indonesia.

Japan taxpayers owe money to themselves, as the Bank of Japan owns half of JGBs outstanding. Yields on 10-year JGBs are well below 10-year US Treasuries.

I suspect the Fed should build a larger balance sheet, to avoid over-leveraged US taxpayers.

Another oddity is that central banks followed the establishment of fractional banks. So, central banks glommed onto the existing fractional banking system to make monetary policy. With some help from Rube Goldberg.

But regulated fractional banks look likely to become a shrinking part of the financial landscape in coming years.

AI version:

"Yes, stablecoins have the potential to significantly reduce deposits in commercial banks, with estimations indicating that up to $6 trillion in U.S. bank deposits could migrate to stablecoins under certain regulatory outcomes. Bank of America CEO Brian Moynihan and other industry leaders have warned that if stablecoins are allowed to offer competitive yields (interest) to holders, they could drain 30% to 35% of total U.S. commercial bank deposits."

QE strikes me as the better option, to avoid over-leveraging of taxpayers and to prevent a suffocation of the economy.

Plus, wipe out all property zoning.

Destiny S. Harris's avatar

Really clear and refreshing, Scott. I appreciate how you cut through the noise and show how much confusion comes from overcomplicating things. The focus on NGDP as the signal that actually matters makes a lot of sense.

Douglas Tyler's avatar

The push for simplicity resonates. I do wonder if part of what makes this regime feel harder for investors is that even when the dominant variable is identified, volatility and path dependence make capital allocation outcomes less forgiving than in prior cycles.

Mate Maras's avatar

This is great btw AD = C + I + G = GDP = M*V = P*Y

Todd Ramsey's avatar

"Monetary policy and NGDP have almost no effect on long run economic growth."

Doesn't the Fed have the ability to constrict real output through too-tight monetary policy? For years on end?

Are you defining "the long run" as long enough that prices would adjust downward and real output would resume growing?

Scott Sumner's avatar

"Are you defining "the long run" as long enough that prices would adjust downward and real output would resume growing?"

Yes.

Arturo Macias's avatar

They do not understand it, because it is false.

Not all monetary transaccions are part of GDP. People buying/selling a house produced 10 years ago are part of MV, but not of GDP, nor PY.

Scott Sumner's avatar

The versions used in textbooks are generally identities, in which case it cannot possible be false. Thus textbooks define V as:

V = P*Y/M

Arturo Macias's avatar

But there is substantial definition of V: the amount of money transaccions in a period. Otherwise there is nothing to “understand”. I would say that velocity from the perspective of a bank is “payments/deposits”.

Scott Sumner's avatar

Long time readers know that I find debates over semantics to be stupid and tiresome, but I agree that people define velocity in different ways. I will say that my definition is the one that is almost always used when people publish data on velocity over time, for instance at a data site like FRED:

https://fred.stlouisfed.org/categories/32242

Arturo Macias's avatar

I don’t think this is semantic. Velocity is first an economic concept, then a ratio. In fact between central bank money and NGDP there are two multipliers: the banking system makes deposits out of high powered money (the bank multiplier) and then the public makes transactions by using deposits to make payments (the proper velocity).

The majority of payments become NGDP: when you buy things produced in US in the same year), but people using money to buy foreing goods or used cars are not making NGDP, while my understanding is that their transactions are part of money velocity.

Scott Sumner's avatar

Hmm, you say it's not semantic, and then disagree about the definition in the link I provided. Saying I'm using the wrong definition is exactly what the term "semantic" means.

This is a waste of time.

Here's what FRED says:

"Velocity is a ratio of nominal GDP to a measure of the money supply (M1 or M2). It can be thought of as the rate of turnover in the money supply--that is, the number of times one dollar is used to purchase final goods and services included in GDP."

Arturo Macias's avatar

But that is the problem! What about dollars not used to buy/sell goods that are GDP?

The fact that you have two sides (money transactions) and GDP (value creation) is what makes the concept useful. There is a complex dynamics on bank multipliers and velocity.

Arturo Macias's avatar

The concept of the “velocity of circulation” of money was used in the XVIII century monetary debates, far before GDP. Petty and Hume were prolífic users, before GDP was even considered, and the concept in its classic aception was used by Allais, perhaps contributing to add confusion to the field.

In fact, the first thing I did when I had my first bank statement in Excel was to compute my own “money velocity”, that being young and overpaid was obviouly too high…

Scott Sumner's avatar

Yes, just as the term "gay" meant one thing in the 1890s and something very different in the 1990s. Meanings change over time.

William Ellis's avatar

Maybe the reason NGDP targeting never gets a shot is because some people benefit from booms and busts.

Scott Sumner's avatar

I doubt that. It's pretty clear to me that the Fed regrets its mistakes.

william stepp's avatar

Y = C + I - (G + T) Production minus State Depredation

Max's avatar

All fair - in my macro courses in the mid-80s, my profs insisted that sustained budget deficits should increase aggregate demand but may crowd out investment. We did not talk much about whether it might crowd out consumption, presumably because the increase in G was not offset by taxes which might have reduced consumption. A lot obviously depends on what central banks are doing at the same time.

matklad's avatar

Oh, yes, they do! Will take more time to digest, but just the first two already put in sharp focus things that were fuzzy for me before:

> Real shocks do not cause big jumps in unemployment. Period, end of story. Even I’m surprised by this fact, but it is evidently true.

Accompanied with two extremely convincing graphs (and I guess COVID is a good example of a magnitude of a real shock that does cause a jump).

> But imagine a kingdom where for some strange reason people made contracts to supply 100 “units of wheat” 12 months in the future, where units were left unspecified.

And the reason for stickiness of prices is immediately clear!

Thank you very much for the pointer (and for writing those articles in the first place, obviously!)

Scott Sumner's avatar

Thanks, glad you liked it.

Kevin Lacker's avatar

I'm curious, you keep returning to the theme that "reasoning from a quantity change" is bad. Is there some underlying principle here that is applicable past just economics? Or is the rationale fundamentally about these intersections-of-opposite-sloping-curves that seem to mostly happen in economics where you can interpret anything two ways.

Scott Sumner's avatar

It depends how broadly you define "economics". Does it include sociology? How about never reason from a prison population change? You might have more people in prison because you've become tougher on crime, or because you've become less tough, causing the crime rate to soar.

The exogenous/endogenous distinction seems appropriate here.

matklad's avatar

> In a perfect world […] 3 chapters on macro

What would be the closest to those three chapters in our actual world? I skimmed through the Midas Paradox and the Money Illusion, and while some ideas intuitively made sense to me (NGDP driving business cycle, markets for financial assets as indicators of monetary policy), I must confess that many specifics went over my head (how we _actually_ measure NGDP? How close our measurement is to reality?), and I also failed to get a coherent big picture of how things work. (My level of knowledge is roughly https://www.core-econ.org/project/core-the-economy/)

Kee's avatar

Fantastic summary, Professor Sumner. Have you ever considered writing a very short economics book for high-school students? (Speaking as a parent of one!)

Scott Sumner's avatar

Unfortunately, I'm retired from writing books. The one's I wrote previously did not do well.

Shelby Dean's avatar

i suppose increase/decrease in money supply equals same increase/decrease in nominal spending. So, central bank determines level of nominal spending/ngdp.

Scott Sumner's avatar

They are only equal when velocity is constant. But velocity usually changes over time and the supply of money needs to offset those changes.

Moss Porter's avatar

I understood your elegant equation immediately.

But I always thought I was 1 in a million like my Ma always insisted

Scott Sumner's avatar

All my commenters are special. :)

Max's avatar

That I spent too much time studying fiscal levers for macroeconomic policy and not enough attention to the monetary side? I think it's easier for lay undergraduates (and my university econ went no further than intermediate micro and macro) to understand how changes in C, G, S and T affect growth and trade; monetary policy is more abstract.

Scott Sumner's avatar

I'd say it's easier for a lay person to THINK they understand AD = C + I + G.

But does more G make AD go up, or does it make C + I go down?

Moss Porter's avatar

According to the outline of your macroeconomics course it would depend on the monetary policy and specifically the change in nominal income

Moss Porter's avatar

Correction:

According to your fundamental equation, cetirus paribus, C and or I will go down