A lot is being written of NGDP targeting, the central bank, Ben Bernake, the Sumner Critique, and what is coming to be widely seen as a dereliction of duty by the monetary authorities in maintaining a healthy economy. And with this general line of thought and criticism I am sympathetic at worst and more often harmonious. Yet I do have a little sympathy for the position of Ben Bernake, and I will try to use a simple metaphor to explain why that is while also explaining why I fundamentally do support NGDP targeting and largely agree with the Sumner Critique.

Imagine a central banker with an experimental bent, a curious disposition, and an unconstrained sense of mischief (call him Chairman Rooted). Chairman Rooted, like all worthy central bankers, monitors all goings-on within his sphere of influence from a vast wall of LCD screens in his central banking lair, a cavernous vault dug at unconscionable expense into an active volcano. From there, our mercurial Chairman can not only create money with the Dollaratizer – he can make money appear, in both hard and soft forms, instantaneously, anywhere in the world. So on this fortuitous day, our Chairman sets the Dollaratizer to the simple setting of $1, and points it on the sidewalk in front of wandering Bob Nosherstein, and pulls the trigger, chuckling with glee and anticipation.

Bob Nosherstein, strolling down his local commercial boulevard, listening to “This American Life” as he people-watches, was just thinking, as he so often does, “gee, I’d love a candy bar, but all I thought to bring on this particular stroll were my house keys and my iPhone” when – lo and behold! There before me lies a pristine, lightly-sizzling bank note worth one whole dollar! Bob, figuring himself lucky today (and clearly not an economist), quickly scoops up the freshly-baked note and swerves quickly into his local grocer, where after a short bout of analysis paralysis selects his favorite candy bar, takes it to the register, and hands it to the clerk. In exchange for this one dollar, which did not exist before this transaction, Bob Nosherstein has acquired his Snickers.

And here is where we leave Bob to his delicious fate, and where our story and Chairman Rooted’s experiment really begins. For now the economy has, circulating through its tremendous rushing river of transactions, one more dollar than it did before. And as every good economist knows, MV = PQ, so Chairman Rooted’s dollar, tossed into the equation by Bob’s sweet tooth, means an increase in MV; so clearly, either P, Q, or both must change.

But how?

That is, in fact, the trillion-dollar question; and how you answer it will tell you, perhaps more than anything, how you view our current economic conundrums and how you would opt to solve them. So let’s break it down.

Bob bought a candy bar from his local grocer (let’s call them “Waifsay”); that is one more candy bar that would not have been bought, were it not for Chairman Rooted’s shall-we-say-unconventional methods. So what happens? Well, assuming Waifsay operates in a conventional and profit-seeking way, the uptick in demand for Snickers will result in an equal uptick in supply; ie, they will attempt to increase their Snickers inventory to meet what they believe is new demand. And eventually, this increased demand will reach the good people at Mars, who will, seeing the possibility of selling one more Snickers than usual, try to make one more Snickers.

So out of the path of this dollar, a bit (let’s say $0.03) has become Waifsay profit; a bit (another $0.03) has become Mars profit; and now the rest ($0.96) will go towards trying to assemble the material and labor necessary to make a new Snickers.

So what’s in a Snickers? Well, you have nougat, caramel, peanuts, and milk chocolate; you have wrappers; you have the time and labor it takes to assemble it; and eventually you will have the cost of shipping it to the Waifsay in Springfield, where Bob Nosherstein lives. I will skip nougat because nobody really knows what nougat is; let’s discuss caramel, peanuts, and chocolate; in their raw ingredients, these break down to sugar, peanuts, cocoa, oil, and milk. So those are the five commodities, not counting nougat (because, seriously, what is nougat) that will see an increase in demand by some amount (for argument’s sake, let’s say $0.10 each) when the demand for Snickers goes up by a buck.

So what will happen? Immediately, you will probably see that dime represented in a slight price increase, since the supply of these commodities cannot be increased instantaneously. However, the providers of all these commodities, seeing that price increase, will attempt to increase supply. There are three possible outcomes to that attempt:

1) Failure

2) Success at no opportunity cost

3) Success at some opportunity cost

An example of #1 might come with milk; perhaps the cows simply cannot be milked more, and it may take years to increase the milkable cow stock enough to match that demand. In this case, the only thing that will happen is an overall increase in the price of milk.

An example of #2 might be peanuts; a peanut farmer in Georgia, seeing the price of his crop tick up, might spot an open patch of unused arable land on the edge of his field, and say “gosh maybe I should plant more peanuts.” And eventually supply increases and the price returns to its prior equilibrium.

An example of #3 might be cocoa; perhaps seeing a price increase, a farmer decides to plant more cocoa in lieu of something else; or perhaps needs to hire an extra farmhand to help plant the cocoa, taking away his ability to invest elsewhere, etc; so while this does increase the cocoa supply and stabilize the price, it may have secondary effects elsewhere.

So while all these demand increases will result in some short-term price increases, some will “settle” quickly back to their original price level due to supply increases sooner and some will not.

Then you have wrappers, which are mixed plastics and metals. They are petroleum-intensive, and therefore you are mostly going to see an increased demand result in an increase of price for supply-constrained oil. Let’s call that another nickel.

Then you have the labor and transportation, our remaining $0.39. This is all “kind of, you know, it depends” sort of stuff: does making more bars require running more machines, and if so does Mars have more machines or need to buy it? Do they have some slack in their existing labor supply or will they need to hire more? And how do they transport? This is where things start to get complicated, but it will probably involve some combination of a) tapping unused economic capacity; b) tapping something of which supply could increase; c) tapping something (probably petroleum) tightly supply-constrained.

Do you see where all this is going? Firstly, we can see that the single dollar Chairman Rooted created will, due to the magical power of V, raise NGDP by more than $1.00, as it will flow into different hands and institutions who will then pass it along. But more importantly is the fundamental question of what you might more colloquially term “inflation” versus “stimulus,” but I will call “P-share” versus “Q-share” – for each dollar inserted into the economy at a specific point, how much of that increase in MV (as you have probably noticed by now I am holding V constant; that need not be the case but for the purposes of this exercise I think it is a relatively safe assumption) results in an increase in “P” and how much results in a increase in “Q?” You can also read “MV” as aggregate nominal demand and “PQ” as aggregate nominal supply, which is further clarifying – will that new dollar result in increasing “real” supply or only the price level of the existing supply?

As we can see with the Snickers bar, the answer is probably almost always going to be a combination of “both, but by how much, who knows?” The answer depends, almost entirely, on supply constraints, or supply elasticity. Adding new money to the economy will always result in increasing the demand/MV side of the equation; and since MV=PQ and AD=AS the right side of the equation must go up; but how is an open question. It is possible to imagine scenarios where the P-share of such a monetary injection was 100 and the Q-share was zero – for example, our kooky Chairman Rooted could decide all he wanted in the world was to swim in a pool filled with petroleum and purchased some barrels of oil. It is also possible, though less likely, to imagine a monetary injection of Q-share=100; one could hire an au pair, for example, but that would probably still involve some share of that person’s purchases finding their way to a supply constraint somewhere in the economy.

If you ask me, then, what the crux of the monetary issue is, my answer would not be “would NGDP targeting increase AD?” It absolutely would. The question is better phrased as “what is the Q-share of that AD increase?” Implicitly, folks like Scott Sumner are arguing that the overall medium-term Q-share of a more dovish/stimulative/aggressive/pro-growth monetary policy would be high, my gut guess is at least 50 and probably closer to 75, where some of the increase in MV would be felt as inflation but the majority would be in the creation of new goods and services and an increase in employment and living standards. And that may very well be true.

But then again, it may very well be not. The scarcity issue is, fundamentally, the crux of almost all economic issues (and perhaps the definition of economics itself), and if you reframe “scarcity” as “asymptotically supply-constrained” you can see where this is going, right? If you look at charts of the price of crude oil, you see a tremendous spike in summer 2008, just before the floor falls out underneath the American economy. Since then, global GDP has recovered and surpassed its original level, as has American GDP, but American employment is still far below its peak level, as much of that new growth and employment were in poorer and, crucially, less petroleum-intensive economies. The oil-cost of a new job is nowhere higher than it is in the United States. It is far from impossible to believe that a monetary policy aggressive enough to put America back to work at mid-00s level would result in such a massive corresponding increase in the price of oil that the higher employment level would be unsustainable or the stability of the currency would be badly damaged or both.

Now, maybe this wouldn’t happen. Maybe the price and feasibility of alternative energy sources has changed the equation; maybe the 2008 oil price spike was a speculative bubble unrelated to fundamental global supply and demand; or maybe you think the price of oil should be higher and that will stimulate growth in alternative energy research and usage, create economies of scale, etc. But we don’t know that. All we can do is guess. And, fundamentally, that’s all Ben Bernake and the FOMC can do as well. As it the American price level is stable and American employment is stagnant. Thus, it seems intuitive that we could stand a little inflation to increase employment. But there’s no clear evidence that we wouldn’t see an exponential growth in inflation as we see an arithmetic growth in employment. We could be living in a Malthusian world where the fundamental scarcity is not food but energy. Or we could not be! But we don’t know that.

What we would have to do to find out is experiment; essentially, we would have to intuitively believe that there is enough slack in the economy that we could put lots of people back to work in exchange for a significant-but-tolerable increase in prices, and bet on that hunch by doing it. If I were Chairman Rooted and I were running the FOMC, I would make that bet; and so would, I’d bet, Scott Sumner, Paul Krugman, Matt Yglesias, Ryan Avent, Evan Soltas, Tyler Cowen, David Beckworth, Karl Smith, and innumerable others. And so far Ben Bernake has been unwilling to make that bet, which has consigned many, many people to unemployment, poverty, and misery. But I do, at the very least, empathize with a hesitation to make such a bet. Currently, I am an armchair analyst of the highest order. And even Nobel Prize-winning economists, when not in power, have the comfort of strongly advising counterfactuals and condemning the status quo. But if Ben Bernake announced a 5-6% NGDP-growth target (with some slack on the upside for catch-up) and began some asset purchases to back up his target, yes, we would definitely see a spike in AD/MV and a commensurate spike in AS/PQ. But what if after six months, a year, two years, the P-share of that increase was consistently above 50? Above 60? Above 70? One could see inflation at 5%+ levels even as unemployment remained above 7%. And what if you had a “criticality” as Karl Smith likes to say, and unemployment begins falling, but inflation started to creep towards double-digits? Energy costs are a funny thing, since they’re a major component of the cost of everything, including the cost of labor. At what point does Bernake have to yank on the brakes and induce another recession to keep the price level remotely stable?

I think in a world where all scarcity is slowly-but-surely ameliorable – ie, in a world where overall supply can, over time, rise to meet increasing demand – NGDP-targeting is by far the best way to manage a currency and an economy. But in a world governed by one or more fundamental supply constraints – ie, where medium-to-long-term elasticity is at or near zero for a commodity necessary for basic functioning of the economy – all NGDP targeting might do is raise the price level. This may be something you want to do anyway, as it would effectively erode the real debt levels of the population relative to income and prices even as it hurt creditors, which might be something we need – but it is different from the argument that we are simply lacking enough AD to maintain full employment.

The real solution to this problem, is, of course, eliminating energy scarcity. But until we do, we will always have to look at even the best plans for AD management as a double-edged sword.

Now I will go have a Snickers.