You shall not crucify mankind upon a cross of 2% annual CPI growth
February 24, 2015 7:22 PM   Subscribe

In the years since the financial crisis, central banks across the world have struggled to stimulate adequate aggregate demand. Most mainstream economists agree that this is due to inherent impotence of monetary policy at the zero lower bound, although some, including some central banks, now say that the effective lower bound, though finite, is below zero. The Swiss and Danish central banks are currently testing that idea.

However, a growing number of writers are advancing the idea that this impotence is no theoretical requirement, but an artifact of the current monetary policy tool kit, and particularly inflation targeting. Scott Sumner and Nick Rowe are two of the more prominent advocates of this claim. They suggest that rather than targeting an annualized measure of inflation (whether or not the actual target is the stated one), central banks should have a Nominal GDP Level target (NGDPLT). Nominal, meaning that prices are not attempted to be adjusted for inflation. Level, meaning that errors carry forward, so if the target is missed by 1% this year, next year's target growth will increase by that same 1%.

While no central banks have explicitly adopted such targeting yet (although some suggest one already implicitly did), at least one European finance minister is talking about related ideas in the form of NGDP linked bonds, and Japan's Abenomics are certainly in the same vein.

And if you want to play along at home, there is now, thanks to Sumner and funded by Gabe Newell, a quarterly US NGDP growth rate prediction market. Janet Yellen has, as of yet, failed to mention it.
Sumner, not one to rest on his laurels, is offering yet more radical targets.

Meanwhile in the US, Fed Chair Yellen pushed back on a bill sponsored by Senator Rand Paul to "audit the Fed", which some see as fundamentally a means of reducing the Fed's operational independence.
posted by PMdixon (31 comments total) 30 users marked this as a favorite
 
What? Are you implying the Swiss will pay me to buy a car? Is it that if I have a zero balance, I will be sucked into some alternate reality inside a small safe deposit box?
posted by Oyéah at 7:27 PM on February 24, 2015


Oyéah: "Is it that if I have a zero balance, I will be sucked into some alternate reality inside a small safe deposit box?"

No. Rather, if you have a positive balance, you will be awarded negative interest rates. It seems crazy; anyone faced with negative interest rate could plausibly hold the money in a mattress at a better return.
posted by pwnguin at 7:35 PM on February 24, 2015 [3 favorites]


pwnguin: But even that isn't costless. It's expensive to have people guard your giant pile of cash or gold dubloons or whatever, especially when risk is factored in.

If you need to spin it, call it a fee for FDIC backing.
posted by phrontist at 7:42 PM on February 24, 2015 [1 favorite]


The Swiss have a pretty weird problem in that too many people would like to leave their money in Swiss banks, which causes CHF to appreciate against EUR, which hurts exporters, so solutions are gonna look pretty weird, yeah. Although at any scale there's definitely non theoretical, non zero costs to holding cash, either through theft or security.

Random fun Nick Rowe links that don't really pertain:

Negative valued money (this is more cute than anything else)

Did inflation targeting destroy its own signal

Do the Greeks need Greek banks? (Probably?)
posted by PMdixon at 7:45 PM on February 24, 2015 [3 favorites]


What does it mean, in the "radical targets" link, to "target" labor compensation? Does that mean, reduce wages and compensation for everyone? The article keeps saying "stabilize" but I have no idea what that means in this context.

I will admit to being mostly at sea when it comes to economics jargon here. Inflation targeting, as per one link, seems to be about "deciding that x rate of inflation is ideal/allowable." But then it goes on to talk about NGDP allowing "errors" to carry forward. An error meaning, the percentage outside of that allowable range of inflation? And what does that mean, for the average checking account holder/401k account holder/wage-earning laborer?

And what does any of this have to do with "auditing the Fed" and Rand Paul?

It's an odd thing to be talking about high-level economics at the same time as we talk about the economic policies of a person like Rand Paul, who seems to understand these things even less than I do.
posted by emjaybee at 7:57 PM on February 24, 2015 [2 favorites]


... as we talk about the economic policies of a person like Rand Paul ...

Known to his closest friends as 'Kruger'?
posted by jamjam at 8:17 PM on February 24, 2015 [2 favorites]


What does it mean, in the "radical targets" link, to "target" labor compensation? Does that mean, reduce wages and compensation for everyone? The article keeps saying "stabilize" but I have no idea what that means in this context.

Most central banks operate under some broad mandate of "price stability," which they are either explicitly or implicitly given some leeway to interpret. Most macro theory assumes, and empirical evidence mostly confirms, that a central bank can choose a single nominal value about the economy (meaning dollar terms, so NOT number of cars/housing starts) and make it be whatever they want, up to some error. This is the target. Since the 90s, there has mostly been a consensus around inflation targeting, meaning that some preferred price index grows at some preferred rate. The central bank then sets interest rates and buys and sells assets to try to achieve that goal.

Wages are sticky (hard to adjust) downwards, which is why in a recession there are more layoffs than pay cuts. Sumner claims, and I think I tend to agree, that in practice this means that stabilizing (targeting, say, a 5% growth path) total nominal labor compensation would amount to stabilizing total hours worked (ie avoiding layoffs). What happens instead, presumably, is effective wage depreciation during recessions via inflation of non labor inputs. From a welfare perspective, this seems preferable to the status quo of mass long term employment. Since labor compensation is the biggest component of GDP, this is going to end up in a broadly similar place.

The difference between level targeting and rate targeting is that, assuming symmetric errors, the former lets me make good predictions about the cost of things in 2030 and the other doesn't. Because growth is multiplicative, errors (difference between target value and actual) grow, not cancel, so that the average distance from the straight line growth path grows as the square root of time. Level targeting lets people plan better over longer spans of time.

Secretly there's actually quite a bit to do with auditing the Fed. The Pauls are basically deflationists; they think monetary policy should involve zero or negative price level growth. However, price level includes the price of labor, so that is a transfer of future productivity away from the future workers and to current asset holders. All monetary policy is war between debtors and creditors.
posted by PMdixon at 8:23 PM on February 24, 2015 [17 favorites]


Thanks for this. I've been following Sumner for a while, and like his work a lot. My sense is that NGDP targeting is much more attractive to internet kooks like us than to central banks, though I don't know exactly why. Certainly its the case that as a nerd on the internet if I see a cool, halfway plausible idea I'll be tempted by it, and the massive, boring macro-econ required to understand why that cool idea wouldn't really work are beyond me.

This is precisely what I have in common with gold bugs and Rand Paul.

The other reason to worry about Rand Paul is because NGPD level targeting would require the Federal Reserve to commit to unlimited money creation, and Paul wants to rein those powers in. Now: it looks to me like a credible commitment to NGDP level targeting would mean that the Fed didn't have to do the things it's been doing like massive quantitative easing, and so actually the Fed's balance sheets would be healthier. But the tricky thing is that speculators looking to break an NGDP level target could always attack the Fed's independence through people like Paul, arguing (also plausibly) that the Feds actions involve an unlimited--i.e. potentially infinite--commitment. That means that the Fed's ability to make a credible commitment is limited by Congress's buy-in.

If it turns out that Sumner and Rowe are wrong about the Fed having all the tools it needs, then that too could potentially be pretty scary. "Looks like it would work to me" is hardly enough of a reason to steer a $17 trillion dollar economy onto the rocks.
posted by anotherpanacea at 8:37 PM on February 24, 2015


Ah God, I still have a lot of work to do tonight and it appears you're getting ready to take all comers in defense of this post. But still, I must strongly object to your comparison between inflation rate targeting per se and the gold standard (for those who don't know, the title of this post is a reference to William Jennings Bryan's Cross of Gold speech) - for one thing any low-but-positive inflation target is qualitatively different from the gold standard which is at least anti-inflation and even deflationary in many circumstances.

For another thing your post completely fails to mention the fed's dual mandate, which not only targets a low-but-positive level of inflation, but also simultaneously targets a low unemployment rate. Because that is the real drawback of too-low inflation, is money hoarding, economic slowdown, and the accompanying unemployment and hedonic misery. So one could view the dual mandate as a goal to minimize unemployment while avoiding runaway inflation; viewed that way it's not so unreasonable after all.
posted by Joey Buttafoucault at 8:42 PM on February 24, 2015 [7 favorites]


hardly enough of a reason to steer a $17 trillion dollar economy onto the rocks.

What was the reason in 2008?

Seriously, the language related to econ discussions, not the tech talk, but the just-so morality tales, is beyond reform, and often cloyingly condescending in its transparent blame avoidance and shifting.
posted by dglynn at 8:48 PM on February 24, 2015


Excellent post.
posted by four panels at 8:48 PM on February 24, 2015


It's almost like monetary policy isn't enough, and you'd need fiscal policy too. Nah.
posted by wuwei at 9:06 PM on February 24, 2015 [7 favorites]


What was the reason in 2008?

My sense is that the Federal Reserve didn't know what to do: perhaps NGPD targets would have worked (Bernanke had published theoretical work in this direction) but the FOMC didn't (and don't) have their ducks in a row for it. Bernanke called for and expected Congress to respond, repeatedly telling them that tight federal fiscal policy (that is, too low spending, not running high enough deficits) was exacerbating the crisis and slowing the recovery.
posted by anotherpanacea at 9:24 PM on February 24, 2015


Ok I'll be that guy. So...I don't understand the articles very much and don't even understand the first paragraph of the FPP so if someone could explain this like I'm five I'd be your friend forever okthnxbye.
posted by zardoz at 9:37 PM on February 24, 2015 [7 favorites]


Great post.
posted by So You're Saying These Are Pants? at 10:06 PM on February 24, 2015


a person like Rand Paul

I think it's possible to be too generous.
posted by 1adam12 at 10:19 PM on February 24, 2015 [2 favorites]


"anyone faced with negative interest rate could plausibly hold the money in a mattress at a better return."
This is why there is a shortage with Swiss Franc notes with high face value.
posted by yoyo_nyc at 10:20 PM on February 24, 2015 [1 favorite]


I have posted it before and I post it again. The problems - diminishing returns, dependence of energy and economic growth, limit of energy resources, and exponential laws - make solutions provided by a central bank look like a remote idea.

"Anyone who believes exponential growth can go on forever in a finite world is either a madman or an economist."


Limits to Growth–At our doorstep, but not recognized
http://www.resilience.org/stories/2014-02-12/limits-to-growth-at-our-doorstep-but-not-recognized

Wealth And Energy Consumption Are Inseparable
http://www.declineoftheempire.com/2012/01/wealth-and-energy-consumption-are-inseparable.html

"There is No Steady State Economy (except at a very basic level)"
http://ourfiniteworld.com/2011/02/21/there-is-no-steady-state-economy-except-at-a-very-basic-level/

Galactic-Scale Energy
http://physics.ucsd.edu/do-the-math/2011/07/galactic-scale-energy/
posted by yoyo_nyc at 10:38 PM on February 24, 2015


Hi Yoyo-nyc the hypertext link procedure in metafilter always gets me as will, I wonder why they don't just let you write in HTML like Wikipedia ? Will read your links with interest.
posted by Narrative_Historian at 10:52 PM on February 24, 2015


[Linkified yoyo_nyc's urls ]
posted by taz at 12:19 AM on February 25, 2015 [1 favorite]


But still, I must strongly object to your comparison between inflation rate targeting per seand the gold standard (for those who don't know, the title of this post is a reference to William Jennings Bryan's Cross of Gold speech) - for one thing any low-but-positive inflation target is qualitatively different from the gold standard which is at least anti-inflation and even deflationary in many circumstances. 

Gold standard creates terrible welfare outcomes, and inflation targeting is vastly superior.
In practice though, there's a significant contingent of bank governors that would be quite happy to implement opportunistic disinflation. See the Bundesbank crowd, as well as Fed's Plosser and Bullard. Level targeting makes them explicitly argue for that, not just sneakily hide it in the residuals.

For another thing your post completely fails to mention the fed's dual mandate, which not only targets a low-but-positive level of inflation, but also simultaneouslytargets a low unemployment rate.

Yeah, I was trying to keep a more global perspective. Right at the moment tho, dual mandate is giving them an excuse to talk up summer liftoff without any real rise in inflation expectations. I think the dual mandate, in practice, just gives the members more room to reason backward from their preferred course of action.

It's almost like monetary policy isn't enough, and you'd need fiscal policy too. Nah.

This should be a really puzzling and unintuitive result, though, at least as regards unemployment. Clearly the state fisc can be spent wisely or poorly, and at too high or low a level, but it seems really weird to say that unless the state employs them, large numbers of people will sit at home even though they would rather exchange labor for stuff. It makes much more sense to me that we have too many commitments attached to stuff already, so we can't give it to people in exchange for labor. In that case, we just need to depreciate the value of those existing commitments so that we can give some of it to people in exchange for labor. This is what monetary policy does.

This is why there is a shortage with Swiss Franc notes with high face value.

I thought this was a long standing phenomenon driven by money laundering, no?
posted by PMdixon at 4:53 AM on February 25, 2015 [1 favorite]


It makes much more sense to me that we have too many commitments attached to stuff already, so we can't give it to people in exchange for labor.

PMdixon, could you clarify what you mean by this? I'm not particularly well-versed in economics, so I may be misreading you; are you saying that inflation may be positive here because it reduces the real value of savings, thus making it more attractive to use that money to employ people? If so, is there good evidence that the vessels most capital is held in are actually fixed to nominal value, because that seems suspect to me.
posted by thegears at 5:17 AM on February 25, 2015


are you saying that inflation may be positive here because it reduces the real value of savings, thus making it more attractive to use that money to employ people? If so, is there good evidence that the vessels most capital is held in are actually fixed to nominal value, because that seems suspect to me.

I think the first is a standard assumption in this kind of work: it's the paradox of thrift idea that drives most progressive macro theory. But I do wonder about the second: clearly a lot of investments are relatively inflation-neutral. The stock market is going to be relatively indifferent to nominal GDP (though not exchange rates) and commodities, real estate, etc. will all respond. The way I always hear the story about NGDP is that this is a way to respond to sticky wages: make people take pay cuts rather than lay people off, so productivity doesn't fall when weird things happen with money for complicated speculative reasons. But this won't, at all, effect the distribution of returns between capital and labor: it'll force the returns to labor down to make up for shortfalls in capital returns.

Now that starts to sound a lot less progressive. Sure: labor will do much worse during a recession or depression than it would under an inflation regime. But why not force capital to take the cut? Why not go after savers who aren't getting returns in nominal dollars? This is where I think some of the larger problems with Fed policy and helicopter drops come through.

Say you find a way to depreciate savings: is the response always and only to redirect savings to riskier labor-based investments? That is: if someone told you that your retirement savings would depreciate slowly but safely (as inflation does), would you really invest it in venture capital and hedge funds that would put lots of folks to work, knowing that businesses fail and workers don't care about your profit margin? A lot of people would rather have their money in safe assets than in growing assets, and for good reason: they need the money more than they need the returns. That's why I still have a savings account even though I don't make any interest at all on it.

Now: most money is not like my savings account, held in a close-to-cash vessel. But that doesn't mean chasing yields to higher risk/return ratios is always a good idea. The problem with risk is that sometimes you lose! You don't want your pension invested in domestic manufacturing IPOs or whatever industrial policy will put people back to work: you want it invested safely so you can afford to get old.

So yeah: maybe NGPD level targeting is a good idea. But you can see how the savers (and their ally, Rand Paul) have an effective argument that poor and middle-class workers should oppose macro policy that gives money to banks so that all workers will have to take an unnegotiated pay cut.
posted by anotherpanacea at 6:05 AM on February 25, 2015 [1 favorite]


Wuwei: It's almost like monetary policy isn't enough, and you'd need fiscal policy too. Nah.

FWIW, this is a point that both Yellen and Bernanke have repeatedly made in their testimony to Congress and elsewhere - in the somewhat deferential, cagey way that they sort of have to, but basically they've said "we can't do everything and these games y'all play with shutdowns are not helping anything."

PMDixon: This should be a really puzzling and unintuitive result, though, at least as regards unemployment. Clearly the state fisc can be spent wisely or poorly, and at too high or low a level, but it seems really weird to say that unless the state employs them, large numbers of people will sit at home even though they would rather exchange labor for stuff.

I guess I'm puzzled about whether (and why) you're puzzled. Fiscal policy affects aggregate demand, either directly (by purchasing stuff and hiring people) or indirectly (giving loans or tax breaks to houesholds and firms to do the same). If there's a shortfall of aggregate demand for whatever reason, the government can make some of that up. But maybe I'm not understanding your point.

If so, is there good evidence that the vessels most capital is held in are actually fixed to nominal value, because that seems suspect to me.

Well, most bonds are nominal, not real (the government issues TIPS, which are inflation-indexed, but most Treasury bonds for instance are nominal). They're a promise to pay back a dollar tomorrow, not a promise of a real return of some percentage. To the extent that the marginal investor is indifferent between putting money in a nominal bond or investing it somewhere else, targeting higher inflation would reduce the real return on the nominal bond and push that marginal investor to invest in something else, hopefully something that puts people to work. That's the idea, anyway.

To add to the collection of links: Larry Ball, who sort of hops between the Fed and Johns Hopkins (like a lot of the JHU econ faculty) published this a few years back: The Case for Four Percent Inflation. Ball is a very clear writer and I think it should be more or less accessible to the lay person unlike a lot of academic economics research.
posted by dismas at 6:48 AM on February 25, 2015 [1 favorite]


(also the Fed targets PCE inflation, not CPI!)
posted by dismas at 7:14 AM on February 25, 2015


Fiscal policy affects aggregate demand, either directly (by purchasing stuff and hiring people) or indirectly (giving loans or tax breaks to households and firms to do the same)

But the Fed moves last, so anything Congress does will just reduce the Feds' willingness to boost inflation, and aggregate demand will fall back to where it was previously.

Notice that right now we have a 2% inflation target, but it's not a level target. That means we fail to hit the target every quarter, and fall further and further behind. Yet if we went above the target: boom! The Fed would act to push inflation down below it immediately. So it doesn't matter what Congress does on the fiscal side: inflation is capped at 2%.
posted by anotherpanacea at 7:15 AM on February 25, 2015 [1 favorite]


But the Fed moves last, so anything Congress does will just reduce the Feds' willingness to boost inflation, and aggregate demand will fall back to where it was previously.

Right, conditional on the Fed viewing the 2% target as a ceiling rather than a midpoint they should be hitting on average over the business cycle and also being hawkish as hell. Which may be the correct model. It seems to me that part of the point of inflation targets is that they should be symmetric; many countries have a band around a central target and that's how they talk about policy. But the US was only implicitly targeting 2% inflation until a year ago and has been only sort of slowly embracing the kind of transparency that has been more common in other inflation-targeting countries. At least in the US, central bankers used to think mystique was important, and while I think that's changing over time, it's not changing overnight. Point is: I don't know what Janet Yellen actually thinks. I'm not sure if the FOMC would be so quick to cut off inflation in light of some fiscal stimulus. In any case, it's sort of academic since Congress doesn't seem interested anyway.

I'm not sure, but I think (some) of the policymakers who are eager to raise rates aren't doing so because they see 2.5% inflation as so much worse than 1.5% inflation, but because they are worried about asset bubbles. I suspect that's part of the eagerness to get rates away from the ZLB.

I'm not against price level targeting per se. It makes sense as a form of forward guidance and works well in models. But there's a political and institutional dimension to all of this that doesn't get captured in models, which is that literally no country has tried it and central bankers, especially at the Fed, are nervous about experimenting. (and Congress is also apparently pretty nervous about it). Formal inflation targeting was something that was tried and studied for more than two decades before the Fed formally said "we are targeting 2% PCE inflation" so I wouldn't hold my breath waiting for price level or NGDP targeting in the US. Maybe if New Zealand tries it first.
posted by dismas at 7:46 AM on February 25, 2015


I agree with everything you say here. But I'm not sure how it responds to the point that fiscal policy can't trump Fed policy.
posted by anotherpanacea at 9:42 AM on February 25, 2015


It's amazing how if you don't pay people more and you don't pay people at all (by only investing in work proxies like houses, instead of workers themselves) then people don't have money.

But I guess we can't talk about that.
posted by effugas at 7:23 PM on February 25, 2015 [1 favorite]


Anotherpanacea: I guess my point was a long-winded way of agreeing with you. But I don't know if the Fed will choke off growth just because it can.
posted by dismas at 8:59 AM on February 26, 2015


Thanks! That makes a lot more sense.

I think my concern is: the Fed will choke off (fiscal policy-led) growth because it has the wrong idea about how macro policy should work.
posted by anotherpanacea at 10:16 AM on February 26, 2015


« Older As a society, we have now passed peak bae.   |   Great Ideas (With Wheels) Newer »


This thread has been archived and is closed to new comments