You are currently browsing the tag archive for the ‘Scott Sumner’ tag.

quickly...before i die...take my magical sword...and set forward guidance for the kingdom...yes, i know interest rates are at zero, but i know you have the power, dammit...quickly, before i am succeeded by republicans

So just around this time last year I wrote the following, which I will quote in its entirety because it’s awesome:

Let’s model a kingdom. In this model, the kingdom is a closed economy, and (very importantly) it is “well-normed” – it has strong norms relating to governance and society that tend to be widely honored and respected.

This kingdom is governed by two individuals: the king, and the wizard. Most formal power, as well as the titles of head of state and head of government, is vested with the king. The king has formally unlimited powers to tax and spend, raise armies, and adjudicate disputes, but in practice is limited by norms, sense of duty (symbolized in a sworn oath to serve in the interest of the whole kingdom and its subjects), and the patience of subjects; therefore, the king tends to maintain inherited intuitions to which their power has been delegated, like courts and military bureaucracy. The crown is hereditary – the first-born child of the king (this is a gender-progressive kingdom) inherits the crown, and in the past, though there have been occasional hiccups, most transfers of power have been peaceful and orderly.

The king must retain a wizard, who is charged in vague terms with independently securing the safety, security, and prosperity of the kingdom. The wizard bears a hat that grants them vast yet mysterious magical powers. Unlike the crown, which is symbolic, the wizard’s hat is in fact where the magical powers are vested, and is not hereditary. When the existing wizard dies, the king selects the next wizard, who receives a lifetime appointment. Extremely strong norms dictate that the king select whomever is widely acclaimed the wisest scholar in the kingdom, regardless of their personal feelings towards that individual or inclination to select an ally as wizard. Often the wizard will survive the king.

As stated above, the wizard has vast powers, but they are mysterious and to some extent ill-defined. There is no user manual for the wizard’s hat, and often throughout history wizards have surprised themselves with the consequences of exercising their powers. Therefore, norms have developed that the wizards exhibit strong restraint in exercising their powers, even in times of emergency. Extremely strong norms have also developed against the king making formal or open requests of the wizard, as well as against the wizard interfering in the quotidian or terrestrial business of the king. In the past, there have been some violations of this norm in both direction, but for the most part it tends to persist. Consequentially, the wizard tends to be reclusive, speak carefully and opaquely, and avoid commitments to use their powers. There is much dispute among the subjects of the kingdom to exactly what the wizard is doing or could be doing, and when the wizard ought to exercise their powers.

Your assignment: model the governance and economy of this kingdom.

Obviously this is kind of an allegory. And I’m going to just blow it wide open by saying the king is the fiscal authority and the wizard is the monetary authority. And while it doesn’t quite match up with how the modern world works, I think it has some implications.

Now there is great sturm und drang in the econoblogobloid about whether Keynesianism or Monetarism is “right” and what “monetary offset” is and I just don’t get it at all because to me it all makes perfect sense together.

My theory is this – monetary policy dominates fiscal policy, but doesn’t always, or even usually, exercise that dominance to determine point outcomes. Instead, it usually determines range outcomes, and within those ranges fiscal policy exerts influence.

So for example if the monetary authority declares “the rule is that inflation will never exceed 2%” that means a whole range of outcomes are possible; just not those where inflation exceeds 2%. So fiscal policy can, by being expansionary or contractionary, have a substantial influence on inflation, unemployment, and GDP, just within the range of outcomes that are possible with below-2% inflation.

If the monetary authority declares “we will print $100 billion and buy bonds with it every month until unemployment hits 3%” that means a whole range of outcomes are possible, but all those outcomes must account for an additional $100 billion of “high-powered money” in the economy every month. So fiscal policy can, by being expansionary or contractionary, have a substantial influence on inflation, unemployment, and GDP, just within the range of outcomes that are possible with regular large monetary infusions.

If the monetary authority declares “we are targeting the path of NGDP and we will never let it deviate ever ever no matter what ever” that means a whole range of outcomes are possible; just not those where NGDP deviates from its price path. So fiscal policy can, by being expansionary or contractionary, have a substantial influence on inflation, unemployment, and real GDP, just within the range of outcomes that are possible with rock-solid 5% NGDP growth.

If the monetary authority declares a whole bunch of stuff, some of which is concrete, some of which is fuzzy, and some of which is muddled, well…many outcomes become possible. But many are not!

So the story of 2013 is not one where monetarism was right because the Fed got exactly what it wanted; it’s one where fiscal contraction couldn’t generate outcomes outside the bounds set by the Fed. But it does mean that outcomes could have been better had fiscal policy been less contractionary. Counterfactuals are hard, but that’s a reasonable one.

Now, I’m a believer in the power of monetary policy, so I think it would be great if the Fed set an NGDP target and committed to not only a growth target but a path target and that they were going to overcompensate to get us back to the pre-recessionary path. I even think an optimal NGDP target is something more like 7%, not Scott Sumner’s girly-man 5%. But what I don’t believe is that absent such direct, confident, dominating policy guidance from the Fed, that Fed policy still generates pinpoint outcomes or even tight ranges of outcomes. The economy improved in 2013. It could, and should, have improved more, and that’s on both the Fed and the Congress.

Now, as much as I really, really don’t like Arnold Kling, there are also some PSST dynamics out there that constrain the power of monetary policy, which is why the auto bailout was a really good idea, but barring massive systems collapse I think monetary policy can do pretty much all the heavy lifting of macroeconomic policy but that conditional on monetary policy fiscal policy can have some room for maneuver.

Note that this is pretty much 100% what both monetarists and Keynesians think when we’re not at the ZLB; all I’m saying is that it is also true at the ZLB, and that monetary policy really doesn’t change that much. We just think it does. Which is part of the problem, because money illusion is a real and serious thing, which goes back to why our NGDP target should be higher. In general countries that are better at avoiding recessions have a somewhat higher tolerance for inflation than we do.

Inline image 1

Responsible prudent savers. Totally ants.

Scott Sumner, lingua in bucca, swapped Formica for granite countertops and writes:

PPPPS. Yes, granite is very durable, which makes it investment . . .

. . . and of course saving too!

This, of course, goes to one of my hobbyhorse points – the exceptionally fuzzy line between "consuming" and "saving." The better way to view the world, IMHO, is one in which, beginning at some arbitrary point, we apply our time and existing stuff, through the media of institutions, to make more stuff, and that stuff varies in function, quality, durability, etc.

This also reminds me of a tangential point – young people, on average, probably do not travel anywhere near the optimal level. Travel can be expensive, but when you are young there are two factors that mitigate strongly in favor of travel. Firstly, you are most able to enjoy it. Let me tell you right now that future 50-year-old me would have had a way tougher time enjoying Mehrangarh than the present 26-year-old me. Secondly, travel is a perpetuity, and while most of their utitlity is illiquid, that’s as much an advantage as disadvantage – nobody can expropriate, steal, damage, or tax it.

This is much the same as education, which is why education debt is a bad idea, but while a college degree can be so expensive most people simply cannot fund it out of current consumption, an unforgettable month-long jaunt through India for two can cost less than $5000 including airfare, adequate lodging, and all consumption. Which is a lot, but if you have a young cohabitating pair with no kids making at least a combined $70-80k, as long as their rent isn’t too expensive and neither is drowning in debt it’s perfectly possible to save adequately for such an adventure over less than a year. And you could make a very strong argument that traveling while young is just as much "investment" (in something that produces a lifetime-long flow of happy memories and increased knowledge and wisdom, as well as fun stories to share with others) as "consumption."

So Ashok and I sparred a bit on Twitter re: the meaning and effect of taxation and spending (and probably pestered the heck out of James Pethokoukis and Joe Weisenthal in the process). I’m not sure how to embed Twitter conversations (if anyone knows how, I’m all ears), but the long-and-short of it is that the actualities of taxes and spending are weirdly different from the optics.

The trick is to remember that every policy change is a change from some baseline. So, from whatever the baseline currently is, there is no fundamental or economic difference between:

1) Cutting taxes by X on some activity, and
2) Spending X subsidizing that activity

assuming that they are both funded identically (though identical tax hikes, spending cuts, or debt incursions).

Now, in practice, there will be differences. Scott Sumner’s thought experiment about the society that taxes 100% of GDP by taxing 100% of income then writing welfare checks equal to taxed income demonstrates that, since we would expect that society really would look different than the one that taxed nothing at all (if only because such a program would have some overhead). But those differences would be based in behavioral economics, not classical or neoclassical economics.

And the same in real-world examples. There would definitely be differences between these two alternative scenarios:

1) A 2% payroll tax cut (debt-funded).
2) A check mailed to every American for the exact same amount (debt-funded).

But those differences would be instutional, not economics (the check-cashing industry, for example, would obviously prefer the second policy to the first). But there’s no reaosn to think they would "crowd out" (or for that matter, "crowd in") different activities.

The real point is, as Matt Yglesias says, the tax share of GDP is a very poor to think about the “size of government.”

As you might have noticed, in my initial minimum wage-related salvo against Don Boudreaux, I used a somewhat unusual graph:

 

fedminwagepopfull

Needless to say, it did not come from nowhere. In fact, it came from a lot of thinking and a decent amount of work, partially inspired by, curiously enough, Don Bourdreaux’s “Catalog” project, as well as by Scott Sumner’s work in general as well as his frustration with conceptual understandings of inflation. Rather than beat around the bush any more, though, I’m going to tell you what I’ve created, then explain it and defend it, then provide examples:

I have created a new econometric index. It consists of dividing the nominal price or value of something across time by nominal GDP per capita and representing it as a percentage. This divides out the currency unit, and so is an index measure of quantity, not price. I am calling the index, as well as the index unit, the “Percappy” (plural “percappies”).

Now – why?

It is theoretically and empirically sound. The Percappy tells us what share NGDP-per-capita could obtain a single unit of a given thing at a given time. It requires only three measurements – GDP, population, and the nominal price/value at a given moment in time – none of which need to be additionally weighed, balanced, or adjusted in any way. It does not need to be controlled for inflation over time, since inflation would affect equally both numerator and denominator. It does not need to be controlled for PPP or exchange rates across national borders as long as correct local prices are used.

It tells us something new and useful. The Percappy tells us, essentially, what share of per-capita national income it would take to acquire a single unit of some good, service, or financial product. This allows us to easily compare living standards (again, controlled perfectly for inflation) across time as well as across national borders, even diagonally (for example, you could compare other countries’ present consumption frontiers to the United States’ current consumption frontier).

It is analytically powerful. With a minimum of computation and weighting, using only easily-accessible publicly-available data, many different things can be compared – the prices of goods, services, commodities, financial products, and even more (as we shall soon see) can be compared while controlling for any number of factors that tend to complicate such comparative analyses.

It is simple to understand. Unlike inflation, which is conceptually challenging and whose definition is not universally agreed on, both the idea and the process behind the Percappy is very simple and easily understood – it answers the question
“how much more or less of some thing can be bought in different times and places?” by dividing prices by per-capita national income.

What it doesn’t tell us (and other limitations). It tells us very little about inflation or exchange rate fluctuations unless a large amount of data is computed and compared. The Percappy index is powerful because it easily divides away these factors to look at “objective” standards of living; but the trade-off is that it can tell us very little about the monetary, financial, fiscal, and other policy-related factors that may drive those changes. It also tells us nothing about quality. Also, historical prices can be hard to come by.

Let’s look at some examples. Firstly, a simple one – the price of a McDonald’s hamburger over time (note: log scale; all prices are per-unit and all measurements in %; the formula is 100P/[GDP/POP])

chart_1

Unsurprisingly, we are richer in terms of McDonald’s burgers than we used to be. What about some other delicious items?

chart_1 (1) chart_1 (3) chart_1 (2)

 

Interestingly, while Oreos and Hershey’s Bars seem to both be trending downwards (albeit not as sharply as McDonald’s burgers) Cornflakes seem to have bottomed out in the early-to-mid ’90s and have since climbed back to 1960s levels.

What about some other consumer prices people care about?

chart_1 (6)

FRED has only been keeping track since 1990, but clearly we’re not doing well on that score.

But the true power of this is to tabulate more than just common consumer prices. What about the value of the S&P 500?

chart_1 (7)

What about the average dividend of the S&P 500?

chart_1 (8)

And, to be fair to the goldbugs, what about gold?

chart_1 (5)

Hopefully even these examples, as simple and unprocessed (and, thanks to Google Spreadsheets, amateurishly presented) as they are, show some of the power of this index. Essentially, it can “crunch” almost anything, including (at least in theory), other indices (I’m thinking about feeding it the Case-Shiller index).

I am hoping other people think this project has some value – if you do, you can help! Here’s how (bleg time):

1) Data! Find historical prices is harder than you might think! If you can send me time-series price data of just about anything I would be quite grateful – or, you can do it yourself! It’s pretty easy. This could be easily crowdsourced.

2) Mathematica. I got this software and I love it, but haven’t quite figured it out yet. Especially the “get large volumes of data into it so you can crunch with with panace” part. Any Mathematica aficionados out there want to send me tips?

Here’s hoping people find this useful, helpful, interesting, worth picking up on or helping out with, or some combination thereof.

Join 3,848 other followers

Not Even Past

Tweetin’

RSS Tumblin’

  • An error has occurred; the feed is probably down. Try again later.