Confusion and Context

Tracking down Y = C + I + G + NX, I came across an old Steve Roth post: Why Does Y Equal Real GDP?. Roth and correspondents get into a lot of discussion that I want to skirt. My motivating question was: Does the Y represent real GDP or nominal GDP (or does it perhaps represent real output, which is not really the same as any GDP)? Short answer: It seems to be ambiguous at best. That'll do for now.

While I was reading Roth's post and some of the comments I got distracted by a different thought. Let me take you through it.

Roth opens by looking at two macroeconomic identities from a Nick Rowe post. The first of these is the National Income Identity:

Y=C+I+G+NX
C is consumer spending. I is business investment spending. G is government spending. And NX is net exports.

Later in the post, Roth quotes Saturos:

Matt Yglesias (http://www.slate.com/blogs/moneybox/2012/05/13/fun_with_accounting_identities.html) has a new post in which he takes Scott Sumner’s version: MV = C + I + G + NX. That might be the best approach of all – it shows you that all the changes in “income accounting” variables that get reported on the news must all be manifestations of fluctuations in the overall volume of spending, MV.

(Yglesias says Y represents "real output" but doesn't say whether he distinguishes "output" from GDP. The difference? Output is output; GDP is a measurement.)

It stops me cold when Saturos refers to "the overall volume of spending, MV."

Now... Nobody's identifying any FRED series here, so everything is much like the meaning of Y: Ambiguous at best.

Usually, though, M is some quantity of money, maybe FRED's M1 or M2, and V is the value you get when you divide nominal GDP by the M you are using. So MV (or M multiplied by V) is really

The M's cancel out, and you're left with nominal GDP. (This does not answer Steve Roth's question; rather, it seems to be what gave rise to it.)

Back to Saturos. Saturos says MV is the overall volume of spending. And we have just seen that MV is equal to nominal GDP. Therefore, Saturos is saying nominal GDP is the overall volume of spending. That's what stopped me cold.

Nominal GDP is NOT the overall volume of spending. It is the overall volume of final spending, with none of the "intermediate" spending. Nominal GDP is just the cream on top, not the whole bottle of unhomogenized milk. Got it?


Farther down, in the comments, Steve Roth replied to Andrew. Roth wrote:

GDP = Total dollars spent = MV
Again, this is not correct. GDP is not total spending. GDP is final spending only.

Total spending includes both final and "intermediate" spending. GDP includes the final spending, but excludes the much larger intermediate spending. Why count only final spending? Because final spending is the cream, that's why.

I'm not picking on Roth and Saturos. Everybody makes this mistake. I've even seen it in the New York Times. But here's the thing: I'm in the middle of trying to define a measure to use as context, so that I don't have to use GDP all the time. I want to move from using GDP -- the sum of final spending -- to using a much broader measure. For my "context" variable, I want to use the sum of total spending, not the sum of final spending.

People who think GDP is a measure of total spending will never understand.

An Inconvenient Context

GDP is a measure of final spending only. Not all spending. For example, GDP excludes things that can be taken as corporate tax deductions. Things like the "cost of goods sold". If you buy stuff in order to sell it, that is not "final" spending.

A carmaker buys tires so that the cars he sells can be driven off the lot. Maybe he buys 4000 tires and sells 1000 cars. The cost of the tires is included in the price of the cars, obviously. Let's say all 1000 cars are included in GDP. So now GDP already includes the cost of the 4000 tires the carmaker bought. So you don't want to add the separate purchase of 4000 tires to GDP, because then GDP would include 8000 tires. That would be double-counting: We know there were only 4000 tires involved. So economists say the purchase of the tires was a preliminary (they say "intermediate"; I say "preliminary") transaction, not a "final" transaction. And they do not add the preliminary purchase to GDP.

In our scenario, the manufacture and sale of those tires did happen. Yes, the value of all that work is included in GDP when the car sales are included in GDP. However, there are many transactions involved in the production and sale of the tires that are not separately added to GDP. That is because GDP is a measure of what we have produced. It is not a measure of the economic activity that was required for production.

GDP is a measure of what we have produced. It is not a measure of the economic activity that was required for production.

If you listen for it, you will often hear people say GDP is the size of our economy. You will hear them say GDP is a measure of all economic activity. The latter claim is most certainly incorrect. GDP is a measure of what we have produced. It is by no means a measure of all the economic activity required by the production process.

GDP is the cream floating on top, in a glass bottle of milk that's not homogenized... if you're old enough or bold enough ever to have seen such a thing.

Source: Udder Farm Milk Cream & Cheese Company
via FlavourCrusader

For the sake of argument, so to speak, I am throwing together data to create a number to use as a context number in place of GDP. I don't want to figure just the cream for context. I want to figure everything in the bottle.


In order to describe where I'm going, let me begin by describing where I start. I start with GDP. GDP is "final" spending. The cream on the top. It includes most or all of consumer spending, most or all of business investment spending, and most or all of government spending. (And, yes, net exports.)

Wikipedia points out that the government component does not include transfer payments, and the business investment component does not include purchases of financial products. The former (I think) are counted in GDP as part of the consumer spending component; the latter (Wikipedia says) are saving, not investment.

But let me put it the way I put it the other day:

Back in the late 1970s when I got my three credits in macro, they said GDP equals consumer spending plus business investment plus government spending plus net exports.

See it? GDP includes business investment, but not all of business spending.

Almost missed that, didn't you, in the flurry of details about transfers and financial products. If you take all of business spending and put it in two piles, the pile of business investment spending would amount to about 18% of the cream floating at the top of the milk bottle. The other pile of business spending, the pile that's not investment, completely fills the bottle below the cream.

Got it now?


I want to take final spending -- GDP, the cream on the top -- and add to it the non-final spending that represents actual economic activity, but is written off as business expenses, filling the milk bottle in the process.

I have to do this by poke-and-hope, as I've never seen it done by someone who might actually know what they're doing. I'm thinking businesses write off their non-final expenses, but they also write off (or at least they depreciate) their final expenses. So I think if I look at total business tax deductions I will get both final and non-final spending all in one number. I can live with that.

But the final part of that number is already included in GDP. If I take GDP and add total business income tax deductions (as an estimate of total business spending) then I will be double-counting final business spending. So I need to start with GDP, subtract out the business investment component, and then add total business tax deductions. And this will give me a number that represents the dollar value of the economic activity that was required for the production of GDP.

In words it sounds complicated, but the arithmetic is simple:

GDP - I + Business Income Tax Deductions
where I is the business investment component of GDP.

In order to figure the numbers for economic activity required for the production of GDP, I need the numbers for GDP. That's simple enough; I can get them from FRED. I need the numbers for business investment. For this I can use Gross Private Domestic Investment, also available from FRED. The third thing I need is the series of numbers for business income tax deductions. For this I can turn to the Historical Statistics for data through 1970.

After 1970 I can use various editions of the Statistical Abstract. If I find the right table in the Abstract I can pull out numbers for four or five consecutive years. Then I need a later edition of the Abstract so I can grab another batch of numbers. I really want a one- or two-year overlap so I can see that there is continuity in the data. I sure don't want my graphs to show data revisions that I don't know about.

Gathering these numbers is the "inconvenient" thing, in case you were wondering.

Actually, I had this all done in time for the post of 26 June. But at the last minute I noticed a problem. I was using numbers for corporate business instead of all U.S. business. Hey, you know, corporations dominate; so if I only count corps I get the dominant effect.

Yeah, but if I only count corps, I may misinterpret the growth of the corporate economy as the growth of business. I think non-corporate business faded as corporates rose, so that the growth of business spending would be less than the growth of corporate spending, and less than my first series of numbers were showing. But I'm not sure, and I can't be sure until I get a second batch of numbers completed, numbers that include proprietorships and partnerships as well as corporate businesses.

That's what I'm doing now.


Man! That takes hours.

I got numbers for 1959 thru 2008. That's decent.

Okay, here's what I got:

Graph #1: Federal Spending relative to Final Spending (blue)
Federal Spending relative to Final + NonFinal Spending (red)

I'm just gonna leave you with that. I have too much time in this already.

Reflecting on Context

As the recent posts have been considering context when evaluating economic data, and as my Saturday morning post is not yet (Friday evening) writ, it occurs to me to "reblog" an older post on the topic:

Federal Debt Held By Federal Reserve Banks

F, D, H, B, F, R, B. It's easy to remember: Just look at the title of this post. FDHBFRB is the name of a FRED dataset.

There is another dataset too, with the same letters, plus an N at the end. N is for "new", I suppose, suggesting that the original set of capital letters represents a discontinued series. It does.

Here's what I looked at first:

Graph #1: Discontinued (blue) and New (red) Fed Holdings of Federal Debt, Relative to GDP

Fed Holdings (of Federal debt) relative to GDP. What's wrong with this picture?

Debt went up, that's what's wrong. Debt went up *way* more than GDP since the 1950s. GDP is a small denominator. GDP makes Federal Debt Held By Federal Reserve Banks look bigger than it is.

Here's a better measure:

Graph #2: Discontinued (blue) and New (red) Fed Holdings of Federal Debt, Relative to TCMDO

Fed Holdings relative to Total Credit Market Debt Owed.

By the standard of the 1950s and '60s, Fed Holdings since the 1990s should have been twice as high as they were. By that standard, Fed Holdings are low yet today. Even with that big spike there at the end.

Fed holding should be twice as high, or Total Credit Market Debt Owed should be half what it was since the 1990s. Or some combination of the two.

This is the Arthurian policy recommendation: more Fed holdings, and less credit market debt. Why? Because you can use credit for just about everything these days. But you can't use credit to pay off debt.

Federal Reserve holdings of Federal government debt is a measure of how much money the monetary authority has put into the economy. What should we compare that to? GDP? Why? Because we always compare everything to GDP? I need a better reason.

I compare those holdings to the amount of debt that the banking system has generated from that money. This is not a context at random. It is the most relevant context possible, from my point of view.

A Convenient Context

FRED Blog asks How big is the federal government? They show a graph of "Federal Government Current Expenditures" relative to GDP:

Graph #1: Federal Spending relative to GDP (from the FRED Blog 23 June 2014)

They write:

The graph shows the huge government buildup during WWII, almost doubling in 1942, and its equally impressive contraction thereafter. Since then, the size of the government has fluctuated between 17 percent and 23 percent of GDP...

Not my idea of fluctuation. For fluctuation see Graph #4 in yesterday's post.


Their question doesn't sit right with me. The question "How big is the Federal government?" is what you would ask if you were arguing that government is too big. It's a popular question. I suppose that's why FRED Blog addresses it. But it's not a good question. Myself, I'd rather try to get people asking better questions.


"One way to determine the size of the U.S. Federal government is to look at its expenditures," they write. "Of course, population and the economy have grown, so it’s a good idea to use a ratio to measure expenditures. For example, you can divide expenditures by GDP, and this is exactly what is shown here."

Reminds me. A few months back I remarked

why they always show debt relative to GDP is beyond me.

Two people, whose knowledge and thoughtfulness I respect, replied. Geerussell wrote:

I assume it's because looking at anything relative to GDP is convenient shorthand to get a sense of scale...

Jazzbumpa seconded that remark:

geerussell is right. It provides context.

I think those were knee-jerk reactions. I mean, I know it provides context. And if you think of GDP as "the size of the economy", it would seem to provide a very good context indeed.

But what if our economy is having troubles? I mean, what if economic growth has been slowing for three or four decades? Is "the size of the economy" still a good context in the face of secular stagnation? I don't think so. A slowing economy, as context, makes other things that are similarly slow appear perfectly normal. A slowing economy, as context, makes things appear to be growing that are slowing, but slowing somewhat less quickly than GDP.

One could argue it is the slowing growth of GDP that makes Federal spending appear to be growing too fast. It would be an easy argument to make.

Anyway, the FRED Blog graph uses the one data series that economists always use as a basis of comparison: GDP.

Oh ye of little imagination!

The purchase of the pebble

Yesterday I showed that gross corporate income is just a little less than twice the size of GDP:

Graph #1: Gross income to U.S. corporations is almost twice the size of U.S. GDP.
The gap between the red and blue lines is all corporate tax deductions.

Gross corporate spending is only slightly less than gross corporate income.

Corporate spending is bigger than GDP? How can that be?? Ha! I've been waiting for the chance to tell this story.

I go for a walk one day and find a pretty pebble. I clean it up and put it in my pocket.

Later I show the pebble to my friend B. B likes it, and offers me a dollar for it. Done!

B paints a little picture on the pebble and sells it to C for $2.

C puts it in a pretty box and sells it to Big D Stores for $3.

Customer E buys it for $4 and gives it to her boyfriend.


The "final spending" in that little scenario is the $4 that Customer E pays for the pebble. It's "final" because E is the consumer of the product. All the spending that came before was "intermediate" or preliminary spending -- the $1 and the $2 and the $3. There was $6 of preliminary spending and $4 of final spending, and a total of $10 in spending.

Why is "final spending" important? Look at the income generated in the scenario: I made a dollar, and B made a dollar, and C made a dollar, and Big D made a dollar. Four dollars of income was generated by the transactions. That's the same as the four dollars of final spending, Customer E's purchase of the pebble. That's why it's important.

But there was also $6 of preliminary spending in the scenario -- half again as much as the final $4 spending. The preliminary spending is much more than the final spending. That sort of thing is similar to what happens with corporate spending: the $6 is a tax deduction, and the $4 is taxable income. The numbers are different, of course. But the concept is the same.

I hope this puts your mind at ease.


Back in the late 1970s when I got my three credits in macro, they said GDP equals consumer spending plus business investment plus government spending plus net exports.

See it? GDP includes business investment, but not all of business spending.

Back in the late 1970s, they said the business investment part of GDP is called Gross Private Domestic Investment, GPDI. I don't know why I remember that, but I do. Anyway, if you take GDP and subtract out GPDI, you've taken the business portion out of GDP.

They told us (again, I don't know why I remember this) business investment is about 17% of GDP. That was a long time ago, of course; it could have changed by now. Here's what FRED has on it:

Graph #2: Gross Private Domestic Investment as a Percent of GDP

About 17%. A little higher since the mid-1970s. Call it 18%.

Anyway, we can subtract this business investment spending out of GDP. And then we can add the total corporate deductions of U.S. corporations. Businesses get a tax deduction for pretty much all of their spending. I'm using corporate deductions as a measure of total business spending.

It's crude, I know. But I don't know a better measure.

Doing the crude thing, this is what I get:

Graph #3: GDP (red) and GDP plus the rest of Corporate Spending (blue)

Graph #4: GDP as a Percent of "GDP plus the rest of Corporate Spending"

Okay. Now we're ready. Yesterday's post and this one are just some background info I wanted to pass along before we get to tomorrow's post.

See you tomorrow.

Grit your teeth...

... while we work our way to the interesting stuff at the end.

Here's what I did:
1. Google U.S. business tax deductions data historical...
2. Get to irs.gov...
3. Search irs.gov for business tax deductions data historical...
4. Find Table 13...
5. Prepare to download Table 13.

The file I got was named histab13e(2).xls. You might not have the (2) in the name if you download it. I think I downloaded the thing once before, so I have (2).

The file contains corporate tax data for the years 1990-2010. Among the major categories: Number of returns, Total receipts, Total deductions, Income subject to tax, Total income tax, Tax credits, and Total income tax after credits.


I made a copy of the original page (Histab13E) and deleted a bunch of rows. Deleted most of the sub-categories, in case you don't have the appetite. I deleted none of the corporate tax credit categories, in case you want to get your blood pressure up.

Here's the zoho:



I made a copy of the file and reduced it even further, then added values for nominal GDP from FRED. Then I made a graph showing gross corporate income relative to GDP (blue), net corporate income relative to GDP (red), and corporate income tax owed (yellow):

Graph #1: Corporate Gross Income, Net Income, and Income Tax

The blue line says that gross income for U.S. corporations was almost twice the size of the U.S. GDP. Betcha didn't know that.

Oh, and the gap between the red and blue lines? That's all tax deductions.

Nick Rowe has me ROFL

Nick at Macromania:

Let's start with a very simple model with no commercial banks. Individuals hold Fed currency. They have a desired stock of money, and the difference between the actual and desired stock is excess money.

"The difference between the actual and desired stock of money is excess money."

Anybody here think they have "excess" money? Raise your hand.

The whole is greater than the sum of its parts

Livio Di Matteo in What's in a Name?:

We all know that the word “economics” comes from the Greek “okionomia” which refers to the thrifty management of household affairs. By extension, the origin of the term “economy” is closely related to the same term as it is from the Latin “oeconomia”, which is again from the same Greek “okionomia”.  From all this, it is not difficult to see an economy as simply the agglomeration of individual households when it comes to the European language tradition.  In a sense this nicely encompasses both our micro and macro traditions, as macroeconomics simply becomes the study of the sum of many individual household behaviours.

No no no no no no no: "macroeconomics simply becomes the study of the sum of many individual household behaviours." No.

That's what people think? No wonder the economy is such a mess.

Growth Modeling: The nonlinear thing

At Yahoo Answers, Simplicitus said

Total factor productivity is the otherwise unexplained productivity left over after all the individual factors (labor, capital, etc.) are taken out.

See? That's why I say TFP seems like an error term, a correction value, an adjustment to make the answer come out right. I never forget, my chemistry teacher in college joked one time that we can "multiply by zero and add the right answer." That's not exactly what TFP is, but it is pretty much how I used it for yesterday's graph.


In Total Factor Productivity (PDF, 5 pages) Diego Comin writes:

Total Factor Productivity (TFP) is the portion of output not explained by the amount of inputs used in production. As such, its level is determined by how efficiently and intensely the inputs are utilized in production.

TFP growth is usually measured by the Solow residual. Let g_Y denote the growth rate of aggregate output, g_K the growth rate of aggregate capital, g_L the growth rate of aggregate labor and alpha the capital share. The Solow residual is then defined as g_Y − α∗g_K − (1−α)∗g_L.

This is starting to get familiar. The growth of output minus weighted values of the growth of capital and the growth of labor leaves a "residual" because the growth of labor and capital do not fully account for the growth of output. The difference, the discrepancy, is said to be due to changes in the efficiency of labor and capital, and it is called TFP, Total Factor Productivity.

So if I get labor and capital and TFP from FRED, and use an appropriate alpha, I should be able to plot a line very similar to Real GDP. Did it yesterday, in fact.

That's why I dwell on the Solow Growth Model. I think I can use it to create a simulation of the economy in a spreadsheet. That's one of those things I try to do every once in a while, create a spreadsheet where each row represents a year, and the numbers for one year are used to generate the numbers for the next year. Doesn't sound difficult, but I've never got a satisfactory result. Maybe the Solow model will help.


One of the basic ideas I try to convey on this blog is the idea that there is a particular level (or a limited range of levels) of finance that is best for economic growth. Too little finance, and money for economic expansion is unavailable. Too much finance, and the cost of accumulated debt undermines growth. When you hear people talking about the "nonlinear" effects of debt, this is exactly the phenomenon they mean. (You heard it here first.)

But I can't really show what I think happens unless I can simulate a growing economy. I think I can take the Solow growth model and add something to simulate a financial sector and a growing accumulation of debt. So then we can see if accumulating debt has the effects I describe.

And my financial sector will be built on the stable, predictable economy of the Solow model, so that changes arising from the addition of a financial sector will be quite obvious. But don't expect to see it tomorrow.

Meanwhile...

Growth Modeling: I chickened out

I found a PDF titled Creating a Macroeconomic Model Using Real Economic Data, only five pages (plus a page of questions I didn't look at).

The file is SolowProjectNotes.pdf, and (backtracking the URL) it comes from Professor Rodney Smith of the Department of Applied Economics at the University of Minnesota. From his home page, click Teaching to get a list of Professor Smith's handouts and notes on his Solow Project and other stuff.

The SolowProjectNotes PDF is great! Smith takes you step-by-step and explains everything along the way. I spent most of Saturday going through the PDF, gathering data and entering calculations in a spreadsheet. It's one of those things where you have to put a lot of hours into it before you can make a graph to see if you're anywhere near right. I have patience for that kind of work, but it was most of Saturday.

I was almost ready to make a graph in the spreadsheet, and I chickened out. My brain got full, I couldn't do any more, I didn't have any time to check my work, it was just too much. So I stopped work on that project.

But you know what? I learned some stuff along the way. Here's the first formula Smith presents:

Here's what I understand (I hope I got it right):

"Y" is real output, or real GDP. The little "t" means "for the year we're figuring numbers for" or something like that.

"B" is a constant that scales the right side of the equation to equal the left side.

"K" is capital (and K-sub-t is the amount of capital in the year we're figuring). The K-sub-t is raised to the power of alpha, I think that little fish-looking thing is alpha. I'll get back to alpha in a minute.

// EDIT 28 SEPT 2014: WELL, THE NEXT PARAGRAPH IS WRONG. "A" IS LABOR PRODUCTIVITY, NOT TOTAL FACTOR PRODUCTIVITY. I'M NOT SURE WHAT THOSE THINGS ARE YET. BUT I THINK THEY'RE NOT THE SAME.

"A" is Total Factor Labor Productivity, and "L" is labor, or the number of people in the workforce. A-sub-t is multiplied by L-sub-t, and the result is raised to (one minus alpha).

// EDIT 28 SEPT CONTINUED: I'M NOT CHANGING THIS PART. I DID USE TFP FOR A TO PRODUCE THE GRAPH BELOW AND IT SEEMED TO WORK PRETTY WELL. SO MAYBE tfp AND LABOR PRODUCTIVITY ARE SIMILAR??

You can use "gross fixed capital formation" or "gross domestic fixed investment" or something similar for K, and "Total Factor Productivity" for A, and "Civilian Labor Force" for L, and you've got most of the numbers to fill in the equation.

The "alpha" we can use the value Professor Smith gives, 0.371 for the U.S. Then, one minus alpha is 0.629; Smith gives that number also. So now we've got everything for  the right side of the equation, everything except B.

I took FRED's Real GDP and put it on a graph. Then I took investment and factor productivity and labor numbers from FRED and arranged them according to the equation. I plugged in Professor Smith's values for alpha and one minus alpha. And I set B equal to one.

I graphed the thing, then fiddled with the value of B until the calculated line got close to the Real GDP line. Came up with B = 1.25. Here's the graph:

Graph #1: Real GDP (blue) and a Solow Simulation (red)

Nice.

I don't know how they figure Total Factor Productivity. Maybe it's an error term. Maybe there's a discrepancy in the calculation and they use TFP to reduce or eliminate the discrepancy. In that case what I'm doing here would be circular and silly. But even if that's true, I know more now than I did before.

All in all, pretty neat.

Growth Modeling: The Asymptotic Decline of Output

From The Solow Growth Model (PDF, 10 pages), unattributed, but the URL leads to Ana Sayfa of Dokuz Eylül University in Turkey:

The Solow Growth Model is a model of capital accumulation in a pure production economy: there are no prices because we are strictly interested in output = real income. Everyone works all the time, so there is no labor/leisure choice. In fact, there is no choice at all: the consumer always saves a fixed portion of income, always works, and owns the firm so collects all “wage” income and profit in the form of all output...

This model, then, is a model that captures the pure impact savings = investment has on the long run standard of living...

Ingredients: Consumers and Firms. All consumers own the firms, so consumers receive all output, and therefore all profit and rent.

Since there is no government (no taxes), there are no imports/exports (no trade), consumers receive everything from the firms, and there are no financial markets, savings is simply investment (the only place consumers can put their “money”, which is actually output, is simply back into the firm)…

Let's take this model that captures the pure impact that "savings = investment" has on the long run standard of living, let's take it and tweak it a tad. Let's NOT suppose there are no financial markets.

Just suppose.

That means there is a little less productive investment than there is savings, so financial savings accumulate. And total spending (consumption plus productive investment) is less than total income. Right away you can see an asymptotic decline because, if we always spend a little less than we receive, then the amount we spend and receive must eventually approach zero. (Factors like inflation and population growth disguise this trend but do not alter it.)

We continue to define spending and income as defined in the Solow model: "Spending" is spending on the production and consumption of output; and "income" is the income generated by the production and consumption of output.

Because financial markets exist, total spending is less than total income. The difference is the interest that accrues to accumulated savings. This difference, this interest, is income to financial markets. To productive markets, it is a cost.


The assumption that financial markets exist has two effects on the Solow Growth Model: It makes growth slower and costs higher. The rate of growth is reduced, because productive investment is reduced. And productive sector costs increase, because income accrues to savings.

Growth is slower and costs are higher, because financial markets exist. Sounds like a stagflation problem, don't you think?. Lucky for us, this is only a model.

Growth Modeling

From The Solow Growth Model (PDF, 18 slides) by Roberto Chang, Econ 504, Rutgers University, September 2012:

Once you assume the existence of the aggregate production function Y = F(K, AL), it is clear that the growth of output can be due to growth of A, K, or L.

As usual, investment is assumed to equal savings. The key behavioral equation of the Solow model is that savings equal a constant fraction of output...

Sometimes it is useful to make a few assumptions. But you want to be aware of the assumptions you're making. I suppose if you're an economist they taught you that stuff. People like me, we have to trip over it before we find it.

Watch your step.

Not excessive spending

At David Stockman's Contra Corner, on the History page: America’s Bubble Empire And The Echoes Of Imperial Rome:

The characters on stage are familiar to us — consumers, economists, politicians, investors, and businessmen. They are the same hustlers, clowns, rubes, and dumbbells that we always see before us.

But in today’s performance, they are doing something extraordinary: They are the richest people on the planet, but they have come to rely on the savings of the world’s poorest people just to pay their bills.

No.

Does it make sense that we are the richest people in the world, and yet so much in debt? Maybe, if we all lived in splendor. But we don't.

This is not the problem:

They routinely spend more than they make — and think they can continue doing so indefinitely. They go deeper and deeper in debt, believing they will never have to settle up. They buy houses and then mortgage them out — room by room, until they have almost nothing left.

Even this is not the problem:

They invade foreign countries in the belief that they are spreading freedom and democracy, and depend on lending from Communist China to pay for it.

That all comes down to nothing more than excessive spending. Oh, stick China in there, and war, and the argument starts to take on a life of its own. But prune off the emotional crap, and it all comes down to excessive spending. That is the argument. It is wrong.

Does it make sense that we are the richest people in the world, and yet so much in debt? No. So how did it happen?

We use debt credit for money. That's how it happened.

Uptrend to 1966, then stable to 1982, then uptrend again.

Hey.

On 11 June I showed this graph:

Graph #1: Total Factor Productivity (TFP)

I wrote:

Uptrend to 1966, then stable to 1982, then uptrend again. I know I've seen that pattern before. I just can't remember where.

I'll find it eventually.

Browsing my test & development blog a few days later, I came upon these notes:

Saturday, May 3, 2014
It turns a corner in 1966

I have to think about what this graph shows:

...

Dunno, but look at the second graph at SRW's "Not a monetary phenomenon"

http://www.interfluidity.com/v2/4561.html

This one, the second graph today, I'm subtracting number-of-employees from gross-domestic-product. You can't do that. It's vaguely, vaguely related to Okun's law. But you can't subtract people from GDP. It's mixed units. (So, that graph never made it to this blog till now.)

Whatever.

Here's the second graph from Interfluidity #4561:

Graph #3: "RGDP divided by the number of workers in the labor force" -- SRW

(See? You can DIVIDE GDP by number-of-workers. But you can't SUBTRACT people from GDP. Remarkably, though, similar patterns emerge both ways.)


Here's Total Factor Productivity (from Graph #1 above) in blue, left scale, and the Interfluidity graph in red, right scale:

Graph #4, Comparing Graphs #1 and #3

Now that's interesting.

Okun's law defines a relation between employment and output. SRW's graph tests that relation. Looks like variations in that relation are related to Total Factor Productivity.

SRW says the baby boom's entry into the workforce outpaced the advance of technology, and is responsible for the flat spot there from 1966 to 1982. Seems to make sense. The problem I have with the idea is, when I look at that flat spot I can't help but think "that's not my fault. I didn't cause the flat spot." Subjective, yes.

I'm wondering now how they calculate TFP. They probably back into it, plugging results into some calculation and coming up with an immaculate number. The results that SRW's graph shows could easily be responsible for the pattern that TFP shows.

I'm wondering now how they calculate TFP.

All cars are suddenly identical? Please, make them all like this:

Source: Jalopnik

What I love about Nick Rowe? His analogies. This is from Bank runs, keynesian multipliers, monetarist cold potatoes at Worthwhile Canadian Iniative:

Car banks. Suppose the Jalopnik devil waved a wand and made cars a fungible asset. All cars are suddenly identical, and will always be identical. Car banks spring up. You give your car to the car bank. In return, the car bank pays you rent/interest, and promises to pay you one car on demand. Since most of the time most people don't need a car, car banks become fractional reserve car banks. There's no profit in having cars sit idle. Owning the bank's promise to pay you one car on demand is as good as owning a car, provided the bank can always deliver on that promise.

I read that, I know we're talking about money and banking. But it helps to think of it in terms of cars. It's like seeing with fresh eyes. But here's something:

Since most of the time most people don't need a car, car banks become fractional reserve car banks. There's no profit in having cars sit idle.

That's the "loanable funds" thing. The car bank gets to loan out the cars to other drivers because it *has* the cars, because we deposited them in the car bank.

Yeah, it works with cars. It works with cars, because a bank cannot create a new car simply by lending it out. It has to actually have a car before it can lend it. That's not true for money. Banks create money by lending it out. (Actually I'm starting to mull that one over. But I'm not ready to talk about it yet.)

Banks create money by lending it out, so they don't need to get our deposits in order to lend. I understand that.

So I have to ask, again: Why do banks want our deposits?

Sometimes something strikes you

I clicked Random Eyes (see sidebar), then clicked random a few times and ended up with M2 velocity:

Graph #1: The Velocity of M2 Money
Click Graph for the FRED Page via Random Eyes

I've seen it before, of course. But this time it looked almost flat to me until that sudden, large uptrend that runs from 1991 to 1995.

Those dates ring a bell for ya? They do for me. That was the time of a strong increase in M1 money growth, and a fall in the debt-per-dollar ratio. With a strong increase in M1 growth, I think M1 velocity probably fell -- just the opposite of M2 velocity. That's odd, if true, and it's worth a look:

Graph #2: Velocity of M1 (red) and M2 (blue) Money
Shown on Right and Left Scales Because the Numbers are Quite Different

I left the data box visible, and maybe you can see just at the end of the 1991 recession, a one-pixel-wide, slightly darker gray vertical line. Where that line crosses the red and blue plot lines FRED has added pink and blue circles to highlight the first quarter of 1991 on those plot lines.

Sure enough, beginning in the first quarter of 1991, M1 velocity (red) trends down while M2 velocity (blue) trends sharply up. M1 continues down to fourth quarter 1993; M2 continues up till first quarter 1995.

Velocity of money is calculated by dividing GDP by the quantity-of-money number. GDP is GDP, the same for the red line and the blue. So the difference between the two lines is due to differences in the M1 and M2 money numbers. Compare growth rates:

Graph #3: Growth Rates of M1 (red) and M2 (blue) Money

As expected, M1 money (red) shows a big growth spike beginning just after that 1991 recession, while M2 money growth (blue) slowed to near zero. There is a similar red spike in the mid-1980s but the blue line is much higher then. So the "gain" in M1 (red) was not as big in the mid-1980s as in the early 1990s.

There was a change in M1 accounting in 1994 that reduced the amount of money reported as M1 money:

Sweep programs create distortions between reported data on the monetary aggregates and accurate measures of the money stock.

In order to correct for these distortions, Graph #4 duplicates Graph #3 using M1 that has "sweeps" added back in:

Graph #4: Growth Rates of M1 + Sweeps (red) and M2 (blue) Money

For the period from the mid-1990s to the most recent data, the two lines are noticeably closer on Graph #4 than #3. But the big difference between the two lines in the early 1990s remains.

To see the difference in growth rates, I took the two lines from Graph #4 and subtracted the one from the other. I'm trying to see how much faster (or slower) the one money measure was growing, than the other. I subtracted M2 from M1. That means that when the difference is above zero M1 is growing faster, and when the difference is below zero M2 is growing faster:

Graph #5: Growth Rate of M1 + Sweeps less Growth Rate of M2

The big gains for M1 show up as large mid-1980s and early-1990s peaks on Graph #5. There are two much smaller peaks in the 2000s, and some unevaluable activity since 2008. Other than that, the blue line is almost always below the zero-line: M2 almost always was growing faster than M1.

What makes this more interesting is that M1 money is a component of M2 money. If you take M2 and split it into the money we expect to spend and the money we hope to save, M1 is the part we expect to spend. The rest is money we're trying to save. So basically, Graph #5 shows that our savings have almost always grown faster than our spending-money, except during those two massive spikes and a few brief interludes.



If you take Potential GDP, which is a measure of best-case GDP, and put its growth rate on Graph #5, it looks like this:

Graph #6: Growth Rate of Potential GDP (red) Overlaid on Graph #5

The general trend of Potential GDP -- the red line -- is downhill. The one notable exception to this downtrend is the uptrend created (says I) by the massive spike in M1 money growth in the early 1990s.

Why? Because no matter what you want to do in this economy of ours, it takes spending money to do it.

Try typing "asshole" and see what you get

In the sidebar at La Bocca della Verità, an invite to RealClearPolitics:

Type Bigotry Into Google, See What you Get

How could I resist? But I did, though. I didn't type anything into Google. I just clicked the invitation. Despite their run-on sentence.

They show this image:

"Why is 'right-wing' a correlation to bigotry?" the accompanying text asks. "Why is bigotry a descriptor to 'right-wing?'"

That's why I'm here. Why is "right wing" a correlation to bigotry? What the hell does that even mean? I guess they mean Why is "right-wing" correlated to "bigotry"?

It isn't, of course. Not in the example sentence, anyway.

The other one means even less: Why is bigotry a descriptor to 'right-wing?' A "descriptor"?? I looked that one up on Google:

A descriptor is a word used to describe or identify something. So: Why is "bigotry" used to describe or identify "right-wing"?

Again: It isn't. If the word "bigot" was used to identify or describe a right-winger, you could say "bigot" and people would know you're talking about a right-winger. But the actual phrase used is "right-wing bigotry". And that surely implies that there is also left-wing bigotry. Otherwise, you wouldn't have to say which side you're talking about.


But actually trying to understand what the words mean, that's not really part of the game anymore. The game now is to cultivate outrage.

Decline of civilization.

Happy-go-lucky economics

In a PDF dated February 2000, Dea... // No, wait. Let me start again

In a PDF from the year 2000, Dean Maki of the Federal Reserve wrote:

Household debt is at a record high relative to disposable income. Some analysts are concerned that this unprecedented level of debt might pose a risk to the financial health of American households and ultimately lead them to curtail their spending...

A high level of indebtedness among households could also lead to increased household delinquencies and bankruptcies, which could threaten the health of lenders...

And then there's this:

In stark contrast to the view of growth of consumer credit as a negative force in the economy, a consensus seems to be emerging from recent research that consumer credit growth is positively related to consumption in future periods. Little evidence has been found that household debt service burdens are negatively related to future consumption, though some theoretical models suggest a more complex relationship may be at work. Specifically, high debt service burdens could make household consumption more sensitive to a drop in income...

It strikes me that the economic research seems to focus on current events without regard for macroeconomic principles. As long as everything's going along fine, the consensus emerges that everything is fine. Nobody looks at the subtle changes that could be a sign of troubles -- nobody looks at them as potential troubles, anyway -- because everything is going along fine.

It's happy-go-lucky economics.


From Maki's conclusion:

Consumer credit growth is often viewed in the press and on Wall Street as a negative force in the economy that will cause future consumption to slow. The available research on this question suggests quite the opposite: Strong growth in consumer credit tends to be associated with positive future growth in consumption...

From this point of view, high debt burdens are not a negative force in and of themselves; they should only be viewed as a problem to the extent that the expectations of future income on which the borrowing was based were too high.

Maki equates "growth in consumer credit" with "high debt burdens". Oh, these two are related, to be sure. But they are not the same. To equate "credit use" with "debt" is to equate "buy now" with "pay later". If you cannot distinguish between the two, you cannot understand the economic problem of our time.

Let's have some fun

At FiveThirtyEight: The Slow Death of American Entrepreneurship by Ben Casselman. From the article:

Americans started 27 percent fewer businesses in 2011 than they did five years earlier, according to data from the Census Bureau. As a share of all companies, startups have been declining for more than 30 years.

Declining since the late 1970s. Interesting.

Among possible causes of the decline, Casselman points out that "The U.S. economy is also increasingly dominated by large corporations". By coincidence, perhaps, Casselman also mentions "the long and well-documented decline in family-owned businesses."

Trends like that don't sit well with me.


I tried to figure out what data was used for Casselman's graph, with no success. But I found some more graphs at BLS. Graphs are always fun.

The number of new business establishments (establishments that are less than 1 year old in any given year) tends to rise and fall with the business cycle of the overall economy. As shown in chart 1, the number of new establishments for the year ending in March 2010 was lower than any other year since the series began.

This graph starts out with a pretty good increase. But it starts in 1994. The pretty good increase is what we got during the good economy of the latter 1990s. If you look at the graph without thinking of that, you might imagine the same strong uptrend stretching back in time another ten or fifteen years. I doubt it.

Another feature of this graph: The drop associated with the Great Recession started in 2006. That's noteworthy.

Next, Chart 5:

As shown in chart 5, the period from 1993 to 2006 was marked by an increase in the number of births and deaths, indicating a higher amount of business “churn”—that is, new business establishments entered and old establishments exited the economy in greater numbers. Since the most recent recession began in December 2007, births have experienced the steepest decline in the history of the series. New establishments are not being formed at the same levels seen before the economic downturn began, and the number is much lower than it was during the 2001 recession.

Sure enough: new establishments are not being formed, and the number is much lower. But doesn't it strike you as odd that this is all we get in the way of commentary from the BLS? I'm not the policymaker. I'm the critic. I'm supposed to be the guy saying "Hey, the economy isn't doing well." BLS is supposed to do a little more than that, don't you think?

I also don't like it that they say things like "the steepest decline in the history of the series". The history of the series? "The period from 1993 to 2006" is only about a dozen years. The series isn't old enough to have something you could call a history. When you're old as I am, you'll understand.

Like the second graph (BLS Chart 1), the third one (BLS Chart 5) shows an early-onset decline of establishment births, 2005 or 2006. But the peak in employment gains from startups came even earlier. BLS Chart 6 (not shown) indicates that

The number of jobs created from establishment births peaked in the late 1990s and has experienced an overall decline since then.

Trends like that don't sit well with me.

Notes on a Sunday morning

From the 8th of June.

The wife watches the Sunday morning news commentary. George Stephanopoulos is on at the moment. We have republicans attacking the President for retrieving a soldier from the Taliban. This is presented on TV as a struggle between republicans and democrats. That's wrong. The actual struggle is much bigger and far more interesting.

The economy -- you remember, we do economics here -- the U.S. economy has been deteriorating for a long time. This deterioration is a challenge in the Toynbee sense: During the life of a "civilization" there is a series of "challenges" which are met with "responses" by a "creative minority". As long as the response provides a satisfactory solution to the challenge, the civilization advances. When the response is inadequate, civilization goes into decline. It's a remarkably simple analysis.

The U.S. economy has been in decline for a long time -- since the 1974 recession, at least. This decline is our challenge. To the rest of the world, it seems the U.S. is weak and vulnerable. That is the motivation behind attacks like the ones that took down the Twin Towers and damaged the Pentagon.

As Toynbee put it, a civilization that is not advancing is in decline. We need a response suited to the economic decline that is the challenge of our time. We've been responding for forty years now, without success; internal discord only grows more strident.

It's time to try a different response.

Oh! I've seen this shape before...

Been looking into the Solow Growth Model just a bit, lately, and took a look at Total Factor Productivity:

Graph #1: Total Factor Productivity (TFP)

Uptrend to 1966, then stable to 1982, then uptrend again. I know I've seen that pattern before. I just can't remember where.

I'll find it eventually.