Forget about Sumner

Rummaging thru my old posts this morning I noticed a pretty neat graph of debt growth, then moved on. An hour later I noticed myself quoting Scott Sumner:

Krugman makes the basic mistake of just looking at time series evidence, and only two data points: US growth before and after 1980. Growth has been slower, but that’s true almost everywhere.

Sumner disputes the significance of Krugman's statement. But he agrees that US economic growth has been slower since 1980. When I saw that, the bell went off in my head again. Backtracking, I was lucky enough to find the graph I saw earlier:

Breaking the 54-year period into two equal parts --

Total (Public and Private) Debt relative to GDP

A 25% increase in the first period, and a 100% increase in the second period. Debt grew four times as fast in the more recent period. Relative to GDP.

Maybe it was the growth of debt that caused the slowing of economic growth.

Nah. Sumner says forget about debt.

Not only government debt, let me add

... for any level of the interest rate, a higher debt load means that the government will permanently need to spend more money just to pay the interest on the debt. This is not a matter for arcane debate, but rather is a consequence of the most basic arithmetic.

- James Hamilton

And not only government debt.

Not Net Interest

Following up on the idea that working with "net interest" is bad economics, I figure today it should be safe to look at some gross interest numbers.

Interest Income compared to Corporate Profit:

Graph #1: Interest Runs Much Higher Than Profit From the 1970s to Maybe Eight Years Ago

Comparing them as a ratio, Interest relative to Corporate Profit:

Graph #2: Interest Relative to Corporate Profit

Interest and Profit run about equal (at 1 on the vertical scale) for most of the 1960s. For most of the 1970s interest was mostly on the high side of one and a half times the size of profit. Interest jumped to about 4 times corporate profit for most of the 1980s, then gradually worked its way down to near where it was in the 1960s.

Finally, compare employee compensation to the sum of interest and corporate profit:

Graph #3: Compensation of Employees relative to (Interest + Corporate Profit)

Labor income ran at around 2.8 times capital income in the 1960s. Labor income declined (relative to capital income) all through the 1970s, when rising wages were said to be the driving force behind inflation. I don't see it.

In the 1980s and '90s labor income generally stayed in the neighborhood of 1.6 times capital income.

There was a sudden jump around the time of the 2001 recession. Then labor income fell until it was not much more than equal to capital income. But then, the good news: Since 2007, labor income has only increased, relative to capital income. And what good times these past ten years have been!

Hm. How did I miss that?

There's no such thing as net interest

It's not even on the list!

If I have $10 interest income and $100 interest cost, I have negative $90 interest income, net. But there's no such thing as "negative income". If it's negative, it isn't income. You can do the math and get that answer. You can be $90 short. But you cannot have "negative $90 income". Because "income" is not a math word like add or subtract. Income is an economic concept.

You can't have negative income. If it is negative, it isn't income.

We can take my negative $90 and add it to your positive $100, and call it $10 "net" interest income. But I didn't have any net interest income, and you had $100. It is some kind of fallacy of composition to add my negative and your positive together and come up with $10 "net". And anyway, it doesn't work that way. Not at the macro level.

Figuring net interest is like a big party where we all get together and have drinks and talk about our finances, and every time two people are talking and one of them has positive net interest, he gives the whole positive amount to the guy with negative net interest.

Then the guy who gave money away has achieved "zero net" and he can go home. The other guy refigures his finances and looks for somebody else to talk to. And the party doesn't end until nobody has negative net interest. When that happens, we take and add up all their positives, and we call that "net interest".

The economy doesn't work that way.


If I have $10 interest income and $100 interest cost, the $90 difference comes out of my other income. Not out of someone else's positive net interest.

Suppose I have negative $90 interest income and you have positive $100 interest income, and between us we have $1000 of other income. Add my negative $90 and your positive $100 together and you get "net interest" income of $10. So other income is $1000 and net interest is $10 and we want to say the net interest is one percent of the other income.

But really, I paid for my $90 interest shortfall out of other income, and so we should figure between us we have not $1000 of other income but $910. And actual interest income was not only my $10 and your $100, but also the $90 out of other income that I paid to a third party. So that's $200 of interest income relative to $910 of other income, and that is a lot more than one percent.


You can't add my micro to your micro and come up with macro. It doesn't work that way.

Figuring net interest assumes that no one has to pay interest costs out of their other income until after everyone with positive net interest income has given away everything above "zero net" to someone with negative net interest income.

It would be nice to have so much interest income that you could use it to pay your interest expenses, and have money left over for your other expenses. But most people don't. Maybe banks do, but most people don't pay interest expenses out of their interest income. We scrape together income from all available sources to pay whatever bills we have to pay. Figuring net interest understates financial cost and creates a false picture of our economy.


Figuring net interest assumes that everyone almost always has enough interest income to cover their own interest expenses. That's just bad economics.

Similarly, every time someone says private debt doesn't matter because debt is money we owe to ourselves, or because one person’s debt is another person’s asset so the world has no net debt, it is a fallacy of composition. Or some kind of fallacy, for sure.

Position and Behavior, and Mack and Mike

"The sum of micro is macro"

No.


"The answer that macro is the sum of micro is technically correct." -- Quora

No, it is not correct. And adding the word "technically" doesn't make it correct.


"The whole (macroeconomics) is but the sum of the parts (microeconomics)" -- the same Quora

No. If Macro is nothing but the sum of Micro, then there can never be a fallacy of composition. But that fallacy arises. Therefore it cannot be true that Macro is nothing but the sum of Micro.


Again at the same Quora link, we read that macroeconomics "studies the functioning of economy as a whole", while microeconomics "analyzes the behaviour of individual components like industries, firms, and households." Lets start with that.

Take an "individual component" of the economy: a firm, say. The behavior of this firm includes employing people to make things. It also includes selling what it makes. Thus the firm's behavior affects not only its own position but also that of its employees and that of its competitors. It also affects labor markets and resource markets and the market for its product, to some degree: less for a small firm, more for a large firm.

The influence of our firm's behavior, to some degree changing the position of its employees and its competitors, will in some measure affect the behavior of those employees and competitors. To a lesser degree it will affect the markets in which it participates and, to a still lesser degree, other markets and the participants in those markets.

Now, consider all of these influences arising from the behavior of our example firm. Do these all have perfectly linear effects? I mean, if a $1 decision by our firm leads in sum to a one-cent change in the behavior of the rest of the economy, does a $100 decision lead to a $1 change for the rest of the economy? It would have to, if Macro is nothing but the sum of Micro.

And then, does a $100 billion dollar decision by our firm lead to a $1 billion change? It is highly unlikely. It is highly unlikely that once you work out the effects of a one-dollar change, you can scale your numbers up to billions and have everything stay in perfect proportion.

But if things don't stay in perfect proportion, then the influences arising from the behavior of our firm are not perfectly linear.

And if those influences are not perfectly linear, then the sum of micro is not macro.

The definition of randomness

When flipping a fair coin 21 times, the outcome is equally likely to be 21 heads as 20 heads and then 1 tail... Assuming that a change in the probability will occur as a result of the outcome of prior flips is incorrect

- Wikipedia

Have you been paying your share?

Monetary Interest Paid per Capita is now about $8000 annual.

Graph #1

Have you been getting your share?

Monetary interest received per capita is now about $8000 annual:

Graph #1

It's almost ten years since the interest rate went to zero

Graph #1

This is 2018. It's almost ten years since the interest rate went to zero. However, after a brief intermission, interest income resumed increase in 2010:

Graph #2

And after falling since 2007, the cost of interest is still almost 15 percent of GDP:

Graph #3

The cost of interest is still about two and a half trillion dollars a year:

Graph #4

You want it all, don't you. You want that interest income for yourself. Yeah, me too.

The thing is, though, it creates problems for the real (non-financial) economy.

Some graphs

Graph #1: Employment in Finance as a Percent of Employment

Graph #2: Average Full-Time Pay in Finance

Graph #3: Average Full-Time Pay in Finance and Elsewhere

"the financial ability to make work–life balance choices"

My view is that the economy is going downhill and has been since maybe 1966. That colors everything I write.

So I have to respond when I see this, from Mary C. Daly of the San Fransisco Fed:

U.S. labor participation diverging from international trends

Source: OECD: rates for ages 25–54.

This chart compares the percentage of prime-age workers in the labor force in Germany, Canada, the United Kingdom, and the United States. In these other advanced economies, labor force participation of prime-age workers has increased overall and now stands far—several percentage points—above the rates observed in the United States.

Which raises the question—why aren’t American workers working?

The answer is not simple, and numerous factors have been offered to explain the decline in labor force participation. Research by a colleague from the San Francisco Fed and others suggests that some of the drop owes to wealthier families choosing to have only one person engaging in the paid labor market (Hall and Petrosky-Nadeau 2016). And I emphasize paid here, since the other adult is often staying at home to care for house or children, invest in the community, or pursue education. Whatever the alternative activity, some of the lost labor market participation seems related to having the financial ability to make work–life balance choices.

"Whatever the alternative activity, some of the lost labor market participation seems related to having the financial ability to make work–life balance choices." When I see that, I want to puke. Because it says people are opting out of the labor force because they can afford to opt out of the labor force. This economy is just so damn good that people don't need the work. Puke, puke, puke.

This economy is so good that people don't need the work? I don't see it. But I dunno, I could be wrong: I'm retired. I opted out of the workforce. I can't deny it. I didn't have to retire. I could have stayed at work till I got fired for falling asleep at my desk. Not that there's anything wrong with that.

Maybe. Maybe the economy is so good that people don't need the work, and that's why US Labor Force Participation is down. But I don't think so.

But this is definitely an interesting idea, the idea that labor force participation responds to economic conditions. The idea that the participation rate comes down when people can afford not to work, and goes up when people can't afford not to work. This idea can be used to explain a lot of the increase in the participation rate between 1965 and 1990, say.

Graph #2: US Civilian Labor Force Participation Rate, 1948-2018

Labor force participation went up, we're often told, because women entered the workforce. But nobody ever says that many of those women entered the workforce because hubby was no longer bringing home enough bacon. Or that they split up over money.

That part's always left out. Just like the years before 1990 on Mary C. Daly's graph.

"Q"

From Returns for Domestic Nonfinancial Business (PDF) by Sarah Osborne and Bonnie A. Retus:

Tobin’s Q, or simply “Q,” is the ratio of financial-market valuation of corporate assets to the current-cost value of the assets. A Q ratio above 1 indicates that financial markets value corporate assets above their replacement cost; as a Q ratio rises above 1, companies may be more inclined to make direct investments in plant and equipment. A value of Q below 1 indicates that the financial markets value corporate assets below the replacement cost; as Q falls below 1, companies may be more inclined to buy other companies for their capacity rather than make direct investments.

I don't know these things. That's why I capture them. If the quote is correct, then:

Q is above 1 when financial markets highly value corporate assets. And Q above 1 is good for investment. So a "high" stock market should be good for investment.

Q is below 1 when financial markets put low value on corporate assets. And Q below 1 (a "low" stock market) is good for mergers and acquisitions. Well that makes sense I guess.

Who knew? (Everybody but me, probably.)

How does this work, exactly?

Graph #1

• The secondary market knows what the Fed will do ahead of time?
• The Fed raises rates based on the secondary market rate?
• Pure coincidence?
• Other?

On explaining productivity -- a follow-up

Me:

Cobb-Douglas is a "production" function. A supply side function. Demand concepts like the Kaldor-Verdoorn Law are ignored. Schneider's "simple growth accounting decomposition" shows clearly that slow productivity is caused in part by slow technological growth and in part by capital shallowing. From this, one could conclude we need policies that boost tech growth and policies that enhance capital deepening. But to reach such conclusions without also considering the demand side would be a grave error.

In my notes just now, I came across this excerpt from 2016, from John Mauldin:

GDP growth has only two basic components: growth in productivity and growth in the workforce size. That’s it. There are two and only two ways you can grow an economy: increase the (working-age) population or productivity.

Therefore—and I'm oversimplifying quite a lot here—a recession is basically a decrease in production (as, normally, population doesn't decrease). Two clear implications emerge: The first is that if you want the economy to grow, there must be an economic environment that is friendly to increasing productivity.

Mauldin -- a very sharp guy -- leaps from a conclusion arising from growth accounting to a very specific conclusion about the policy needed to boost productivity. That leap is exactly the kind of grave error I had in mind.

On explaining productivity

An oldie by Patrick Schneider at the Bank Underground: There are two productivity puzzles, 17 Nov 2016.

Two productivity problems: Schneider says there is a level problem and a growth rate problem. His analysis hits home because "levels and rates" are almost as central to economic analysis as "supply and demand".

Six graphs. He looks first at "Labour productivity relative to trend" long term, then just at the years after the GFC where a gap opens as productivity falls behind the trend. In the remaining graphs, Schneider analyzes the gap and finds that

... the level puzzle is almost all technology (Chart 5), but the growth puzzle is due to both slower technology growth and less capital per person employed (Chart 6).

In an update he adds:

Using the updated capital services series (chart below [previously Chart 6]), slower capital deepening can explain nearly two-thirds of the growth puzzle from 2011-2016 (compared to one-half previously).

It was somewhere along in there that it suddenly struck me: This is nothing like what I say about productivity. I say productivity will be rising soon, if it isn't already. I say productivity will rise when the economy starts to grow, and that the economy will start to grow when accumulated private debt starts to grow faster than GDP. And I say things are moving along right on schedule.

But I'm no economist. My expectations of improving productivity arise from graphs: from similarities visible in different time periods and different debt ratios. I look at debt-to-money ratios, debt-to-income ratios, and private-to-public debt ratios.

Schneider's evaluation of the productivity problem arises from "an aggregate, constant-returns-to-scale, Cobb-Douglas production function".

Schneider links to a page by Robert Solow, where Solow opens by saying "it takes something more than the usual 'willing suspension of disbelief' to talk seriously of the aggregate production function." Schneider uses it anyway. He rearranges the Cobb-Douglas and finds that

we can attribute changes in labour productivity to growth in either technology or capital depth, and we can also attribute deviations from trend to deviations from the trends in these two terms.

Based on the production function, Schneider attributes the fall in productivity to problems of technology and capital deepening. And his graphs seem to confirm this.

It is interesting stuff. I like the graphs. But is it right to take a formula some guy developed back in 1927, rearrange it, and conclude from the rearranged formula that there is no other factor influencing productivity? You would have to put a lot of faith in that one formula.

In the link Schneider provides, Robert Solow reminds us that the economy consists of production and consumption, both:

the aggregate production function is only a little less legitimate a concept than, say, the aggregate consumption function...

Supply and demand. Or as Keynes said:

there cannot be a buyer without a seller or a seller without a buyer.

Pretty much the whole trouble with economics today is that some people favor the supply side and others, the demand side. Some are Big-Endians; others, Little-Endians. Meanwhile, every egg comes with big and little ends. Every transaction involves supply and demand.

Cobb-Douglas is a "production" function. A supply side function. Demand concepts like the Kaldor-Verdoorn Law are ignored. Schneider's "simple growth accounting decomposition" shows clearly that slow productivity is caused in part by slow technological growth and in part by capital shallowing. From this, one could conclude we need policies that boost tech growth and policies that enhance capital deepening. But to reach such conclusions without also considering the demand side would be a grave error.


I don't mean to be overly hard on the Cobb-Douglas or the people who use it. I think that most valid explanations are valid, even conflicting ones. I like Schneider's story. I think it explains a lot, and provides good insights. But I don't think it tells the whole story. Here is his first graph:

Productivity Runs Flat Since 2008

Here is my graph from yesterday:

Non-Federal Debt Runs Flat Since 2008

See the similarity? Cobb-Douglas doesn't explain it.

GDP Growth has been slow since 2008 because credit growth has been slow

Note the sudden change around 2008

Note the very gradual increase since 2013. Gradual acceleration. We're just starting to feel it now. Unfortunately, that means debt is going up again.

// The FRED data

Debt Service: First the Silver Lining, then the Cloud

Back in early January, Edward Harrison observed Consumer credit: largest gain in 16 years and well ahead of expectations:

Economic data coming out of the United States continue to show a robust consumer-led expansion.

Harrison gave some numbers, and was optimistic:

Analysts see this not as reflective of distress but buoyancy as the Conference Board’s measure of consumer confidence hit a 17-year high in November. Nomura, for one, released a note saying: “This appears consistent with a strong labor market with low unemployment and elevated consumer confidence, which we expect will continue in the near term.”

Moreover, since consumer spending makes for almost 70 percent of the economy, these numbers bode well for economic growth figures to be released at the end of the month...

Then he dropped the other shoe:

The dark cloud in all of this is the fact that this is debt-fueled consumption.

Yep. But if these signs of economic vigor were fueled by debt, then they should have been predictable -- as predictable as the trouble that will arise from growing consumer debt.

More predictable, even. You might have looked at Household Debt Service two years back and noticed it rounding the bottom:

Graph #1: Household Debt Service as of 3 March 2016

You could have selected the round bottom data, just after the sharp "V" that bottoms out in 2012Q4, and put a trend line on it:

Graph #2: As Above, with Trend based on 2013Q1-2015Q3

You might have said to yourself, "Hey! That thing's going up!"

If you look at where the trend line is in the first quarter of 2018, it seems unrealistically high. But now (two years later) we can ask: How did your prediction turn out?

Right on track:

Graph #3: Same as Graph #2, plus Current Household Debt Service Data (red)

The current data (red) is a surprisingly good match to the path predicted by the trend line.

Of course, the good match does not guarantee that our economy will soon be vigorous again. Still, as Ed Harrison was saying, "Economic data coming out of the United States continue to show a robust consumer-led expansion."

I expect vigor. Vigor until Debt Service costs get too high, and then we lose the vigor.

Why?

You can see Debt Service going up in the early 1980s, after the 1982 recession, when Reagan got that massive spike in Real GDP growth. You can see Debt Service going up in the mid-1990s, when Bill Clinton got what Alan Greenspan called "the new economy":

Graph #4: Household Debt Service (blue) and Real GDP Growth

Debt Service costs went too high in the Reagan years, and that interfered with vigor. In the Clinton years, Debt Service costs stayed between 11.0 and 11.5 until the end of the decade, and the economy's performance was impressive enough that Greenspan gave it a name.

The two datasets are on different scales. However, in both the Reagan years and the Clinton years, the economy regained vigor while Debt Service costs were low, and then economic growth remained vigorous until the blue line went above the other.

Debt Service costs went low again after 2010. Now, though, they are picking up again, and the economy is improving. You could have predicted it two years ago.


I expect our economy to show increasing vigor over the next couple years as Household Debt Service continues to follow the trend line up toward the 11.0 level. Above 11.0, I worry that Debt Service costs will become burdensome and cause growth to slow. I guess we'll have to wait on that and see how things go.

Meanwhile, if you want to play along, take a look at productivity. It too is rising.

You say it's a "given". I say it was created by policy in the first place, and to create new policy now to aid and abet your so-called "given" is pure folly.

Wordy Bill has posted Part One of his response to a German critic of MMT. I figured I'd read it, as it should be an overview of key points of controversy from the perspective of one of the mainstays of MMT thought. Bill says:

The aim of fiscal policy is not to deliver a particular fiscal outcome (surplus or deficit). Rather, it is to ensure that the discretionary government policy position is sufficient to ensure full employment and price stability, given the spending and saving decisions of the non-government sector.

He emphasizes the word "given".

The spending and saving decisions of the non-government sector are a "given", he says. In other words: Those decisions create an economic environment, and the aim of fiscal policy is to ensure full employment and price stability in that environment.

I disagree.

The spending and saving decisions of the non-government sector arise largely in response to policy. Those decisions may seem a "given" at the moment. But it's not like the government had no hand in creating them. You know goddamn well, for example, that economic policy encourages saving. It has been policy since the beginning of time, almost, to encourage saving. So when you look at the amount of money that has been tucked away as saving in one form or another, what you are looking at is a result of policy: the decisions of individuals (or groups or businesses or whatever) in response to policy.

That's not at all the same as what Bill says, that those decisions are a "given".

This comes up because I recently quoted the last bit of Chapter 10 from Keynes's General Theory. Here again is his last sentence:

We have to accept [the sufferings of unemployment] as an inevitable result of applying to the conduct of the State the maxims which are best calculated to “enrich” an individual by enabling him to pile up claims to enjoyment which he does not intend to exercise at any definite time.

We do not have to accept the spending and saving decisions of the non-government sector as a given. Those decisions are an inevitable result of policy. If the result turns out to be a problem, then the policy was bad, and we should change it.

It is a great error to accept as given, that which is not a given.

Are US trade deficits caused by high US labor costs?

Lotta noise in the news lately about the Trump tariffs. My wife, at dinner, says something about the tariffs compensating for US wages being higher.

But I'm not sure US wages are still "higher". Higher than wages in China, maybe, but China is a latecomer to our trade deficits problem. So I have to stop and wonder if higher US wages really are the problem. Because, you know, for me it's always the cost of finance that creates our problems.

At a glance, all the Google search results mention labor. Like my wife, everybody thinks US wages are the cause of our trade deficits. But that's why I do graphs.

To see for myself.


The trade deficit developed in the 1970s...

Graph #1: US Trade Deficit as Percent of GDP

... just about the same time that wages inexplicably started falling behind:

Graph #2: via EPI

Odd, isn't it? Looks like the trade deficit developed when we didn't have the income to "buy American" because our wages were too low. Not because our wages were too high.


Nobody ever points this out, but if wages were too high, wouldn't that be good for wage earners? When I look at the economy I don't see "good for wage earners".

You've heard of "labor share" -- labor share of income. You might have heard people say labor share has been falling since the year 2000:

Graph #3: Labor Share

Since 2000? Looks to me like it's been falling since 1960.


At FRED a search for Costs per unit of real gross value added of nonfinancial corporate business: turns up 10 series. Five of them show quarterly data. Among these are "Compensation of employees (unit labor cost)" and "Unit nonlabor cost".

If a cost is not a labor cost, it's a non-labor cost, right? So if I take labor cost per unit, and add in non-labor cost per unit, then I have the total cost per unit.

Labor cost per unit relative to total cost per unit shows decline from beginning to end:

Graph #4: Unit Labor Cost as a share of Unit Total Cost

Is high labor cost the cause of our trade deficit? I am inclined to say no.


Here, maybe this is a better picture. Compensation of employees as a share of Gross Domestic Income:

Graph #5: Employee Compensation as a Share of Income in the US

There is a flat in the 1950s and '60s -- you have to have lived in the 1950s and 60s -- and employee compensation goes high after the flat, peaking in 1970. Since then it's all downhill. Faster down since 2000, just like Labor Share.

Maybe the big spike in compensation, 1965-1970, made US goods and services too expensive to compete in global markets, creating our trade deficits?

Maybe. But compensation has been trending down since that peak in 1970. Meanwhile, trade deficits have been getting worse, not better. So there does not seem to be a clearly defined "high employee compensation causes trade deficits" relation.

Anyway, since the early 2000s, employee compensation (as a share of US income) has been lower than it was in the 1950s and 60s. And the trade deficits just get bigger.

Are US trade deficits caused by high US labor costs? I don't think so.

"the possible positions of equilibrium"

"I shall argue that the postulates of the classical theory are applicable to a special case only and not to the general case, the situation which it assumes being a limiting point of the possible positions of equilibrium."

-- from Chapter One of The General Theory by J.M. Keynes

The "possible positions of equilibrium" can be understood as different average levels of employment in different periods, due to the economy being in different equilibrium states. A "limiting point" is a minimum or (in this case) a maximum. The level of employment which corresponds to the "special case" described by Keynes is the highest level: the one he called "full employment".

But it is hard to see equilibrium levels of employment on a graph like this:

Graph #1: Unemployment

The cycles may be obvious, but is the equilibrium is not.

Business cycles are like the economy breathing. Even at rest, you're breathing. Even with the economy at equilibrium, the business cycle is going through its phases. As a result, it is not easy to see when the economy is in equilibrium. If there is such a thing.

Is there such a thing as economic equilibrium? Some people say no. You could look at the unemployment graph and say no. And yet, the economy can be "good" for thirty years or more, or "not so good" for thirty years or more. Those could be 30-year periods of equilibrium, even if there are multiple business cycles in each period.

Anyway, Keynes thought the economy experienced equilibrium. The economy could settle into a low-employment equilibrium or a high-employment equilibrium, he said. And who am I to disagree with him.


I was looking at the debt of Domestic Nonfinancial sectors relative to GDP, a US version of the multi-nation data Stephen Cecchetti looks at in The real effects of debt, and I noticed something:

Graph #2: Non-Financial Debt relative to GDP

Three periods of equilibrium. Three different levels of equilibrium in my lifetime. And between times of equilibrium, disequilibrium. Times of undeniable economic calamity: One that ended with a "great" recession. And one that started with inflation so severe that it still informs policy today.


If you want to see equilibrium in the economy, don't look at employment. Look at finance. Why? Because finance is the source of disequilibrium.

"Fixed-weight" problems

RE: Karl Whelan's A Guide to the Use of Chain Aggregated NIPA Data PDF from June, 2000.

The U.S. Department of Commerce in 1996 switched from "fixed weight" calculations to "chain-weight" calculations for converting nominal values to real values. Solved one problem, created another.

I want to understand the problem with the newer method. But I have to start by trying to understand the problem with the older method. Because the problem with the older method, as Whelan describes it, is unbelievable:

While the fixed-weight methodology has the advantage of simplicity and ease of interpretation, it also has a number of undesirable features. Most importantly, the growth rate of a fixed-weight measure real GDP depends on the choice of base year. Take 1998 as an example: The growth rate of fixed-weight real GDP in this year was 4.5 percent if we use 1995 as the base year; using 1990 prices it was 6.5 percent; using 1980 prices it was 18.8 percent; and using 1970 prices, it was a stunning 37.4 percent!

In a footnote, Whelan adds: "These figures actually understate the true pattern."

Yikes. How is this even possible?


I went to ALFRED to see the RGDP data immediately before and immediately after the change:

Graph #1: Last Data Before the Change (blue) and First Data After (red)

Same two series, indexed to the start-date of the red line:

Graph #2: Same Data, Set Equal at Start-of-Red

And as a ratio:

Graph #3: After the Change, relative to Before the Change

Yeah, I don't see anything there. The revision made Real GDP higher, as you would expect. That's our main way of improving the economy these days: revise the data. That's the fallback strategy whenever theorists come to resemble Euclidean geometers in a non-Euclidean world and their deep divergences of opinion destroy the practical influence of economic theory.


Whelan explains:

The reason we get higher growth rates for real GDP when using earlier base years is the well-known problem of "substitution bias" associated with fixed-weight indexes. Categories with declining relative prices tend to have faster growth in quantities; the further back the base year the larger is the weight on these fast-growing categories and so the faster is the growth rate of real output.

He adds:

Similarly, for a given base year, the growth rate of a fixed-weight quantity index tends to increase over time as the output bundle becomes increasingly expensive when measured in terms of the base year's prices. This problem became more severe after the mid-1980s because of BEA's decision to measure computer prices according to the hedonic method...

Yeah, this part I don't get. Why does the growth rate increase as the output bundle becomes increasingly expensive? That doesn't sound right. Giving an example of when it supposedly happened doesn't explain anything.

But the "Categories with declining relative prices tend to have faster growth in quantities" part, that makes sense. If the price of beef goes up more than chicken, people switch to chicken. If beef and chicken prices both go up, people switch to beans.

I'm still troubled by this. I still postpone belief in the truth of Whelan's claim that the "fixed weight" growth rate of real GDP in any given year depends on how distant the base year is from the given year.

I'm still ruminatin'.