## August 05, 2008

### What does the GDP deflator measure?

As Menzie explained yesterday, it isn't inflation. Since there still seems to be some controversy about this issue (e.g., Rich Karlgaard, Instapundit, and Reuters), let me take a stab at it as well.

Rather than walk through the detailed accounts of the U.S., let me invite you to calculate for yourself nominal GDP, real GDP, and the GDP deflator for the economy of Islandia, a pleasant South Pacific locale where they import oil and grow coconuts. In 2007, Islandia produced 500 coconuts, which residents sold to themselves for \$1 each, and imported 1 barrel of oil, which cost \$100. The oil was paid for by borrowing from foreigners.

Ready to calculate GDP? We want the dollar value of domestically produced final goods and services, to wit, the dollar value of the 500 coconuts. Got the answer? Very good. Nominal GDP in Islandia for 2007 was \$500. If you wanted to describe that in real terms, you'd call it 500 coconuts. You don't count the oil in either nominal or real GDP because Islandia didn't produce any oil.

Here are the numbers for 2008. We grew 510 coconuts, sold them for \$1.01 each, and still imported 1 barrel of oil, paying \$125 for it. So nominal GDP was \$515.10 (a 3% increase) and real GDP was 510 coconuts (a 2% increase). The change in the implicit GDP deflator would be the change in the ratio of nominal GDP to real GDP, namely, +1%.

But wait a minute, Islandia's pundits decry. How can your crummy accounting claim that inflation was only 1%? Last year we bought 500 coconuts and 1 barrel of oil for \$600, but this year if we tried to buy the same thing it would cost us \$630. The inflation rate, they tell you, is obviously 5%, not 1%. You must have intentionally cooked the books, they charge, just to make the economy appear better than it is!

You patiently try to explain that imports aren't included in GDP, and that's why the numbers came out the way they did.

But they're not going to believe you.

Technorati Tags: , , , ,

Posted by James Hamilton at August 5, 2008 10:00 AM

digg this | reddit

And then the islanders would ask why you're parroting the separate, official figure of 3% for inflation, when there's this website that claims the government understates the inflation rate, which, it claims, is really about 5.1%.

Posted by: Silas Barta (formerly Person) at August 5, 2008 10:32 AM

All well and good. But let's say 1 barrel of oil is required to produce 500 coconuts in 2007.

In 2008, 1 barrel of oil is required to produce 480 coconuts, so in 2008 we need 1.0625 barrels to produce 510.

Or do coconuts just grown on trees?

Posted by: aaron at August 5, 2008 10:41 AM

"You patiently try to explain that imports aren't included in GDP, and that's why the numbers came out the way they did."

I thought GDP = C + I + G + (Ex - Im)?

Posted by: Ben at August 5, 2008 10:42 AM

Silas Barta (formerly Pearson)-- I would be very interested to see a publicly available spreadsheet that allows anyone else to try to reproduce these claims. I can go through the BEA calculations by myself line by line, but I have no idea how these "alternative" numbers are produced. Why should I then doubt the BEA but trust the shadow statistics?

Aaron, the calculations are a little more complicated (but actually bring out some very important additional points) when we focus on value added rather than coconuts growing on trees.

You're certainly correct, Ben. In this case, for 2007, we have C = \$600 and Im = \$100, so GDP = \$500.

Posted by: JDH at August 5, 2008 10:53 AM

Actually, "Islandia" is "Iceland" is Spanish, so it wouldn't be on the South Pacific!

Posted by: Natalio Ruiz at August 5, 2008 11:22 AM

"Why should I then doubt the BEA but trust the shadow statistics?"

Because the shadow statistics better reflect the effect current economic conditions have on peoples bank accounts.

Posted by: anon at August 5, 2008 11:33 AM

I get that a lot of the inflation we are seeing is imported inflation. Still, 1.1% seems awfully low and decellerating sharply when all other measures of inflation are significantly higher and accelerating, What was PPI annualized in 2Q, something like 14%, CPI 9% or so. I would have expected the deflator to be lower, but still trending upwards. Something more on the order of 2% than 1%. The point about services inflation (largely non traded) in the previous post (sorry forgot who made it) was a good one, and that seems to be running at about 3% or so in the 2Q. Insert 2% GDP deflator and the growth in 2Q looks significantly weaker (although the inventory decline is very good news looking forward).

Posted by: Dirk van Dijk at August 5, 2008 01:38 PM

"Because the shadow statistics better reflect the effect current economic conditions have on peoples bank accounts. "

Is anecdotal evidence really best way to choose which statistics to believe?

As economists have pointed out many a times, people pay more attention to price increases (most recently food and energy) than price decreases (clothes, computers, cellphones, televisions) and pay more attention to things they buy frequently (food and energy) than things they buy infrequently, even if they spend the same percentage of income.

Its really not all that surprising people think inflation is understated. And don't forget that people are more likely to complain that inflation is understated for them than they are to mention that inflation actually seems overstated when accounting for everything.

Also, CPI came in last month at 5% year over year, so even the government numbers are showing unacceptably high numbers. And if you don't trust the deflator... reread both Jame's and Menzie Chinn's articles!

In short, trust a price index based on the method it uses, not on what "seems" right.

Posted by: JohnS at August 5, 2008 01:38 PM

I see two fundamental flaws with the GDP deflator:

1) There are no imports in GDP, yet negative import prices are part of the deflator.

2) The deflator does not account for intermediate goods or services which are crucial for the functioning of our actual economy.

Posted by: Charlie Stromeyer Jr at August 5, 2008 01:51 PM

Try it this way?

The GDP deflator tells you what it would cost to purchase a unit of GDP off the shelf.

But that's not what we consume, a unit if GDP is very far away from the typical market basket consumed by households. There are reasons we want to know what a unit of GDP would cost, but not for purposes of measuring the cost of living.

That's what the CPI and PCE measure for us.

Posted by: Mark Thoma at August 5, 2008 02:25 PM

1:51 pm:

Import value is included in C + I + G.

The amount of domestic product in C + I + G is
C + I + G - M

So it is consistent that the GDP deflator includes negative import prices.

E.g. imported oil inflation is subtracted from total inflation in C + I + G; high oil inflation will reduce the deflator, other things equal.

Posted by: JKH at August 5, 2008 02:35 PM

I am glad that you consider this issue more, as I believe it is important. I'll return later with a more thorough analysis of the issue.

Posted by: Stefan Karlsson at August 5, 2008 02:36 PM

Actually, use C + I + G + E above
instead of just C + I + G;

same argument

Posted by: JKH at August 5, 2008 02:37 PM

JDH: Flipping out numbers in the government's calculations with realistic ones, won't make them redeem my TIPS for more dollars. Believe me, I've tried.

The reasons you should reject the government's numbers -- despite being checkable -- are that:

1) They subtract out alleged quality improvements but don't subtract out quality degradations. I *still* can't find the lines where they've factored in flimsier cereal boxes or less meat content in pepperonis. And I keep getting "no results" when I search for how they factor in promptness of service. (I admit I could be using the wrong keywords.)

2) Actually finding changes in product quality would require lab testing, which the BEA doesn't do.

3) CPI ignores elasticity of the various goods. A gain in food/energy prices while luxury cars goes down, is not equivalent to a gain in luxury car prices while food/energy prices go down, even if their magnitudes and weightings make them come out the same.

JohnS: Do you notice a common characteristic of the things increasing in price, and of the things decreasing in price? Food and energy are things you have to buy often and are hard to hard to stop buying or substitute away from, while computers and cellphones are closer to luxury goods and can easily be put off until later.

So, when an economist says to me, "Consumers aren't really losing purchasing power because electronics cost less", what I hear is, "Oh, you're struggling to pay for food and gas? No, no, you're just IMAGINING that because, like, if you were to smash your iPod for no reason whatsoever, the new one ... um, would cost the same, but ... store more music and have a prettier screen!

Posted by: Silas Barta (formerly Person) at August 5, 2008 02:40 PM

OK, OK, I give. I now understand why the deflator is a measly 1.1% - it especially helped to know that housing costs (including materials costs, which have also plunged) are included in the calculation. That deflation alone would explain alot.

But you're never going to get me to buy in to the OER portion of CPI. :-)

Posted by: Jim D at August 5, 2008 02:40 PM

Table 4
(Price Indexes for Gross Domestic Product and Related Measures: Percent Change From Preceding Period) illustrates the tale nicely. Personal consumption expenditures went up by 4.2% which is much closer to the CPI; durable prices fell by 1.6% and residential investment (houses) prices fell by 2.2%; imports up by 28%.

Posted by: Oleg at August 5, 2008 04:25 PM

JKH, how would you interpret what Erwin Diewert wrote which is:

"Indeed, it is the fact that import quantities have *negative weights* in the GDP deflator that causes it to be unsuitable as a measure of general inflation."

Posted by: Charlie Stromeyer Jr at August 5, 2008 06:14 PM

The U.S. doesn't produce coconuts, however. They produce an incredibly wide variety of raw materials, intermediate and final goods, including oil.

This breadth of production means there is a very strong, very stable long term relationship between CPI and the GDP deflator over time. Sure, it doesn't measure inflation, but our composition of production doesn't change overnight, and so the deflator (and it's relationship to CPI) shouldn't change much either, quarter-over-quarter. The 2nd quarter deflator just blew this relationship up. It's in the wrong direction and too small by several orders of magnitude.

Similarly, there's a pretty strong relationship between employment and GDP over time. Seven months of declining employment decidedly does not indicate that real GDP growth is going to move from -0.2% to 1.9%. And you can lump pretty much every other economic indicator in this boat (construction spending, industrial production, orders).

The only statistic that's out of line here is the deflator. It's simply wrong. The only question is whether it was by accident or design.

Posted by: ts at August 5, 2008 09:25 PM

Touche Silas Barta-

You have itemized several nagging faults with BEA/BLS/Census Bureau hedonics and weighting methodology. Thanks for injecting some common sense, as well as the desire for scientific methodology and transparency.

Dr. Hamilton - As long as the Econobrowser is on a roll regarding US government economic statistics... How about a discussion on alternative measures of economic progress: HDI, GPI and ISEW?

Posted by: MarkS at August 5, 2008 09:57 PM

Charlie Stromeyer Jr:

The deflator is intended to be a measure of inflation within the scope of what it is intended to capture.

In this case, that scope is GDP.

GDP includes what is purchased domestically, excepts that it excludes import content - i.e. it excludes imported inflation.

"General" means what the person using it wants it to mean. If he wants it to include import content, then yes, the GDP deflator misses this.

The point is, its intended to miss it, because its intended to cover GDP, which also misses it.

Other measures capture it, as has been explained.

Posted by: JKH at August 6, 2008 02:56 AM

JKH, if import prices rise then the immediate effect of this rise is to reduce the GDP deflator.

I still don't understand why this is supposed to make sense intuitively. Perhaps we should agree to disagree, or we could just defer to Erwin Diewert's view since he is one of the world's top experts on economic indices.

Posted by: Charlie Stromeyer Jr at August 6, 2008 05:24 AM

As promised, I have written a lengthy response to you (and knzn and Brian Wesbury) on this issue which can be read here.

Posted by: Stefan Karlsson at August 6, 2008 05:56 AM

Charlie Stromeyer Jr:

Not quite what I meant.

Simplified scenarios:

If import prices rise, and the price rise is exactly passed through to the final price, then the CPI increases but the deflator is unchanged.

If import prices rise, and the price rise is more than reflected in the final price, then the deflator increases and the CPI increase equals the sum of the import price increase and the deflator increase.

If import prices risk, but the final price doesn't fully reflect it (i.e. domestic margins decrease), then the deflator decreases, and CPI increases by the sum of the negative deflator effect and the positive import effect

Posted by: JKH at August 6, 2008 06:09 AM

Isn't this example exactly why you should use a consumption deflator (e.g. PCE price index or CPI) rather than a production deflator (e.g. GDP deflator) when considering the welfare of the countries current inhabitants?

Posted by: Anonymous at August 6, 2008 06:15 AM

Silas,

Your point about TIPS argues, in general terms, that a given measure of price change, the GDP deflator, should be perfectly applicable to a specific purpose, such as a TIPS calculation. It shouldn't be, unless it is the intended purpose. The GDP deflator is not intended to work in TIPS calculation, since another inflation measure has been specified in the contract. We have more than one inflation measure because they are meant to fit different needs. Why in the world should the GCP deflator work in a TIPS calculation when the TIPS contract specifies a completely different inflation measure? What you are going to find, if you look far enough, is that there is a deflator for just about every aggregate, and for most broad categories of goods and services. None of them can fit every purpose. That's why there is more than one measure.

There are similar problems with your other complaints. You want promptness factored into inflation measures, but that's just you. I'd like environmental harm factored in, but that's just me. These are idiosyncratic wishes and there is no need - no possibility, in fact - for BLS to account for all ideosyncratic wishes. Same with price elasticity. Putting aside the problems with creating a system that accounts for elasticities, this is just another idiosyncratic wish on your part, not some absolute of price comparison that BLS has ingnored.

If transparency in BLS calculation is a problem, then it's a problem, but it might be addressed by getting in touch with BLS statisticians. Very often, they are willing to describe the processes they use if you ask. If it can be addressed in that way, then transparency is no longer a problem. You may object to the method, but if the objection is ideosyncratic, why should the rest of us care?

ts,

What you say is correct, up to the point that you insist that CPI and the GDP deflator should track well together every quarter. I'm not big on "should" and this is an example of why I'm not. The calculations are there to see. "Should" argues either that the figures should be jiggered to make the two track quarter by quarter, or that there is some mistake or cheat at work. The answer is "no" in both cases. The two figures should track over time, but not every quarter. This is, broadly, the same argument that Fed officials and others make when they say that divergences between core and headline inflation measures should not persist over the long haul. They are troubled that the divergence has lasted so long, but note that "long haul" qualification.

Posted by: kharris at August 6, 2008 06:18 AM

This is why economists are irrelevent. Oil which is the prerequisite for all production goes up in price 25%, but you raise the prices on your nuts by 1%. Any business run this way will be broke in a few years, which is exactly where the USA will be if energy costs continue to rise

Nathan

Posted by: Nathan Walters at August 6, 2008 06:33 AM

MarkS: Thanks :-)

One correction I should make, though. To be consistent I should have said, "They subtract out alleged quality improvements but don't *add in* quality degradations."

kharris: I understand that CPI and GDP deflators are different measures. That's why I said: "And then the islanders would ask why you're parroting the separate, official figure of 3% for inflation, ..." I was not critiquing the GDP deflator, or believing that GPD deflator is used for TIPS with my criticisms, and that should have been clear to you.

As for your trivializations of my CPI crtiques: My "promptness" criticism is not idiosyncratic. It's quite simple:

If a producer finds prices have gone up and therefore replaces some of the sugar in his product with a filler, but keeps charging the same nominal price, then you should be able to see how this is a stealth price increase (charging the same for less stuff), and why a hedonics adjustment should count it as having gone up in price.

Likewise, if service quality goes down -- as identifiable in hard, objective metrics -- this is an instance of degrading a product by filling it with a lower-quality input, in this case (much as it may feel weird to say this) lower-quality labor.

If you're going to subtract off quality improvements, while not even watching for quality degradations, you might as well ditch the whole hedonics thing, because you're throwing in a ratchet effect: stuff will get cheaper when it gets better, but not more expensive when it gets worse. Something could fall in quality 99%, get counted as being "the same price", and then double in quality, and this will be counted as an overall 50% drop in price, when in reality, you're paying the same for 2% of the original quality.

Now, I can understand why you want government metrics to factor in environemental damage, but even if the government measured this, it shouldn't go into CPI, because that damage is distributed to all people, whether or not they use a given product (hence the problem). Also, it is not a way a producer can generally conceal price increases.

If it should be included, it should be in some modified GDP metric so that a decrease in environmental damage shows up as a gain in adjusted median income.

The fact that you can think up a metric for which you didn't even put it in the right category, does not mean instances of product degradation should be ignored on the grounds that caring about that particular measure is idiosyncratic. Concealing cost increases is concealing cost increases.

Posted by: Silas Barta (formerly Person) at August 6, 2008 06:54 AM

JKH, the GDP deflator is affected by relative movements in exports and imports, so what happens to the deflator if import prices rise significantly more than export prices?

Also, what Stefan Karlsson says in his post above seems to make sense to me or does anyone disagree with him?

Posted by: Charlie Stromeyer Jr at August 6, 2008 07:18 AM

JohnS:
"Is anecdotal evidence really best way to choose which statistics to believe?"

No, but the rate at which people save money (for those who actually try to save) is quantifiable.

And if people can't afford the reduced-price "occasional" goods because necessities are costing them a lot more, then those goods should not be given very much weight in a measure.

Black box or not, I'm going to trust the model that better-fits my situation. Not a "sense" of what someone else tells me my situation is.

Posted by: anon at August 6, 2008 08:53 AM

As one of the vast non-economist unwashed masses, the biggest problem I see with this way of calculating things is that it ignores the reality of the world.

When the government says "GDP growth is fine", that means to us plebs that our bosses and leaders will say "all is good, stop complaining, you're just a bunch of whiners". And they'll continue gorging themselves on the last drops of punch in the bowl.

Meantime, we see all around us in the empty offices, quiet malls, stagnant wages, increasing layoffs, that all is not good. Growth, which to us mere mortals really means more job opportunities, has come to a standstill.

So, it seems to me that rather than justifying the math behind the current determination of GDP, and backing the govt's "all is well" message with numbers, economists should be extremely concerned that the whole system is completely out of touch with reality.

It didn't end well for the last bunch that said "let them eat cake"...

figuring out how to make the calculation

Posted by: foo at August 6, 2008 09:01 AM

FWIW, Randall Forsyth writing in "Current Yield" in Barron's agrees with both Stefan Karlsson and I:

http://online.barrons.com/article/SB121763145053806001.html?mod=9_0031_b_this_weeks_magazine_market_week

Posted by: Charlie Stromeyer Jr at August 6, 2008 09:02 AM

Silas,

If in one place, you make a point and in another you don't make it, it isn't my job to figure out which you mean. I was responding to this from you --

"Flipping out numbers in the government's calculations with realistic ones, won't make them redeem my TIPS for more dollars. Believe me, I've tried."

This certainly seems to imply that the topic at hand - the GDP deflator - ought to have something to do with TIPS.

"Simple" is not the opposite of "idiosyncratic". I addressed your request that promptness be assessed in making inflation calculations by saying the request was idiosyncratic. Nothing you said addressed that point. Your response is to discuss filler in sugar, not promptness and not the general value of including promptness in price calculations.

You seem to have concluded, wrongly and with no evidence, that I am opposed to accounting for quality degradation in price measures. I am not. Your initial arguments was faulty and I pointed that out. If the example of accounting for environmental quality seemed outrageous, that was more or less the point - that too, is faulty. I wasn't going for technical accuracy. Sorry you missed that.

There are problems with a number of government economic measures, and those problems should be addressed. Making up weird requirements (even if you later decide you don't really mean it) and asking that inflation measures be adjusted to your own liking don't contribute to improving government data.

Posted by: kharris at August 6, 2008 09:16 AM

Silas Barton, you are exactly right about service - and about the impaired hedonic adjustment system. For instance, one can easily collect stats that show an increase in canceled flights from airlines. In fact, there was a tremendous spate of them this spring, coming all at once. From the hedonics perspective, we have a one time inflationary spike, then. But this simply isn't reflected in the basket of goods and services having to do with inflation. From a hedonics perspective, things like natural disasters are also inflation causers.

Economics often seems like the art of segregating the sets of objects which one wishes to statistically represent until you get your desired picture. Certainly all of those free traders, who refuse to recognize within country trade as being any different from between country trades, suddenly remember that difference when they are doing cost of living statistics. Of course, what we need is not stats telling us what economists think is inflation, but stats tell us simply how much more or less expensive it is to buy a car, heat a home, send the kids to college, get a medical examination, etc. Knowing this would be good and useful information for us as citizens, in order to plan our lives. Economists, however, do seem intent on making sure such information is obscured, in part because economists are, for the most part, the theologians of the plutocracy. Who get all tearful that the average citizen doesn't remember the wonderful deflation in the price of video games from Japan, but instead remember silly things, like the stagnation of their wages or the slowing of the rate of their wage increases. Silly yahoos!

Posted by: roger at August 6, 2008 09:30 AM

One of the ways i have always tried to explain the problems many people have with imports in the GDP accounts is to point out the GDP is an estimate of production. But we do not measure production.
Rather, we measure consumption and adjust that for changes in inventories and trade to only indirectly derive a measure of production.

This is in line with the comment we are all guilty of using that consumption accounts for two-thirds of GDP. No it does not. Consumption accounts for zero percent of production. Consumer spending account for two-thirds of consumption, not production.

Posted by: spencer at August 6, 2008 10:11 AM

Spencer-

Regarding your last paragraph - nicely put. Furthermore the main reason that living standards have improved so much since the industrial revolution is that we do not consume all that we produce.

Posted by: Rich Berger at August 6, 2008 10:23 AM

I welcome this continuation in the discussion of the concept of inflation.

It seems to me that the GDP deflator and the CPI do roughly what they are supposed to do, but I do not think that either of them represent what one would expect "inflation" to mean. As its name implies, I would expect inflation to be about an overall expansion of the price structure by an increase in money supply. I would therefore not expect it to be adjusted for substitution or quality change.

If such adjustments are to be made, then it is only fair to adjust historical statistics a long way back, and come to different conclusions about trend growth etc. In the mid-nineties, it used to be inferred from Fed speeches that their acceptable rate of inflation was 2%. OK, so now post-Boskin, it should be, say 1%. But somehow, it does not seem to work out that way.

Posted by: RebelEconomist at August 6, 2008 10:42 AM

roger: Thanks! :-) And remember folks, if you like my posts, don't forget to click on my name to check out my internet website.

kharris: Wow, you took pretty much everything in my post out of context. Let's see if I can't reconnect the moving parts.

- You're right that exerting effort in reading my posts is not your job, but then, nothing here is really anyone's job. Still not an excuse to chop my posts into meaninglessness.

- If you looked at what I replied to in the TIPS remark, I was replying to JDH's defense of the statistics that "they can be audited." I replied that that doesn't make them more credible, since proving the auditable numbers wrong doesn't change the payout on TIPS, i.e. supposed hedges against inflation.

- I didn't claim that idiosyncratic was the opposite of simple; I was claiming that my justification for measuring tangible components of service quality, was simple, in that it followed directly from the claimed purpose of collecting inflation data. The interest in wait times is therefore not idiosyncratic, but one of the better metrics that can capture changes in service quality, and thus, inform hedonic adjustments.

- My discussion of sugar/fillers was to help you understand the impact of quality measures in determining changes in price, and how they can be hidden. I then bounced off that discussion to show its counterpart in how changes in the price of *services* can be hidden. (And by the way, if you seriously saw the mention of sugar in my response to a discussion of promptness, and concluded from that that I was necessarily off-topic, that seriously calls into question your ability to handle higher-level concepts. Please tell me you made a more respectable error than that.)

- I did not conclude that you were opposed to including quality degradation in price measures, just that you did not appreciate the necessity of one metric I mentioned (service promptness) is accurately accounting for quality degradation.

- Making an outrageous counter-proposal would, in no way, help establish my original proposal to be unjustifiable.

- Your counterproposal acutally wasn't in and of itself outrageous, it's just that the measure even at best wouldn't be applicable to inflation -- it belongs as a component of GDP, and, I should note, even there doesn't need to be factored into the offiical numbers, since even if they did calculate it, it's easy for anyone to add on to create whatever metric they want for whatever purpose. (i.e. you don't have to worry about untangling a complex relationship between the two)

- I never "made up some weird requirement and later decided I didn't mean it".

- I didn't claim government inflation measures should be adjusted "to my own liking". Rather, I identified a specific, measurable, typical way that price changes are not being captured.

Whew!

Posted by: Silas Barta (formerly Person) at August 6, 2008 11:02 AM

A better example would have been coconuts and a fertilizer that is essential to grow coconuts. In this case had fertilizer doubled in price, I have feeling that \$515.10 worth of coconuts would not have gotten you the 510 initially assumed by the Island's BEA in the first round of its Real GDP calculation, but rather only 470 as the higher price of fertilizer increased the price of the final good, hurting demand. Thus, rather than an increase in GDP from 500 to 510 coconuts, there was a recession to 470 (same nominal GDP in both cases).

At the end of the day, we need to look at the numbers and ask ourselves if they make sense and 1.1% does not. Should the GDP Deflator be as high as that implied by the CPI? Probably not, but somewhere in between would cause another revision to GDP, which was my point all along...

Posted by: Jake at August 6, 2008 11:14 AM

Maybe this helps:

Consumption is C=Cd+Cm (consumption of domestically produced goods + foreign imported goods)

and Cm = M

Thus, when we talk about Y=C+X-M we are really saying that:
Y=Cd+Cm+X-M canceling out the effect of imports and giving you domestic production.

Imports enter both positively and negatively. To capture the effect of higher import prices on consumers, a deflator would need to maintain Cm in the equation (not subtract M). That way consumption (C) would include what is bought from abroad.

Posted by: celo at August 6, 2008 12:50 PM

I do not understand this whatsoever. Let's say the entire domestic economy is composed of selling widgets, which producers purchase component parts from foreign manufacturers valued at 97 cents. Then the domestic company assembles and packages the products, which add another 2 cents in cost. The widgets are sold at \$1, thus generating a penny of increased wealth. The country sells \$1B in widgets. Is GDP \$1B, or is it \$1B, less \$0.97B (i.e. subtracting the cost of the foreign components)?

Posted by: Tax Lawyer at August 6, 2008 01:14 PM

Tax Lawyer, GDP for your economy would be \$0.03 B.

Posted by: JDH at August 6, 2008 01:22 PM

celo, what happens if you consider these as primary inputs used by domestic producers:

imports
labor inputs
capital inputs
natural resource and land inputs

like Erwin Diewert does?

Posted by: Charlie Stromeyer Jr at August 6, 2008 01:29 PM

That is weird. My post is under "Jake", while the post under "Celo" is someone else's.

Very weird.

Posted by: celo at August 6, 2008 02:41 PM

And now it posted under "Charlie Stromeyer Jr"

For the record, my (Celo) posted this:

Maybe this helps:

Consumption is C=Cd+Cm (consumption of domestically produced goods + foreign imported goods)

and Cm = M

Thus, when we talk about Y=C+X-M we are really saying that:
Y=Cd+Cm+X-M canceling out the effect of imports and giving you domestic production.

Imports enter both positively and negatively. To capture the effect of higher import prices on consumers, a deflator would need to maintain Cm in the equation (not subtract M). That way consumption (C) would include what is bought from abroad.

--Celo

Posted by: celo at August 6, 2008 02:43 PM

Ok, I'm stupid. The name comes *after* the post.

I'm going home now...

Posted by: celo at August 6, 2008 02:44 PM

JDH:

to the extent that these are reflected in the final price of the commodity, it should show up in the PCE or CPI deflators. They expenditure method is measuring final expenditures. The problem is that by including the (-M) term, you are removing the price of final purchases of imported goods.

Posted by: celo at August 6, 2008 02:48 PM

Readers who, as I did, found Menzie Chinn's reference to Diewert's Palgrave review of inflation measurement useful may also like this more recent review paper:

A review of reviews: Ninety years of professional thinking about the consumer price index
by Marshall Reinsdorf and Jack Triplett

http://www.nber.org/chapters/c5068.pdf

Posted by: RebelEconomist at August 6, 2008 02:53 PM

kharris,

Give readers fixated with cost of living an article tangentally related to cost of living, and they'll fill this space with complaints of inadequate accounting for cost of living.

JDH,

Thanks this is fun.

Posted by: KevinM at August 7, 2008 10:47 AM

Tax Lawyer, yes, GDP in your economy is \$0.03 B, because you actually pay for your imported components. The most important assumption in this post is "The oil was paid for by BORROWING FROM FOREIGNERS." If your imported components were also financed by borrowing from foreigners, the answer to your question would be \$1B

Posted by: Anna at August 8, 2008 09:41 AM

No Anna, to pay for the imports by some mechanism other than borrowing or selling assets would require offsetting exports. My answer to Tax Lawyer (GDP = 0.03) assumed that net exports = -0.97.

Posted by: JDH at August 8, 2008 09:48 AM