David R. Henderson  

Feldstein's Insight on Standards of Living

Left and Right: A Socratic Dia... Cold James Buchanan...

In a recent op/ed in the Wall Street Journal, my former boss at the Council of Economic Advisers and Harvard economist Martin Feldstein points out that the data on real incomes in the United States systematically understate its growth. The article is titled "We're Richer Than We Realize," WSJ, September 8 (September 9-10 in the print edition.)

An excerpt:

Consider how the government handles manufactured products when their quality improves. Statisticians track a large number of products. For each, they ask the manufacturer two questions: Has the product changed since last year? If so, how much more does it cost to make this year's model than it would now cost to make last year's model?

If there is no increase in the cost of production, the government concludes that there has been no increase in quality. And if the manufacturer reports an increase in the cost of production, the government assumes that the value of the product to consumers has increased in the same proportion.

That's amazing! I knew, and have written about, the fact that the government understates improvements in quality. I had not known how naively the government did that.

It reminds me of something I did know and reported on in "The Digital Economic Revolution," Red Herring, September 1, 1997. I wrote:

Which brings us back to the government data. To compute labor productivity in an industry, the federal government's Bureau of Economic Analysis divides the output of an industry by the number of people employed. Not bad for, say, copper mining, where tons of copper mined is a pretty decent measure of output. But how do you think the federal government, with all its high-powered analysts and its multimillion-dollar budgets for gathering data, measures productivity in the banking industry? The number of transactions per employee? Or maybe the per-employee value of deposits and loans, adjusted for inflation?

Neither. The Commerce Department's august Bureau of Economic Analysis measures output of banking by the number of people employed in banking. This means that if the number of banking employees rises by 10 percent, then the government's data crunchers simply assume that output rises by 10 percent. Therefore, the banking industry's productivity growth is zero, not by observation, but by definition.

Of course, productivity in banking is growing. According to surveys by the Bank Administration Institute, the number of checks processed per hour, a measure of bank workers' productivity, rose from 265 items in 1971 to 825 in 1986, a rate of increase of 7.6 percent annually. Presumably computers were a factor in this productivity growth. And as noted by Martin Baily, an economist at the Brookings Institution, and Robert J. Gordon, an economist at Northwestern University, the per-check processing costs for electronic funds transfers (EFTs), which were made possible by the information technology revolution, are a fraction of the cost of conventional check processing. EFTs still constitute only a small percentage of transactions, but as this segment grows, productivity will increase.

Marty points out another factor that understates growth:
There are other problems that cause the official statistics to underestimate the true growth of real income. A basic government rule of GDP measurement is to count only goods and services that are sold in the market. Services like Google and Facebook are therefore excluded from GDP even though they are of substantial value to households. The increasing importance of such free services implies a further understatement of real income growth.

Sadly, a number of commenters on the Journal's site failed to get his point. I'll quote three and, rather than tell you what's wrong with the commenters' statements, leave that as an exercise for the reader. Remember that these statements are made by people who, it is clear from tone, think they're challenging Marty's argument.

matthew kimball writes:

Another example is food has gone up and quality hasn't, gas has gone up, home price have gone up, clothing has gone up. All the essentials have gone up.

Gregory Weinman writes:
Telling me the LG G6 I bought to replace my LG G4 is far better is immaterial because both cost the same.

John Callahan writes:
Mr. Feldstein sounds much like New York Fed President William Dudley did half a decade ago--out of touch with the average American. Mr. Dudley was in Queens touting improvements in technology in regards to the cost of living- "Today you can buy an iPad 2 that costs the same as an iPad 1 that is twice as powerful.... You have to look at the prices of all things." The residents of Queens were far more concerned about rapidly rising grocery and gasoline prices and rightfully so. As one resident noted, "You can't eat an iPad."

HT@ Francois Melese.

Comments and Sharing

COMMENTS (15 to date)
LK Beland writes:

From his WSJ article:
"My own judgment is that the true annual growth rate could exceed the official figure by two percentage points or more, implying that the true annual rate of real per capita income growth during the past two decades has been much more than double the official 1.3%."

I thought that typical estimates from overestimation of inflation indices were of the order of 0.5% to 1.0%. 2% seems very high to me. Surprisingly, he is arguing that the Obama administration has overseen an economic boom.

Right Wing House Music writes:

I presume the point is that assuming that real incomes are stagnating based on flawed government data, and making policy decisions from that flawed data can lead to negative outcomes.

However, I would like to point out that if the entire concepts of "inflation" and "consumer price index" are based on faulty assumptions about productivity and product quality, then most of macroeconomics needs to be thrown in the trash and rebuilt from the ground up.

Micke writes:

I for one am impressed that matthew kimball, Gregory Weinman and John Callahan managed to publish their comments using only food and rent, as technology apparently isn't a part of their lives.

BC writes:

"Free" services like Facebook and Google do earn revenue from advertising. If we assume these firms are profit maximizers, then isn't it fair to conclude that they don't charge users for their service because the advertising revenue they get as a result of not charging users (and, hence, having more users) exceeds what they would have been able to collect in user fees? In that case, don't the advertising revenues already include the value-add to users? In fact, the advertising revenues over-estimate the value to users because users, by assumption, would not have been willing to pay as much to use the services. An analogous situtation might be where a firm gives away an iPad as part of a promotion. No consumer buys the iPad, but the firm does, so the iPad gets counted in GDP. Can't we think of advertisers as buying search services from Google and giving away those services to Google's users?
Do advertising revenues get counted in GDP, perhaps as investment instead of consumption?

Matthew Opitz writes:

It is impossible to calculate GDP in terms of utility because utility is subjective. If a nation starts off the year owning 100 apples and ends the year owning 120 oranges, has that nation gotten richer? If you really like oranges, you'd think yes. If you really like apples and hate oranges, then that year was actually a catastrophe.

Prices, however, are objective (contra the subjective theory of value) over the long term. The classical economists called these long-term objective prices "natural prices" because they saw that supply and demand fluctuations evened out around these centers of gravity. These were the prices that made it typically profitable to produce a commodity assuming a typical cost of production.

If a market price was above the natural price for a commodity (possibly due to a surge of demand outstripping supply), the market price was bound to come down as more investors flooded into that line of business. Conversely, if the market price was below the natural price for a commodity (possibly due to a sudden unfashionableness of the commodity), the market price was bound to come up as investors fled this unprofitable line of business and reduced supply. Even if the demand changes (i.e. perceptions of the utility of these commodities) was permanent, prices would still converge (or rather, fluctuate turbulently) around the natural price, albeit with supply permanently adjusted upwards or downwards, as economist Anwar Shaikh has described in his latest book "Capitalism:
Competition, Conflict, Crisis."

So, the only GDP measure that matters is nominal GDP. Ah, but WHICH nominal GDP? Venezuelan pesos aren't very useful, right? Correct. In fact, anything that isn't "alpha-money" (the strongest, most relatively-appreciating money over the long-term) doesn't tell you much. Measuring GDP growth and profits in terms of alpha-money is the only logical baseline to use because it tells you whether society (or any individual producer) is wealthier now (in an objective sense) versus the counter-factual opportunity cost of if that society/producer had simply hoarded the alpha-money over that period of time.

In other words, if you make profits or see growth in terms of Venezuelan pesos but make losses when measuring your returns in terms of U.S. dollars or ounces of gold, then that tells you that you would have been better off hoarding dollars or gold. This holds true even if you can now buy what appears to you to be greater utility with that later amount of Venezuelan pesos. Most economists would probably say then that you've made a "real profit" or seen "real growth in real GDP." Nonsense. You've lost money in an objective sense in terms of alpha-money. You would have been able to buy even more utility if you had just hoarded alpha-money!

So, what is alpha-money! There are strong theoretical reasons for why it must be a commodity-money (see Sam Williams's "Critique of Crisis Theory" blog for more details on that), and empirically we know that it has been gold for the past 500 years at least of world history. There are strong theoretical reasons for why bitcoin, for example, cannot possibly be an alpha-money over the long-term, despite having had a good run recently. The biggest reasons is that it has no objective, socially-necessary cost of production. Its cost of production is entirely arbitrary, based on the algorithm and how many miners are trying to take advantage of it, and thus it has no determinate "natural price" to give its market price an anchor during speculative jitters. (Gold, on the other hand, has an objective "natural price" and thus, despite being a nearly useless material, has a stable anchor for its exchange-value).

So, the only way that I will ever measure profits or GDP is by first converting all currency prices into golden prices (i.e. commodities being worth x many ounces of gold). If you do that, you correctly see that the 1970s were a disaster for U.S. GDP, the 1980s and 1990s were a great recovery, and the Great Recession was likewise another great disaster up until about 2012, when our golden GDP has started to recover since then.

Jon Murphy writes:


"Do advertising revenues get counted in GDP, perhaps as investment instead of consumption?"

As I understand it, advertising is assumed to be part of the final price of the good, so it is not directly counted in GDP. Furthermore, since the "final price" of Google et al is 0, then the assumption would be that there is 0 advertising, and thus no contribution to GDP.

robc writes:
most of macroeconomics needs to be thrown in the trash and rebuilt from the ground up.

I agree with the part before the "and".

Mattb writes:
Telling me the LG G6 I bought to replace my LG G4 is far better is immaterial because both cost the same.

This quote from Gregory Weinman is very strange.

Looking online, the LG G6 new is about $650. The LG G4 new (you can still find them some places) is about $200 now. If the LG G6 wasn't better and he didn't want the new features of the G6, why did Gregory buy it? He could have saved $400 by replacing his G4 with another G4.

Even given that, though, the new and improved phone still costs the same now as the G4 did when it came out? I don't see what his complaint is? He gets more for the same price. Inflation adjusted, the price is even lower.

Nick writes:


As you point out, the price FB would be able to charge users (in the business judgement of FB)is less than the amount it could receive in advertising revenue. That does not mean value to consumers is overstated or understated - it simply means that FB can make more with advertising than it can offering a no advertising platform with user fees. Facebook runs a near ZERO marginal cost business and entry is a serious concern.

David R Henderson writes:

@LK Beland,
I thought that typical estimates from overestimation of inflation indices were of the order of 0.5% to 1.0%. 2% seems very high to me.
Good catch. I didn’t quote that part because I didn’t find it persuasive. I should have done so and then criticized it as you do. Here’s The Concise Encyclopedia of Economics entry on Consumer Price Indexes, written by my Hoover colleague Mike Boskin. His estimate in the range you cite: 0.8 to 0.9 percentage points per year.
Surprisingly, he is arguing that the Obama administration has overseen an economic boom.
Yes. I don’t remember Marty arguing that in, say, 2015. :-) Although, truth be told, he may have and I missed it.

David R Henderson writes:

@BC, Jon Murphy, and Nick,
Re advertising, I’m not sure about the answer. I’ll think about it. If I come up with something, I’ll post as a new post rather than as a comment.

Vivian Darkbloom writes:


The following article written by a BEA economist addresses the issue:


Floccina writes:

BTW IMHO it would be better to state the annual rate of real per capita income growth in dollars rather than in percent becuase it is not more likely to bet expontial than liniar.

Also on gasoline, engines have been getting more efficent over time. A car that once got 10 MPGs now gets 20MPGs.

Glen writes:

Facebook is probably not a good example of a 'free' service. The users are more akin to consultants hired to do market research and paid in trade.

"Gregory Weinman writes:

Telling me the LG G6 I bought to replace my LG G4 is far better is immaterial because both cost the same."

This quote might be a jab directed against the methodology being described rather than the article. Just saying.

Comments for this entry have been closed
Return to top