Wednesday, May 20, 2015

The Problem With GDP Data Collection

My favorite Keynesian economist, Martin Feldstein, the George F. Baker Professor of Economics at Harvard University, and the president emeritus of the National Bureau of Economic Research, writes at WSJ:
Today’s pessimists about the economy’s rate of growth are wrong because the official statistics understate the growth of real GDP, of productivity, and of real household incomes. Government statisticians are supposed to measure price inflation and real growth. Which means that, with millions of new and rapidly changing products and services, they are supposed to assess whether $1,000 spent on the goods and services available today provides more “value” or “satisfaction” to American consumers than $1,000 spent a year ago. Even more difficult, they are tasked with estimating exactly how much it costs now to buy the same quantity of “value” or “satisfaction” that $1,000 could buy a year ago.

These tasks are virtually impossible, and the problem begins at the beginning—when an army of shoppers go around the country at the government’s behest to sample the prices of different goods and services. Does a restaurant meal with a higher price tag than a year ago reflect a higher cost for buying the same food and service, or does the higher price reflect better food and better service? Or what combination of the two? Or consider the higher price of a day of hospital care. How much of that higher price reflects improved diagnosis and more effective treatment? And what about valuing all the improved electronic forms of communication and entertainment that fill the daily lives of most people?

In short, there is no way to know how much of each measured price increase reflects quality improvements and how much is a pure price increase. Yet the answers that come out of this process are reflected in the CPI and in the government’s measures of real growth. This is why we shouldn’t place much weight on the official measures of real GDP growth. It is relatively easy to add up the total dollars that are spent in the economy—the amount labeled nominal GDP. Calculating the growth of real GDP requires comparing the increase of nominal GDP to the increase in the price level. That is impossibly difficult.

The measurement problem is particularly severe for new products. Consider a new drug that improves the quality of life, reducing pain or curing a previously incurable disease. The ability to buy that new product means that a dollar is worth more than it used to be, and that the properly measured level of real GDP is higher. The official method of calculating the price index doesn’t incorporate this new product until total spending on it exceeds some threshold level. It is then added to the government’s price calculations, but only to record whether the cost of the drug goes up or down. The main effect of raising well-being when the drug is introduced is completely ignored. The same is true of other new products.

The result is that the rise in real incomes is underestimated, and the common concern about what appears to be the slow growth of average household incomes is therefore misplaced. Official statistics portray a 10% decline in the real median household income since 2000, fueling economic pessimism. But these low growth estimates fail to reflect the remarkable innovations in everything from health care to Internet services to video entertainment that have made life better during these years, as well as the more modest year-to-year improvements in the quality of products and services.

While changes in officially measured real GDP statistics don’t fully capture increases in Americans’ standard of living, year-to-year fluctuations are still useful as one indicator of business-cycle conditions.

Note: Feldstein is not saying there is not a business cycle, he is merely pointing out that there are some secular trends that improve the standard of living for many.  -RW


  1. There are secular trends that improve life and there are secular trends that degrade life. So how does he know the improvements outweigh the degradations? His observations do not support his conclusion. Further I would argue he is correct when he writes: "In short, there is no way to know how much of each measured price increase reflects quality improvements and how much is a pure price increase." If he is speaking about the government. However, individual consumers can and do adjust their purchases based on their perception of changes in value to them. And government's inflation of the money supply, whether it exceeds the rate of growth of productivity or not always degrades the individuals life. The governments inflation of the money supply is always theft.

    1. Brian: Similarly, and more simply, we could argue that as a consumer, I dont need to know any of the complicated duplicitous measurements that brainwashed economists can contrive. I could more accurately take note of my higher expenses, and roughly guesstimate my personal inflation.

      Who is to say what basket of goods is THE MOST correct?

      But more to the point, as production efficiency increases, prices will go down. As the market expands for a certain product (LCD TV's), more manufacturers enter the arena, and again, prices will go down. Thus in this healthy deflationary cycle, employment will likely go up, yet national measurement of GDP will deceptively appear to go down, perhaps being labeled a recessionary contraction by the brain dead economists and media.... any comment?

      Charles N

  2. Also, we can add to the uncertainty presented by GDP measurements and data collection the following questions:

    1-Since most GDP measurements include government expenditures, one could argue, that the aprox $4 trillion budget (federal expenditures), which is theft from the productive sectors, is much less efficient at contributing to national prosperity, and jobs, than the generators and owners of this wealth would have achieved (unless you're a Keynesian:) My unsupported guess: it loses up to 75% of its economic horsepower,

    2-There are several alternate measurements of productivity, but the most interesting was proposed by Rothbard. Private Product Remaining (PPR).

    Charles N