GDP Has a Glaring Blind Spot

Written by Noah Smith via Bloomberg View on 12/15/17 

When you read words like “output,” “economic growth” or “national income,” you’re almost always reading about gross domestic product. This measure, more than any other, has come to be equated with national prosperity and living standards. We use it to compare countries with each other, and also to compare the present with the past.

But GDP misses many important things. It neglects the value of leisure time, and of all the activities that people do outside of their jobs -- cook food, take care of kids or listen to each other’s problems. Others note that it neglects to consider environmental QuickTake GDP and economic inequality. A few have even suggested supplementing GDP with measures of national happiness.

But I want to point out a more prosaic and boring, yet also more subtle and insidious problem with GDP -- it fails to account for quality differences. And these can be large.

One component of GDP is consumption, which measures how much money people spend on goods and services -- when you buy a $500 phone, you’re adding $500 to GDP. If I spend $500 for a low-quality flip-phone and you spend $500 for a top-of-the-line iPhone, we’re both adding the same amount to national income, but the consumer probably derives much more utility from the better product.

Presumably, of course, people would spend more on the higher-quality version, but not necessarily. People might be bad at estimating the quality of a product. This is especially true for things that last a long time, or which take a long time to show their value. For example, suppose two new brands of car come onto the market -- one that’s well-built and will be running fine 10 years from now, and another that will be ready for the junk yard in four years. Since the brands are both new, consumers can’t distinguish between the two based on reputation. The cars look and run the same, and cost the same, so people pay the same, even though one is of superior quality.

Or consider education. The cost to go to a private school is paid upfront, but the benefit can only be seen years down the line. If a school is well-established, you can look at the performance of its alumni, but schools open and close all the time.

I think about this problem a lot when I visit Japan. The clothes I buy in Japan tend to last much longer than those I buy in the U.S., and yet I end up paying the same for both. While my experience might be anecdotal, I would bet that no government agency ever bothered to test clothes for longevity before incorporating their price into the GDP numbers.

Japan in general provides a good case study in how quality differences can go unnoticed in the statistics. Officially, Japan is considerably poorer than the U.S.

But there are lots of ways that these numbers fail to capture how life in the two countries is different. Some are the kind of things that never get counted in economic statistics -- Japan has much less crime, for example, but also a more punishing corporate culture. But others represent differences in the quality of goods and services that people pay money for -- things that GDP ideally should measure, but doesn’t in practice.

Tables in Japanese cafes, for example, tend to be spotlessly clean, thanks to the tireless efforts of the employees. Go to most Starbucks in the U.S., however, and crumbs and other detritus will cover many of the tables. One would think that the value of clean tables would make Japanese customers pay more for their coffee, but in practice it’s almost impossible to tell whether this actually happens.

Many of the little quality differences  -- Japan has longer-lasting products, cleanerbuildings and streets and food that’s healthier but still tastes great -- won’t show up in the relative GDP numbers. Economists can correct for relative price differences by measuring the disparate prices of non-tradable goods and services -- an adjustment known as purchasing power parity. But PPP will tend to miss a lot of these quality differences.

Economists occasionally try to use a more sophisticated technique called hedonic estimation, which  tries to gauge how much people are willing to pay for certain characteristics of a good or service, and then looks at how those characteristics change over time. But this method can miss a lot of important things -- it’s unlikely, for example, that government statistics agencies are sending people around to check how clean the tables are at Starbucks. Also, it has trouble dealing with the issue of product longevity, or with things like education whose value only becomes apparent down the line. Even when it works, hedonic estimation is much more useful for comparing GDP across time within a single country than for comparing between countries.

So while GDP is a good measure of the really big economic differences between rich and poor countries, it’s not very good for comparing the true economic well-being of different countries at similar levels of development. If you want to know whether Japan, the U.S., Germany, South Korea or some other rich country is truly wealthier, you can’t just look at the GDP numbers -- you also have to live there and see for yourself.