Is Gross Domestic Product grossly deceiving?

Why GDP Growth is an Outdated Model


Paul Donovan, UBS’ Global Chief Economist, examines the origins of GDP and its uncertain future in today’s global landscape.   

The world in 1934 was a very different place from the world today. The US monetary union had imploded the year before, and was shakily being pulled back together with preparations for reforming the Federal Reserve. Trade was increasingly subject to protectionism. Domestic service was the largest single occupation for women in the UK – housework was an activity that was paid for, at least for the middle and the upper classes.

This was the economic world into which the forerunner of gross domestic product (GDP) was born. Gross national product (GNP) was presented in a report to Congress in January 1934 although it was actually an interpretation of an earlier British concept. The report was an instant bestseller – Americans in the 1930s had a reverence for economics and economists that is sadly all too often lacking in the world today, in our view.

The principle behind GDP was simple enough. Calculate the value of everything produced in an economy over a set period of time. That meant that there had to be a market price for the economic output. Leisure, housework (not done by a servant), or do-it-yourself did not count towards GDP because there was no market value for it.

War substantially increased the incentive to calculate GDP. GDP allowed governments to work out the potential for tax revenues (a motive with a history dating back to the 17th century). The Second World War was a total war, requiring complete mobilization of economic resources. For a country to mobilize all of its economic resources, the government had to have some idea of what those resources were in the first place.

Prisoners Of War, WWII

After the war ended, GDP really caught on. By the 1950s, all major economies were publishing the data, and have continued to do so in a largely unchanged manner ever since. Over 80 years after that rush to buy copies of the first GDP report, governments are still using the same philosophy in measuring the size of their economies. So what is wrong with GDP as a measure? There are two problems with the use of GDP. The first is whether GDP captures what it is supposed to capture. The second is whether GDP is what we want to measure at all.

It is not practical to actually measure all economic activity in a timely manner. GDP is a series of educated guesses as to what is happening in real time, and those guesses are revised when additional information comes in. Much of the data relies on surveys which sample the population and find out what they are up to.

The problem with any survey-based data in a world where “customer satisfaction” is sought at the click of an email is that people have stopped filling in forms or surveys; if they do fill in surveys, they are less likely to answer all the questions; and if they do answer a question, they are less likely to answer accurately. This makes real time economic data less and less reliable.

Further accuracy problems arise as the world stands on the cusp of dramatic change in how and where people work, and how and what people consume. The World Economic Forum summarises this as the “Fourth Industrial Revolution.”

If real-time GDP does not rapidly reweight its assumptions amidst revolutionary change, then real-time GDP data will measure a world that no longer exists.

This all means that GDP data is being revised more frequently than in the past, and the revisions to the data are larger than in the past. However, financial markets do not acknowledge this deterioration in GDP data quality. Markets still react to the initial GDP data release, which is by definition the least accurate data available. Because the process of revision can take years, investors tend to ignore the subsequent improvements to the accuracy of the data.

Big data, sadly, does little to improve matters. GDP is supposed to measure total economic activity. While many forms of big data are available, most of the numbers that are generated tend to be web-based and skewed towards the economic activities of larger companies. In developed economies, small companies (what Americans call the “mom and pop” businesses) account for around 60% of private sector GDP and around 70% of private employment, and these companies are less likely to be represented in big data. Transaction-based big data excludes the consumption of those citizens who lack regular access to the internet, or who do not have access to electronic money (which means anyone who does not have a bank account). These limiting factors mean that lower income groups and the elderly tend to be significantly under-represented in big data statistics.

Even if GDP could be made more accurate, there is still the question of whether GDP is the right statistic for policy-makers and investors to prioritise. GDP assumed importance as a measure during a period of all-out war, when governments were seeking to maximise resource utilisation with little thought beyond the immediate objectives of the war effort. GDP was never really intended to measure well-being or sustainability. However, governments and media are inclined to try to torture the GDP data into representing the standard of living in an economy (when it does not).

GDP only measures economic transactions. At the turn of the century, in Japan around 2% of those aged over 65 lived in some kind of care facility, while in Denmark the proportion was almost 15%. Culturally, old-age care in Japan has taken place within the family (although this is changing). This means that old-age care in Japan is less likely to contribute to Japanese GDP; unpaid family assistance is not recognised in the data. The same care in an institution in Denmark will add to Danish GDP.

The value of leisure is also largely unrecognized by GDP. Unless someone is paying for their leisure activity, there is no recognition of the value of leisure. This undermines the usefulness of GDP as a measure of living standards. Involuntary leisure (unemployment) may not be desirable, but many people in employment believe that being allowed leisure from work enhances their standard of living. There has been some attempt to include illegal activity (the “shadow economy”) into GDP calculations, but this is really just a means of incorporating economic transactions that the government is unable to tax, and does not address the issue of non-economic activities that nonetheless increase well-being.

Perhaps the most significant criticism of GDP today is that it does not recognise environmental constraints on growth. China’s high level of GDP growth has been also generated dramatic pollution. Does an increase in GDP really raise the living standards of the citizens of a smog-bound city? But pollution is just the tip of the iceberg. GDP today is created by the unsustainable use of natural resources. We are boosting GDP today by spending on environmental credit, and through the excessive consumption of or long-term damage to environmental resources we are denying future generations the ability to enhance their standard of living.

The mindless pursuit of GDP today can damage living standards and undermine GDP tomorrow. Credit has to be paid back at some point, after all.

The problem with GDP is therefore that it is not very accurate in real time, it is becoming less accurate as the world changes, and we believe it is not designed to measure living standards properly. The pursuit of GDP growth may also do lasting damage to economic sustainability via the overuse of environmental credit – which will have to be paid back in the future. Still, GDP is not likely to disappear. The fact that it is the near universal standard for measuring economic activity means change will be hard to bring about. The fact that investors continue to react to the initial GDP print despite its many problems suggests that for financial markets immediacy is prized over accuracy. The data is flawed, but it seems we are stuck with it.