GDP is flawed – just not the way most people think


At Oxford university, many students regard attendance at lectures as optional. So teachers who fail to enlighten or entertain end up talking to empty rooms. A malicious fellow student measured lecturing performance by computing the ratio of attendance at the start of a course to attendance at the end. The highest score was earned by the hapless teacher of a first-year course on national income accounting.

Few universities now offer such a course. They have responded, or pandered, to student preferences, and the economics curriculum has moved on. Not necessarily in a good way; national income accounting remains central to economic statistics, and hence to economic policy, but is no longer well understood. Diane Coyle has bravely attempted in a recent book to make the subject once more accessible, and even interesting.

When economists talk about economic growth they are measuring the change in gross domestic product. Lay criticisms of GDP are often based on the indisputable observation that there is more to life than economics and material goods. GDP omits work done in the home, mostly by women. It values expenditure on war and nursing care on the same basis. It records the despoliation of the environment only by reference to the amount spent despoiling it, and then includes the amount spent to clean it up. It does not tell us how happy we are or how fulfilling are our lives.

These objections are valid but largely beside the point. GDP is a measure of the productive performance of the economy. How that productive potential is used is an important subject but a different subject, and only partly an economic question. It is a poor criticism of a thermometer that it does not tell us how comfortable we feel. Yet GDP is not a physical fact like temperature but an artificial construction. Its measurement is conventional and subjective. We should ask whether GDP is a good measure of what it is intended to measure.

GDP is gross, so makes no allowance for depreciation. If there is a lot of shortlived investment – as in information technology – output is overstated if you include such expenditure as investment (which Americans are inclined to do) and understated if you write it off as incurred (which Europeans tend to do). GDP is measured at constant prices but what do you mean by the constant price of a piece of software? These different conventions matter a lot to the answers you reach.

GDP is domestic, so you measure what is produced within a country’s boundaries regardless of whom it is produced by or for. The combination of gross and domestic means that you include the total value of output, less operating costs incurred in that particular year. So resource producers look richer than they are. The Yes campaign in the run-up to Scotland’s independence referendum has pointed out (correctly) that an appropriate attribution of oil revenues would show that Scotland is today one of the world’s wealthiest countries (measured by GDP per head); but Scots would make a mistake if they thought that calculation showed independence would make them better off.

And national income accounting cannot handle the financial services sector. Reported output of financial services rose dramatically during the 2008 financial crisis. This nonsensical result arises because the measurement of financial services output is strongly influenced by the margin between average bank lending and borrowing rates, which increased sharply. When someone confidently quotes the contribution of financial services to national income, you can be sure they have no understanding of the esoteric concept of “financial services indirectly measured” (don’t ask). Only a few people in the depths of national statistics offices do. This problem casts doubt on the validity of reported growth rates both before and after the crisis.

It once puzzled me that many economists in the financial sector forecast and discussed GDP without knowing what it was. I have since realised the job of market pundits is not to forecast GDP but to forecast what the statistics office will announce is GDP, and that is not at all the same.

Yet reality will always break through. In Ireland, almost all the problems I have described came together to make the country appear, from its GDP statistics, better off than it was. The resulting hubris left it worse off than it need have been.

Print Friendly, PDF & Email