The Demise of GDP Is Premature
By Anthony de Jasay
Replacing GDP with more sophisticated yardsticks could prove to be the slippery slope to daddy-knows-best.
For the basics of GDP, GNP, and measuring well-being, see Gross Domestic Product, by Lincoln Anderson in the Concise Encyclopedia of Economics.
If GDP will not grow as one would wish, use an alternative that will. Referring to the teaching of the communist sociologist Edgar Morin, Mr. Sarkozy reminded public opinion that gross national product was not the same thing as well-being, and that it took many other variables than material production to make a civilisation and a high quality of life. On the league tables of these other variables, France ranked much higher than on the GDP table.
To drive home the point, he invited the two most prominent economists of the worldwide Left, Amartya Sen and Joseph Stiglitz, both of impeccable technical credentials and a high political profile, to co-chair a committee to be charged with working out measurements of components of well-being and civilisation that would tell us more than plain GDP. That GDP does not tell a complete story of the good life is not news. It has been common knowledge for at least a generation. A working party of the OECD has been studying the problem for years, but one more committee must surely be a good idea.
GDP has many venial sins but it has one great virtue. Like any other index number, it suffers from the necessity of adding apples and oranges and subtracting bananas from the sum, operations that are intrinsically meaningless. For operations that are not nonsensical the apples, oranges and bananas have all to be artificially converted into numbers on the same scale. Their sum can serve as an index. In most indices, the number assigned to the apples, oranges and bananas are ultimately matters of opinions and judgments. GDP as an index number, on the contrary, has nothing to do with the opinions of the indexer. It is the “objective” result of the “subjective” choices of billions of consumers and producers who jointly determine world prices. GDP is based on market prices, not on opinions about how apples and oranges ought to be valued from a moral point of view. It would be impossible for a left-leaning statistician to overweigh welfare provisions or public ill-care and a right-wing one to overweigh police services over and above the money spent on these services which, though not sold in a market, employ labour whose price depends to a great extent on the value the market puts on labour of various types. GDP, for all its faults, does not admit ideological massaging. It can hardly be manipulated to favour Left or Right.
Admittedly, GDP can come up with silly results, but those are small-scale. If people start taking in each other’s washing instead of doing it themselves, GDP will go up. The Swedish welfare state used to be mocked for mothers looking after the small children of other mothers and grown children taking care of other people’s aged parents instead of their own, all of which lifted GDP. In a tiny sheikdom, GDP per head may go sky-high as the oil is pumped out from beneath the desert sand, but no depletion allowance is deducted to account for there being that much less left to pump out in future years. In the Northern hemisphere, a mild winter depresses GDP and a hard one boosts it. There is a bit of double counting in most national income accounts, and adding in the “black” economy is educated guesswork. But these vices do not vitiate the concept, nor its honesty in conveying good and bad news alike.
What should we expect Sen-Stiglitz amendments to do? For one, they might decide that if money GDP goes up by 2 per cent, quality-adjusted GDP is higher if all incomes go up by 2 per cent than if all incomes go up by 1.8 per cent and the remaining 0.2 per cent swells the incomes of a few thousand households by 30 per cent. Mr. Sen is intellectually fastidious enough not to say that the former distribution is actually bigger than the latter. But he might say that he likes it better and that most people do, too, and that this should in some way accounted for. Distribution would in any case affect quality and may be awarded an index number.
It would be no surprise if the Sen-Stiglitz committee also composed an index number out of the several indices that seek to measure the achievements of popular education, perhaps building into it a special allowance for the occupational or income category of parents. Another obvious candidate for an adjustment factor for the quality of life could reflect the state of health of the population and the success of efforts to care for the sick. Questions of equality and inequality might well enter into these calculations, too.
A quality-adjusted GDP could very well also reflect the length of time couples living together stay together, the ratio of urban parks to built-up areas, data on rubbish collection and disposal, the average time the gainfully employed population spends travelling to and from work, the rigour of food safety standards, crime statistics, the suicide rate, pub drinking hours and the regulation of industry and commerce in the public interest.
The list could go on and on, depending on the productivity and ambitions of the Sen-Stiglitz committee and the brief they are given. Some of the potential items on it are trivial or risible, but most do have a serious impact on how people get on with their lives and how propitious are the circumstances for them not to be too discontented even if GDP growth gives them little or nothing from one year to the next.
Nothing much can be done about many of the items on the list; they are facts of life, time and place. Others, however, can be modulated, pushed back or forth and generally shaped by public policy.
Precisely here lies the ground for fresh concern. When faced with the choice between the status quo and some potential improvement that it is the legitimate business of government to tackle at a cost spread so wide and thinly over the entire society that nobody much notices it, the bias of public policy in most societies is to go for the improvement. The effect can be good in detail but unexpectedly and puzzlingly bad in the aggregate. Welfare states tend to suffer from this syndrome. It is not new, and has been noticeable even before the replacement of bare GDP data by sophisticated adjustments and allowances for quality has come to the center of media attention and political “spin”.
Now, however, a new element will redouble the activist bias. When, as in the case of modern France, one welfare provision, one administrative intervention and one extension of the area of public policy follows another and each of the several hundred laws and decrees passed each year brings a putative improvement yet GDP for that very reason grows more and more sluggishly, opinion in the country becomes dubious and may even flirt with some liberal ideas. But if the race for improvement along a hundred fronts produces heartening improvement in the quality-adjusted neo-GDP that is destined to bury the orthodox one, the hand of activist government is incomparably strengthened. Not only does Daddy know better what should be done, but the numbers start saying so, too. This looks like the ideal start of a long slippery slope. On the whole, it may be safer to face embarrassing truths and unflattering league table positions, and not hasten the demise of the old and plain GDP.
The State is also available online on this website.
For more articles by Anthony de Jasay, see the Archive.