Cooking the Books 1: Lies, damned lies, and statistics
Marx argued that by his day economics had declined from a genuine attempt to understand the workings of capitalism as, for instance, with Adam Smith and David Ricardo, into a thinly disguised apology for capitalism. Just look at any modern textbook – with their claim that scarcity is eternal because human wants are infinite and their introduction of “entrepreneurship” as being as essential to production as natural resources, labour and machinery – and you can see how right he was.
In one field of economics, however, there has been some progress since Marx’s day: in measuring national income and output. Since the 1940s governments have published figures for these. Everybody has heard of “GDP” (Gross Domestic Product) and “growth” (which is the increase of GDP from one year to the next) and Marx would surely have loved to have had such figures. But there are shortcomings here too.
In particular how to treat government spending has been a problem, as was again highlighted by a report published at the end of January by Sir Tony Atkinson (who as plain A. B. Atkinson used to usefully concern himself with statistics on the unequal distribution of wealth ownership) for the Office of National Statistics on “The Measurement of Government Output and Productivity for the National Accounts”.
Government Output? What output — surely the government as such produces nothing, just consumes the output of the sector of the economy where labour-power is applied to materials that originally came from nature to produce commodities for sale? Yes, that’s just the point. But the figure for GDP (which is supposed to be a measure of total output) is obtained by counting spending by individuals on consumer goods and services, by firms on new equipment, and by the government, and adding them together. Including government spending in this involves double-counting. This is recognised to some extent in the official figures in that the payment of pensions, the dole, income support and the like as well as the payment of interest on the National Debt are excluded from government spending as being “transfer payments”. But, logically, so too are the wages and salaries of government employees, yet these are not excluded. Similarly, other government spending (as on purchasing computers and bombers) is already included via the amount for indirect taxes that enters into the prices of what individuals and firms buy.
The artifice that the government statisticians found to get out of this has been to treat government spending as productive, as resulting in a “product” (education, health care, administration, law and order, “defence”, though not “social security”). As a result, national output is inflated by as much as 20 percent. The former state capitalist countries of Russia and Eastern Europe did not count such government spending as productive, i.e. did not double count it as part of output, and when they adopted the same national accounting system as in the rest of the capitalist world their GDPs jumped by 18-24 percent depending on the country.
Why did statisticians in the West go down the road of pretending that all government activity is productive? Probably because national accounting developed at the same time as Keynesian economic policies were first applied (and could even be said to have been developed to underpin them), and Keynes attributed a key role to government spending. Keynesian economics is now rejected, but not the counting of government spending as an addition to output. Instead of getting itself out of this hole by abandoning this statistical practice, the government decided to keep on digging and appointed Sir Tony to come up with ways of measuring government “output” in such a way as to be able to take into account increasing “productivity” (the politicians’ “value for money”).
If the Atkinson report’s recommendations are accepted this will inflate GDP even more by artificially (and arbitrarily) increasing yet further government “output”. Talk about cooking the books.