Kenneth Rogoff, Carmen Reinhart and the spell of magic numbers

Bom artigo sob o complexo de inferioridade de alguns economistas e a eterna busca pelo numero mágico...Modestia a parte não tenho esse problema e sou muito feliz por ser economista, digo, graduado, mestre e doutor em economia. Sem desmerecer os amigos engenheiros e advogados, acredito que a graduação é fundamental na formação do economista.

Humanity will never cease its futile search for magic numbers. The latest example of this quixotic quest is to be found in the saga of Harvard professors Kenneth Rogoff and Carmen Reinhart, who suggested that a debt to gross domestic product ratio above 90 per cent was bad for economic growth. Those conservative politicians and their counsellors who based their advocacy of strict fiscal policy on this magic ratio have retreated with egg on their faces.
This episode has lessons well beyond the immediate fiscal debate. Generations of economists worried that their subject was not regarded as a hard science, a syndrome labelled as “physics envy”. Milton Friedman complained that it had not unearthed key numbers or ratios that could be regarded as constants. He claimed to have discovered one in his statement of the relationship between inflation and the money supply (“inflation is always and everywhere a monetary phenomenon ... ). But without denying the connection, Friedman’s rule does not work in the way that key natural science constants do. Even if one has decided how to measure both the quantity of money and the price level, there is no stable ratio between the two. Friedman’s own work showed that the ratio not only varies in the short term but over several generations.
This is a long way from the physics constant “g”, the acceleration the Earth imparts due to gravity, although even it has to be treated carefully. A stone thrown in a vacuum will fall at an acceleration of 32ft per second every second. But a stone thrown from your window will not fall as fast because of the friction of the air.
The search for a physics-type relationship in subjects such as economics is of long standing. After the second world war it was sometimes proclaimed that if government spending rose above 25 per cent of GDP there would be trouble. This assertion gained credence. Yet there were all sorts of difficulties in testing the assertion. Did the 25 per cent ratio cover capital as well as current spending? If it covered both, countries with a large state sector would be doomed even if a large part of the expenditure was in capital investment deemed essential for growth. But other paradoxes appeared if public investment was removed from this ratio. Money spent on recruiting science teachers appeared as dangerous current spending, while similar sums diverted to extending school lavatories appeared as virtuous capital expenditure. In any case state spending, however defined, has for many decades been well above 25 per cent. It may or may not have slowed down growth, but it has not had the predicted disastrous consequences.
Looking through some old papers the other day, I found a more specific example of unfortunate attempted quantification. In 1970, a widely quoted study by economics professors Richard Lipsey and Michael Parkin proclaimed that if the UK could only tolerate an adult unemployment rate of 2.1 per cent (then regarded as high), wage increases would average only 3 per cent per year, thus putting an end to the inflation problem. Unemployment in fact rose to several times the proclaimed ratio without this magic effect. But it was only in the Great Recession starting in 2008 that earnings increases dropped to 3 per cent or less.
I would be completely misunderstood if anyone thinks that I am intending to pour scorn on quantified research. But it is worth reminding ourselves of Aristotle’s dictum that each subject has its own degree of precision and it is a mistake to look for a similar degree in all of them.
An example of a limited but useful research finding is known as “Okun’s Law”, or more accurately “Okun’s Rule of Thumb”, named after Arthur Okun, a former chairman of the Council of Economic Advisers. This states that a 3 per cent increase in output (above trend) is associated with a 1 percentage point fall in the unemployment rate – the remainder of the increase is accounted for by rises in hours worked, labour force participation and productivity. Okun would have been the first to emphasise that the exact relationship depends on the country and time period under consideration. More important: the relationship applies to short-term fluctuations due to business cycle or policy changes and would be unlikely to apply to long periods of stagnation or inflationary pressure.
The general moral is that economic analysis and policy would benefit from a less credulous acceptance of each purported research finding. It would pay to wait until a number of different studies employing different techniques point in the same direction and have survived professional criticism. In the meanwhile we can go quite a long way with a few well established elementary generalisations and inferences from them.

Samuel Brittan

Fonte: FT