The Nobel prize, and the ONS
Two US economists, Thomas Sargent and Christopher Sims, have won the 2011 Nobel prize, "for their empirical research on cause and effect in the macroeconomy."
You might ask: what other kind of macroeconomics is there? Aren't economists always concerned, in one way or another, with cause and effect?
The answer to that question is yes. But it is often hard, in a complex world, to distinguish between the two. Especially when the 'facts' provided by statisticians are themselves up for debate, as we discovered, once again, last week in newly revised GDP data from the Office for National Statistics (ONS).
To take a topical example of cause and effect, we might observe the government announcing budget cuts, and we might also observe the economy slowing down. But are the cuts causing the slowdown, as Ed Balls believes, or are both of these developments a symptom of the global financial crisis and its aftermath, as the Chancellor believes?
Or, we observe a very low level of bank lending to businesses in the UK; but is that because of a lack of loans being made available by banks, or a lack of demand on the part of companies? Which is cause, and which is effect?
In their own different ways, Sargent and Sims produced crucial tools for economists to answer these questions, which are still being used today. That makes them more than worthy of today's prize. But the assumptions underpinning their approach have fallen out of favour recently - among economic commentators, if not the profession as a whole.
Put it another way, this is not one of those times when the Nobel committee has tried to 'capture the mood'. Quite the contrary.
Put briefly, Sargent and Sims were part of the generation, in the 1970s, that took the economic profession in a much more mathematical direction, with a much greater focus on the role of expectations. Sargent, in particular, was concerned with the internal consistency of economic models: the results had to follow, with unbroken logic, from first principles.
There wasn't a lot of room for messy, ad hoc assumptions to make the model more closely fit the facts. Nor was there much room, in the early models, for financial crises - or the aspects of human nature that make them more likely. In that sense, you can put them firmly on the 'fresh water' side of the arguments within economics that I have discussed here in the past (see my posts of 17 July and 2 November 2009).
It would be unfair to say, as Paul Krugman famously said of this school, that they "mistook beauty for truth". They were far too sophisticated for that. (Indeed, in some of his more recent work, Sargent has tried to incorporate more realistic definitions of human rationality.) But both Sargent and Sims are definitely on the side that says beauty - logical consistency - needs to come first. Otherwise, in their view, economics as a science might as well pack up and go home.
What - on earth - you might ask, does this have to do with the ONS? Well, let me take you back to where I began. Whatever you think about their approach, the Nobel committee has decided to honour Sargent and Sims for their "empirical research on cause and effect in the macroeconomy".
But whether you're a fresh water economist or a salt water one, empirical analysis - working out what's actually going on in the economy - depends on knowing the facts. And last week's revised GDP data revealed that, at crucial times in our recent economic history, the facts that policy makers had in front of them were completely wrong.
I mentioned - on the day the figures came out last week - that growth in the lead up to 2008 had been revised up, and so had the total loss in output during the recession. What I did not mention was the sharp revisions in the quarterly data, some of which were questioned by economists at the time.
In the summer of 2008, the Monetary Policy Committee (MPC) was still debating whether to raise interest rates to confront inflation. It turns out that the recession was then firmly under way: the latest data show GDP shrank by 1% in the second quarter of 2008. But the MPC didn't know that. They had an ONS estimate in front of them which said the economy had grown by 0.2%.
And remember the argument that raged two years ago, when everyone was expecting the economy to come out of recession in the third quarter of 2009? At first, the ONS shocked everyone by declaring that the economy had shrunk by 0.4%. Economists at Goldman Sachs, among others, disagreed.
The ONS stuck to its guns, even in later revisions. But now it says that the economy did grow in that quarter - by 0.2%. Growth in the fourth quarter of that year has also been revised up, from a first estimate of 0.1% to a much healthier 0.7%. Those are pretty big revisions, much larger than the long-term average revision cited by the ONS at the time.
Does this mean we can expect this year's estimates to be revised up as well? Kevin Daly, one of the Goldman Sachs economists who first questioned the 2009 figures, does believe the 2010 figures will be revised up further, to bring them more into line with the strong company survey data and other indicators that he favours over the ONS version of events.
Growth in 2010 has already gone up from 1.4% to 1.8% with last week's revisions. But I'm afraid that's the end of the good news. He says those same surveys now suggest that growth in 2011 will be revised down.
This is not to suggest that the ONS is rubbish - or worse than other official statistical bodies. Statistics get revised all the time, all around the world. But it does underline a lesson of past recessions that the ONS were reluctant to accept at the time of that 2009 debate: that they get it most wrong when the economy is 'on the turn'. The revisions are always largest when the economy is just going into recession, or just coming out - in other words, when policy makers most need to know what is going on.
Whether you like your economics beautiful, like the 2011 Novel Laureates, or messy, like John Maynard Keynes, you don't get far without reliable statistics.