Today has opened with a revision of French national income statistics for the last quarter of 2011. In the event La Republic recorded an unchanged quarter on quarter growth rate of 0.2% but the year on year rate reduced to 1.3% from 1.4%. No doubt the media will report the change as something important when in fact it is well within the expected error range and is a clear example of spurious accuracy. It may mean nothing at all. Even worse using Gross Domestic Product as a measure of economic well being has many and varied flaws.
What is Gross Domestic Product?
Gross Domestic Product is the market value of goods and services produced by workers and property in the country concerned, regardless of their nationality. So it is defined with geographical boundaries.
Gross Domestic Product = consumption (personal) + government spending + business investment + net exports (exports – imports)
Seems reasonably simple How is it counted?
We immediately hit a problem in that their are three ways of measuring this and you have probably already guessed that they do not always agree. I examined the figures for Portugal a few months ago and they differed at the widest point by 4%. When you consider that 4% growth or reduction is very significant we have a problem. The best we can hope for is that the differences remain fairly static, although in the “expect the unexpected” era which we now appear to live in that is a long way from being convincing to me.
The three ways of measuring GDP are using expenditure in the economy,using output and using income. As discussed above they do vary and there is debate over which is the more accurate. In the UK a previous Chancellor of the Exchequer tried to silence this to some extent as when investigating the differences I was told that Nigel Lawson had instructed the statisticians to use the output measure and adjust the others to it. Why was I asking? In the United States research had suggested that in the credit crunch era using the income method was more timely and accurate.
In general the numbers published are for the output measure and I wonder how many of those who quote them know that there are in fact two alternatives.
For those wondering about the US situation then using the income method gives you a deeper recession and a stronger recovery but we are still in a worse position (about 3% worse). As economic policy depends on where we think we are we enter into a zone where it may be (frankly probably is inappropriate..)
For real economic growth we need to measure inflation
Here we have another problem as one of the most common complaints by the general public is that inflation is not measured well and is understated. Some economists (which include me) do think that there are problems here. If we just think of the UK then we have two measures of consumer inflation which are CPI (Consumer Price Index) and RPI (Retail Price Index) and they are usually different often by a wide amount.
To get a measure of real output we need to take the output we measure and use an inflation measure to take out the effect of inflation. Actually we do not use the two used above we use what is called a GDP Deflator. This adjusts the nominal figure to give us a real one.
The problem is how accurate this is and there are doubts to say the least. For the UK the 2010 number was 2.86% and whilst it covers the whole economy and not just consumer inflation I have less faith than I used to have that we can rely on it. The catch is that if it was higher then real economic growth was lower (perhaps as indicated by the income measure that some countries use….).
So added to the problems above we have another one.
Immigration/ Emigration is another issue
I do not wish to discuss these as to whether they are a good or a bad thing (although in general they are good) but they are something we need to measure. Otherwise if you have a lot of immigration into a country you will record higher GDP numbers due to their activity but whilst the country may be better off individuals may be worse off. Accordingly the numbers are often measured per person to allow for this effect.
Seem simple? Er well no as if my country is any guide where we have no real idea of how much immigration there has been over the past decade or so we are much less clear about its effect. As a consequence we are also less clear what our population is. So we can say that as we stand per person or as it is usually expressed per capita GDP has a high error count.
The Reverse is also true
If we look at Greece, Ireland and Portugal we see countries which are likely to be seeing net emigration right now. Individuals have mentioned they are part of this on this blog. It is often forgotten that GDP falls will include this factor.
The accuracy of trade figures is shocking
Using net exports or exports minus imports gives us perhaps the biggest problem for statistics which is that they are very inaccurate and can be revised years and even decades later. Just as an example the figures for the UK current account in the first quarter of 2011 have been revised favourably by £800 million this morning. When you see that exports were revised up by £2276 million and imports by £1476 I hope you see that we are dealing with a number (net exports) which is the difference between two numbers we measure poorly and even worse we are never entirely sure we get right.
Should we measure GDP anyway?
Having hopefully unsettled your view on whether we can measure GDP I now intend to raise some problems about whether we should measure it. One issue was put well by the French economist Alfred Sauvy.
Marry your cleaning person, and you will make GDP drop!
This highlights the issue of what I will label household production. Consider an economy where everybody cuts their own grass, now consider one where they cut their neighbour’s grass and get paid for it. Suddenly GDP has surged and yet if you argue that you are likely to cut your own grass more carefully then you are left with the view that economic well being has probably just deteriorated.
We have a similar problem with voluntary work particularly if it replaces paid work. On a stand alone basis very little if any will be captured by GDP statistics and if it replaces paid work then we will see a recorded fall in economic output. Yet voluntary work can be valuable and if provided by people with specific skills it may be very valuable.
Accounting for public work is almost impossible and this part of government spending has lots of issues. As I am a fan of London’s parks lets us look at their upkeep. Our only way of measuring this is by what they are paid which may bear little or no relation to work done. i see examples of hard work and also no work and can only conclude,sadly, that my profession has no way of telling between the two.
GDP sometimes tells us the wrong thing
The earthquake,tsunami and subsequent Fukushima nuclear power station problem that hit Japan just over a year ago is a case in point, which I raised at the time. The destruction of the event was not recorded by GDP statistics but any reconstruction efforts will be. So Japan as a nation is worse off but GDP numbers will show an improvement over time.
Another example is the problem of issues such as crime. An increase in crime will make everyone worse off but a response involving more police will increase GDP. If we move onto consider the legal profession we have a deeper problem. What do they produce? As we struggle to cope with that we realise that we are in an era where their significance has increased but also at a time where many would argue that their influence is often far from beneficial. Their influence on our political class may well be an example of institutional capture.
Resource use is another problem. If we look at the UK we have used up a lot of the gain from the natural resource that was/is North Sea Oil. This will have boosted UK GDP over time but there is no deduction anywhere for using up a finite resource. If we think of say deforestation it is easy to argue that economic activity here boosts GDP but is usually a bad move. Another example of this was the (in)famous Exxon Valdez oil spill in Alaska where the highly expensive clean-up effort provided a boost to US GDP because there is no measure used for damage to the environment.
The Financial Sector
I have left perhaps the worst problem of our times for last. Let me give you a simple sounding question. How do we measure the output of the financial sector? As we look around and see some factors ignored and others at danger of double,triple or even worse counting we should be increasingly unsettled. Millions then billions and more often latterly trillions pass before our eyes.
I am left with the view that the financial sector was overvalued on our way into the credit crunch and that it has discernibly shrunk since. Has our economic growth measure allowed for this? No! Mostly this is due to all the government and central bank intervention creating false markets in so many areas. You may wonder if this is a reason for some of the intervention……
So when we thought we were well off we were not as well off as we thought we were and now it is worse. Not exactly a cheery conclusion but the truth I think.
It is plain that GDP statistics do not stand up to the analysis that they are often used for. We can see that they are not accurate enough and measure many of the wrong things. Accordingly we are left with the view that as we do not know where the economy is at any point in time those who argue for what is called economic “fine-tuning” will only be correct by chance if they use official statistics. In fact they may turn out to be a malevolent influence as they may be acting in the wrong direction. My tutor at the LSE Willem Buiter put it well I think.
I know I know nothing; but at least I know that
Now if we bring this fully up to date how significant does it seem now that UK economic growth in the last quarter of 2011 dropped from -0.2% to -0.3%? That is what has been released whilst I was typing this.