Monday, 11 March 2013

The Problem with GDP

1         Introduction

Measuring Gross Domestic Product (GDP) is important. It tells us how well our economy is doing, both in its own terms and in comparison to the economies of other countries. In the UK, for instance, 2012 started with two successive quarters of negative growth, which is technically considered a recession; and although the economy then picked up a bit in the third quarter, boosted by the Olympics, a further decline in the fourth quarter meant that GDP across the whole year remained flat. In contrast, the US economy grew by around 3.5% in 2012, while Spain experienced an overall contraction of 1.4%. On this basis, therefore, one might say that, in 2012, Britain did a little better than Spain but significantly worse than the USA.

We also use GDP to measure relative indebtedness. At £1,149 billion, the UK national debt, as of 1st March 2013, stood at around 73% of GDP, while our annual fiscal deficit – the additional money the government has to borrow each year to make up the difference between tax revenues and public expenditure – was just under 8%. This is only slightly worse than the Spanish position, where the national debt also stands at 73%, but where, through a substantial programme of spending cuts, the government has managed to reduce its annual deficit from 9% in 2011 to 6.7% in 2012. In contrast, the US national debt has now reached 106% of GDP; and although its annual deficit, at 6.85%, is only slightly higher than Spain’s, it is clearly in the worst position of the three.

Even though one can always interpret the data in different ways – as I have done here – GDP is thus quite clearly an important international benchmark, and one against which so much else is measured, determining not only whether a government’s economic strategy is deemed to be succeeding, but its credit rating, its ability to borrow money, and the domestic interest rates its banks are consequently forced to charge. Given such enormous significance, it seems only sensible, therefore, to ask how accurate these quarterly measurements of national economic performance are, and whether they provide a true and meaningful representation of what is actually going on in the economy they attempt to encapsulate. 

2         Technical Issues & The Odd Bit of Jiggery-Pokery

So let’s start with the question of accuracy, where there are two main issues, the first of which derives from the way – or rather ‘ways’ – in which GDP is calculated. For there are three of them.
  1. The first method of calculation is based upon expenditure and uses the formula:
GDP = C + I + G + (X – M)
where C = private consumption; I = gross investment; G = government spending (not including expenditure on pensions or benefits); X = exports; and M = imports. 

  1. The second is known as the ‘production’ or ‘value-added’ approach and is based on the gross value of an economy’s output minus the cost of intermediate consumption. Take, for example, a baker who uses flour to bake break. The difference between the value of his ‘output’ (the price he charges for the bread) and the cost of the flour that is intermediately consumed, is what is known as the baker’s ‘value-added’ component. GDP, as calculated on this basis, is the aggregate of all these value-added components across the entire economy.
  3. Finally, there is the ‘income approach’. This aggregates all the wages, salaries, corporate profits, interest payments, and earning from unincorporated businesses, deducts any subsidies received and then adds back indirect taxes and depreciation.
In the UK, the Treasury and the Office of National Statistics (ONS) use all three of these methods and, indeed, attempts to reconcile them. The problem is that the figures which each method produces seldom tally. When he was Chancellor of the Exchequer, Nigel Lawson, as a consequence, instructed his officials at the Treasury to favour the ‘value-added’ approach and, if necessary, to ‘massage’ the figures generated by the other methods to make them consistent. And this is still the practice today.
Not that this in any way invalidates the exercise, of course. What it means, however, is that the analysis is not quite as objective and scientific as some people tend to believe, and that the choice of method used or favoured is already, to some extent, a political decision, which can, and usually does, affect the outcome.

Still, you might say, this is just a minor wrinkle. The second problem, however, is slightly more serious. It results from the fact that each of the methods used requires the collection, organisation and analysis of huge amounts of data – more in the case of some methods than others – most of which is unavailable at the time when the quarterly GDP figures are initially published. A lot of the data therefore has to be estimated. During the last thirty years, as a result, only five sets of the quarterly figures issued for the UK have not had to be revised at a later date. 

That’s not to say, of course, that we should discount – even less ignore – the figures when they first come out, especially as the UK Treasury and ONS are remarkably accurate even in their first cut, seldom being out by more than ±0.5%. This is not the case, however, in all countries. Over the last five years in Portugal, for instance, the margin of error has been ±4%, which some might say makes the figures almost meaningless, and casts doubt even upon the revised figures when they are published six months later. 

Nor is this the only respect in which one has to be wary when comparing international GDP data. For there is another very important complication which has to be taken into account when calculating changes in GDP from one year to the next: inflation. To appreciate why this should be so, suppose, for example, that a country’s ‘Nominal GDP’ – its GDP unadjusted for inflation – grew by 2% in 2012. One might think that this was quite good. If, however, inflation during the year was running at 2.5%, this would mean that the country’s ‘Real GDP’, adjusted for inflation, actually decreased by 0.5%. 

This then leads to even further complications.

In Figure 1, I have set out the relative growth in both ‘Real’ and ‘Nominal’ GDP for the UK economy between 1997 and 2011. You will note, however, that in 2009 the two lines actually come together and cross. This is because the ONS, from which I obtained the data, just happened to use 2009 prices to calculate ‘Real GDP’ for each of the previous and subsequent years. However, it could, just as easily, have chosen 2005 prices, for instance, or any other year’s prices as the basis of this calculation, which would then have given us a completely different set of figures and a completely different graph. 

For researchers, like myself, collecting data from different sources, this can be very confusing, especially as not everyone states whether their data is real or nominal, and fewer still identify the base year. It can also, therefore, lead to mistakes, which, given the political sensitivity of the subject, may not always be entirely innocent. 

 Figure 1: UK GDP 1997-2011 Adjusted and Unadjusted for Inflation (Source: ONS)

Factoring inflation into the equation also raises another issue. For in addition to now needing to find out whether the figures presented are nominal or real – and, if the latter, the year used for comparing relative prices – it can also be important to know the way in which inflation, itself, is calculated, particularly if the method used includes some elements of substitute pricing – for which, I suspect, some explanation may be required.

Normally, inflation is measured in terms of a fixed collection or basket of purchased items. Occasionally, however, these items have to be replaced with something more appropriate. The VCR cassette recorder, for instance, which may well have been included in the official shopping basket in the 1990s, is now very probably a Blu-Ray player. However, as Mitch Feierstein points out in his book, Planet Ponzi (Bantam Press 2012, pp76-81), even countries such as the USA are not above a little jiggery-pokery in this regard, using substitute pricing to keep inflation – or at least the official index of inflation – artificially low. They do this by arguing that if the price of orange juice goes up, consumers may well decide to buy apple juice instead. They therefore feel justified in replacing orange juice with apple juice in the monitored shopping basket, thus keeping the figure for inflation lower than would otherwise be the case. In fact, Feierstein believes that US inflation is actually now running at 10% higher than the official figure. What this also means, therefore, that US GDP, adjusted for inflation, is almost certainly growing far less strongly than the 3.5% for 2012 mentioned in my opening paragraph. Indeed, if Mitch Feierstein is right, it very possibly means that the USA has been in recession for the last three years.

3         Deeper and more Serious Problems

Despite all this, however, nothing I have said so far seriously undermines our use GDP as the principal measure of a country’s economic performance. All it means is that we should be a little more careful, professional – and yes, a little more honest – about how we prepare, present and use this data. There are, however, a number of ways in which, due to the way it is calculated, GDP can be far more seriously misleading. 

The first of these results from the fact that there are types of economic activity which none of the methods of calculating GDP are able to capture. These not only include goods and services traded on the black market and the proceeds of crime, but services rendered for payment in kind – or without payment at all – within a variety of social structures, some of which are more important in some countries than in others.

In Mediterranean countries, for instance, where the extended family still has an important role, grandmothers are far more likely to look after their grandchildren than in most northern European countries. Because, no money changes hand, however, their contribution to the economy is not included in GDP. In most Scandinavian countries, in contrast, where children are usually placed in private child care centres or state-run nurseries while their parents go out to work, the value-added contribution of this child care industry is included, thus making these countries technically that much richer. In fact, there is a joke going round among economists – well-known for their thigh-slapping sense of humour – that if only Greek parents could be persuaded to pay their children’s grandparents to look after their offspring, all of Greece’s economic problems would be solved. GDP would rise; national debt and the fiscal deficit would consequently fall as a percentage of GDP; and the government might even able to start borrowing money on the bond markets again.

Of course, this is just an amusing fantasy. What makes it amusing, however, is that because of the way GDP is measured, it appears to make sense. Moreover, there are people who actually take this kind of boot-strap economics quite seriously. They are called Keynesians; and they include many actual economists and most left-wing politicians. They are only able to indulge this fantasy, however, because of something that is included in GDP, but arguably shouldn’t be: the public sector.

At this point, I suspect, you may well ask why the manner in which something is paid for should have any bearing on whether or not it is included in GDP. What difference does it make, for instance, that whereas in the USA health care is largely supplied by the private sector and paid for through insurance, in the UK it is mostly both supplied and funded by the state? Am I saying that it is OK for the US health care industry to be included in USA’s GDP, but improper for the National Health Service to be included in the UK’s? Obviously not. And yet, there is a problem. For whereas industries operating in the private sector our principally driven by market demand, the expansion or contraction of industries operating in the public sector is almost entirely a political decision.

As a consequence, opposition politicians are able to argue that, instead of contracting our public services, the UK government should be expanding them, in that this would bolster growth. As, indeed, it would. For being included in GDP, a decision to employ more police officers, nurses and child-minders would immediately increase this all-important measure, very probably taking us out of recession. Apart from further adding to the fiscal deficit and laying up even more problems for the future, however, this is all it would do. For the Keynesian claim that it would also stimulate growth in other parts of the economy is unsupported by any historical data, as can be seen quite clearly if we look in more detail at the GDP figures for 1997 to 2011, previously presented in Figure 1

In Figure 2, I have broken down the headline figure for Real GDP into three categories:

  1. Public expenditure – excluding pensions, welfare benefits and interest payments – based on nominal figures taken from the ONS Blue Book and adjusted for inflation using 2009 pricing.
  3. The contribution of the Financial Services Industry, as outlined in ‘Financial Services: Contribution to the UK Economy’, by Lucinda Maer and Nida Broughton, using data obtained from the ONS, based on 2011 pricing, which I have converted to 2009 pricing for consistency.
  5. The residual contribution of the private sector, based on the figures obtained from the ONS and used in Figure 1.
Note that the graph in a compound layered graph, in which it is the width or depth of each band that represents the contribution of each category to the economy.

 Figure 2: Real GDP 1997-2011 Broken down by Category

The most important thing to note here is that the underlying shape of the lowest band, representing the non-financial private sector, is exaggerated by each of the bands above it. This is because, from 2001 to 2008, the public sector and the financial services sector grew far more steeply than the non-financial private sector, and continued to expand even after the latter fell off the cliff in 2008. One can see this even more clearly in Figure 3, where I have recorded the annual growth rate for each of the three sectors in tabular form.

Figure 3: Annual Growth Rates 1997-2011 in each Category

Here one can see that, adjusted for inflation, between 2001 and 2008, the public sector – or that part of it included in GDP – grew by an average of 5.89% per year: a total increase of 58% over the eight years in question. Far from stimulating the private sector, however, once financial services have been taken out of the equation, the private sector more or less stagnated, with average annual growth of just 0.84%. 

Part of the reason for this is that, although the government was already running an average annual deficit of £35 billion per year throughout most of the period – from 2002 to 2007, to be exact – in or order to fund this massive expansion in the public sector, it raised overall taxation to one of the highest levels in UK history: almost 40% of GDP. Despite the windfall taxes it was getting from the banking industry, which also grew dramatically throughout this period – from 5.8% to 10.4% of GDP – its overall effect on the private sector was therefore one of depression.

At this point, you are probably wondering why no one spotted it at the time. Well, to some extent, they did. It was just that, in the middle of a banking boom everybody was having it so good, no one was prepared to listen. More to the point, if one looked only at the headline figures for GDP, it appeared as if the economy was doing well. Between 2001 and 2008, in fact, overall growth, adjusted for inflation, averaged 2.51%. Indeed, it was argued that the strength of the economy both justified additional public expenditure and proved that this public expenditure was not an unaffordable burden. What we all therefore failed to see was that two thirds of the headline growth comprised a banking sector boom in which most of the deemed value was illusory, and public expenditure itself. The result was that when the banking bubble finally burst, bringing down with it hundreds of private sector companies which had their overdrafts and other forms of credit suddenly cut, the government not only lost a large percentage of its tax revenues, but found itself obliged to fund an enormous public sector it could no longer afford. Nor could it raise taxes any higher than they already were to bridge the gap. In less than a year, as a consequence, the annual deficit rose from £35 billion to £160 billion.

And it is this that is at the heart of our continuing financial difficulties. Many people think that it was the bailing out of our banks that was the cause of our debt problems. But as I pointed out in ‘The State of UK Banks & The Long Haul to Recovery’, the amount of money actually used in this way was relatively small, with most of it provided in the form of loans to our banks to replace inter-bank lending, which, at that point, had more or less completely broken down. The real problem was not, therefore, the banking crisis itself, but the unbalanced structure of our economy which it laid bare: a stagnant private sector overly dependent on financial services; and a top-heavy public sector for which we no longer had the tax revenues to pay. And this is still very much the state we are in today.

Our political problem is that we still don’t seem to understand this. As a consequence, people constantly hark back to the halcyon days before 2008, and look around desperately trying to find ways to return to 2007 levels of growth. They don’t realise that the growth we experienced during that period was largely chimerical and that the way it was obtained – by encouraging a banking bubble and pumping money into the public sector – is not something we should try to revive, even if we could.

Instead, we need to do two things. Firstly, we need to reverse the unsustainable expansion in the public sector that occurred between 2001 and 2008. During that period, more than a million extra public sector jobs were created, taking total employment in the sector from just over 5 million to more than 6 million. Those jobs now have to be ‘uncreated’ – which is a lot harder to do, and will be met with far greater resistance than their creation. We also have to tackle the burgeoning cost of a welfare system which is still causing public expenditure to rise despite cuts in those parts of the public sector contributing to GDP. That, however, is the subject for another day. 

Secondly, we need to start supporting the ‘real’ economy: that part of the private sector which actually produces goods and services that people want to buy. Unfortunately, the best way to do this – apart from leaving it entirely alone – is to cut taxes: something for which there is limited scope while we are running such a huge deficit. As a result, this whole process of rebalancing the economy is going to have to be gradual, and will probably take as long as it took to unbalance it in the first place: which probably means another five years, at the very least.

The good news is that it has already started happening. If you look at the growth figures for 2011, as shown in Figure 3, you will see that while, as result of cuts to public expenditure, the public sector contracted by 2.79%, the non-financial private sector grew 2.56%. Because of the cuts in the public sector, however, and their immediate effect on overall GDP, net growth for the year came out at less than 1%, which everyone greeted with scorn. For again, most people don’t see below the headline figures. They don’t therefore realise that the growth in the non-financial sector was actually much better than the headline figures suggest, and that it was only dragged down by cuts in the public sector which are absolutely essential to reduce the deficit.

The problem for the government is that this is now going to go on happening quarter after quarter. I haven’t yet seen a breakdown of the GDP figures for 2012, but my guess is that the recession that technically occurred during the first six months of the year was simply the result of two consecutive quarters in which the necessary cuts in the public sector exceeded the growth in the private sector, which still may have been substantial. For given the growth in employment during this period, it is almost inconceivable that the private sector actually contracted.

Indeed, it is only when one looks below the headline GDP figures that the employment figures make any sense at all. For over the last two years, despite a double-dip recession, employment has continued to rise, hitting record levels. Indeed, not a month goes by without economists appearing television, scratching their heads over how this could be possible in a stagnant economy. The answer, however, is that the ‘real’ economy isn’t stagnant. Significantly, more new businesses are currently being created than during any previous period in our history, many of them by young people, who, tired of trying to get a job working for someone else, have decided to ‘do it for themselves’, working 60-70 hours a week, paying themselves a pittance while still living with their parent or a supportive spouse or partner: factors which ensure that this is another form of economic activity which hardly registers in the GDP figures, but which is actually the life-blood of our recovery. 

The really disappointing fact, however, is that none of this registers with our news media. In a few weeks, GDP figures for the first quarter of 2013 will be published, and they are widely predicted to show another quarter of negative growth. Taken together with the negative figures for the last quarter of 2012, this will therefore be widely hailed as the third phase of a triple-dip recession. Again, it will be said that the government’s economic strategy is not working and that we need a ‘Plan B’. Not once, during all the television debates that follow, however, will anyone ask whether the headline figures for GDP might not be giving us a false impression. Even less will anyone dare to question whether, simply because someone at some point decided to define a ‘recession’ as two consecutive quarters of negative growth, this means that a recession is necessarily what is going on in this instance. Fort this, of course, would be to suggest that the term, itself, requires some redefinition, specifically to exclude periods of negative growth, where the real economy is still expanding, but in which this is offset by a political decision to make necessary and economically sensible cuts in public expenditure. 

And this, of course, is the real problem with our use of GDP: that having lost track of what it really measures, it has become something of a shibboleth; something which, in itself, cannot be questioned. Instead of seeing it for what it is, a handy economic indicator, which nevertheless needs to be looked at more closely before we start drawing any definite conclusions, it is used as the first and last word in any debate, a non sequitur to which most audiences, lacking any deeper understanding, have little choice but to nevertheless demur. My worry is, however, that continually using it in this way will eventually tip the political balance; and that the proponents of the fantasy economics that got us into this mess in the first place will once more be given the reins. And what they will do, of course, is quite impossible to rationally foresee.