The genius of GDP is that it somehow manages to squeeze all human activity into a single number. You could think of it as like pushing a large frog into a small matchbox. Still, at its best, GDP is a brilliant and useful insight. It provides a snapshot of one version of reality, a number that policymakers can act upon. But if the beauty of GDP is aggregation, that is also its biggest flaw. No single number can capture all that is worth knowing in life – even if you’re an economist.
Think of it like a car dashboard. There’s a fuel gauge that tells you how much petrol you have left in the tank and a speedometer indicating how fast you’re going. Perhaps there’s another display that tells you what music is playing. All three give you valuable bits of information. You cannot, however, combine them into a single number that tells you anything meaningful. They are in different dimensions.1
To create a single number requires measuring everything in the same unit, which in economics means converting everything into dollars and cents. When it comes to something hard to price – say volunteer work, life expectancy, clean air or a sense of community – you must either figure out a way of attaching a dollar amount to it or just forget all about it. The idea that all things can be priced stems partly from the work of Jeremy Bentham, the man whose missing head, you’ll recall, is worth ten pounds. The theory of marginal utility, a foundation of modern economics, states that the price of a good or a service reflects the additional satisfaction gained from consuming an extra unit of that product. It is this reductionism that allows economists to convert everything into those complicated mathematical models that occasionally cause your eyes to glaze over.
But prices cannot be a proxy for everything. That means much of what we care about as human beings is either left out of our economic calculations or converted, using some jiggery-pokery, into a dollar amount that can be included. What about the other dimensions? One solution, instead of aggregating everything, is to go the other way – by disaggregating.2 That might seem like a retrograde step, like uninventing the wheel. If GDP was a brilliant invention – which it was – then why on earth would we start taking it apart? If you don’t like the idea of scrapping the wheel, think of it like this: how about liberating that frog from the matchbox?
The problem with growth as measured by GDP is that it has become the overlord of measures. It is the number we use to define success. Of course economists and policymakers look at dozens of data points – unemployment, inflation, net exports, retail sales, house prices and wages – to arrive at their models and predictions; Alan Greenspan, the former chairman of the Federal Reserve, used to examine men’s underwear sales as a proxy for economic activity. But in popular discourse GDP is king. Remember, economic growth is synonymous with GDP. To test out the supremacy of GDP, all you need to do is imagine a politician saying the following: ‘I propose shrinking our economy in return for X.’ No one angling for elected office would say such a thing, no matter what X is: flame-retardant public housing, a fairer society, a two-day working week, free pizza. The idea of not maximising growth has become almost unthinkable in modern political discourse. That is because growth has – almost without us knowing it – become all we care about.
This book does not argue – as some do – for GDP to be scrapped. However flawed, it is a powerful measure and a useful policy tool. As many rightly point out, economic growth often correlates with the things we do care about – education, health, freedom to choose how we live – because higher incomes and a higher tax base can provide the resources to pay for these things. Richer countries generally serve their citizens better than poor ones, as long as they are properly organised and have strong democratic institutions pushing for fairness and equality before the law.3
But this book does argue strongly for two things. The first is easy and quite within reach of any reader. Let’s call it scepticism. Anyone who has read this book – and if you’ve made it this far, thank you – should have a better understanding of what our growth statistics capture and what they don’t. What is in there that arguably should not be: pollution, crime, long commutes, missiles, long working hours. And what is not in there that arguably should be: good jobs, green space, decent healthcare, any measure of sustainability. So next time you hear the economy has grown by such-and-such, it is worth quietly chuckling to yourself and reflecting on the limited nature of that information, as well as the abstract nature of the economy being described.
I would argue that scepticism is valuable in itself. After all, as a journalist I’ve pretty much made a career out of it. But it gets you only so far. To say simply ‘I don’t believe in this number’ is a nihilistic way of looking at the world. Which brings me on to the second argument of this book and the question that always comes up in any discussion of growth and GDP: ‘What would you replace it with?’ None of the alternatives are robust enough or broad enough to supplant GDP altogether, so the answer is this. Rather than replace GDP, we should add to it, so that we can flesh out a more nuanced view of our world.
Being of the TV generation, I think of it like this. If I switch on the evening news and the newscaster is putting on a grave voice and headlining five or six numbers, which numbers should we select? Or, if you get economic news alerts on your phone, which are the numbers you really couldn’t live without? The idea is that the measurements in question should tell us something important. We the public should be invested in them, much as we are in growth as measured by GDP. Which numbers make the grade?
Here are a few suggestions.
This is so obvious it feels almost embarrassingly late in the day to mention it. A startlingly easy adjustment to make to GDP is to turn it into a per-capita figure by dividing it by the number of people in the country. Yet this is rarely done – at least not in normal public discourse. All too often growth is expressed in absolute terms, with no account taken of population expansion.
If your growth rate is 2 per cent but your population is also growing at 2 per cent then, on a per-capita basis, you are going precisely nowhere. Investors often get excited about the growth rates of some developing countries, forgetting that much of it is simply the result of high birth rates. The easiest way of increasing the size of an economy is simply to add people. If Donald Trump wants 3 per cent growth, it is very easy to achieve: all he has to do is knock down that unbuilt wall and invite in 10 million new people to America each year. What he really wants, however, is 3 per cent per-capita growth, which is quite another thing – and much harder to achieve.
Economists struggle to imagine how a country can possibly progress if it is not forever adding people to its workforce. That’s why so many talk about Japan, with its mildly shrinking population and positive per-capita growth rates, as being in a ‘demographic death spiral’.4 Economists are so wedded to the idea that the economy must always be expanding that they find it hard to break the logic of ‘just add people’. If Thomas Malthus thought more and more people would be the death of human civilisation, modern economists think the reverse.
Yet unless we imagine the world’s population increasing indefinitely, we really must begin to imagine a world where the economy eventually stops expanding – in rich, mature economies at least.5 That does not mean that income per capita necessarily needs to stop rising. And that is what ultimately counts. Reporting growth on a per-capita basis is a small but important step in putting people – rather than some abstract notion of the economy – at the centre of our policymaking.
Top 10 countries by GDP (purchasing power parity) per capita in 2015, published by IMF
Country |
GDP (PPP) per capita ($) |
Qatar |
132,870 |
Luxembourg |
99,506 |
Singapore |
85,382 |
Brunei |
79,508 |
Kuwait |
70,542 |
Norway |
68,591 |
UAE |
67,217 |
Ireland |
65,806 |
San Marino |
62,938 |
Switzerland |
58,647 |
|
Bottom 10 countries | |
Country |
GDP (PPP) per capita ($) |
Eritrea |
1,300 |
Guinea |
1,238 |
Mozambique |
1,192 |
Malawi |
1,126 |
Niger |
1,077 |
Liberia |
875 |
Burundi |
831 |
Democratic Republic of the Congo |
767 |
Central African Republic |
628 |
This goes one better than GDP per capita. It has one big advantage and one little one. The big advantage is that it is a median and not a mean. The mean is the average, but people often misunderstand what this implies. In a society with a total income of 100 consisting of four people, if one person earns all the income and the others earn absolutely nothing, then the average income is 25. The chief statistician of the Financial Times, Keith Fray, tells a joke at his staff presentations. ‘Bill Gates walks into a bar. On average, everyone in the room is a billionaire.’ I laughed out loud when I heard it, but then that’s me. The ‘humour’ (oh those statisticians!) stems from the fact that the average is merely the sum total divided by the number of people, in this case those in the bar. It can give a distorted picture of what we think of as typical. That’s because it doesn’t tell you anything about how income is distributed. As our societies grow more unequal, this is a bigger and bigger problem.6
In contrast to the mean, the median lines everybody up – which sounds ominous, I know – and picks out the person in the middle. The median gives you a rough-and-ready idea about how the typical person lives. You could call it a number for the precarious middle class. In this age, when the middle class feels forgotten and misrepresented, that’s a big advantage.
The little advantage of median income is that it deals with income rather than production. Income is a more intuitive concept for most people: what do I have to live on rather than how many forklift trucks were produced this year? In national income theory, production and income should be the same, since we buy what is produced and produce only what can be paid for, but it doesn’t always feel like that.
Steve Landefeld, the man in charge of US GDP for twenty years, is also a fan of medians. But, he says, they tell a story that might make some politicians uncomfortable. ‘Given the political sensitivity of those numbers, that could be something a statistical agency would be a little nervous getting into,’ he told me. ‘But I think there is a rising consensus in the US and other countries that it’s something we need to do.’
Even the notion of median income has its limits. In addition to finding out how the typical person or household is doing, it is right to shine a light on those left behind. There are several ways of doing this. After all, it took clever academic sleuthing to uncover the shocking fact that less-well-educated, middle-aged white Americans have shortening lifespans. One rough and ready measure of inequality is the Gini coefficient, which is often expressed on a scale of 0 to 100.7 Invented by Italian statistician Corrado Gini in 1912, at one end of the scale 0 represents a society of perfect equality in which everyone earns the same; at the other 100 is an economy in which one person earns everything. Relatively equal societies, such as those in Scandinavia, have a Gini coefficient of below 30, with the most unequal society in the world, South Africa, at 63.8 The US has a Gini coefficient of 41, the UK 33 and Germany 30.9 One could perform the same exercise for wealth or assets – which generally reveals greater inequality still.
Another way is to disaggregate income according to percentiles of the population. This reveals how income is shared and how the median fares against rich and poor. This chart tells the story of what has happened to US wages from 1980, for those at the top, in the middle and on the bottom strata of society. It also highlights just how far the typical wage is from the average.
Yet another way of encapsulating inequality in a single number is to compare one segment of the population with another, say the top 10 per cent by income against the bottom 10 per cent. Here’s how a select number of wealthy countries look according to that ratio, with the most equal at the top. For example: in relatively egalitarian Iceland, the top 10 per cent of the population earns 20.6 per cent of total income against 4.1 per cent by the bottom 10 per cent – for a ratio of 5 to 1. Meanwhile in Mexico, where the top 10 per cent earns a whopping 36.4 per cent and the bottom 10 per cent a paltry 1.7 per cent, the same is 21.4 to 1.
Iceland |
5.0 |
|
Germany |
6.7 |
|
France |
6.9 |
|
Canada |
9.3 |
|
Japan |
10.4 |
|
UK |
10.6 |
|
Turkey |
13.1 |
|
US |
18.3 |
|
Chile |
20.6 |
|
Mexico |
21.4 |
|
OECD11 |
GDP measures the flow of income, but we also need to know about the stock of national wealth. Otherwise we’ll get Bill, the banker, mixed up with Ben, the gardener. In the US the Bureau of Economic Analysis has been compiling net domestic product numbers for years, though these receive a fraction of the attention – to put it mildly – of GDP. Net domestic product (NDP) is calculated by subtracting the depreciation of capital goods such as roads, airports and housing from GDP. If a nation is adding to its capital stock, the NDP will rise. If not, it will fall. The gap between NDP and GDP gives you an idea of whether a country is running down its capital to achieve an unsustainable boost to current production.
Brent Moulton, who served for many years as the number two at the Bureau of Economic Analysis, thinks there is a strong case for putting greater emphasis on NDP. ‘One of the fundamental questions about a nation is: is its wealth growing or shrinking?’ he told me when we met at G Street Food in Washington, a sandwich joint that, confusingly, was not to be found anywhere near G Street.12 (Directions as well as measurements can be misleading.) Figuring out wealth, he said, is particularly important in countries that are dependent on natural resources.
But even a country like the United States, if it is to prosper, needs to nurture its stock of assets, whether its infrastructure, its universities or its natural resources. Whether or not countries are adding to their wealth should be a key question of national accounting.13 Only countries with a strong net domestic product can raise their standard of living over the long term.
Well-being
No serious advocate of GDP would claim it is a measure of well-being, yet few would deny that in public discourse it has morphed into a proxy of just that. What constitutes well-being is an intensely subjective judgement and thus impossible to measure objectively. But the attempt is worthwhile. That is because going through the exercise of trying to define and measure what well-being means helps societies prioritise. Trying to measure the unmeasurable – whether happiness or some broader measure of societal well-being – focuses the mind.
If Maryland’s Genuine Progress Indicator were given the oxygen of publicity and political backing, it would become the subject of noisy public debate. Imagine if the governor said he was going to change policy dramatically because the GPI had fallen. Some people would celebrate the indicator’s values, cheer when it went up and worry when it went down. Others would argue that the GPI measures the wrong things and should be jettisoned for something better. Either way, seeking to get a handle on well-being would spark debate about what, as societies, we are striving for, and provide a benchmark against which those goals can be measured.
CO2 emissions
I leave the most controversial to last. CO2 levels are the simplest way of monitoring natural capital because measuring carbon does not rely on trying to put a price on a forest or a watershed, an exercise that is fraught with difficulty. CO2 acts as a proxy for pollution. ‘If we’re frying our planet,’ says Joseph Stiglitz, ‘you want to know that we’re frying the planet.’14
This suggestion highlights something that should have become glaringly obvious over the course of this book. Statistics are not neutral. Nor are they boring. They are hugely political. If we bother to measure something, it is because we think it is important and want to influence it. For those who accept that high levels of CO2 are bad, the objective is to control or reduce it. But for those who don’t accept that there is any correlation between human activity and climate change, trying to limit CO2 emissions is a waste of time and money. More, it is destructive of what really counts: growth. That is essentially the argument Donald Trump made in pulling out of the Paris Climate Agreement.
The majority of scientists disagree with that assessment, saying that human activity is having a discernible impact on global temperatures. My position is this. I have no particular knowledge of climate-change science, just as I have no great understanding of aeronautical engineering. But when I board a plane, I trust that the engineers have got the art of keeping the plane in the air more or less down to a fine art. That’s also my conclusion for the overwhelming number of scientists who warn about the dangers of climate change. I don’t know, but I trust them. And if they’re right, it seems eminently sensible to track the level of CO2 and then do something about it.15 What’s the worst that could happen? More breathable air?
These are my suggestions. But they might not be yours. You may want to measure happiness or the regional distribution of income, or emphasise broader measures of unemployment.16 You may like Alan Greenspan’s underpants.
Whatever your preferences, there is always a trade-off between too much information and too little. As soon as you disaggregate numbers or come up with new things to measure, you have to weigh their importance. Many of the outcomes we might seek – a clean environment, healthy lives, safe streets, higher income, job security – are on different dimensions. Whenever you form an aggregate you lose information. On the other hand, if you don’t form an aggregate you can’t process or handle all the information that you do possess.17
One solution to this problem is the dashboard. Using the analogy of a car dashboard, or better yet one in an aeroplane cockpit, the idea is to monitor multiple things at once. Perhaps the ultimate dashboard concept is the OECD’s Better Life Index, which was launched in 2011 after a decade of work.18 This compares thirty-eight countries across eleven different ‘topics’ from housing and environment to security and income. The beauty of the index is that you can personalise it, giving whatever weight you like to each category. It is a sort of studio mixing deck of well-being. Say you value work-life balance above all else, simply turn that dial up to maximum and see which countries come out best. (Answer: the Netherlands.)19 The index is less of a policymaking tool and more of a way of illustrating the trade-offs between different desirable outcomes.
Another strong recommendation of this book is to improve how we measure public services. We tend to undervalue state provision of education, healthcare and roads. That is because they are provided free of charge. The UK’s Delivery Unit, which benchmarked everything from train delays to school exam results, was an imperfect but valuable step in the right direction. We should build on it. If we don’t measure the contribution of the public sector properly, the likelihood is that we will privatise it into oblivion, although that is precisely what many countries – steered by the invisible incentives of growth – are seeking to do.
In poorer countries many years of fast growth may be necessary to bring living standards up to acceptable standards. But growth is the means to achieve desired goals, not the end in itself. In the richer world the whole idea of growth – at least as conventionally measured – may need to be overhauled. In economies where services dominate, goods and services tailored to our individual needs will be what determine the advance of our societies. These could be anything from genome-specific medicines to personalised care or tailored suits. That is different from more and more stuff, an arms race of growth. Instead, it means improvements in quality, something that GDP is ill equipped to measure. Some fifty years ago one US economist contrasted what he called the ‘cowboy’ economy, bent on production, exploitation of resources and pollution, with the ‘spaceman’ economy, in which quality and complexity replaced ‘throughput’ as the measure of success. The move from manufacturing to services and from analogue to digital is the shift from cowboy to spaceman. But we are still measuring the size of the lasso.20
The quality issue also answers some of the underlying concerns about our seeming compulsion to acquire more and more material possessions, whatever the costs to our personal lives or the environment. The story of the Eurostar with free Chateau Petrus hints at how wealthy societies might ‘grow’ in future by improving the quality of the experience. Quality – whether well prepared, locally grown food, personalised medical care, more cultural and outdoor activities, individually tailored products or better design – is lower-carbon than quantity. Standard growth accounting leads us inexorably down the garden path marked MORE.
Some of the measures we might hope to emphasise – whether net domestic product or a broad definition of unemployment – already exist. Others, such as measures of natural capital, quality or well-being, are very much works in progress. We need to put more money and more clever people on to the task if we are to invent killer measurements to match the ingenuity of GDP. Accountants are only messengers. (Of course, if the accountant works for PwC and hands Warren Beatty the wrong envelope, delivering the wrong message can have embarrassing consequences.21) Accountants need support so that they can bring us useful messages. In a shameless bid to win the professionals over to the cause of this book, here is a plea: give them more money and power.
The UK contributes about £127 million each year to the particle accelerator at CERN, the European Organisation for Nuclear Research in Geneva. ‘For that, we get a share in measuring the mass of the Higgs boson to I don’t know how many decimal places,’ says Nick Oulton of the London School of Economics, referring to the elementary particle. ‘All very exciting stuff. But the Office for National Statistics’ budget – and that’s not just for GDP, of course, that’s for registering births, marriages and deaths and to track migration, employment, unemployment, all the components of GDP as well as GDP itself – its budget this year is £173 million. The natural sciences have convinced everyone that if you want answers, you have to spend a lot of money.’ 22 The same should be true of cracking the conundrums raised in this book. If GDP was the Manhattan Project of economics, then it is high time for its Moon landing and its Mars mission.
As well as money, national accountants need political cover. Statistics are controversial. They don’t always bring convenient news. Attempts in the US during the Clinton administration to introduce regular ‘green accounts’ were stymied by a Congress responding to the interests of big business, including the coal industry.23 Bill Clinton said, ‘We need not choose between breathing clean air and bringing home secure paychecks. The fact is, our environmental problems result not from robust growth but from reckless growth.’ Congress thought otherwise. The Bureau of Economic Analysis was threatened with a 20 per cent funding cut should it pursue green accounting – which was quietly sidelined. ‘Statisticians really can’t be expected to carry that kind of water,’ says one bureau insider, still bruised by the encounter two decades later. ‘This has got to be a societal decision to do these things. The statistical agencies can do all their homework, advance a proposal. But if democracy says no, then it’s no.’24
The gap between what economists tell us and our own experiences – the one that Nicolas Sarkozy said was so dangerous – is two-fold. It is also somewhat contradictory. On the one hand, GDP may actually be underestimating how well off we are. Because it is poor at measuring innovation, bald figures on income and production cannot capture the huge advances that have been made in health, technology, comfort and access to knowledge. GDP may be making us all more miserable than we have any right to be by under-representing what we already have. On the other hand, GDP overestimates some aspects of our lives. Though the economy is supposedly in constant forward march, many feel left behind, marginalised, abandoned and trapped in a never-ending race to pay for the goods and services that are supposed to define our lives. In these circumstances, simply increasing the size of the economic pie and then hoping we’ll each grab a decent slice is not a satisfactory or sustainable policy.
If we need better numbers, there’s an opposite and equal truth. We cannot govern by numbers alone. Rulers rule with rulers: they measure. But not everything can be costed and quantified. Not everything looks better because it is adorned with a dollar sign. That is one of the lessons of the recent political backlashes in which voters have rejected politics-as-usual.
The invention of GDP has given rise to a class of technocrats and economists who implement policy for the good of the economy, but not always for the good of the rest of us.25 They have inherited a Newtonian view of what an economy is, as though it were a rational and predictable system, ‘a singular entity with well-defined mechanical relationships between different moving parts connected by metaphorical pipes, cogs and levers’.26 Too often it is thought of as something outside human experience. As one unorthodox thinker wrote, ‘Mathematics brought rigor to economics. Unfortunately, it also brought mortis.’27
Before the invention of GDP, the word economy never figured in political discourse, something almost unimaginable today. No one thought about it as a separate entity. Until 1950, the idea of an economy never once appeared in a UK political manifesto with its modern meaning. All that changed with the invention of GDP. GDP was like a back door through which economists sneaked on to the public stage – and into the halls of government and bureaucracy.
Economists can bring valuable discipline to policymaking. But theirs is not the only view. There are competing ways to govern. It should not be necessary, as the venerable British Library has done, to justify its activities on the basis that, for every £1 it receives in public funding, it creates £4.40 for the economy.28 Nor should a well-known children’s charity need to encourage dads to read to their children on the basis that improved literacy will add 1.5 per cent to GDP by 2020. Some things – safe streets, good jobs, clean air, open spaces, a sense of community, a sense of security and well-being – are good in themselves. So is the love of books and reading. Sometimes more income will help us achieve what we want. Sometimes it will not. But more income – more GDP – should never itself be the goal. At most, it should be the means by which we achieve our ends. As Simon Kuznets himself asked, ‘What are we growing? And why?’
Our priesthood of economists depends on ‘the idea that there are economic laws we cannot contradict any more than we could contradict physical laws: that however much we should like to, we cannot go against the logic of “the economy”’.29 And yet, sometimes we can. And sometimes we must. The economy is not real. It is merely one way of imagining our world. Gross domestic product is not real either. It is just a clever way of measuring some of the stuff that we humans get up to. Growth was a great invention. Now get over it.