GDP: a wealth of meaning is crammed into those three letters. Gross domestic product is our principal index of economic welfare. It is the subject of conferences and hand-wringing editorials. It is the metric by which the success or failure of government policy is judged.
Such is the ubiquity of GDP in public life, it is easy to forget it is not a naturally-occurring phenomenon but a human invention, like chocolate cake or the internal combustion engine. And like an old car trundling along the motorway, GDP is struggling to keep pace with the demands of the modern world.
As investment shifts towards service sectors and digital assets, GDP is failing to provide an accurate picture of how economies are performing. It does not account for the distribution of economic gains, which might explain why Western governments failed to anticipate the populist votes of recent years. So is it time for GDP to be replaced? Or does it remain an indispensable tool, despite its flaws?
“If you are trying to value spending and income, GDP remains the most relevant statistic. But it is not perfect,” says Stewart Robertson, senior economist for the UK and Europe at Aviva Investors. “It is quite easy to measure the number of widgets coming out of a factory, but quantifying the economic value being produced by digital and creative industries is much more difficult. In an economy that is evolving in these directions, it is legitimate to ask whether GDP is still fit for purpose.”
A brief history of GDP
GDP is a relatively recent creation. The first comprehensive national accounts were devised by the Russo-American economist Simon Kuznets, who worked in US President Franklin D. Roosevelt’s administration in the 1930s.
Roosevelt’s predecessor Herbert Hoover had relied on patchy data such as stock market indices and freight-car loadings to measure the economic impact of the Wall Street Crash of 1929. But Kuznets devised a more accurate metric, which could indicate a nation’s entire output in a single number. He and his small team criss-crossed America, visiting factories and farms to conduct interviews and collect data. In 1934 he presented his first report, which revealed a shocking fact: the US economy had almost halved in size since the crash.1
In the run-up to the Second World War, the US government found GDP a useful way to measure the nation’s economic capacity – and the likelihood of military victory. Kuznets worked in the Planning Committee of the War Production Board and may have influenced the timing of America’s entry into the war, having concluded the US would be better positioned if it delayed its involvement.2
Around the same time, the British economist John Maynard Keynes was doing similar work across the Atlantic. Keynes’s crucial contribution was to include state spending in calculations about the scope of the economy. Building on Keynes’s theories, governments in the post-war period used national accounts statistics to create econometric models to project the impact of policy decisions using so-called ‘fiscal multipliers’. GDP became a key tool of macroeconomic management.
Growth and welfare
GDP is failing to provide an accurate picture of how economies are performing
There are three main ways to measure GDP: production, income or expenditure. The most familiar method uses expenditure, according to the formula GDP = C + I + G + (X – M), or consumer spending plus gross investment plus government spending plus exports less imports. But this seemingly-simple equation hides many complexities.
For example, consumer spending estimates are subject to a seasonal adjustment to smooth out fluctuations over the course of a year. Like other components of demand, GDP numbers are adjusted for inflation using a general price index or deflator. Further tweaks are needed to compare how different economies are faring; hence conversion rates for purchasing power parity (PPP).
The final GDP figure therefore involves a lot of adjustments, estimates and guesswork based on previous findings, says Robertson. “You measure a nominal value and derive a price index, and use the price index to establish a real value. The process is quite convoluted with plenty of scope for error. In 1996, the Boskin Commission in the US discovered the consumer price index (CPI) overstated inflation by 1.1 percentage points a year, which meant output was higher than previously thought."
Despite these issues, 20th century governments found GDP a neat way to gauge the health of their economies and a workable proxy for overall living standards. Strong growth in the post-war era came hand-in-hand with increased employment, higher incomes, a greater range of consumer goods and widespread innovation – what British prime minister Harold Wilson called the “white heat” of technology. Higher GDP-per-capita also proved to be positively correlated with lower infant mortality and longer life expectancy, according to research from Nicholas Oulton at the London School of Economics (see figure 1).3
What’s in and what’s out?
Nevertheless, GDP is an imperfect index for a nation’s overall welfare. In his recent book The Growth Delusion, Financial Times journalist David Pilling criticises what he calls the “cult of growth”, arguing that policymakers’ focus on GDP is problematic. He points out GDP growth depends on a perpetual increase in human consumption, which threatens to cause irreversible damage to the environment.
Pilling also points to the fact that GDP includes all sorts of morally-dubious practices in its measurements. Kuznets believed GDP should only incorporate activities that contribute to human welfare, but nowadays war, organised crime and natural disasters can all boost output.
Combined with GDP-based spending targets, these inclusions can lead to absurdities. In 2015, the FT reported that a rise in illegal drug trafficking and sex work had added £9.7 billion to UK GDP the previous year; on the same day it ran an editorial discussing Britain’s pledge to keep its defence expenditure at two per cent of GDP. Joining the dots, a reader’s letter wryly observed: “If only prostitutes worked a bit harder, the army could have a few more guns.”4
Even as it includes the proceeds of illegal work, GDP leaves out a range of activities society deems beneficial, such as unpaid social care, childrearing and housework. In 2000, British economists calculated that total unpaid household labour was worth £877 billion, or about 45 per cent of all the country’s economic activity for that year.5
Aside from the discussions over what to include, there can be problems in retrieving the relevant data, especially in emerging economies where vast swathes of commercial activity are invisible to central government statisticians.
In 2014, for example, Nigeria announced the results of a three-year statistical survey that discovered the economy was 90 per cent bigger than previously thought. The country leapfrogged South Africa as the continent’s largest economy overnight; its debt-to-GDP ratio fell precipitously and its markets immediately began to attract more foreign investment. But everyday life changed little for the people running small rural businesses whose economic contribution had been belatedly recorded in the national accounts.6
The intangible economy
Measuring GDP is also becoming more difficult in advanced economies as services replace manufacturing as the key driver of growth. It is much easier to measure the output of a factory that produces countable objects than a service-based company specialising in consulting or product design.
GDP involves a lot of adjustments, estimates and guesswork
The growth of investment in intangible assets such as data, design and expertise only exacerbates this problem. Many cutting-edge companies now rely on digital platforms with scarcely any physical presence: Airbnb, the world’s largest accommodation provider, owns minimal real estate; Uber, the world’s biggest taxi company, has hardly any vehicles in its portfolio.
Jonathan Haskel, a professor of economics at Imperial College London who joined the Bank of England’s Monetary Policy Committee in September, explored the rise of the intangible economy in his book Capitalism Without Capital, co-authored with Stian Westlake.7 He says GDP fails to accurately record the activities of intangible-focused companies.
“If we are going to track the economy via GDP, particularly when it is turning up or down, we need to better measure intangibles. And since many intangibles are about product upgrading – think of spending on design and branding – we need to be sure we are measuring new products and their prices when we look at GDP.
“When we look at wider well-being, all these factors come in as well: if we are to measure consumers’ welfare and the reach of their purchasing power, we need be sure we are capturing all that they can potentially buy,” Haskel adds.
It is likely that some of the economic contribution of technology companies is going unrecorded, which could mean we are underestimating overall growth. Then again, aspects of the intangible economy are probably bad for GDP. Jobs that would once have been performed by paid employees are being automated. Airbnb has no need for hotel clerks or concierges, for example; customers make their own bookings and carry their own bags to their rooms – in economic parlance, these roles have moved outside the ‘production boundary’.
As Will Page, director of economics at music-streaming service Spotify, put it: “The goal of disruptive technology companies, in the statistical sense, is to reduce GDP. To wipe out transaction costs, which are being measured, and to replace them with convenience, which is not being measured. So the economy is shrinking but everyone is getting a better deal.”8
The rise of populism
But is everyone really getting a better deal? Since the turn of the century, new technologies have created a range of covetable consumer gadgets that have made our lives more convenient, but dissatisfaction with the state of Western economies is growing. The Brexit vote in the UK and the election of Donald Trump in the US were widely attributed to a sense of futility among voters who had been excluded from the fruits of economic growth.
In relying on GDP as their chief economic indicator, governments may have missed underlying trends. GDP was never intended to measure economic distribution: it simply measures the size of the cake, not how it is divided up. According to some experts, policymakers’ failure to anticipate the rise of populism may have been the result of their dependence on top-line GDP measurements amid the fast-changing dynamics of the modern economy.
Illegal drugs and sex work contributed £9.7bn to UK GDP in 2015
“If the Office for National Statistics in the UK had put more resources into collecting regional and finer geographical statistics in the past, we would have known that some parts of the country simply haven’t benefited from GDP growth for about 10 years or more; it was all very concentrated in the southeast,” says Diane Coyle, Bennett professor of public policy at Cambridge University and author of GDP: A Brief but Affectionate History. “We think we are only measuring what we see, but in fact it’s the other way around: we see what gets measured.”
In a recent paper written with Benjamin Mitra-Kahn, Coyle argues for a two-stage reform to GDP: as an initial measure, GDP should be amended to incorporate intangible assets, remove unproductive financial investment and adjust for income distribution. Eventually, GDP should be replaced with a “dashboard” that records “access” to six key assets: physical assets, natural capital, human capital, social and institutional capital and net financial capital. Coyle and Mitra-Kahn argue this approach would help avoid the sort of “complacency about economic performance” we have seen in recent times.9
“We ought to pay attention to the distributional question even if we don’t change the statistics,” says Coyle. “If we ask what are the sorts of assets people have access to, we would start to think about other things: What is the transport infrastructure available to people in areas of low income? What are the schools like; are they able to build up the human capital to give people the life chances they need? You can think about distribution in a much more empowering way if you have these kinds of figures.”
Alternatives to GDP
Other experts say GDP is worth sticking with, despite its flaws. In 2017, Coyle and Mitra-Kahn’s paper shared the inaugural Indigo Prize in Economics with an entry written by Haskel and colleagues, which makes the case that many of the mooted alternatives to GDP throw out the baby with the bathwater. While GDP is flawed in many ways, it still provides policymakers and investors with a useful barometer of how economies are performing and therefore a basis for efficient capital allocation.
Haskel and his team argue GDP fulfils two key principles of measurement alternatives do not. First, it avoids ‘double counting’ by only measuring value added at each stage of production (for example, it counts sales of sandwiches, but subtracts the bread used to make them). Second, its reliance on prices ensures items are accorded different weights depending on their relative economic importance at a particular point in time.10
Haskel’s team argues we should hold onto GDP, but reform it to address its principal flaw; namely its failure to account for intangible assets and wider human welfare. To ensure the use of ‘free’ intangible assets such as apps and websites are recorded in the national accounts, Haskel’s team advocates using surveys to determine how much consumers would pay for them. Such surveys have found users would be willing to pay US$14 per month to preserve their access to Facebook and as much as US$1,300 per month for search engines.
GDP could also be extended to encompass broader human and environmental wellbeing. New variables could be added to national accounts to give a more-comprehensive picture of overall welfare without compromising GDP’s usefulness as an indicator of growth, investment and economic performance. For example, leisure time and life expectancy could be measured as a supplement to the consumption figures, on the basis that more income is of no use if you have no free time in which to spend it.11
There have been attempts to put these kinds of reforms into practice. In the US, the Maryland state government now refers to an index called the Genuine Progress Indicator (GPI) before making its budget decisions.12 The GPI supplements GDP figures with variables such as leisure time and unpaid housework, and subtracts so-called ‘regrettables’ including pollution and time spent commuting.
Similarly, in 2013 the Australian Bureau of Statistics launched a platform called Measures of Australia’s Progress (MAP) to track education, health and social trust alongside traditional economic variables.13MAP showed that while the economy and per capita income had increased over the previous decade, social trust had stagnated and the health of the natural environment had regressed.
Despite these innovations, it is likely to be some time before the majority of states, economists and investors end their reliance on GDP, says Coyle. “There’s a lot of interest in change at the moment, but it’s a bit like having a technical standard, like driving on the left side of the road. Nobody is going to switch until anyone else switches. If politicians started saying GDP is not important, all the newspapers would say: ‘Well you’re only saying that because it isn’t growing.’
Users would be willing to pay $1,300 to use free search engines
“So there needs to be some kind of consensus and enough intellectual firepower behind switching to something else, as was the case when GDP was invented during the Second World War and immediately afterwards. The debate about what we would switch to is still going on,” Coyle adds.
For now, GDP is the best metric we have for the state of economies, the flow of investment and per capita wealth. To paraphrase Winston Churchill’s famous observation about democracy, GDP remains the worst way of measuring economies apart from all the others that have been tried from time to time.