How seriously should policymakers take initial GDP estimates?
Gross domestic product is the key measure of the health of an economy. But how much store should we place in a figure that seems difficult to pin down?
On June 6 2017, Stats SA’s quarterly GDP update made for grim news. Two consecutive quarters of negative growth had pushed the economy into a technical recession. Nine months later, however, the national statistics agency revised its numbers and — just like that — the recession had never happened.
Last week’s GDP announcement revealed nothing as dramatic: Stats SA didn’t magic away SA’s 2018 recession. But it did revise its figures going back four years.
Growth was revised up from 1.3% to 1.4% for 2017, and down from 1.3% to 1.2% for 2015, and from 0.6% to 0.4% for 2016. For 2018, it was revised down by 0.1 percentage point in each of the first two quarters — to -2.7% in the first and -0.5% in the second — and third-quarter growth was revised up from 2.2% to 2.6%.
GDP, the estimated value of all the goods and services produced in a country over a specified period, is a key metric used by policymakers, economists and investors to judge the health of the economy. The figure is used to determine how an economy has grown or contracted over time — and thus informs whether policy should boost or rein in spending, for example. It also allows for comparison between economies.
Given that last week’s revisions spanned every sector and every quarter, it raises the question of how much store we should place in initial GDP figures. How much confidence can we have in a metric that seems difficult to pin down, yet informs the country’s economic policy trajectory?
"There’s nothing unusual about this," says Econometrix MD Azar Jammine. "To shift the figure a decimal or two doesn’t make much of a difference."
Similarly, Investec economist Annabel Bishop says Stats SA’s revisions to GDP data since 2000 have reflected fractional increases or decreases of about 0.1% from a GDP growth perspective, adding about 0.6% to overall growth in these 18 years combined.
SA is not unique in doing this: countries around the world restate their GDP on a monthly, quarterly or annual basis. In fact, initial estimates of GDP are always subject to revision; it can take two or three years to reach the most accurate figures.
It’s actually impossible to calculate the economy because economists don’t even agree on what aspects are important and what should be included in the measureDawie Roodt
Stats SA’s deputy director-general for economic statistics, Joe de Beer, says the agency updates its data in real time as new information becomes available. So the GDP figure may be adjusted as more accurate values are obtained for the sectors that contribute to the national product, or to account for seasonal fluctuations, or as tax income is reconciled.
"We have financial statements coming in months [after releasing GDP numbers] and that alters the figures," he says.
In a 2017 paper on GDP revision, Asanda Fotoyi, an economist with the nonprofit Trade & Industrial Policy Strategies research organisation, says SA revises its GDP for previous quarters and on an annual basis when it makes its quarterly GDP announcement. Major revisions are made every five years, at which point previous estimates are also restated at constant prices.
"GDP estimates … are needed for the monetary policy committee to set interest rates and the National Treasury to set budget limits, but because of the need for timely data, policy decisions are often based on preliminary estimates, which are later revised as more comprehensive data becomes available," says Fotoyi.
She cautions against "an overreliance on initially published preliminary estimates that are subject to change".
There’s also the question of measurement, given that much economic activity — in the informal sector, for example — goes unrecorded. And there’s disagreement about what should be measured in the first place.
"It’s actually impossible to calculate the economy because economists don’t even agree on what aspects are important and what should be included in the measure," says Efficient Group chief economist Dawie Roodt.
Others take issue with GDP itself. Political economy professor Lorenzo Fioramonti, for example, believes using a single statistic to measure the wellbeing of society has resulted in disastrous policy decisions.
Fioramonti argues that quality of life doesn’t necessarily improve with higher growth, while the potentially costly environmental effect of growth-driving industries is not factored into GDP calculations.
What it means
Last year’s revision to GDP numbers undid SA’s 2016-2017 technical recession
"Moral principles such as equity, social justice and redistribution," he says in his book Gross Domestic Problem, "are only taken up by policymakers if they comply with the GDP-led development model."
Roodt acknowledges the flaws in GDP, but he says the measure remains important as it is a readily accessible number that quantifies the health of the economy. "The measure of GDP is used to determine our fiscal deficit or current account as a ratio of GDP — and those are figures investors and credit rating agencies are paying attention to."
Similarly, Jammine believes society does "put too much emphasis on the GDP figures, but you do need to have a focal point". Thereafter, he says, economists also consider other indicators, such as unemployment, the level of inequality and the wage gap as measures of economic health.
For Fotoyi, circumspection is required, so that "not too much confidence [is] attributed to interpretations of the initially published estimates".
"When looking at quarterly GDP data, users are encouraged to focus on the trends rather than the actual growth rate figures," she says.