Economists are used to analysing data that is weeks old by the time it arrives. GDP, unemployment and inflation figures are all released with a considerable time-delay.
The first estimate for US GDP, for example, is released a month after the end of the quarter to which it corresponds. The figure is then revised, with a second estimate a month after the first, and a third estimate a month after that.
This release schedule works during periods of calm. However, when policymakers are having to navigate periods of volatility, real-time data (or as close to real-time as possible) is essential. This means decisions can be made based on current developments, not what was happening a month ago.
COVID-19 has therefore, brought the importance of significant, timely data into sharp relief.
The pandemic has forced institutions such as central banks to catch up with private sector economists, by placing more weight on high-frequency, immediately available data.
The Bank of England uses data on credit card transactions, footfall on high streets and Google searches for terms relating to economic activity (such as “unemployment” or “estate agent”) to model demand in different sectors of the economy. These indicators are used to establish monetary policy decisions. The key is that these indicators are available quickly, so they show what’s actually happening in the economy.
Economists at Harvard and Brown universities have created a dashboard to detail the state of the US economy in real-time. This means the effects of the pandemic can be tracked quickly. Consumer spending, job vacancies and unemployment claims are mapped down to the county level, meaning the impact of the virus on local economies can be seen. Other indicators often used to show the health (or otherwise) of the US economy include demand for gasoline, restaurants and flights. These indicators all clearly show a sudden drop in demand as lockdowns were imposed, with a modest bounce back as restrictions have been eased. Demand for restaurants, however, has fallen back again in states that have recently seen surges in infection rates.
Economists also aggregate these high-frequency indicators into single indicators designed to provide a general overview of the whole economy - in the same way GDP does, but updated weekly.
The Bundesbank has recently began releasing a weekly activity index (WAI). Created in response to the pandemic, this measure of real activity is constructed using a number of indicators that cover different sectors of the economy: electricity demand, searches for welfare support, and consumer confidence surveys.
The Federal Reserve Bank of New York has also been releasing a Weekly Economic Index (WEI). Built in a similar way to the Bundesbank’s WAI, it uses ten different daily or weekly data sets, covering consumption, the labour market and output. It has been carefully calibrated so changes in the WEI can be interpreted as changes in quarterly (inflation adjusted) GDP.
Comparing GDP growth and the WEI, the WEI appears to be instructive about the trend of lagging economic measures such as GDP are heading.
Acting fast and with precision
Having timely indicators allows policymakers to act decisively. Alongside timely data, the availability of granular data on different sectors of the economy means interventions can be tailored to have maximum impact. If, for example, credit card data shows consumer demand is lagging, measures can be introduced to boost spending (such as cutting VAT).
Using their data, the economists at Harvard and Brown find that spending by households in high income areas fell by substantially more than that of households in low income areas (see chart). The fall was particularly severe in sectors requiring face-to-face contact (such as restaurants and beauty salons).
This in turn had knock-on effects. The new data showed that workers commuting to high income areas to work in these face-to-face services were more likely to face redundancy or cuts to their pay than other workers.
This was to be expected: high income households have greater discretionary spending, with a greater proportion spent on non-essential services.
But the dynamic also differs from previous recessions. Rather than being constrained by falls in their income, high income households were put off by fears of the virus. Spending on certain services not requiring face-to-face contact actually rose (e.g. gardeners).
The Harvard and Brown economists also found that simply reopening an economy did little to boost consumer spending. States that reopened earlier saw did not see any immediate boost to spending. On the contrary, spending followed a similar pattern to states which reopened later.
This adds weight to the argument that economies will only completely return to normal when a vaccine or treatment is found. This will allow for confidence to return: what’s good for public health is what’s good for the economy.
In the meanwhile, governments need to do everything they can to support workers whose businesses are hit by reductions in spending. This means providing welfare payments to workers, and support to businesses. Ultimately though, workers may need to be helped to find jobs in sectors that are growing. Policymakers now have the data to quickly identify which these sectors are.
Written by Thomas Schiller, Ad-Hoc Economics