Another way to look at the recession: energy consumption is down





Note: Estimates represent 90% of national electricity consumption drawn from the grid, adjusted for season, time of day, day of week, weather and holidays. Source: Author’s Collection and US National Weather Service Automated Surface Observing Systems

New data on electricity consumption over the past three weeks suggests a sharp decline in US economic activity comparable to that of the Great Recession. This may already be the deepest downturn since the Great Depression; it is certainly the fastest.

These numbers are important because our official statistics cannot keep pace with the abrupt economic changes caused by the coronavirus shutdown. All those shuttered shops, silenced factories and darkened office buildings are not yet counted in the government’s official economic figures, which take months to collect, process and report.

But the evidence of the strong economic development is seen in a large and rapid decline in electricity consumption in recent weeks.

The numbers come from a new electricity-based metric that Steve Cicala, an economics professor at the University of Chicago, designed to track the state of the economy and how it’s changing on a day-to-day basis. The idea of ​​tracking electricity use, he says, stems from the observation that most economic activities require electricity.

Cicala’s findings are consistent with similar analysis from the US Energy Information Administration and reports from regional power providers. “In terms of this magnitude of event, I don’t think we’ve had anything like this on the grid in recent history,” said EIA analyst April Lee.

Mr Cicala said his indicator was useful in times of rapid economic change, adding: “While it’s not a perfect measure, it certainly helps close the gap so we can get the most complete picture.” .

The accompanying graphs show steep declines in electricity consumption in high-density cities, rural areas and industrial centers.

In New York, electricity demand fell 14% from its recent average as restaurants, theaters and offices closed.


A similar decline is evident in eastern Wisconsin and Michigan (the regions share the same energy market). On Tuesday, electricity demand was down 12%.


The idea of ​​tracking the economy through electricity consumption when more reliable indicators are not available is not new. For example, many analysts see China’s electricity consumption as a way to verify figures that provincial authorities are compiling under pressure to meet government quotas.

Although the United States has an array of reliable official economic statistics, none are well suited to tracking a downturn like this, when the unemployment rate can change as much in two days as it typically does on a whole year.

It is likely that this electricity data – which can be produced with a lag of around a day and is available for different parts of the country – will also provide the first clear signal that the current free fall is ending.

So far, the clearest picture of our economic deterioration has come from the number of people filing for unemployment insurance each week. But the government’s economic bailout has transformed the unemployment insurance system in ways that could skew those numbers in the weeks to come, making it difficult to spot a turning point.

For example, if the government’s small business loan program incentivizes employers to keep workers on the payroll, unemployment insurance claims could plummet, even as the labor market continues to deteriorate. Alternatively, the growing generosity of benefits could cause other employers to lay off inactive workers, which could lead to an increase in initial jobless claims for reasons unrelated to underlying labor market trends.

Consumption of electricityand the last recession

One way to assess the accuracy of this new economic indicator is to replay the signals it would have sent during the previous recession. Lehman Brothers filed for bankruptcy in September 2008, heralding a sharp turn in an already shrinking economy, but gross domestic product figures released a few months later initially suggested only a moderate drop in production. It took several months for the government to release revised figures revealing the true extent of the economic carnage.

In contrast, electricity consumption confirmed almost immediately that the economy had collapsed. In addition, electricity consumption fell the most in the regions of the country that later data would confirm had experienced the largest drop in employment.


Even though electricity consumption closely followed the last recession, this does not guarantee that it will start again. Today’s social isolation economy is very different from the pre-pandemic economy. The loss of production resulting from factory closures will be clearly visible in the reduction in electricity consumption, but other changes – such as differences in the energy intensity of working from home rather than in the office – could give more ambiguous signals.

How bad could this get?

Mr Cicala’s data also shows that electricity consumption has fallen sharply in the European Union in recent weeks. The impact of the coronavirus was initially greater in Europe, and so was the initial economic effect. If the United States continues to follow Europe’s path with a slight delay, it will experience a further decline in economic activity in the weeks to come.

But even worse results are possible. Electricity consumption in Italy – the epicenter of Europe’s coronavirus outbreak – has fallen more than three times as much as in the United States, suggesting a particularly steep economic decline.




Source: European Network of Electricity Transmission System Operators

Electricity consumption in selected areas

The attached figures show electricity consumption in a few markets across the country. The geographic regions covered may seem unfamiliar; they reflect the electricity service areas in the different regions of the country.

As you read these graphs, be aware that the technical adjustments needed to separate the effects of seasonal and weather patterns on electricity consumption from changes induced by economic activity are inaccurate, especially with daily data. Therefore, some daily shifts (especially weekends) may still reflect idiosyncratic factors like social distancing. In addition, changes in the use of electricity in certain regions can be more or less drastic depending on the industrial structure of the local economy.


The electricity data we report comes from hour-by-hour measurements of total electricity consumption in power supply areas. In most of the United States (and much of the European Union), utilities purchase electricity from generators in auctions that take place every few minutes, and the price and quantities purchased in these markets are made public almost immediately.

Electricity consumption is higher during the working day than in the evening; higher on weekdays than on weekends and major public holidays; higher in summer and winter than in autumn and spring; and higher when cold weather leads to increased heating or when warm weather leads to increased use of air conditioning. Cicala’s numbers adjust for these factors, effectively removing their effects, to better reveal underlying movements in electricity consumption that are more likely to reflect changing economic conditions.

The result is an indicator that can track changes in economic activity on a day-to-day or week-to-week basis, as electricity consumption rises and falls with the intensity of household use, commercial and industrial.

As useful as these figures are, over longer periods – say months or years – it is better to rely on more standard economic measures, which are less likely to be skewed by changes in the size of the population. industry or changes in energy efficiency.

Mr. Cicala will continue to update and publish these electricity consumption figures daily on the University of Chicago Energy Policy Institute website.

Comments are closed.