Quantifying Lost Economic Activity: COVID-19

The government orders to shut down business and stay at home have had an impact and will continue to have an increasing impact on the economy.  New unemployment claims already reached approximately 10 million in March.  Politico states, “Economic forecasts that predict unemployment will exceed its historic 25 percent peak during the Great Depression are becoming routine, and the number of jobs lost in a mere three weeks now exceeds the 15 million…”[i]  Just how much is the economy being slowed resulting in these horrific unemployment numbers, which will continue to increase each week? 

This article employs “use of electricity” as a surrogate or proxy for economic activity to answer said question.  Electricity usage is captured every hour and recorded immediately, so it is temporally a better measure of economic activity than other indicators such as gross domestic product (GDP), which severely lag and take a month to collect samples and then analyze.  Use of electricity is measured in megawatt-hours and referred to as electric load or demand.  Specifically, the electric utility demand for the PJM Regional Transmission Organization area, which covers the following 13 states and District of Columbia is examined; Delaware, Illinois, Indiana, Kentucky, Maryland, Michigan, New Jersey, North Carolina, Ohio, Pennsylvania, Tennessee, Virginia, and West Virginia.  PJM serves over 65 million people in ensuring electricity is reliably supplied to all homes and businesses in its territory, which is ultimately distributed to end-use customers by their local utility.[i]  The electric demand under mandatory economic shutdown is compared to the forecasted demand absent the mandated shutdown.  The resulting reduction in electric demand is reflective of reduction in economic activity, as the primary driver of electric demand and electric demand forecasting is economic activity.   The PJM Demand Forecast Supplement states, “The foundation for the PJM Load Forecast is a regression model that examines the dependent variable, load, versus a number of independent variables including end-use (which broadly captures economics and end-use indicators)…”.[ii]  As of April 13, 2020, electric demand, and correspondingly economic activity are down 10.86%.  Importantly, as the shutdowns wear on, demand and economic activity are projected to be down 18.80% by May 15th if businesses are disallowed from reopening.

DatePercent Down
March 29, 2020-7.14%
April 13, 2020-10.86%
May 15, 2020-18.80%

Downward Spiral Trend of Economic Activity

PJM provided the data[iv] on “actual load” versus “expected load” if the Coronavirus had not occurred resulting in government shutdown of much of the economy and restriction of personal activity.  PJM compared the loads being experienced under Coronavirus to expected loads based on 25 years of history.  They did this on a “weather normalized” basis, which controls for temperature impacts.  The hourly load data starts March 1st and runs through April 13th.  I analyzed the data and Coronavirus load impacts for this article using regression analysis and controlling for ambient temperature, as well as the type of day, such as weekday. 

Electricity usage is a good proxy for economic activity.  As businesses temporarily shut down or reduce hours, they use less electricity because equipment, lighting, and manufacturing processes are turned off.  Correspondingly, less electricity is used for gas pumps at gas stations since less driving occurs.  These are just a few examples of how electric load is reduced as economic activity is reduced.  As businesses go out-of-business, they shutter and use even less electricity as all electric service, such as lighting, refrigeration, and other basic facility functions are switched off.  The latter takes time to occur as businesses temporarily stay afloat but eventually deplete on-hand cash reserves.  Also, the increasingly ramped up government restrictions on business activities continue to drive load further and further down.

The mandated stay-at-home orders also continue to drive electric demand down since people as consumers purchase less and do less.  This slowing in consumer demand reduces electricity at manufacturing sites and retail establishments.  Consumers purchase less and do less in general also reducing demand for goods and services.  Electrical demand at homes might go up slightly at first since increased appliance and computer usage occurs, but it is more than offset by reduction in load at businesses and schools to where these individuals used to travel.  Unemployment will continue to rise drastically, with over 15 million increase over the past three weeks alone.1   If people lose their homes or rental residences due to unemployment or reduced employment, load at these residences will be reduced if there are a lack of replacement owners/tenants due to mass unemployment.  Loss of residence takes time to occur due to emergency funds, limited government bailout, and legal/regulatory hurdles, so expect a continuation of economic activity drop.

The graph below shows the percent reduction in actual electric load from forecasted load absent Coronavirus, and hence indicates reduction in economic activity, compared to where it is forecasted to be under normal conditions.  The trend line in red clearly shows load and economic activity dropping off over time.  The horizontal line at 0 indicates where the difference (delta) would be between actual load and forecasted load if the Coronavirus shutdown had no impact on economic activity.  The following link to Analysis Details describes the details of how these expected values were statistically calculated using Stata based on load data from the PJM website.

Figure 5


[i] “Unemployment claims near 17 million in three weeks as coronavirus ravages economy,” by REBECCA RAINEY and QUINT FORGEY, Politico, 2020/04/09

[ii] https://pjm.com/about-pjm/who-we-are.aspx

[iii] “2020 Load Forecast Supplement,” PJM Resource Adequacy Planning Department, January 2020

[iv] “PJM COVID-19 Near Term Load Analysis,” 2020/04/14, PJM Interconnection, https://pjm.com/-/media/committees-groups/pandemic/postings/covid-19-near-term-load-analysis.ashx?la=en

Leave a Reply

Your email address will not be published. Required fields are marked *