Author: Andrew Berger-Gross
In a world awash in economic uncertainty, we rely on dependable sources of information to guide our decision-making. Data from government agencies such as the U.S. Department of Labor’s Bureau of Labor Statistics (BLS) offer a reliable and unbiased reading on the condition of our economy. However, these data represent estimates and should be interpreted with caution when they are initially published, as they are often released in preliminary form and are subject to revision.
Each year around this time, the BLS publishes revised job growth and unemployment and labor force statistics for all states. These annual revisions incorporate information that was not available when the data were first published. We report on these revisions every year; see last year’s article for an example. While revisions are necessary to ensure our understanding of the economy is as accurate as possible, they can result in changes that are sometimes modest, and sometimes dramatic.
In this article, we summarize recently published revisions to North Carolina’s job growth estimates.
The COVID-19 crisis led to a historic plunge in employment levels in spring of 2020, followed by a rapid improvement in labor market conditions. Revised figures show that the number of jobs in North Carolina declined slightly less during the worst of the crisis, and ended the year slightly higher, than indicated by the preliminary data [Figure 1]. While notable, these topline revisions do little to change our understanding of the impact of COVID-19 on North Carolina’s economy.
Figure 1
However, when digging beneath the headline numbers, we find that data revisions materially affected trends in certain sectors of the economy. For example, North Carolina’s manufacturing sector added jobs during the second half of 2020 at a faster pace than initially estimated, ending the year only 4% below its February 2020 employment level versus the 6% reported in the preliminary numbers [Figure 2].
Figure 2
These revised data also affect our understanding about how COVID-19 impacted our state’s Metropolitan Statistical Areas. Preliminary estimates indicated that metro areas in North Carolina were hit harder than the state’s non-metro areas [Figure 3]. Revised data show that metro-area employment was higher than initially estimated, suggesting that the employment impact of COVID-19 was equivalent across the urban and rural portions of our state.
Figure 3
What do these revisions mean for data users? How can we understand what is happening in the world around us when economic statistics that are reported today are subject to revision at a later date?
It is important to acknowledge that no data point is perfect. Any effort to gauge economic conditions in real time is bound to be clouded by uncertainty. We recommend that data users consult a wide range of high-quality information sources to guide their decision-making. If all these indicators point in the same direction, then you can be relatively confident in the story they are telling. If these indicators contradict each other, then you should exercise caution before drawing conclusions about the state of the economy.
We also recommend that you be judicious when interpreting month-to-month movements in economic data and, if possible, focus instead on long-term trends. Monthly data are often noisy, subject to revision, and represent only a snapshot of the economy. Long-term trends are much more stable, less affected by data revisions, and provide more reliable information about what is happening in our economy.