The following Q&A details the data sources that Emsi is using during this economic crisis, as well as other sources that could be useful to you outside of Emsi tools.
What are the best pieces of Emsi data to look at for information about the impacts of COVID-19 shutdowns?
The Job Posting Analytics report found in Analyst and Developer (or the Job Posting Competition report in Talent Analyst) is currently the best source for real-time data, and the interface has been modified to expose daily postings with a 1-2 day lag at the most. March JPA was released in Analyst on 4/3/2020 for the US, UK, and Canada. Additionally, see the new publicly available JPA Dashboard on the Emsi site.
The Impact Scenarios in the Input-Output model (found in Analyst and Developer) are a good place to go to get an idea of the impact that job losses will have on a particular region. You can choose an industry you’re interested in, find out how many jobs there are currently, and plug any job loss information you find from the internet into the I-O model. The model will then show you the impact of the loss of those jobs on other jobs and regional earnings. In the absence of traditional LMI that’s current enough to speak to the COVID-19 situation, I-O is a good way to model the impacts.
How to model short-term impacts in the Input-Output model:
Nation: Ignore induced multipliers; focus on initial, direct, and indirect.
State & smaller: Watch for Type II multipliers coming soon in Analyst. These are tailored to model short-term impacts in state and smaller regions.
What traditional LMI data used by Emsi will be the first to catch up with COVID-19 changes?
The most up-to-date government source is LAUS, which is the source of our unemployment data. The latest county-level data available from LAUS is for January. Data for March will likely not be released until the second week of May.
Is there any traditional LMI data not used by Emsi that I can use that might be helpful?
Weekly Unemployment Insurance Claims: The weekly unemployment insurance claims data, filed by displaced workers, may be helpful for anyone trying to understand the magnitude of the hit to the workforce in the United States, and how that impact is changing on a weekly basis. This is the source of the data you are currently seeing referenced in news about unprecedented levels of unemployment claims.
From the link above, select the latest Unemployment Insurance Weekly Claims Report. A PDF will be downloaded. The first line of Page 4 of the PDF will show UI claims for the prior three weeks:
CES Monthly Employment Counts: The BLS’s Current Employment Statistics dataset measures monthly employment counts. This data comes from a monthly survey of about 145,000 businesses. It is available at the national level with varying levels of NAICS detail. National data for the prior month is released about a week into the subsequent month (e.g. March was released on 4/3/20):
Data is also available at the state and metro level with no NAICS detail (total employment). State/metro data for the prior month is released at the end of the following month (e.g. February was released on 3/27/30):
This data will differ from Emsi’s QCEW-based employment counts since QCEW and CEW collection methods are completely different. However, the benefit of CES in this situation is that it shows month-to-month employment, with significantly less lag time than QCEW, although QCEW’s counts are ultimately more accurate. For more information about the relationship between Emsi employment and CES employment, see this article.
What datasets is Emsi researching and considering adding to help during this time?
Emsi is currently working to publish monthly unemployment data when it is released every month, rather than waiting to publish several months at a time with each datarun.
Emsi is investigating the possibility of publishing both the UI claims and monthly CES data outlined above.
How will Emsi’s employment projections show COVID-19 effects?
It is important to remember that Emsi’s projections are forward projections of past employment trends. They do not take into account any legislation, current events, or non-employment factors. Changes in recorded employment will affect projections.
Since Emsi’s employment projections are based on historical employment data with slight adjustments for BLS and state projections, they will only change as historical employment data reported in the BLS’s QCEW dataset changes. The effects of the coronavirus layoffs will begin appearing in QCEW 2020Q1 data, which is due to be released by the BLS on 9/2/20. This data will be incorporated into Emsi’s 2020.4 datarun, due to be released on 10/9/20.
The effect on projections will likely be slight initially but will increase over time as more quarters of 2020 QCEW data become available. As more data comes in showing job loss, each datarun’s projections will likely decline progressively in comparison with the prior datarun’s projections (e.g. an occupation with 10-year projected job growth of 100 in the 2020.3 datarun might show growth of 90 in the 2020.4 datarun, 50 in the 2021.1 datarun, and so on).
How are openings and replacements data affected by COVID-19?
Replacement data comes from the Bureau of Labor Statistics and is published every two years. This dataset is simply a national-level estimate of the percent of each occupation that needs to be replaced every year. Emsi multiplies Emsi job counts by these replacement rates to arrive at the number of replacement jobs. The next update to replacement rates is not due to be released until Fall 2021, and it is too soon to say what (if anything) the BLS may do to try to model the effects of the COVID-19 shutdowns on replacement needs.
Emsi’s Openings counts are simply replacement jobs plus jobs due to growth. Openings will reflect COVID-19 changes to the extent that they show up in employment data (from QCEW, effects beginning to show 10/2020) and replacements data (possibly no effects shown; data from BLS released late 2021).
If you have additional questions concerning Emsi data during this time, please contact your account manager, or use the Submit a Question feature on the right.
Submit a Question
Let us know what specific questions we can help you with (we may even add your question to our knowledge base).