EMSI Data Update: Four New Categories

For the past year EMSI has been gearing up for a big update and improvement to how we present our labor market data.

Up to this point, we’ve made our data available in Analyst (EMSI’s web-based labor market tool) in two formats: Covered and Complete. This has enabled us to serve customers who need the data to remain similar to state-issued numbers in the Quarterly Census of Employment and Wages (QCEW), which is our Covered set, and those who want to see a comprehensive view of their economies, our Complete set.

One of the most valuable things that we do with our data is unsuppress QCEW with a high degree of accuracy. We have to do this because, in its raw form, QCEW hides firm-specific numbers for confidentiality reasons. However, the Covered set contains more than just unsuppressed QCEW; it also includes military, rail workers, and adjustments for records not classified in QCEW. And our Complete set was and is comprehensive, but it has actually contained three other employment categories in addition to QCEW. Wouldn’t it be nice to see all four categories on their own? Wouldn’t it be nice to show numbers in the tool that match QCEW? We thought so.

On July 10, our users can start to view employment data in four sets that can be mixed and matched.

Explanation of Employees & Proprietors

Before we get to the four new categories, let us make two important distinctions that will help us better understand the data. Our datasets are primarily concerned with those workers who are classified as either “employees” or “proprietors.”

  • Employees – Employees are often referred to as simply “wage and salary workers.” This includes workers receiving wages and salaries, as well as those working for commission, tips, pay-in-kind, and other similar forms of payment. These workers can be employed by any public or private organization. Government workers, regardless of the industry or agency (federal, state, and local) are also considered employees. People who work for all incorporated private organizations (for-proft and non-profit) are also considered employees.
  • Proprietors – Broadly speaking, any worker who does not fall into the “employee” category will be considered a proprietor. In technical terms, anyone who files Schedule C (Profit or Loss from Business), Schedule F (Profit or Loss from Farm), or Form 1065 (U.S. Return of Partnership Income) is considered a proprietor. These are people who work for their own unincorporated business, practice, or farm. It is important to note that people who work for their own incorporated businesses are considered wage and salary workers for their own companies, and are thus not considered proprietors. In addition, all partners in a business partnership are considered proprietors and counted individually.

4 New Categories

Now, here are the four categories. Again, we are really pleased to be able to bring these to you. They might be a tad more to learn, but this is a better way to break the data out and will alleviate some common data issues. Deacon James, VP of Data Products at EMSI, said of the changes, “The four classes provide clearer distinctions between types of workers, allowing our customers to find exactly what they need. I have been looking forward to releasing these additional worker breakouts for some time now.”


This is simply unsuppressed QCEW and will closely resemble what you’d get from your state LMI shop. This means that any employment number that QCEW publishes will show up in our tool as the exact same number. Anyone who needs EMSI numbers to stick closely to Bureau of Labor Statistics QCEW numbers will have employment numbers that match.

Advantages of this set: If you are working on projects (like grant applications) where it is more helpful to have numbers that resemble what you’d get from your state or if you just want to look at typical payroll employees, we recommend using this set.

2. Non-QCEW Employees

This set is other groups who are not captured by QCEW, but who still count as employees. This includes railroad, military, some non-QCEW federal government workers, UI-exempt non-profits, and a few other miscellaneous categories. This is particularly helpful for evaluating those tricky military and government sectors that can dominate regional economies.

Advantages of this set: When paired with QCEW, you get a complete picture of employment in the region.

3. Self-Employed

These are self-employed workers who count their self-employed work as their primary source of income. Being able to include it along side our more traditional datasets will be very informative.

Advantages of this set: A growing number of industries (see below) are composed of the self-employed. If you are just using QCEW to evaluate the industries, you might be missing a lot of jobs.

4. Extended Proprietors

These are workers who are counted as proprietors, but classify the income as peripheral to their primary employment. Many industries (primarily oil & gas extraction, finance & insurance, and real estate) include people who are considered sole proprietors or part of a partnership, yet have little or no involvement or income in the venture. And an increasing number of people fall into this category (e.g., those who do freelance work on the side, like writers or musicians) and now you will be able to see who they are more clearly.

Advantages of this set: This set is most useful when combined with the other three categories to get a complete picture of all proprietor and employee income (IO & general economist geeks).

Which One Should I Use?

Now, realizing that we are giving users access to MORE data also means that we need to help folks understand which set we would recommend that you use. Here is a simple way to think about it. If you previously used the Covered dataset, just switch to QCEW. If you used Complete make sure you have QCEW + Non-QCEW + Self-Employed selected. You might notice that we aren’t recommending that you use the extended proprietors for the “complete” set. This is because they don’t really help you get the best picture of the primary worker activity. If you do include them, just be aware that you are including a lot of jobs (particularly in real estate, oil & gas, and finance & insurance) that are not actually worker activity.

Here the simplified way of thinking about it.

Here are a few more pointers based on common things people look for in the data and how you can use the data in combinations.

  • I want to look at the best jobs people can go “get.” If you want to primarily consider jobs people would go and apply for (think becoming an employee rather than being self-employed), categories 1 and 2 are what you should be looking at.
  • I just want to look at salary data. In this case, it is just like the previous section. We recommend you use categories 1 and 2.
  • I want to look at ALL jobs in my economy. If you want to see all jobs, categories 1-3 is the way to go.
  • When Can I Use #4? Categories 1-4 should only be used for non-labor analysis (flow of money, input/output analysis). This would typically be for more advanced users. Feel free to contact us if you have more questions about input-output modeling.

Other pointers:

  • Colleges or organizations providing salary data for the most part will want to use categories 1 and 2 together as a default.
  • If you are familiar with or are held accountable to state LMI data, you should use only data category 1 (QCEW) because it will match what your state has. Note that we have removed suppressions, so there will be some differences. Workforce investment boards or other state agencies should probably use only this dataset at a default.
  • Private businesses, consultants, research groups and/or any group that wants to think about the complete employment picture of an economy should use categories 1, 2, and 3 as their default.
  • Input/Output – If you are using the input-output model or studying the flow of money in an economy, use all four sets together.

Data Illustrations

Below we have included two data illustrations to show how the different sets break out. Both demonstrate the usefulness of the data.

For the first table we’ve included the percentage of total employment each category has accounted for nationally from 2001-2012. The first thing to notice is how the proportion of workers captured by QCEW has declined. While it is still the greatest area of employment, fewer people are covered as payroll employees (think big companies with benefit packages). Where QCEW has declined, the extended proprietors has gained. Again, this is employment that represents a lot of 1099 workers who do things like have a hobby business, invest in real estate, or even referee sports. The proportion of self-employed and non-QCEW workers has remained fairly stable over the past 11 years.

(click to see a more full-screen version)

In the chart below we show what the four datasets look like by industry for each 2-digit NAICS sector. Sectors like manufacturing and admin, support and waste management are dominated by QCEW workers. Sectors like education services, other services, and government have larger numbers of non-QCEW workers. Extended proprietors make up roughly half or more of the agriculture, mining, and real estate industries. Many self-employed workers can be found in agriculture, construction, and the arts.

(click to see a more full-screen version)

Want to see a demonstration?

We know this might not be the easiest thing in the world to grasp. If you need more help or are just interested to learn more, please join us for a webinar or contact us for a demonstration. We will be hosting three identical webinars on July 16, 17, and 18 at 2 p.m. Eastern (11 a.m. Pacific). If you are interested, please register with Penny Rench.


Again, this has been a lot of information to digest, but we are really excited and believe the move to four data categories will help our users. These categories mean that you can choose exactly what numbers you’re showing so the data is much more transparent than it has been. You can pare the data down to simply unsuppressed QCEW, or you can look at only self-employed/proprietor workers. EMSI is committed to assembling accurate, comprehensive data that serves your needs.

“Our goal has always been to provide the best, decision-ready data to our customers,” EMSI CEO Andrew Crapuchettes said. “Adding this new level of granularity and flexibility will allow us to fulfill that goal even further.”

A PDF version of this article can be found here. In addition, we’ve developed this quick fact sheet on the data categories. Please contact us if you have any questions or would like to learn more.

These data categories are available in beta form in Analyst. Users can access the new data by going to their Account page and choosing “2012.2 Class of Worker BETA” under Dataset Version. Users who still need access to Covered/Complete data can choose the two older datasets.

2 Responses to “EMSI Data Update: Four New Categories”

  1. Cathy

    Question about Extended Proprietors–do they work at least 1 hour for $1 in pay? I’m not clear if they are like “silent partner” or if they are actually producing a good or service.


    • Rob Sentz

      Cathy. This dataset covers the same types of jobs as the Self-Employed dataset, but these jobs represent miscellaneous labor income for people who do not consider it a primary job. This category includes minor or underreported self-employment, investments trusts and partnerships, certain farms, and tax-exempt nonprofit cooperatives. It could be working in a small part time job like coaching or it could be investment in a real-estate partnership. Does that help?