New Separations Methodology and Its Impact on Occupational Employment Projections
by Sanjukta Chaudhuri
What are employment projections?
Employment projections are defined by the Bureau of Labor Statistics (BLS) as “…likely employment patterns in the U.S. economy”. This article discusses a major shift occurring in how long-term occupational employment projections will be developed moving forward. The “new separations methodology” has been created and executed by BLS. National long-term occupational projections for 2016-2026 employing the new separations methodology was rolled out by BLS as of October 24, 2017. This article aims to inform interested parties in Minnesota in a preliminary way about the new separations methodology with a planned roll-out at the state level set for 2018.
Occupational employment projections
The process of employment projection consists of a series of steps that produces occupational job openings projections caused by growth and separation.
Step 1. Growth projections of GDP. The employment projections process begins with projecting the future growth of the Gross Domestic Product (GDP). In this step economic growth rate and the Gross Domestic Product and its various components of final demand are projected. This step is the main source for estimating how much output the economy will produce in the future. This in turn is the foundation for projecting patterns of employment that will emerge in the projection year.
Step 2. Industry employment projections. The projected GDP is divided into various components of final demand for products and services. Final demand is then converted into industry level projected demand for their primary output.
Step 3. Occupational employment projections. The industry level projected outputs are converted into the resulting projected industry employment that would be needed to produce the projected output. A predicted distribution of occupations across industries is then used to create occupational employment projections.
Step 4. Occupational openings from growth. The occupational employment projections of the previous step feed into this step as the growth component of occupational job openings.
Step 5. Occupational employment caused by separations. To this is added occupational employment projections from separations. Separations occur owing to labor force exits (retirement, enrolling in school, geographic movement, job loss) and occupational transfers (an incumbent worker switching from one occupation to another).
Step 6. In the final step total occupational openings are calculated as the sum of occupational openings caused by growth and separations.
How are occupational employment projections developed? A simple example
Let’s begin with a highly simplified example to explain the basic principles of how projections of occupational openings will be developed.
Jobs Growth (Stock concept) : Let’s say that in 2016 there were 10,000 accountants. Because of business expansion and growth prospects, in 2017 the economy will need 10,500 accountants. Hence, the accounting occupation will add 500 jobs over the next year caused by growth.
Job openings (Flow concept) : The workforce is not static, it is dynamic, i.e., it is characterized by “churn”. For various reasons current incumbent accountants may leave over the next year. Let’s say that between 2016 and 2017, 400 accountants will retire or drop out of the labor force. So apart from the 500 new openings caused by growth, 400 accountants have to be replaced because of labor force exits. Another 200 accountants will separate from the accounting occupation for some other occupation with better job and career prospects. These 200 vacancies created by occupational transfer will also have to be filled.
Thus now the total number of job openings over the next year in the accounting occupation is actually 1,100, of which 500 or 45 percent are from growth, 400 are from accountants dropping out of labor force or retiring, and another 200 are from occupational transfer. Note that although the number of jobs increases by 500, the total number of job openings is more than double at 1,100. This is because job opportunities arise not just from growth, but also from labor force exits and occupational transfers.
If you now imagine the entire economy consisting of millions of workers across hundreds of occupations, it is easy to understand that the large volume of labor market churn means that the number of job openings is usually much larger than the actual employment levels.
What changes are happening in projections methodology?
The changes occurring concern projection of job openings caused by separations and hence affect only occupational employment projections, steps 5 and 6 of the projections process. Industry projections are not affected. Hence the remainder of the document focuses on changes in occupational employment projections caused by replacement needs.
Employment vs. Job openings: Employment is the total stock of jobs, and the growth in employment is from expansion in the occupation. Job openings on the other hand are not just caused by growth in an occupation but also caused by separations. In other words, job openings include any and all “opportunities to enter an industry or occupation for individuals not currently employed in the industry or occupation”.
Old Methodology: Employment Projections Owing To Replacement Needs
Since the 1990s BLS has used the old methodology for projecting job openings caused by labor market churn. Known as the Replacement methodology, the argument is that replacement needs arise when workers retire, die, or otherwise leave an occupation. It assumed a particular structure of a labor market where workers enter an occupation at a young age, stay in their chosen occupation for their whole career, and then age out or die at retirement, thus creating replacement needs to be filled by younger cohorts of workers. The idea was that job opportunities are only created when workers exit the labor force, thus ignoring the large volumes of inter-occupational flows that also create a substantial number of job openings. There were several disadvantages to the old method.
- It excludes occupational separations as a source of job openings, thus undercounting the number of openings.
- It assumes that labor market behavior (entry rate, exit rate) of younger cohorts will mirror that of older cohorts.
- It posited that a specific age cohort has a replacement need only if more of its members exited than entered. In cases where entry exceeded exit, the replacement need is set to zero, thus undercounting job openings in a second way.
- By focusing only on age, it ignores many other variables that together determine labor market churn and thus job openings.
How the new separations methodology differs from old replacement methodology
- Rather than assume a static nature in the labor market, the new methodology allows for a far more current and dynamic labor market, where job openings are created not just from retirements and exits, but also from incumbent workers separating from their current occupation and moving on to another.
- The new separations methodology takes into account a much broader set of factors that influence occupational separations, including not just age but also gender, race/ethnicity, education, location, industry type, class of worker, and full-time/part-time work status.
- It also expects a high percentage of all occupational openings to arise from separations. The projected number of job openings will increase under the new separations methodology. This is because the new separations methodology counts more job openings, because of its inclusion of occupational transfers as a major source of job openings. The old replacement methodology did not count this. In the example with accountants, the growth in the occupation results in additional 500 jobs, but another 600 openings are created from separations, i.e., 55% of all openings are from separations and only 45% from actual growth in an occupation.
New Separations Methodology in action
The example below is reproduced and extended from Table 1.1 of BLS Employment Projections - https://www.bls.gov/emp/ep_table_110.htm - and shows a highly abridged version of the national occupational employment and openings projections for 2016-2026. Released on October 24, 2017, using the new separations methodology, the following steps break down the numbers using the “all occupations” category as the example.
Employment change caused by growth (10 year projection converted to annual average)
Employment in base year 2016 = 156,064
Steps 1 through 4 (above) produce occupational employment in projection year 2026 = 167,582
Calculate employment change from 2016 to 26 =11,519. This is the employment change from growth.
Calculate percentage change in employment caused by growth from 2016 to 2026 = 11519/156064 X100=7.4%
Occupational openings from growth, annual average = 11,519 / 10 = 1152
Employment change caused by occupational separations (all annual averages only)
New separations methodology uses regression method to project labor force exit rate = 4.7%
New separations methodology uses regression to project occupational transfer rate = 6.2%
Calculate total occupational separation rate (4.7 + 6.2 = 10.9%)
Labor force exit: 4.7% x (156,064+167,582)/2 = 7,549
Occupational transfer: 6.2% x (156,063.8+167,582.3)/2 = 10,040
Total occupational separations = 7,549 + 10,040 = 17,589
Occupational openings owing to growth and separation (annual average) = 17,589 + 1152= 18,740
With the new methodology, 94% of all occupational openings annually will be from occupational separations, and 6% will be from growth.
What comes next for Minnesota Employment Projections?
With the release of national long-term Employment Projections, the stage has been set to roll out the new separations methodology at the state level. As the Labor Market Information office creates and rolls out the Minnesota long-term projections numbers in 2018, watch this space for more concrete occupational employment numbers and their interpretation under the new separations methodology.