Major Change Coming to Replacement Rates

A heads-up if you use occupation replacement rates in your work: new rates are being developed and are planned to be released in two years…and it looks like they will be very, very different. On average, over four times larger.

Right—four times larger; that is not a misprint.

Occupation replacement rates are developed by the Bureau of Labor Statistics (BLS) to describe the number of workers who leave their occupation and need to be replaced by new entrants into the occupation. The rates describe demand due to workers leaving the workforce (such as retiring) plus those moving from one occupation to another. Replacement demand is important in the fields of workforce development and education because—along with growth demand—it helps gauge the future training needs for specific occupations.

When publishing replacement estimates, the BLS includes the caution that the replacement needs are “underestimated” due to limitations in the methodology used for calculating these rates. In turn, in our JobsEQ system which uses these BLS rates, we reference the same caveat.

The Bureau of Labor Statistics has developed a new method that they believe is a more accurate measure for this type of occupation demand. Accompanying this change, the BLS plans to change terminology from “replacement rate” to “separation rate,” partially to help highlight the change in methodology. Regardless of the name change, the BLS emphasizes that the “new method is designed to measure the same concept as the old methodology: workers who leave their occupation and need to be replaced by new entrants into the occupation.”[1]

While the new data have not yet been officially released, the Bureau of Labor Statistics has posted some “experimental” results for comparison purposes. Even though the rates for some occupations change only slightly—dentists, for example—the new rates for most occupations are substantially larger. As one example, the estimated replacement demand for machinists (SOC 51-4041) increases roughly by a factor of four when put in terms of “occupational separations.” Using the old method, replacement needs for machinists during the period from 2012 through 2022 were estimated at about 91,000 workers; using the new method, separations for machinists in the same period are estimated to be over 392,000.

The reason for the large increase from the current replacement rates to the new separation rates, to quote the BLS, is that “the current method undercounts openings because it only accurately measures workers who follow a traditional career path—entering an occupation at a young age, working in the same occupation for many years, then retiring—which is not the case for many workers in most occupations.”[2] The new method is generally described as being “more robust and more statistically sound,”[3] the details of which can be found on the BLS website.

To be complete, there is another technical difference between replacement and separation rates in how each deals with declining occupations, but that change is not as impactful on the change in magnitude in the rates compared to the other methodological changes.

The bottom line is that even though the new official rates may not be available for another two years, this impending change highlights the need to keep in mind the caution about replacement rates: the current rates should be considered only as a minimum measure of training needs due to replacements.[4] It is a caveat of particular importance, given that the degree of underestimation may be severalfold.

Replacement and Separation Rate Comparison for Select OccupationsReplacement and Separation Rate Comparison for Select Occupations


[2] Ibid.

[3] Ibid.

[4] For example, see the BLS description of that here:

The Graying of America

Most people know that the percentage of older Americans is increasing dramatically. What’s less known to the average person is how that graying will impact different areas of the country. The overall demographic shift is illustrated in the chart below that shows four age cohorts. The first three are each based on equal twenty-year spreads: people age 0 to 19, those who are age 20 to 39, and the age group 40 to 59. The fourth cohort is defined as those age 60 and older.

The Decline in College-Aged Students

The “echo boom” of births in the United States peaked in 1990. The children of that peak became college-freshman aged around 2008. Since then, the population of 18 to 19 year olds in the nation has been trailing off. Children born at the peak of the echo boom are now about age 25, and most from that peak are out of college. As a result, the size of the prime college-aged population is on the downswing.

The prime college-attending ages are 18 to 24. This cohort makes up about 58% of the college student population.[1] The population in the United States age 18 to 24 peaked in 2013 at 31,535,000. As of 2015, this population cohort is estimated to have slipped to approximately 31,214,000, a drop of one percent.

This downturn is expected to continue till 2020 when the age 18 to 24 cohort population hits a trough of roughly 30,555,000—a drop of 2.1% from 2015 levels. 

Population projection 2014-2030

Some areas of the country will be seeing more drastic declines than this, while other areas can expect to see no drop at all. Nine states are projected to see growth in the 18 to 24 cohort over the coming five years. From 2015 to 2020, states expected to see this population cohort expand include Utah (+3.9%) and Texas (+3.4%).On the other hand, states forecast to see steeper-than-average declines include Michigan (-6.9%) and New Mexico (-6.8%). 

There is some good news, however, for those wanting to see an increase of population in the college-aged cohort. The number of births in the United States hit a trough in 1997—many children born in that year are just beginning their freshman year in college right now. Following 1997, the number of births began trending upward and peaked in 2007 at a height surpassing that of the echo boom. So while postsecondary schools are facing some unfavorable demographics in the short run, another swell is on its way. 

Research support was provided by Allison Magee and Asim Timalsina.

[1] Based on Fall 2013 enrollment data from the 2014 Digest of Education Statistics.

Why Did Our Unemployment Rate Change?

You may have noticed that your recent historical unemployment rate numbers look different—very different—compared to how you remember them looking just a few months ago. If you’ve noticed this, you’re not alone.

Cost of Living

The cost of living is one of many metrics we examine when assisting site selectors in identifying attractive labor markets for relocations or expansions.