If you google "tech unemployment" (without the quotes), you'll find a number of articles written in June and July of 2019 which refer to tech unemployment reaching a record low. The publications which posted the articles include The Wall Street Journal, Fortune, Yahoo Finance (quoting Fortune), Dice Insights and many others. The original source for the data appears to be the Computing Technology Industry Association (CompTIA), self-described as "a leading voice and advocate for the $5 trillion global information technology ecosystem". The The Wall Street Journal article states:
Tech trade group CompTIA estimates that the unemployment rate for IT occupations in the U.S. dropped to 1.3% in May, a 20-year low, according to an analysis of the latest Labor Department jobs data.
As a result, employers are competing for a diminishing pool of talent. “The demand for tech talent has reached historic levels,” said Tim Herbert, the group’s executive vice president for research and market intelligence.
The CompTIA article gives some prior numbers, stating:
At 1.3 percent tech occupation unemployment is at its lowest rate going back to January 2000, the earliest available detailed occupation-level data from the BLS. The previous low of 1.4 percent occurred in March 2018 and April 2007.
The following graph and table were generated using data from the IPUMS Current Population Survey (CPS) website:
CURRENT POPULATION SURVEY: 2016-2020 Computer and Mathematical Occupations (1000-1299): 2016-2020, grouped by EMPSTAT (percent) Year_Mo Count Employed Unemployed 1 2016-01 4,644,692 97.5 2.5 2 2016-02 4,800,642 97.6 2.4 3 2016-03 4,702,703 97.6 2.4 4 2016-04 4,558,115 97.8 2.2 5 2016-05 4,457,227 97.9 2.1 6 2016-06 4,558,584 97.7 2.3 7 2016-07 4,502,332 97.1 2.9 8 2016-08 4,679,768 97.4 2.6 9 2016-09 4,570,992 97.1 2.9 10 2016-10 4,673,003 96.8 3.2 11 2016-11 4,804,564 97.0 3.0 12 2016-12 4,914,852 97.4 2.6 13 2017-01 4,821,896 97.2 2.8 14 2017-02 4,736,217 97.2 2.8 15 2017-03 4,664,857 98.0 2.0 16 2017-04 4,797,708 97.6 2.4 17 2017-05 4,675,877 98.3 1.7 18 2017-06 4,787,297 97.9 2.1 19 2017-07 4,718,601 97.9 2.1 20 2017-08 4,842,177 97.6 2.4 21 2017-09 4,959,509 97.2 2.8 22 2017-10 4,897,268 97.4 2.6 23 2017-11 5,110,318 97.5 2.5 24 2017-12 5,087,143 97.5 2.5 25 2018-01 5,255,328 97.3 2.7 26 2018-02 5,310,304 97.4 2.6 27 2018-03 5,202,590 98.6 1.4 28 2018-04 5,074,713 98.3 1.7 29 2018-05 5,072,349 97.7 2.3 30 2018-06 4,981,258 98.0 2.0 31 2018-07 5,053,226 98.1 1.9 32 2018-08 5,192,383 97.5 2.5 33 2018-09 5,076,657 98.0 2.0 34 2018-10 5,171,995 97.9 2.1 35 2018-11 5,244,786 97.6 2.4 36 2018-12 5,266,714 98.0 2.0 37 2019-01 5,228,613 97.7 2.3 38 2019-02 5,502,596 97.7 2.3 39 2019-03 5,286,358 98.3 1.7 40 2019-04 5,276,328 97.5 2.5 41 2019-05 5,315,334 98.7 1.3 42 2019-06 5,408,982 98.7 1.3 43 2019-07 5,557,570 98.7 1.3 44 2019-08 5,630,812 98.4 1.6 45 2019-09 5,476,433 97.7 2.3 46 2019-10 5,333,382 97.9 2.1 47 2019-11 5,264,513 97.7 2.3 48 2019-12 5,306,613 97.7 2.3 49 2020-01 5,693,551 97.0 3.0 URL parameters (short)= ?minyear=2016&maxyear=2020&STATE=&geo=NATION&occ=Computer%20and%20Mathematical%20Occupations%20(1000-1299)&empstat=In%20labor%20force&group=EMPSTAT&sortn=2&sortdir=Ascending&color=Set1&geomtype=Line%20GraphThe graph and table show the unemployment rate for the "Computer and Mathematical Occupations" as defined at ACS OCCUPATION CODES (OCC) - 2000-2017 and OCCUPATION CODES: 2011+. This seems to be the "tech occupations" referred to by the CompTIA article resulting in unemployment rates of 1.4 percent for March 2018 and 1.3 percent for May of 2019, identical to those given in the article. However, the graph and table show that this unemployment rate had risen to 3.0 percent by January 2020, the highest level since November of 2016. I could not find any CompTIA article that reports the 3.0 unemployment rate. In fact, a Feb 7th press release is titled "US Tech Hiring Rebounds in January, CompTIA Analysis Reveals" but makes no mention of the 3.0 percent unemployment rate. However, a Jan 10th press release did mention the 2.3 percent unemployment rate in December. But judging from the original google of "tech unemployment" above, few, if any, major publications reported it.
The Feb 7th CompTIA press release does include a footnote that states "[1] The IT occupation data should be viewed as a directional indicator, as there tends to be a higher degree of variance with monthly Bureau of Labor Statistics data at the occupation level." This variance can be better seen in the following graph which shows to tech unemployment rate since 2003:
As can be seen, tech unemployment did drop following the tech crash in 2001 and then grew sharply during the financial crisis of 2008. Since then, it has generally dropped though there appears to be something of a recent uptick. Hence, it does appear that the unemployment data needs to be viewed over several months at least to give perspective. Instead, all of the major publications seemed to focus on the low figure of a single month and then ignore the numbers as they increased in the following months.
In fact, there is an additional limitation in looking at the unemployment rate of a specific occupation versus looking at the overall unemployment rate. If a recent tech graduate is unable to find a job in their field, they may continue to look exclusively in their field for a time but they will likely accept a job outside their field within a few months, even if they keep looking for a job within their field. Likewise, and older worker who is laid off may look for a job in their field exclusively for a time but will likely eventually accept a job outside their field or, if they can afford it, retire. Neither of these groups will show up in the tech unemployment numbers once they find a job outside their field or retire. Hence, the unemployment rates for an occupation give little indication of how many workers may have been driven out of their field.
Another problem is that there is no one definition of unemployment that captures everything that is happening in the employment market. The following graph from John Williams' Shadow Government Statistics shows the official U-3 and the broadest U-6 government measures of unemployment. It also shows John William's "Shadow Stats Alternate" which he says "reflects current unemployment reporting methodology adjusted for SGS-estimated long-term discouraged workers, who were defined out of official existence in 1994." This measure is especially interesting as it suggests that employment never fully recovered following the financial crisis. Of course, this is not an official government measure and is debatable. But even the official government measure suggest that the unemployment rate is not a simple measurement that summarizes the entire employment situation.
