Some commenters on the previous post The Catch-22 STEM Job Market requested a graphical presentation of the data. Note: STEM refers to Science, Technology, Engineering, and Mathematics.
The above bar chart (histogram) shows the results of searching the LinkedIn Job Board for engineer job posts on Saturday, March 9, 2013 at about 8:50AM (PST). The graphic shows the adjusted count of job postings matching combinations of “engineer” with key words and phrases for experience level: “n+ years” where n is a number such as 5, “entry level”, “new college graduate”, and “intern”. The intern jobs are counted in the -1 bin, for less experience than a Bachelor’s Degree. “0+ years”, “entry level”, and “new college graduate” are all counted in the zero bin.
It was found that phrases such as “our company has 10+ years of experience” increasingly contaminate the results above “10+ years” to the point that nearly all job posts that match “20+ years” are spurious. Accordingly, the counts have been adjusted based on the percentage of actual matches for a job experience requirement in the bins above ten (10) years of experience post bachelors degree.
Note that LinkedIn uses some intelligence in matching phrases of the form “n+ years”. For example, “10+ years” will frequently intelligently match job postings with phrases such as “more than ten years of experience” as well as “10+ years” and “10 years”.
The graphic shows that STEM employers appear to mostly seek STEM workers with 2-10 years of paid professional work experience, with a strong peak at five (5) years of experience. Only about 1.9 percent of the job postings are for entry level/no work experience required positions. About 5.5 percent of the job postings are internships, entry-level, or junior level (less than two years of paid work experience). About 90.1 percent of the job posts are for positions with 2-10 years of experience required. Only about 4.2 percent require over ten years of work experience.
The raw data for the graphic in text format and the GNU Octave code used to make the graphic can be found in the appendix at the end of this article. GNU Octave is a high-level interpreted language, primarily intended for numerical computations that is mostly compatible with MATLAB.
The same or a very similar distribution of the number of job postings versus years of work experience required in the job post has been found for a large range of STEM job categories ranging from general categories such as “scientist” to specific categories such as “molecular biology”. It seems remarkably independent of the actual complexity, difficulty, and likely learning curve of the field and subject matter.
The STEM job market appears to be a Catch-22 situation where almost the only way to get a job is to have at least two years, usually 3-5 years, of paid professional work experience. In a number of contexts, employers complain that they cannot find qualified STEM workers and that hundreds of thousands, even millions of job listings such as those surveyed allegedly go unfilled. Yet, this is not surprising. The graphic and source data strongly suggest that STEM employers largely seek experienced STEM workers and rarely hire entry level STEM workers. For example, for employers to be able to hire 10,000 STEM workers with 3-5 years of paid professional work experience, other employers must have hired 10,000 STEM workers with no experience, 3-5 years before. Thus, it is not surprising that employers are unable to find STEM workers with work experience.
A Sampling Issue
Some commenters on previous articles on STEM employment issues have argued that employers primarily fill entry-level STEM jobs through college recruitment and don’t bother to post entry-level positions to job boards such as LinkedIn, Craig’s List, and so on. Hence the data may understate the relative proportion of entry level positions. This is difficult to rule out.
Empirically, employers do post hundreds of intern and entry-level positions to LinkedIn (and other job boards which are similar) as indicated in the jobs posting data used for the graphic.
In certain contexts, many STEM employers loudly claim that they face a shortage of STEM workers of crisis proportions. For example, then Microsoft CEO Bill Gates testified in 2008:
I know we all want the U.S. to continue to be the world’s center for
innovation. But our position is at risk. There are many reasons for
this but two stand out. First, U.S. companies face a severe shortfall
of scientists and engineers with expertise to develop the next
generation of breakthroughs.
Emphasis added. STEM shortage claims like this are not unusual. Presumably, companies facing a severe shortage would scour the planet for prospective employees, leaving no stone unturned in the desperate hunt for the rare STEM worker. Why not advertise entry-level positions on LinkedIn and other job boards as well as through college recruiting offices? 🙂
Of course the qualifying phrase “with expertise” could mean that U.S. companies including Microsoft are seeking only employees with several years of paid professional work experience magically acquired somewhere else as the distribution of job postings on LinkedIn and other Internet job boards strongly suggests. It is common to find qualifications such as “with expertise” or “with special skills” in claims of STEM shortages. This is often followed by an abrupt jump to complaints about the poor quality of public eduction in the United States, seemingly implying that the “special skills” in question are basic science and math skills such as geometry, algebra, trigonometry, and calculus taught in high school and college.
What does the STEM job posting data mean?
In the Hollywood blockbuster Armageddon NASA discovers that a giant asteroid is on its way toward Earth and will destroy the planet unless it is diverted using a thermonuclear weapon which for the purposes of the plot must be inserted several miles inside the asteroid using deep drilling technology that NASA lacks. Actually no one in the world knows for sure how to do it.
In Armageddon, the human race faces a genuine talent shortage of crisis proportions. There is nobody with any experience, let alone 3-5 years of paid professional work experience diverting killer asteroids.
What does NASA do? What would you do? Would you spend years trying to inspire twelve-year-olds to pursue a career in the exciting, hot field of asteroid diversion — finally starting their first job several years after the asteroid has destroyed the Earth? Well, no, you would not. Would you spend several years lobbying to import more non-existent skilled asteroid diverters from Third World nations? Well, no, you would not.
Armageddon being an inspirational Hollywood blockbuster and not the real world, NASA does what any sensible person would actually do. They find the best qualified people they can, which isn’t very qualified, train them to be astronauts in record time, and, of course, blow up the asteroid and save the Earth.
Armageddon is fiction and hopefully will remain so. World War II, for example, was not. What did the United States do after Pearl Harbor? Bemoan its lack of people with 3-5 years paid work experience fighting Nazis, 3-5 years paid work experience building the atomic bomb, or the many other challenging tasks accomplished during the war? Did we whine about how all the people who had been unemployed for years due to the Great Depression had lost their skills and could never work again, leaving us fatally vulnerable to to the virile Aryan Nazis with 3-5 years of paid work experience conquering Europe? Of course, not. The government and companies did exactly what NASA does in Armageddon, they hired the best they could find and trained them when needed — to the extent possible. This is what people do (or should do) in a genuine crisis.
Are STEM employers hiring the best available STEM workers and using formal or informal training to make up for deficiencies? The STEM job posting data suggests not. Instead they appear to be holding out for fantasy employees who are always more appealing than real people.
Understandable but not Based on Evidence
It is perfectly understandable that companies would prefer to hire employees with at least a few years of paid professional work experience. It is lower risk than hiring a new college graduate. In Armageddon NASA would be foolish to turn down someone who actually had 3-5 years paid professional work experience diverting asteroids in favor of Bruce Willis’s team of misfits. But, there is no one. Many “purple squirrel” STEM job postings with dozens of highly specific, nit-picky job requirements are probably hunts for non-existent people.
It is understandable, but it is not based on evidence, the people and skills that are actually available. STEM employers often appear to have no idea of the time needed for a STEM worker to switch from one tool to a closely related tool, always assuming at least three years of work experience is required — somewhat like requiring three years of work experience to switch from driving a Volvo to a Subaru automobile. They also often appear to be seeking non-existent or extremely rare super-programmers or super-scientists who should be able to switch from one tool to another in negligible time, but also require that these imagined super STEM workers will also have at least three years paid professional experience using the the new tool.
Like Bill Gates, most of us are in favor of beneficial breakthroughs. Very few people are looking forward to a world that has run out of oil and other hydrocarbon energy sources without finding a replacement energy source or sources. Genuine breakthroughs have been few and far between in the last forty years. The standard of living in advanced nations appears to be dropping as might be expected if the supply of inexpensive oil is dwindling. Presumably the situation in less advanced nations is worse.
We have already experienced over ten years of a singularly unproductive war, costing over one trillion dollars in direct costs, probably fought over control of the world’s oil. These conflicts are likely to increase in number, intensity, and cost if supplies of inexpensive energy continue to dwindle relative to the growing human population. They are also likely to disrupt oil production and make a bad situation even worse, as occurred with the war in Iraq.
To his credit, Bill Gates appears to be trying to foster breakthroughs and solve big problems through his foundation and investments. Most of us agree with the ultimate stated goals, but will the strategy to get there actually succeed? Not so far.
It is difficult to see how a strong preference for STEM workers with only a few years of experience magically trained somewhere else furthers the cause of breakthroughs that may solve the world’s growing problems. Indeed this strong preference, not a new one, may be a major contributing factor to the slowdown in scientific and technological progress in the last forty years — and the associated slowdown in real economic growth per capita compared to the dramatic growth and progress of the nineteenth and early 20th century. When companies and other organizations fail to fill positions with available STEM workers while pursuing fantasy STEM workers, important work does not get done, projects and businesses fail. In the movies, a killer asteroid destroys the Earth.
© 2013 John F. McGowan
About the Author
John F. McGowan, Ph.D. solves problems using mathematics and mathematical software, including developing video compression and speech recognition technologies. He has extensive experience developing software in C, C++, Visual Basic, Mathematica, MATLAB, and many other programming languages. He is probably best known for his AVI Overview, an Internet FAQ (Frequently Asked Questions) on the Microsoft AVI (Audio Video Interleave) file format. He has worked as a contractor at NASA Ames Research Center involved in the research and development of image and video processing algorithms and technology. He has published articles on the origin and evolution of life, the exploration of Mars (anticipating the discovery of methane on Mars), and cheap access to space. He has a Ph.D. in physics from the University of Illinois at Urbana-Champaign and a B.S. in physics from the California Institute of Technology (Caltech). He can be reached at firstname.lastname@example.org.
Suggested Reading and Resources
IT Talent Shortage or Purple Squirrel Hunt?
March 7, 2013
The Ph.D Bust: America’s Awful Market for Young Scientists—in 7 Charts
The Atlantic Monthly
Feb 20, 2013
U.S. pushes for more scientists, but the jobs aren’t there
July 7, 2012
If there is a skills gap, blame it on the employer.
by Peter Cappelli
New York Times
August 3, 2012
Why Good People Can’t Get Jobs
By Peter Cappelli
Why Companies Aren’t Getting the Employees They Need
Wall Street Journal (Online)
October 24, 2011
LinkedIn March 9, 2013 8:50 AM (PST) Anywhere search Entry Level and Junior Level engineer "entry level" 356 engineer "no experience" 13 engineer "entry level" "no experience" 1 engineer intern 447 engineer "new college graduate" 30 engineer "new college graduate" "entry level" 4 engineer "new college grad" 7 Experience Level engineer "0+ years" 17 engineer "1+ years" 321 engineer "2+ years" 2509 engineer "3+ years" 3488 engineer "4+ years" 1629 engineer "5+ years" 5863 engineer "6+ years" 833 engineer "7+ years" 1246 engineer "8+ years" 1216 engineer "9+ years" 105 engineer "10+ years" 1958 (9/10) referred to job experience, (1/10) referred to number of years the company has been in business engineer "11+ years" 25 engineer "12+ years" 270 engineer "13+ years" 31 engineer "14+ years" 445 engineer "15+ years" 530 (5/10) referred to job experience, (5/10) referred to years in business of company/employer engineer "16+ years" 27 engineer "17+ years" 17 engineer "18+ years" 25 engineer "19+ years" 5 engineer "20+ years" 245 (2/10 employers checked this was an experience requirment. 8/10 referred to the number of years that the employer has been in business) engineer "25+ years" 356 (10/10 different employers checked, "25+ years" did not refer to experience but was boilerplate such as "our company has 25 years experience in xxx") engineer "30+ years" 331 (10/10) refer to years that the employer has been in business, not a job experience requirement Ranges of Experience engineer "0-2 years" 69 engineer "0-1 years" 3 engineer "1-2 years" 159 engineer "1-3 years" 135 engineer "1-4 years" 10 engineer "1-5 years" 39 engineer "2-3 years" 231 engineer "2-4 years" 153 engineer "2-5 years" 207 engineer "2-6 years" 11 engineer "2-7 years" 9 engineer "3-4 years" 74 engineer "3-5 years" 659 engineer "3-6 years" 27 engineer "3-7 years" 30 engineer "4-5 years" 56 engineer "5-6 years" 20 engineer "5-7 years" 266 engineer "5-8 years" 116 engineer "5-9 years" 13 engineer "5-10 years" 250 The Training Gap engineer "will train" 20 engineer "training provided" 8 engineer apprentice 9 engineer apprenticeship 31 Adjusted Numbers (for infographic) -1 447 add interns 0 399 entry level/new college graduate/0+ posts 1 321 2 2509 3 3488 4 1629 5 5863 6 833 7 1246 8 1216 9 105 10 1958 (9/10) referred to job experience, (1/10) referred to number of years the company has been in business 11 25 12 270 13 31 14 222 cut in half due to years in business problem 15 265 (5/10) referred to job experience, (5/10) referred to years in business of company/employer 16 14 17 8 18 12 19 1 20 50 (2/10 employers checked this was an experience requirment. 8/10 referred to the number of years that the employer has been in business) 25 0 (10/10 different employers checked, "25 did not refer to experience but was boilerplate such as "our company has 25 years experience in xxx") 30 0 (10/10) refer to years that the employer has been in business, not a job experience requirement
Octave Code for Generating STEM Experience Graphic and Numbers
stem_data = dlmread('engineer_infographic_data.txt'); bar(stem_data(:,1), stem_data(:,2)); title('STEM POST BACHELORS WORK EXPERIENCE REQUIREMENTS'); xlabel('YEARS OF WORK EXPERIENCE POST BACHELORS'); ylabel('ADJUST JOB POST COUNT'); years = stem_data(:,1); job_counts = stem_data(:,2); interns = stem_data(1, 2); entry = stem_data(2 , 2); junior = stem_data(3, 2); tenorless = sum(stem_data(4:4+8, 2)); senior = sum(stem_data(13:end, 2)); all = sum(stem_data(:,2)); pct_interns = interns / all pct_entry = entry / all pct_jr = junior / all pct_newbie = pct_interns + pct_entry + pct_jr pct_2to10 = tenorless / all pct_sr = senior / all
-1 447 add interns 0 399 entry level/new college graduate/0+ posts 1 321 2 2509 3 3488 4 1629 5 5863 6 833 7 1246 8 1216 9 105 10 1958 (9/10) referred to job experience, (1/10) referred to number of years the company has been in business 11 25 12 270 13 31 14 222 cut in half due to years in business problem 15 265 (5/10) referred to job experience, (5/10) referred to years in business of company/employer 16 14 17 8 18 12 19 1 20 50 (2/10 employers checked this was an experience requirment. 8/10 referred to the number of years that the employer has been in business) 25 0 (10/10 different employers checked, "25 did not refer to experience but was boilerplate such as "our company has 25 years experience in xxx") 30 0 (10/10) refer to years that the employer has been in business, not a job experience requirement