Friday, July 24, 2015

Job Openings Data Misrepresented

Job Openings Data Misrepresented

Think tank personnel cannot read a simple chart, desperate to fool public?

Subjects: shortage of STEM graduates, Adams Nagar, H-1B train your replacement, Job Openings and Labor Turnover Survey, Information Technology and Innovation Foundation ITIF

Adams Nagar, an Economic Research Analyst at the Information Technology and Innovation Foundation (a lobbyist think tank), has completely misrepresented the JOLTS (Job Openings and Labor Turnover Survey) data. On page 2 of "Looking Past the Loopholes: Are H-1Bs an 'Offshoring' Visa?", Nagar claims that there are 5.4 million job openings, but neglects to include that 5 million of those positions were filled in the same period. (This happens each and every month.)

Demand for technology workers has never been higher. Indeed, many of the record 5.4 million positions currently open in the United States are tied to the shortage of STEM workers.

Companies facing costly [H-1B] hiring difficulties face a tough choice: Remain uncompetitive with under-qualified U.S. workers, pay to poach workers from other companies and industries also facing shortages, fill the positions with foreign-born and often U.S.-educated professionals, or leave the position unfilled and forfeit an opportunity for growth.
Unable to interpret the simple Wall St. Journal chart that he linked to, perhaps Nagar is a prime example of an under-qualified U.S. worker that he references.  He holds an M.A. in Political Economy and Public Policy and a B.A. in Economics and Political Economy from Wash. U.

 Job Openings and Labor Turnover Survey (JOLTS)

The JOLTS survey measures actively recruited positions, as Job Openings. It also measures, Hires and Separations, categorized as quits, layoffs with no intent to rehire, discharge of the position or personnel etc. Basically, JOLTS simply measures "churn" or natural turnover in the labor market.

The JOLTS data begins in December 2000, so I decided to take a look at, and chart, the long term averages.   I found it interesting that for all of the years included in the data-set, even during recessions, Hires exceeded Job Openings, with the except being 2015. (This data to June 2015 and June is preliminary.)

The blue line in the chart above is the yearly average of Job Openings which finally returned to year 2000 levels.  The red line, Hires, is currently averaging 369,600 below year 2000 levels.  The yellow line (calculated by subtracting Opening from Hires (implied employment change)), is the amount the Hires have traditionally exceeded Job Openings, this year appears to be an outlier.

Although I hate to use Unemployment data because it does not include discouraged workers,  I used it in the following chart because I wanted to compare those actively seeking employment to the active recruiting format the JOLTS provides in Job Openings.

The above chart adds the JOLTS averages with Unemployment data (U-3) by educational level, this data is only available for ages 25 years and older.

The brown line represents headline unemployment levels (number of persons) with less than a bachelors degree. From 2007 to 2010, this group's unemployment levels grew by 4.95 million. Again, this population is likely to contain discouraged workers, which are not included in the U-3 displayed, but the steeply diving trend in this segment of the unemployed since 2010, makes it appear that the majority of hiring has occurred within the population with less than a bachelors degree.

The green line, represents those unemployed, 25 and older, with a bachelors degree and higher.  Again, because this is U-3 Unemployment data, this represents people actively seeking employment.  This data appears to be quite telling. For the years 2000 through 2008, we see that unemployment levels appear to be closely associated with the jobs which were not actively recruited (the yellow line, Hires vs. Openings). From 2008 forward, a gap has opened up between the non-actively recruited Hires and the formally educated and unemployed population.

Divergence: unemployed college grads and non-actively recruited Hires.
2007 = 202,333
2008 = 245,083
2009 = 711,167
2010 = 956,833
2011 = 1,064,000
2012 = 1,249,000
2013 = 1,227,167
2014 = 1,287,250
2015 = 1,535,433

Note that Unemployment levels for college graduates and above was 559,000 in December 2000 and is 1,307,000 for June 2015. 

Before analyzing what is going on with the college graduates, I refer back to the chart, to point out that prior to 2009, the unemployment levels for those with less than a bachelors degree (brown line) also appears to have had a correlation with the non-actively recruited Hires.

Unemployment: less than a bachelors degree

When even the New York Times reports on it, it is no secret that the majority of recovery jobs have been created in low paying service industries.  Automation is also constantly increasing, so the level of sophistication for some jobs is on the decline.  Additionally, many college graduates are reporting that they are taking wrote service level jobs, bartender, waitress, Uber driver etc.

Unemployment: bachelors degree and above

With the Unemployment levels in this population still 2.4 times higher than they were in December of 2000, there is either, no suitable jobs for them, an unwillingness to hire these workers, or many are not willing to settle for employment that does not fall within their discipline/meet their salary requirement.

An open search for advertised positions on, the premier job website for technical employment yields 86,858 job listings, this is basically identical to pre-recession levels.

Adams Nagar, economics degree and all, must be living in some kind of parallel universe if he thinks that JOLT numbers, which have barely eclipsed December 2000 levels, have anything to do with demand for STEM workers and by extension H-1B visas.

There are already plenty of visas for "domestic" employers, but not if they are used for the "train your replacement" loophole, where those employee's only sin was to be more expensive than a foreign student who is attempting to jump the citizenship line through non-immigrant visa programs. 

"Domestic" employers can't seem to come to terms with the fact that offshore outsourcing firms are the top users of and applicants for the H-1B visa. They are gaming the system, and in actual terms are foreign firms, or whose specialty is arbitraging jobs to foreign nations, benefiting from a program that was designed to help American employers.

Monday, July 20, 2015

Microsoft Uses Women's Soccer Team to Regurgitate Faulty Data

Microsoft Uses Women's Soccer Team to Regurgitate Faulty Data

Microsoft's use of the US Women's Soccer Team and faulty data to try to increase immigration falls to a new low.

Subjects: shortage of Computer Science (CS) graduates, Women's Soccer, H-1B train your replacement, shortage debunked, Computer Science degrees

We gained some traction on "train your replacement" H-1B abuse.  Many thanks to the journalists who covered the stories.

Foremost, a big thank you to ComputerWorld's,  Patrick Thibodeau for his tireless work on covering the H-1B and foreign labor arbitrage topic. "About 500 IT jobs are cut at utility through layoffs and voluntary departures", February 4, 2015.

The New York Times, Pulitzer Prize winner, Julia Preston, who wrote, "Pink Slips at Disney. But First, Training Foreign Replacements", June 3, 2015, and was surprised enough about the comments on the article that she did a follow up article.

The Los Angeles Times
, Pulitzer Prize winner,   Michael Hiltzik, who covered the Southern California Edison labor arbitrage in, "A loophole in immigration law is costing thousands of American jobs" February 20, 2015. (Sorry, no links, The LA Times is currently behind paywall -- I guess I could delete their cookie.).

The Immigration shills waited a while and have now started regurgitating the faulty, data based, immigration attorney funded, H-1B (temporary foreign labor) propaganda.   The use of women's issues and the US National Soccer Team is a new angle though. The article is pretty well identified as propaganda and discussed on Slashdot.
Ten days ago, the United States women’s national soccer team became the most successful team in the history of the FIFA Women’s World Cup by winning the top prize for the third time. However, despite the fact that soccer is not as popular in America as in the rest of the world, it is not surprising that the U.S. women have been dominant in the sport in recent years. The explanation for that success lies in the talent pipeline. With over 1.5 million girls registered in youth leagues, close to 50 percent of all youth players in the U.S. are female and soccer is one of the top three most popular sports among girls ages 5-19.

We only get into the third paragraph before the author whips out the cherry picked graduation data and uses BLS projections rather than actual data.

Unfortunately, the strength in the talent pipeline that we see in female soccer today is not the reality for technology. The U.S. is facing a shortage of Computer Science (CS) graduates. According to the Bureau of Labor Statistics, every year there are close to 140,000 jobs requiring a CS degree, but only 40,000 U.S. college graduates major in CS, which means that 100,000 positions go unfilled by domestic talent. Even more dramatic is that women in U.S. colleges and universities earn only 18 percent of CS degrees."

First, Computer Science undergraduate degrees are not the only degrees that are applicable to Computer occupations.  Additionally, in 2011, the Census Bureau reports that, 1,013,300 (28%) of those employed in Computer occupations held less than a bachelor's degree.

Citizens and Permanent Residents

Computer related degrees conferred 2013TotalFemale-/- Percent Female
Master's Degrees12,7253,30325.96%
Bachelor's Degrees49,2048,72617.73%
Associate's Degrees38,6088,19821.23%

Source: NSF WebCaspar, 11 Computer and Information Sciences and Support Services.

The Bureau of Labor Statistics projections is also rather suspect at 140,000 per year, in fact, since 2009, employment growth in computer occupations is less than 10% of projected, at 13,922.

Source: Bureau of Labor Statistics: OES

By the way, the H-1B initial employment approvals for Computer-related occupations was 80,877 for 2014 alone.  One year of H-1B intake exceeded five years of recession recovery growth of 69,610.

The latest five years of C.S. and C.I.S. graduation data (2009-13), for citizens and permanent residents only (Master's, Bachelor's and Associate) is 444,912 with 92,716 (20.84%) conferred to women. 

After this investment in time and resources, do our qualified women even get a tryout, when temporary foreign workers have a clear hiring preference?

Incidentally, are there any women with temporary visas on the U.S. Women's Soccer Team?

Monday, July 13, 2015

Misleading Employment Situation Report

Here's the quote from the first line of Employment Situation Report for June 2015.  "Total nonfarm payroll employment increased by 223,000 in June, and the unemployment rate declined to 5.3 percent,..."

 These two pieces of data are not even from the same surveys. 

The problem here is that the Unemployment rate is taken from the Household Survey(CPS), not the Payroll report(CES). For the month of June, total employment(CPS) declined by 56,000, yet unemployment edges down by 0.2%. (More than 100% in the decline in the Unemployment rate for June is the result of people leaving the labor force.)

Employment for bachelors degreed and above (25 and older), employment increased by 30,000, curiously that unemployment rate also declined by 0.2%. 

Also, when we read the monthly Employment Situation Report, employment by Industry is taken from the Payroll report (CES) and the April and May reports were revised downward by 60,000. So this seemingly straight forward survey was off by 12% in the previous reports, and again, has nothing to do with the unemployment rate.

If you ever want to stump an idiot who tells you how great the economy is doing and points to the Unemployment rate as proof, ask them, "How many jobs were created that month, and what that number has to do with the Unemployment rate?"

Thursday, May 21, 2015

Debunked: Shortage in Math and Computer Science

Debunked: Shortage in Math and Computer Science

Shortage of Computer Science and Mathematics Debunked -- Professionals crowded out by H-1B

Subjects: H-1B visa, STEM shortage myth, Bachelors Degree, STEM shortage debunked, Math and Computer Science degrees

I am surprised that I had not thought of approaching the STEM shortage myth in this manner before.  Math and Computer Science constitutes 50% of all STEM occupations and roughly 60% of all new H-1B visas are computer related. The H-1B visa program began in 1992; therefore, we can presume that Math and Computer Science occupations were primarily staffed with citizens and permanent residents in 1992.

NCSES ranked colleges conferred 2,317,505 Math and Computer Science degrees and certificates from 1992 to 2013, to citizens and permanent residents as their highest degree.  Using 1992 employment levels as our baseline, we add the new degrees to the 1992 employment levels to find the size of our native Math and Computer Science workforce.

Employment level: Math and Computer Science (1992) = 1,830,800
*Math-Comp Sci (highest degree conferred (1992-2013)) = 2,317,505
Total Math-Comp Sci Workforce = 4,148,305
*College awards: Citizens and permanent residents only

Next, we look at the employment level for Math and Computer Science occupations in 2013 and subtract the workforce number from above to determine if there is a shortage.

Employment level: Math and Computer Science (2013) = 3,696,180
Total Math-Comp Sci Workforce = 4,148,305
Total Math-Comp Sci Workforce Surplus = 452,125 (12.23%)

So where is this shortage? The IT shortage that employers, trade associations and immigration attorneys write incessantly about?  (The occupations targeted with "shortage" adjective has been expanded to STEM occupations, but that topic is for another paper.)

452,125 is a surplus is 9,042 formally trained professionals for every state in the nation, or a national surplus of 12.23%.  For retirement to be a significant factor, a significant portion of those employed in Math and Computer Science occupations would have had to have been 42 years old in 1992. Highly unlikely since IBM's MS-DOS was not introduced until August of 1981.

We also know from the Census Bureau, that a bachelors degree is not mandatory in computer science, 28% of those employed in computer occupations had less that a bachelors degree in 2011.  Aptitude can be proven with coursework, individual initiative, or both.

High School or less
Computer support specialists = 49,200
Computer occupations, all other = 31,800
Software developers = 22,400
Computer programmers = 21,800
Computer systems analysts = 20,800
Computer and information systems managers = 20,100

Some College or Associate’s Degree 
Computer support specialists = 210,700
Computer occupations, all other = 125,100
Computer and information systems managers = 116,500
Software developers = 114,700
Computer systems analysts = 98,700
Computer programmers = 94,300
Network and computer systems administrators = 87,200

Computer occupations less than bachelors degree (2011) = 1,013,300 (28.04%)

Computer occupations employment 2011 = 3,614,046

(In 2011, Computer occupations accounted for 50% of all STEM occupations.)
Source: Liana Christin Landivar, 2013, “The Relationship Between Science and Engineering Education and Employment in STEM Occupations,” American Community Survey Reports, ACS-23, U.S. Census Bureau, Washington, DC.

Detailed by field of degree, the Census Bureau report that only 49% of those with Computers, mathematics, statistics degrees are working in STEM fields.  Under the twisted logic of trade-association, open-borders propaganda, they would have you believe that this is due to these professionals finding better employment in other fields.  The reality is that 51% of these professionals have just as likely been crowded out of their chosen field by temporary foreign workers.

Census Bureau: Bachelor’s Degree for the First Listed Major: 2011
Computers, mathematics, and statistics = 1,879,764

51% not working in STEM = 958,679

Other data sources:

Occupational statistics data:

IPEDS Completions Survey by Race