The fact that the vast majority of them are NOT trying to be Americans is an important one.
...

And, yes, it is a political matter. Even legal immigrants do not have the same foundations of knowledge of the USA's history and concepts of what constitutes individual rights and liberty, and many politicians have been maneuvering for over a century to distort what is taught in government schools to distort or obliterate that history and those foundations in order to promote their anti-liberty, anti-individual rights agendas.

The business executives and academia executives want more cheap, pliant, labor with flexible ethics, while the immigration lawyers and the lobbyists for all of the aforementioned want more money under their control, while the politicians want more votes for their side, and more money and power for themselves and for their families and friends, with little regard to the USA, the US Constitution, or the individual rights on which they are based.

Only a tiny fraction of H-1B grantees are sponsored for green cards....

Only a fraction of green card holders become US citizens.

Foreign students on F visas with OPT and on J and M visas are abused to displace US citizen workers. Guest-workers on H and L visas are abused to displace US citizen workers. Green card holders are abused to displace US citizen workers.

Immigration reform is dead because politicians like Orrin Hatch, Luis Gutierrez, Nasty Pelosi, Harry Reid, Schmuckie Schumer, John McCain, and Lindsey Grahamnesty are so viciously against reform, and much prefer reprehensible immigration law perversion, instead.

STEM is kind of broad, and even includes accountants (the M part, mathematics.) It doesn't just include engineers (that's just the E part.) Aaafrgh.evilmalware!!!!!! Accountantsarenotpartofthemathematicscategory.

AnubisIV wrote:
"Also of note, looking back at the numbers again then, they seem a bit odd if we accept your number. There were roughly 13.1M listed as unemployed last year, and about 10.8M listed as unemployed this year. A reduction by 5M for the reason you cited would suggest that despite losing 5M at the end of the year, they had picked up an additional 2.7M over the course of the year. I don't know what that means or how it matters, but I thought it was interesting."

Curunir_wolf wrote:
"Those numbers are based on unemployment rolls. Congress ended the EUC (Emergency Unemployment Compensation) on December 31, so that took about 5 million people off the list."

Which numbers are based on unemployment insurance claims? Those are different (they do have an "insured unemployment rate" they publish weekly; "The advance unadjusted insured unemployment rate was 2.6% during the week ending February 15, unchanged from the prior week.". They don't break them down by occupation and industry. That's totally separate from the unemployment rate (U3, U4, U5, U6), durations of unemployment, employment/population ratios, labor force participation rates, etc.). It doesn't matter whether or not you're collecting unemployment insurance benefits. If you're unemployed and actively seeking work you're part of these figures. "The Bureau of Labor Statistics does not provide information about unemployment insurance (UI)" See the FAQ and the methods book linked at the top of the FAQ page:
http://www.bls.gov/cps/faq.htm

I've been using numbers from the monthly BLS Household Survey, a statistical survey of about 60K households, and from an un-published quarterly report by detailed occupations based on the same raw data, which they send to media and on individual request to others (just click the data request link at the bottom of most BLS pages)... with lots of caveats about small sample sizes for many detailed occupations making the figures subject to high error probabilities. That caveat is the reason why I give ranges, and I'm basing those ranges on looking at the last 8-12 quarters and the annual data within that time. (BTW, I haven't yet gotten the annual report for 2013, which should have more reliable figures than the quarterlies. Sometimes they don't send that until later, with the new year's 1st quarter report.)

There are CES figures from the monthly Establishment Survey for industry groupings, like the Information Industry (2.135M employed in January), and Software Publishers within that (i.e. for REAL jobs developing software products, of which only 234,300 were reported being employed in December; they're always a month behind). But they throw together all "production workers" which lumps in a lot of other non-management people who are not actually designing and developing software.

Anyway, since the last 2 of the detailed occupation reports showed total aggregate unemployed as 9,263,000 for 2013Q4, and 10,049,000 for 2013Q3, you must be talking about other aggregates from either the Household or the Establishment surveys. So, from the Household survey:
Civilian, non-institutionalized population 16 and older (CivPop16+=CNIP16+*): LNU00000000: 246.915M.
Civilian labor force (CLF16+ =employed+UEASW): LNU01000000: 154.381M
Employed: LNU02000000: 143.526M
Not in the labor force (NILF): LNU05000000: 92.534M
Unemployed and actively seeking work (UEASW): LNU03000000: 10.855M
total not employed: 103.389M
employed/CNIP16+ men: LNU02300001: 63.5%
emp/Emp/CNIP16+ women: LNU02300002: 53.2%
emp/Emp/CNIP16+ total: LNU02300000: 58.1%
emp/Emp/CNIP16+ black: LNU02300006: 52.7%
None of these are "seasonally adjusted", and that's the way I prefer them. (* In one spread-sheet I got started using one abbreviation, and years later when I created another for a different purpose I used another abbreviation; they designate the same things.)

"U6 is specifically the one that counts underemployment!"
...

U6 only counts some kinds of under-employment. Let's see... http://data.bls.gov/cgi-bin/sr... "Series Id: LNU03327709
Not Seasonally Adjusted
Series title: (Unadj) Special Unemployment Rate U-6
Labor force status: Aggregated totals unemployed
Type of data: Percent or rate
Age: 16 years and over
Percent/rates: Unemployed and mrg attached and pt for econ reas as percent of labor force plus marg attached
2014 13.5"

It counts people who want to be employed full-time, but are employed part-time. It counts those "marginally attached" to the labor force. But it doesn't count mechanical engineers or software engineers or biophysicists... who want to be employed full-time, long-term as mechanical engineers or software engineers or biophysicists...
but who are employed full-time, temporarily, as ditch-diggers, as being under-employed. As far as BLS is concerned that's a fully-employed ditch-digger... for now.

"Series Id: LNU03327707
Not Seasonally Adjusted
Series title: (Unadj) Special Unemployment Rate U-4
Labor force status: Aggregated totals unemployed
Type of data: Percent or rate
Age: 16 years and over
Percent/rates: Unemployed and discouraged workers as a percent of the labor force and discouraged workers
Year Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Annual
2014 7.5
Series Id: LNU03327708
Not Seasonally Adjusted
Series title: (Unadj) Special Unemployment Rate U-5
Labor force status: Aggregated totals unemployed
Type of data: Percent or rate
Age: 16 years and over
Percent/rates: Unemployed and marginally attached workers as a percent of the labor force and marg attached
2014 8.6"

"I've been trying to fill a position for six months now, but no qualified person will work for what I'm permitted to offer them."
...

Don't take this personally; I'm just using what you wrote, and how you wrote it, to illustrate a couple of the other problems with the dysfunctional US STEM job markets.

Define "qualified". It's a weasel-word, intended to be as hard to nail to the wall as jello. "Yes, you scored in the top percentile on the SAT and GRE, and have years of experience. Sure, your previous employers often said they loved your work. Sure, you were granted several security clearances in the past. But you only know version 6.8.3, not the 6.8.5 we 'neeeeed'. You're not qualified. We'll have to hire this chap from the 3rd world who just finished his US-government-subsidized 6.8.5 cram-course, instead."

Make that: "The only people able to do the work and willing to work for this firm in this town, aren't willing to do so for less than what they need to pay for a car, home, utilities, taxes... here, especially if you're in Silli Valley, DC, or NYC. Hmmm, maybe we should move to a higher quality of living location."

"This is a natural outgrowth of the old HR saying about attracting "the best and the brightest" but only paying "market" salaries, i.e., 80th percentile talent for 50th percentile pay."

Or, in many cases now, 60th- or 98th-percentile help for 30th- or 40th-percentile total compensation, and then after a few months, we'll dump that batch of cheap labor and bring in more.

"I would prefer to have people who actually are on track to stay here and build a life."
...

I would prefer to see more reasonable numbers of visas, and proper background investigations of each one.

If 20K student visas were issued each year to the very best instead of 300K-516K, and if 1K H-1B visas and 1K L-1 visas were issued each year, and they were all reliably and rationally tested and certified brilliant, exceptionaly knowledgeable, and experienced, I'd prefer that they "stay here and build a life".

But 1M to 1.4M more green cards per year plus 1M to 1.3M more illegal aliens flooding in, 300K-400K J exchange visas (some with guest-work privileges), 300K-425K total H guest-workers, 124K-156K L guest-workers (and their families), 5K-13.2K NAFTA guest-workers, 10K-20K supposedly Einsteins on O visas (including, bizarrely, the occasional, ahem, "outstanding" Dorismar or Bechar-bunny), 3K-4K Australian E-3 guest-workers, 9K-11K vo-tech students on M visas... with no background investigations, some not even so much as interviewed (no, a "check", i.e. a data-base look-up of some of the known and identified worst criminals, does not suffice)... is totally insane.

"immigrants become citizens"
...

Yes, many people who get green cards, i.e. immigrants, become citizens.

Ron Hira at RIT has repeatedly looked at how many H-1B guest-workers are sponsored for green cards and reported that it's a very small fraction. One year, as I recall, it was below 1%.

But it varies widely by employer. A very few employers sponsor a super-majority of their L and H-1B visa-holders for green cards, but I don't think any of the "big guys" -- the firms which apply for over 1K H-1B visas each year -- do so.

"Interestingly, we can see that pretty much all "professional occupations" have a low unemployement rate; only "arts and entertainment" being fairly high."
...

That's the wrong lesson. What the stats show is that each kind of occupation has a characteristic "full employment" unemployment rate, and that the current unemployment rates for STEM occupations are worse (not just in the last 6 years, but over the last several decades), compared with their full employment levels, than the aggregate of all occupations now compared with the aggregate "full employment" unemployment rates (over the last several decades), and actors are doing slightly better, even though their raw unemployment rate is considerably higher.

Unemployment rates, and the flip-side -- employment/population ratios -- have to be examined in context, not as absolute numbers, not even as rigid proportions of unemployed to a specific sub-set "labor force", but as proportions to the reasonable optimal level... and in light of the current legal and regulatory scheme as compared with those of other times.

What's most interesting about the last several recessions is the lack of full recovery. Durations of unemployment (average and median) have been growing longer. Employment/population ratios have been falling (well, for everyone except female-type people, and though their e/p ratios have increased over the last 60 years or so, some who "work at home" would prefer to be able to land a real job while others working outside their homes would prefer to be able for their families to afford for them to "work at home".

What makes "full employment" level or target difficult to discuss is that most people's visceral reaction is that everyone who wants to be should be employed all the time. But transition times from leaving one job and landing another are non-zero. Re-tooling, brokering an employment deal, etc., all take time, and in the interim you're unemployed, regardless of how bright, creative, knowledgeable, industrious... you are. So, there's an inherent "friction"... and then there are disasters, general economic disasters and personal disasters (e.g. injuries). So, no one expects the unemployment rate to reach zero... well except for judges and CEOs considering their own personal situations, perhaps... followed by dentists, veterinarians... The vast majority of STEM professionals are a ways down the list, and actors are in the sub-basement (they tend to cheer up when their unemployment rate is ONLY 20% or so).

Since the 1950s, for instance, the federal government has several times lowered their sights on what constitutes "full employment", i.e. increased the "full employment" unemployment rates. "The United States is, as a statutory matter, committed to full employment (defined as 3% unemployment for persons 20 and older, 4% for persons aged 16 and over)..." https://en.wikipedia.org/wiki/...

"In the Kennedy administration, 4% unemployment was set as an 'interim' unemployment target because they did not want to defend even this [too high] level of unemployment..." --- Lester C. Thurow 1980 _The Zero-Sum Society_ pg 73

see also: Stanley Lebergott "Annual Estimates of UnEmployment in the United States, 1900-1954" _The Measurement and Behavior of Unemployment_ pg 231

The big/significant economic effects are always at the edges where the changes take place. What does changing supply by 1, without changing demand do?
...

What does increasing the labor supply by 136K (number of H-1B visas issued in fiscal year 2012, according to the State Dept. annual report) do?
What does increasing the supply by 2.3M (the numbers of new/initial H-1B visas issued since 1990) do?
What does adding 400K to 600K H-1B guest-workers present in the USA at any particular time in recent years do? (No, the federal government does not have a good handle on the exact numbers. They experimented a few years ago with trying to track people as they entered and as they left, in order to move toward estimating the numbers present in the USA, and gave up after only 2 months.)

Does it make it easier or more difficult for a US citizen in an affected field to get an interview? (more difficult) ... to get a job? (more difficult) ... to negotiate a raise? (more difficult) ... to make long-term economic calculations for decision-making (whether to invest, whether to take out a loan, whether to marry or have children, whether to move for a different job)? (more risky and difficult)
Does it make an employer more or less likely to buy job ads or retain a head-hunter? (less)
Does it make an employer more or less likely to hire or put on retainer an immigration lawyer? (more)
Does it make an employer more or less likely to fly a US citizen candidate across the country for an interview? (less)
Does it make an employer more or less likely to provide relocation assistance to that US citizen employee? (less)
Does it make an employer more or less likely to retain an employee? (less)
Does it make an employer more or less likely to invest in 2-14 weeks of new-hire training that was common before H-1B? (less)

"Except that it is not. There are currently about two million practicing engineers in the USA, and that number is growing by about 70,000 per year."
...

While US citizens have recently been earning between 81K and 97K engineering degrees per year, 41K to 66K "computer and information sciences" degrees per year, 240K to 327K STEM degrees each year, according to the US Dept. of Education, National Center for Education Statistics.

"So we are not "shedding" STEM jobs. The unemployment rate for computer professionals and engineers is about 3% [bls.gov] compared to an overall rate of over 7%."

Let's see. BLS, unemployment rates by detailed occupations, math and computer science, 2013Q4: 3.5%, down from 3.9% in 2013Q3, up from the historical full employment level for these occupations of roughly between 1.1% and 2%, i.e. roughly 1.75 and 3.5 times worse than what an economist would call full employment.

Architecture & Engineering, 2013Q4, 3.4%; 2013Q3, 3%; 2013Q2 & Q1, 3.8%... compared with full employment level running roughly between 1.3% and 2.2%, about 1.4 to 2.9 times worse than full employment.

Sciences in recent quarters had unemployment rates between 3.5% and 4%, compared with full employment levels of roughly between 1.4% and 2.5%, roughly 1.4 to 2.9 times worse than full employment.

The unemployment rate for actors has recently been running between 20.9% and 40.1% as compared with full employment levels of roughly between 10.7% and 16%, roughly 1.3 to 3.4 worse than full employment.

All occupations, recently 6%-7.4%, compared with full employment levels of... 3.5% - 5%, roughly 1.2 to 2.1 times worse than full employment.

Another problem is that BLS does not look at the talent pool for each occupation, only at those currently employed in a particular occupation category divided by the total of those employed in that category plus those whose most recent employment was in that category who are actively seeking work.

A former software architect employed as a pet-sitter is categorized by them as a fully-employed pet-sitter, not as an under-employed software architect. Ditto a former biophysicist teaching the occasional section at the junior college or teaching at a high school.

If the number employed were divided by the number in that talent pool (analogous to what they do with the more general employment/population ratios they do publish) the waste of US citizen STEM talent would be clearer.

"So no, the scientific method works, and has worked for centuries, and will continue to work as long as scientists are rewarded based on [their adherence to] the scientific method."
...

Yes, the more closely we can adhere to that, the better. That's never happened perfectly. After all, as I understand it, scientific journals started as letters among over-lapping circles of acquaintances until someone with a bit of ambition assembled, edited and sent the collections of letters back out to all of the participants. And it was common for friendships and animosities to pre-exist or to develop among these overlapping circles. When there were good editor/publishers, working with well-considered, well-written letters, it worked well.

Unfortunately, despite (because of) "peer review" as practiced today, the scientific method is often abandoned for the sake of politics... as most often is the case with the "warmist hysterics" vs. the "deniers". Factions circle around particular publications, blocking papers from other factions, and making snide remarks and otherwise propagandizing about the others, often disregarding the merits and essential faults of each.

OT1H, no one should be forced to publish sentiments with which he disagrees, or to associate with those with whom he disagrees. OTOH, sustaining the debate as openly and honestly as possibly is the thing.

IMO, it should all get published, with the names of the authors, the (unfudged, un"trick"ed, un-homogenized) data. Dispense with the propagandizing and restrictions against assertions and counter-arguments and counter-counter-arguments from reaching the public light of day.

Dispense with the government subsidies for this political faction and not that, dispense with the "scientists" meetings with editorials and media moguls to plan the propaganda strategies, or at least attempt to get knowledge of all such meetings out to the public as quickly as possible and reported in as much depth as possible.

Let everyone see where data has been jiggered and decide whether those processes are valid or not. Let everyone see the back-slapping and back-stabbing cliques clearly.

Let each person examine and judge each issue to the extent of his ability within his economic means...

I don't see anything wrong with "self-plagiarism". I mean, if X wrote it, X wrote it. It doesn't matter whether X wrote it 1 time or 100 times; it's still X's work. OTOH, I can see how a publisher of one paper might object to material in that paper being re-used in another, because the first publisher won't be able to fully milk it. As a writer not compensated by publishers/producers for some work whose value went to others on a number of occasions that's not tugging at my heart-strings just now.

"No one's saying that the U.S. shouldn't invest more in rural education."
...

I hereby am saying the USA should not expect spending more on education -- whether rural, suburban, city or slum -- to necessarily improve academic results. We already spend horrendous sums in some of the under-performing neighborhoods.

What works is when the locals value academic achievement, when these individuals and family heads see some pay-back coming to "their people" whom they know well. When the school admins, teachers, students and parents place a high priority on academic achievement, when they see that it is possible and that it pays -- both personally and generally -- higher academic achievement results.

In big parts of the USA, UK, and Europe that link between effort, academic achievement and proportional rewards, meritocracy on that basis, has been broken and people won't invest more effort when they see other accessible options.

"And this is bad exactly why?"
...

OK, let's do it this way, then; let's only test Americans in the poorest neighborhoods in which academic achievement is least valued, and test only the children in the wealthiest neighborhoods where academic achievement is most highly valued in China and see how the results come out.

Or we could test all of their students and all of our students and compare, examining the average, median, worst, best, standard deviations.

Or only test the top students in the top US schools and compare them with the top students in their top schools for a change.

Can you see the differences such selective testing produces from universal testing?

3 alternate links. Take your pick:

http://jewishworldreview.com/cols/sowell120313.php3

http://townhall.com/columnists/thomassowell/2013/12/03/a-challenge-to-our-beliefs-n1756023/page/full

http://news.investors.com/ibd-editorials-on-the-right/120213-681322-blacks-working-class-whites-were-similar-before.htm?p=full