Your Ad Here
2013 February 05 Tuesday
Do Highly Skilled Immigrants Make STEM Majors Pointless?

Half Sigma thinks STEM degrees should be avoided due to foreign competition.

Americans are smart to avoid STEM in which they have no comparative advantage over the mass of immigrants and where they may be permanently fired after fifteen years because their skills are considered too old.

I do not buy this argument. Some points:

  • Most of the people who do poorly with a STEM degree studied "S", not "TEM". Science is fascinating to many smart people who try to make a go of it. They mostly end up poorly paid post docs and eventually unemployed. Industrial demand is not high enough to suck up even a majority of the bio and chem Ph.D.s into well paid jobs. This is the market's way of telling you to do something else. The automation of education will further cut into demand for science degrees.
  • Even within STEM's better paid "TEM" the market value of skills varies enormously from one field to the next. Petroleum engineers make way more than mechanical engineers who make more than civil engineers. Software developers used to be paid worse than electrical engineers. But that has reversed. Let the market guide your career choices.
  • People who study computer administration (more "T" than "E") are studying easier and lower paid subjects. The more motivated (e.g. to get lots of Cisco certs and develop lots of skills) and sharper among them can still get over $100k per year in some urban areas.
  • The value of "M" has risen for those who study the right kinds of math. Computer systems are collecting massive data sets. So the value of a statistics degree has risen. Statistics has a big overlap with machine learning and machine learning skills are similarly in rising demand.
  • Moderately intelligent computer programmers who complain about their job prospects are misleading. They suffer from the the Dunning Kruger Effect. They do not know how bad they are. You have to work with a lot of software developers and listen to the best of them talk about much weaker team members to know how extensive this problem is and to what extent the bad developers are oblivious to their poor performance.
  • As Turing Award winning computer scientist Richard Hamming pointed out, what you can accomplish depends on your rate of learning in your career. The best software developers accumulate skills with synergistic value. Those on shallower learning curves with less motivation and talent obscure what is happening with the market for the most talented.
  • If you cut STEM out the list of remaining jobs for smart people is much too short. There are too few slots in the top law schools and law is becoming more automated (and even outsourced) anyway. Medical schools have their slots full as well.

I know far too many affluent techies in the computer industry to think it is a bad place to be. Sure, there are industries where you can rise thru the ranks and make higher salaries as a manager without a STEM degree. A Harvard MBA has value. But how many people can get a prestige MBA? Not many. Plus, for many management positions a STEM degree is a gating factor to even get into the management pool.

The best STEM subjects offer better career prospects than all but a few occupations.

Update: Most of the best paid jobs right out of college are for engineering degrees.

Share |      By Randall Parker at 2013 February 05 09:45 PM 


Comments
Erik Jonsson said at February 6, 2013 2:05 AM:

Randall,

You don't seem to consider the only metric that ultimately matters: the cost of buying enough status to obtain a wife and have kids. This ultimately determines the value of STEM degrees and STEM jobs, not the nominal value of the paychecks they garner.

Based on your blogging, you seem mainly concerned about personal life extension, rather than family and reproduction.

SOBL1 said at February 6, 2013 11:14 AM:

I've seen it time and time again where an American born engineer with moderate social skills becomes the team leader, project manager or face of the engineering or programming team because H1B visa workers have the social skills and extroversion of brooms. Any form of improvisation of self initiative by a native engineer or programmer goes a logn way compared to the asian version of corporate Nazis that are most H1B visa imports.

Dan said at February 6, 2013 12:06 PM:

A STEM degree is a massively good degree to have. Better than ever.

1 - You can earn a decent living in STEM and support a family. This is not trivial for young people now.

2 - In a world of automation, those who can deal with computers and machines are at a huge advantage. Almost all companies will strongly prefer people who are comfortable with tech versus those who 'don't really get computers.'

3 - Someone of unremarkable charisma and creativity can do just fine in STEM# One can't assume such a person could succeed in advertising or sales or whatever else Half Sigma has in mind#

4 - If you work for a STEM company, the average IQ of the company will be much higher than that of society# A rising tide lifts all boats# If the company is full of high IQ types, it can seamlessly enter new markets if others dry up# It will perform well and average people will benefit#

5 - STEM companies drink everyone else's milkshakes# Google is about to drink the milkshake of the friggin' trucking industry with driverless cars, for goodness sakes, having completely consumed the milkshakes of all those cool mad men in advertising#
http://www#youtube#com/watch?v=RKQ3LXHKB34

Better to inside such a company than on the outside.

6 - Even if you will ultimately have a non-STEM job when you are 50, a STEM degree is a great way to get into a company. Young people need hard skills to get in the door. Once you are in the door management, sales, recruiting and all the rest are all available to you.

7 - A STEM degree marks you as smart, or smarter than average anyway. Lots of companies including companies on Wall Street are happy to get STEMs even though the STEMs didn't study finance, because they are smart.

8 - What the hell else are you going to major in? You've got to study something, and is rural sociology an improvement or a step back?

9 - If there is a mix of immigrant Asians and native whites in a tech company, I don't believe that the native whites are at a disadvantage. In fact I think they may at a big advantage in many jobs in the company. They may be better able to move up to management, sales, recruiting or other jobs in the company than immigrant Asians. Whites could actually benefit from a kind of affirmative action in American companies which can't get enough white engineers.

10 - For the time being, tech pay is substantially higher than the national average. That's a bird in the hand. Take it.

aandrews said at February 6, 2013 12:28 PM:

Along with all the "peak jobs" and general decline in innovation and intelligence, there could be another facet to the future that needs to be synthesized: http://www.c-spanvideo.org/program/310316-1


After Words with Jonathan Last
Jan 22, 2013

C-SPAN | BookTV
Author Jonathan Last talks about his book, What to Expect When No Oneís Expecting: Americaís Coming Demographic Disaster, in which he discusses the population implosion in the U.S. and itís impact on the economy, culture and politics. He argued that if the U.S. wants to continue to be a world leader, Americans must have more children. He discussed the countryís demographic changes with Pew Research Centerís DíVera Cohn.

55 minutes | 361 Views


What to Expect When No One's Expecting: America's Coming Demographic Disaster by Jonathan V. Last (Feb 5, 2013)

bbartlog said at February 6, 2013 12:49 PM:

The question you have to ask (if passing up a STEM career/degree) is what your plan B is. Not have a degree and get a job as an assistant manager at Applebee's? Bad move, obviously. Start your own business? OK, that's high stakes, but maybe worthwhile. Get a law degree? Also problematic; that's a very winner-take-most field, and requires a sound inventory of your personal strengths to decide whether it's worthwhile.
Broadly, I'd say that there are two kinds of people for whom STEM isn't a good idea. Those who aren't bright enough and are just going to be marginal (or drop out). And those who have really good interpersonal skills and motivation. For them, tech may not be entirely limiting, but it's not as good as some other career tracks.
For people who are reasonably competent but can't hack STEM, some sort of medical career (RN, LPN, etc.) is probably a decent choice. Not a fun job in general but in most cases decent pay and reliable employment.

Peter said at February 6, 2013 2:08 PM:

"I've seen it time and time again where an American born engineer with moderate social skills becomes the team leader, project manager or face of the engineering or programming team because H1B visa workers have the social skills and extroversion of brooms."

You're talking about middle management. Middle management is often a dead end. Upper mid and higher management positions are far fewer in number relative to overall STEM positions, and the path to upper mid and higher management is often not even through STEM positions, but through MBAs and management, business, finance, marketing, etc. positions.

And Indians increasingly dominate STEM industry management positions. It's basically Jews and increasingly Indians that dominate STEM management.

James Bowery said at February 6, 2013 2:28 PM:

Jonathan Last doesn't say anything about race replacement but he doesn't have to: Obama-type "leadership" will now route government support of fertility to the groups that vote for them. All Jonathan Last has to do is make the case for government intervention in falling birth rates. Politics will do the dirty work in mopping up the white population.

shiva1008 said at February 6, 2013 2:35 PM:

I'm going to do one of those computer administration type master's programs. It's a good segue from my social science undergrad to something more marketable. And like Dan said, once you're in tech, maybe you can use that to jump to something else in tech or combine it with another STEM master's.

Check it Out said at February 6, 2013 3:17 PM:

Hell, I already spent too much time and money getting my degree which says that I have certain skills needed for what I do. Since I'm not a student anymore I think it's time all those years in college start paying off.

Continuing education is yet another form in which the educational corporation is making money off you.

If a kid who's just gotten his B.A. does not have sufficient skills for today's requirements, then he should think about filing a lawsuit against the university, since it did not deliver as it should have for how much it charged.

James Bowery said at February 6, 2013 6:28 PM:

Halfsigma is wrong when he writes:

"Iíve dealt with immigrant-owned businesses, and they are all involved in outsourcing programming jobs to India, or they are body shops providing temporary IT employees who are mostly from their home countries, many of them on H-1B visas. These businesses donít create jobs for Americans, they take jobs away from Americans. But they are good for people who have ownership interests in corporations, the top-level executives and investors."

No they aren't. They are 'good' for those people in the same sense that methamphetamine is 'good' for them. The landscape is increasingly strewn with larger and larger corpses of US companies that bet the company on Asian "brains".

This will keep on happening until the economic rent streams that prop up large corporations dry up substantially and that is guaranteed not to happen so long as there is "continuity of government" at the Federal level.

James Bowery said at February 6, 2013 6:30 PM:

Halfsigma is wrong when he writes:

"Iíve dealt with immigrant-owned businesses, and they are all involved in outsourcing programming jobs to India, or they are body shops providing temporary IT employees who are mostly from their home countries, many of them on H-1B visas. These businesses donít create jobs for Americans, they take jobs away from Americans. But they are good for people who have ownership interests in corporations, the top-level executives and investors."

No they aren't. They are 'good' for those people in the same sense that methamphetamine is 'good' for them. The landscape is increasingly strewn with larger and larger corpses of US companies that became addicted to the short term increases in the bottom line resulting from lowering labor costs with Asian "brains".

This will keep on happening until the economic rent streams that keep large corporations on life support dry up substantially and that is guaranteed not to happen so long as there is "continuity of government" at the Federal level.

Matthew said at February 6, 2013 7:24 PM:

Your assumptions are based on current rates of H-1B immigration of about 65, 000 per year. The new proposal being authored by Sen. Orrin Hatch, who just reelected to his last term has gone full retard, nearly quintuples to 300, 000 the number of H-1Bs corporations will be allowed to import. The devastation will be swift. Is there still a point to getting a STEM degree? Not after Hatch's new law kicks in.

Randall Parker said at February 6, 2013 8:44 PM:

bbartlog,

Yes, what's plan B if not STEM? Law school applications are dropping as the demand for lawyers drops. Med school enrollments haven't gone up much for years. Medical schools are expensive to operate. You can go for a CPA. But lots of financial services are getting automated.

Really, what's the alternative?

Matthew,

What's your alternative?

Randall Parker said at February 6, 2013 8:51 PM:

Erik,

You can work as an engineer in areas where houses are much cheaper. Cities out in the plains states have engineering firms for resource extraction for example. If you live in a small city in Kansas or Nebraska or Iowa you are better off with a STEM degree than without. Housing around Austin Texas is cheaper than housing in North Jersey.

I do not see the downside of being able to design things.

Matthew said at February 6, 2013 9:32 PM:

What's my alternative? Be born to a billionaire, because that's the group Congress is catering to.

Erik said at February 6, 2013 9:36 PM:

Randall,

It doesn't matter if houses are cheaper in those areas if there are fewer single women available, and if there are more single men relative to single women. And those areas you cite have been depopulated of young people, who end up moving to the coasts and major cities for college or after college.

See this map of single male to single female ratios in the US:

http://www.creativeclass.com/_v3/creative_class/2007/04/03/the-singles-map/

The Plains States and other areas involved in the resource industry like Houston have high single male to single female ratios. Austin, Texas has a very high single M to single F ratio. The tech industry areas like Silicon Valley/Bay Area and Seattle have very high single M to single F ratios. The major northeastern cities from D.C. to Boston have the highest single F to single M ratios. The NYC area has the highest single F to single M ratio in the country.

David said at February 6, 2013 10:09 PM:

I'm sorry, but the advice here is very wrong-headed. What does not seem to be understood is that you cannot somehow game the credentialing process by picking your major first and then deciding where you are going to school. Like it or not, credentials matter and the primary credential that counts is the prestige of the university you attend, followed by your GPA and, finally, your major.

Assuming you have the intelligence, the best way to prove that you are smart is to get into a very selective school and end up with a high gpa. A Harvard English major with a 3.9 out of 4.0 gpa, the right internships and recommendations, and job searching within the pool of firms that typically recruit at Harvard will have far better job prospects than an equally smart candidate who got a 3.9 gpa in electrical engineering from a middling state school.

Any proper career path must start with the proper university. Nothing less will do.

Let me illustrate some of the pitfalls of not understanding this process.

I graduated from the University of Chicago. For those unfamiliar with the U of C, this college is known as a pure basic research institution. Its graduate schools are some of the best in the world and they have very few fields of study devoted to anything practical...at least, anything practical a BA/BS could take into the real world. There is, for example, no engineering department by design, since that would distract from the research mission of the school. The closest I ever got to an engineer, in fact, was the crew that set up the physics labs for the students. Nevertheless, the U of C's high ranking, prestige, and intelligent student body attracts some of the top firms in the world despite the lack of practicality.

One of those firms was, of course, Goldman Sachs. Do you know what their requirements were to be considered for their financial analyst program? One sentence: "3.8 gpa on a 4.0 scale, any major, w/quantitative skills." That's it.

Now letís apply Randallís advice. To demonstrate that you are intelligent to a Goldman Sachs recruiter, letís pick one of the "big brain" majors. Any one will do: physics, math, chemistry, statistics and all of the various offshoots like physical chemistry and molecular biology. You can even do a double major in math and economics which should cover GSís requirements.

Do you want to know how many people with these majors or combination of majors actually went to work for Goldman Sachs? Almost none because they could not meet the gpa requirements that Goldman demanded. Why was that the case? To get a 3.9 gpa in STEM at the University of Chicago requires you to be a genius. Since very few are geniuses, the results generated a lot of 3.2ís/3.3ís, well below the cutoff that GS wanted. The typical recruit at Goldman out of the U of C ended up being a humanities/social science major with a 3.9 gpa. Since the U of C requires everyone to know calculus as a condition of graduating, this type of student already demonstrated the ability to ace calculus, thus fulfilling GSís quant requirements.

Yes, GS passed up the math and science majors with lower gpaís to recruit from the higher non-science gpas. Iíve seen math majors get passed up in favor of those who graduated with English degrees. Since I keep track of doings from my alma mater, Iíve seen this typical profile duplicated even at Google, where a U of C student is showcased working there is typically some non-science major. None of them were doing monkey jobs.

I cannot emphasize enough the value and importance of understanding career tracks. Attempts at futurism will likely fail. Start, instead, with looking up the best firms in the fields you want to work in. Backtrack from these firms to the universities from which they typically recruit. Visit those universities and spend time in career and placement services looking up what these companies have required in the past. Select your college and major around that and then shoot for the highest gpa. Anything less is fraught with peril. You do not want to end up like some of the tragedies I know. Simply choosing STEM is not the right path to take.


David said at February 6, 2013 10:17 PM:

There is a related concept that everyone here needs to understand as well. It is called "fit."

djudd said at February 7, 2013 2:16 PM:

FYI: According to the Center for Immigration Studies there are currently 10,000,000 American citizens with at least a Bachelor STEM degree from an American school who do not work in STEM because they couldn't find a job.

How many more millions would be a good goal? I personally know of people with Master's in Engineering laid off from a Fortune 100 company (making record profits on record revenue) who haven't found work in 15 months.

We'll need more than panaceas to deal with the New America

Randall Parker said at February 7, 2013 8:27 PM:

Erik,

So graduate from college, go to a high F:M place (heck, college is a high F:M place), meet a woman, get her to fall for you, and then drag her back to some place where housing is cheap. Most women don't want to stay in NYC their whole life. Kids make them want a house with a white picket fence.

If you have high status you don't really compete with most men anyway. Plus, if you learn how to handle women you really do not compete with most men. I live in what is supposed to be a low F:M area. Yet that's not how it is for me.

djudd,

As I've said above: Some STEM majors, especially some Science majors, have low market value. Does the world need another biologist or chemist? Nope. Definitely not. Yet people keep studying biology and chemistry in droves. The mind reels. And people study easier (and lower paying) engineering subjects like civil engineering in part because they are easier. Petroleum engineering and chemical engineering pay more.

The world has too many physicists too. But people hire physicists to do other things because employers figure anyone smart enough to master physics can do statistics, advanced software development, and other hard tasks.

David,

Goldman Sachs sounds like it has some dumb recruiting rules. They can get away with it. But other firms with better filters are picking up talent GS is missing. Goldman Sachs has laid off thousands and has shifted employment toward lower cost areas, notably Salt Lake City. Trading firms do not just recruit at the big brand name places btw. They also go for some state schools. That's similar to what I see in Silicon Valley.

What do you think are monkey jobs? I know software developers who are making over $200k per year and even higher.

Public profiles of employees of big name firms: The people who are seen outside a firm tend to be marketers, PR, and senior executives. That creates a misleading picture of who really matters inside a company. Some of the people getting paid high salaries are names you've never heard of and they are very technical.

Career tracks and internships: Yes, getting internships at big name places helps. The big companies use summer jobs to size you up. But this is advice more aimed at a pretty narrow age demographic. Once you are out of college there are other tactics to focus on such as doing a great job.

Looking back at all the people I've worked with I did not know where the vast majority of them graduated from college. The people who were respected, made tech leads, given high ranking were more effective. They were more productive.

The average graduate of the South Dakota School of Mines started at a higher salary than the average Harvard graduate in 2012. Companies really need to get things done and they need smart skilled workers to get the stuff done.

Erik said at February 7, 2013 9:39 PM:

Randall,

Look at the map. Those high F:M areas are expensive areas that are hard for STEM workers to buy status in.

Randall Parker said at February 8, 2013 6:48 PM:

Erik,

My daily experience in a high F:M area is that cute STEM girls like me. I like them. It is good. Maybe it is just me. I'm living a charmed life.

But I think if you make it into a prestigious tech firm that's a big signal.

David said at February 8, 2013 11:39 PM:

What does chemical engineering have to do with mining? I thought chemical engineers design chemical plants?

David said at February 8, 2013 11:43 PM:

If this article is about the need for mining engineers, why are they showcasing the schools chemical engineering majors?

David said at February 9, 2013 1:30 AM:

"What do you think are monkey jobs? I know software developers who are making over $200k per year and even higher."

I'm sorry, I did not imply that software development is a monkey job.

Randall Parker said at February 9, 2013 10:41 AM:

David,

Freeport is a mining company. They probably need chemical engineers because the stuff coming out of the ground has low concentrations of desired metals. They've got to run chemical plants to get the stuff out. So it makes sense that a school which trains people for the mining companies graduates chemical engineers.

Ore concentrations are dropping. More of the mines are in less developed countries. Mining companies need to do a large range of things to get the stuff out. Depending on where the mine is located they can need to:
- build roads.
- build a railroad spur.
- build and operate an electric power plant. That means bring in fuel to burn in it too.
- create kitchens and living quarters.
- control local pest threats.
- deal with dangerous local people and corrupt governments.
- operate a chemical plant to concentrate the desired minerals to reduce the volume of what gets shipped.
- run water pumps to keep a mine unflooded. that's true on surface and below ground mines.
- probably lots of other stuff I'm unaware of.

So mining companies need many engineering disciplines. Harvard English degrees, well not so much.

David said at February 9, 2013 11:55 PM:

Randall,

- operate a chemical plant to concentrate the desired minerals to reduce the volume of what gets shipped.

________________________________

And here is the problem with being a chemical engineering major: it is simply not possible to build and operate a chemical plant simply because your company wants to operate and build a chemical plant. CP's are a "nimby" project. Nobody wants to live by one. Consequently, their construction, location, maintenance and improvement is heavily regulated at the local, state and federal level, usually regulated in such a way as to inhibit their construction. At least in developed countries, there is simply not much of a building boom in chemical plants to absorb enough chemical engineers such that supply/demand forces result in high employment rates and high salaries.

What, then, is driving demand for this major? Clearly, a premium is being paid for living in undesirable areas, either within the United States or in foreign, third-world hellholes. I can see plenty of work in the Middle East and East Asia for this kind of activity, but these are not desirable places to live. Whatever premium you make will be absorbed by the higher cost of duplicating a US standard of living. Besides, for how long can you live in a mining town, on an oil derrick, in rural China, or Dubai, before you have to come home? 5 years? 10 years?

Clearly, another factor is companies hiring chemical engineers because the degree signals intelligence. I can see consulting firms like Deloitte, Accenture, McKinsey, Bain Capital all calculating that a person who can handle a ChemE major is capable of handling the work they typically do. In fact, one of my colleagues back in 2000 had a ChemE from Purdue University and the firm had her initially consulting on COBOL projects. She was paid a lot of money, but she was not a practicing chemical engineer. Her salary could not accurately reflect the supply and demand forces for chemical engineering as a practical discipline.

The same logic applies to software developers and other such STEM majors.

Worse, I am not entirely convinced that the "peak oil" signals that you're getting are not being adulterated by government activity. Gas prices are nearly twice what they were at the same point of Bush's presidency. Does this really imply a steep drop in oil production of nearly half or a near-doubling of the world population? That hardly seems plausible. Are we certain that all this fracking business has not been artificially created by government manipulating more easily obtainable sources of oil? Do you really want to advise people to make that bet?

I stand by my analysis. You do not pass up Harvard University to go to the South Dakota College of Mines. If you're smart enough to secure a rent-seeking position, then you're smart enough to calculate that Harvard is a better fit and a more rewarding path.

Randall Parker said at February 10, 2013 11:44 AM:

David,

Very few people can get into Harvard. "Get a degree from an elite school" is not a solution for the vast majority of people smart enough to become an engineer.

Higher cost of maintaining a US living standard somewhere more primitive: I'm guessing some of the corps are providing communities to live in like Aramco has done in Saudi Arabia.

Building up chemical plant infrastructure: Somehow or other Dow, Dupont, Exxon, Chevron, et. al. spend large sums building up facilities in the United States. For example, Molycorp is spending $1 billion to reopen a rare elements mine at Mountain Pass Summit California.

ChemE degree as IQ signal: So then getting a ChemE degree still makes sense. Though getting a degree in computer science is a more sure fire way to get hired in a well paid software engineering job.

Peak oil signals: They seem clear to me. I've read many books and articles on it. The US government is not trying to get people to see Peak Oil as a real problem. I think the USG is afraid people will panic.

The current high oil prices are due to demand rising (especially in Asia) while supply has not risen much. Why has supply not risen much? Cost of extraction has risen. It is expensive to extract from successively deeper waters, from oil shale, and from tar sands.

David said at February 13, 2013 12:00 AM:

Randall,

>>>>>>Very few people can get into Harvard.

I am not saying that everyone should try to Harvard. The audience for my advice is intelligent people with a lot of options. You do not downgrade the prestige of your school. If you get into Harvard, then you don't give that up to attend South Dakota.

>>>>>>Building up chemical plant infrastructure:

Read the Molycorp link very carefully. That rare earth mineral mine was only reopened because of Chinese restrictions on rare-earth minerals. It was closed previously because the Chinese expanded production. This is an example of the political element determining resource scarcity.

>>>>>>>ChemE degree as IQ signal: So then getting a ChemE degree still makes sense.

It makes since out of a prestigious school. Accenture hires chemical engineers out of Purdue to write cobol programs. They will not take just any ChemE.

>>>>>>>Peak oil signals: They seem clear to me. I've read many books and articles on it.

This goes for the rest of this resource writing. See the Molycorp link you posted about resource scarcity being politically determined. See also:

http://www.forbes.com/sites/robertbradley/2012/06/18/energy-companies-are-not-sitting-on-inactive-oil-leases-president-obama/

mencken said at February 13, 2013 8:42 AM:

Erik and David provide needed perspective. "Cute STEM girls" are rare, among other things.

Randall Parker said at February 13, 2013 8:25 PM:

David,

Sure, a Harvard degree has value. Even greater value: the mind capable of getting admitted to Harvard in the first place. I can't find it right now but one study found that people accepted to Harvard who did not attend did as well economically as those who attended.

Rare earths and politics: Well,if rare earths were in plentiful supply in many other countries then China's export restrictions wouldn't have much impact.

Cobol programs and ChemE: Obviously the chemical engineers should study enough software development to know that Cobol's bad language to learn.

The cost of resource extraction really has risen faster than inflation for years.

mencken,

Cute girls are rare. Among immigrant STEM girls cuties are much less rare than among American STEM girls.

David said at February 13, 2013 9:28 PM:

Randall,

>>>>>>>>Sure, a Harvard degree has value. Even greater value: the mind capable of getting admitted to Harvard in the >>>>>>>>first place. I can't find it right now but one study found that people accepted to Harvard who did not attend did >>>>>>>>as well economically as those who attended.

That's interesting, but I doubt it is the case. Harvard does not publish those who were accepted and did not go. It sounds like self-reporting is operating here and the result is impossible to verify.

>>>>>>>>>Rare earths and politics: Well,if rare earths were in plentiful supply in many other countries then China's >>>>>>>>>export restrictions wouldn't have much impact.

My point is that if an American-located mining pit can be opened or closed based on the political or strategic whims of the Chinese government, then why would anyone want to bet a career on that?

>>>>>>>>>>Cobol programs and ChemE: Obviously the chemical engineers should study enough software development to know >>>>>>>>>>that Cobol's bad language to learn.

I remember learning the Cobol/MVS/JCL/CICS language for a consulting gig in the late 90's. One of the more interesting cultural aspects of the software business I noted was how much scorn was heaped on mainframe programmers. The 3GL universe was populated by fuddy-duddies that could not learn anything. Data was procedural spaghetti code with ridiculous GOTO statements, moved from hierarchical databases or flat files. Mainframes were all water-cooled, obsolete behemoths. Just one harsh word after another.

Then I experienced a real production quality mainframe system when I worked on a project for a trust company. The system was built by Andersen Consulting in the mid-90's and it was extraordinarily elegant. It had no GOTO statements. It retrieved data from DB2 using SQL. It had "call...using" functionality that emulated object oriented programming without locking you into the object model. The mainframe never went down even though it handled an immense number of mission-critical transaction data. Hell, the cobol compilers were as fast as the C compilers.

And what was truly amazing is how Cobol is a largely self-documenting language. Someone put a lot of thought into the nature of business data and business operations. Business data does not go beyond two significant digits and mathematical operations are relatively simple. Because businesses are hierarchical and rely on information being passed up and down an organization, cobol had to be english-like, so that even non-cobol programmers could understand how the programmers were manipulating their data. Cobol speed, refinement, power and how it was perfectly designed for business is a testament to good software engineering.

I learned why the 4GL's were heaping so much scorn on the 3GL's. To sell their products, the 4GL's had to create a tear-down mentality. They sold their cost-center services and products as somehow profitable to get companies to do away with their "old" technology. Thus, mainframes were re-invented as networked pc's, and accounting systems were designed, maintained and modified in a system programming languages like C, while no one in mainframe-land ever considered writing a business application in Assembler. The computer operator became a "software guru" and the computer became a mystical, magical device that the guru manipulated with a combination of artistry, super-science and sorcery.

The result was failed project after failed project, so much so that companies are now outsourcing to get rid of all of their computer people. The guys you know, Randall, probably are system architects that preside over high-maintenance, poorly documented code that only they understand, so their job security is assured.

So, yes, software development is a great business. All you need to do is get hired as a system architect on some mission-critical project. Then, you can create poorly documented code that only you understand and can modify. Then, you can hold a company hostage. Or, you can sell upgrades like Microsoft. Or, you can sell personal data like Facebook and Google. Or, you can break into systems and steal, like Adam Swartz.

All of this was the result of current information systems abandoning the old mainframe model.

Randall Parker said at February 16, 2013 2:35 PM:

David,

Cobol is a declining language. My point isn't to denigrate the language. I'm just saying new problems are being solved in C++, Java, C#, a bunch of scripting languages, and some new languages like Go (which has good features for multi-threading btw). Linux kernel devs are coding in C. Apple iOS folks are coding in Objective C (I think). Java has found big niches in web front-ends, Android apps, and some other areas. Cobol? I never hear about it.

Check out a few different measures of language popularity.

But at the upper end of software development knowledge of specific languages isn't as important of a screening filter as understanding of algorithms and substantial accomplishments (in some companies knowledge of specific languages is not even supposed to be a selection criterion). Anyone is who isn't getting interviews that are really IQ tests and tests of algorithm knowledge and general problem solving probably isn't getting interviewed for high end software dev jobs.

Betting on political whims: As compared to what? Betting you are working for a company that isn't going to become an also-ran for some other of a multitude of reasons? The demand for engineers in the resource extraction industries is pretty widely based. Natural gas, conventional oil, oil shale, coal, rare earth minerals, big name minerals (e.g. iron, aluminum), and other purposes.

David said at February 16, 2013 11:38 PM:

Randall,

The language link is interesting but, reading the comments, I see a lot of interest in COBOL. Besides, it's not possible for Cobol to be sidelined so easily.

When you make statements like "But at the upper end of software development knowledge of specific languages isn't as important of a screening filter as understanding of algorithms and substantial accomplishments" it sounds suspiciously similar to how engineering is a broad field that gives it's practitioner a wide array of skills and abilities to function in any environment.

"Understanding algorithms" is a nebulous concept and does not help anyone choose a career path. You may as well tell people that they should be geniuses. How does one train to understand algorithms or pass these tests administered by Google and Apple?

This is the typical problem with science and engineering: the high degree of faddishness involved. Computer science is the worst, with its proliferation of languages.

Randall Parker said at February 17, 2013 9:54 AM:

David,

How to train to understand algorithms: The Cormen, Lieserson, Rivest Introduction To Algorithms is a good start.

Faddishness: Doesn't seem that way to me. We had a burst of scripting languages in the 1990s as some things shifted from being done in strictly typed and compiled languages into interpreted languages that are more suitable for quick results. Made sense to me. But that only went so far. The computationally intensive stuff is still done in traditionally compiled or JIT-compiled languages.

The major thing that happened: languages backed by IBM declined as IBM's OS's declined. Mainframes got replaced by Unix, Wintel, Linux. Linux looks like the long term winner on the server and probably on mobile devices as well. That favors Java and C++ over C#.

What I find most interesting: The JIT compilation of Javascript gives some of the best of both worlds. Fast prototyping and yet fast performance. I expect Javascript to take some server-side market share away from Ruby and Python as a result. What Javascript needs: multi-threading support.


Post a comment
Comments:
Name (not anon or anonymous):
Email Address:
URL:
Remember info?

      
 
Web parapundit.com
Go Read More Posts On ParaPundit
Site Traffic Info
The contents of this site are copyright ©