Debunking the Myth of a Desperate Software Labor Shortage.
August 1, 2001 2:44 PM Subscribe
Debunking the Myth of a Desperate Software Labor Shortage. This recently updated and very extensive study by a member of the CS department at UC Davis CS outlines the issues involved for new and old programmers, including how there isn't really a shortage of programmers in the US, employers are just extremely picky. Also provides suggestions for programmers wanting to keep up with their skills to stay in the industry.
I can attest that it applies beyond programming, as I'm a network engineer. I have skilz up the wazoo, but I'm a generalist, not a specialist. It's been very difficult just getting interviews because employers are being exceptionally picky about specific work experience. Three years ago, the last time I was out, I practically walked into a job sight unseen. Now they say that a four-month gig in something isn't nearly enough -- they want me to have spent the last two years immersed in (whatever).
I'm only 37. They haven't even seen the three gray hairs in my vandyke yet.
posted by dhartung at 3:27 PM on August 1, 2001
I'm only 37. They haven't even seen the three gray hairs in my vandyke yet.
posted by dhartung at 3:27 PM on August 1, 2001
Not having the time right now to wade through this wodge of a document in its entirety, I naturally headed straight for the FAQ. I couldn't pass this one up without comment:
Question: The industry lobbyists say the alleged high-tech labor shortage is due to the failure of our K-12 educational system to develop math skills for engineering careers. Is that true?
The main answer to this question is that the vast majority of high-tech H-1Bs are programmers, not engineers, and programming does not use math.
This from a computer science professor, indeed. Last time I looked (about five minutes ago, I'm a developer) programming consists of an algorithmic succession of logical expressions. Since when was this not a mathematical concept?
I hope this is not representative of the standard of reasoning in the remainder of this report.
posted by normy at 3:46 PM on August 1, 2001
Question: The industry lobbyists say the alleged high-tech labor shortage is due to the failure of our K-12 educational system to develop math skills for engineering careers. Is that true?
The main answer to this question is that the vast majority of high-tech H-1Bs are programmers, not engineers, and programming does not use math.
This from a computer science professor, indeed. Last time I looked (about five minutes ago, I'm a developer) programming consists of an algorithmic succession of logical expressions. Since when was this not a mathematical concept?
I hope this is not representative of the standard of reasoning in the remainder of this report.
posted by normy at 3:46 PM on August 1, 2001
It's worth looking at the Slashdot thread on this report, which notes the author's long history of campaigning against immigration of all sorts, not just the H1-B.
posted by holgate at 3:51 PM on August 1, 2001
posted by holgate at 3:51 PM on August 1, 2001
I just got laid off. My position ends at the end of August. This article depressed the crap out of me.
I think one of the problems is that there is such a glut of real talent out there right now that HR people can afford to be picky. The lure of Internet riches dragged tons of people into the coding field. Now they're all getting laid off and competing for the same jobs.
posted by TiggleTaggleTiger at 3:56 PM on August 1, 2001
I think one of the problems is that there is such a glut of real talent out there right now that HR people can afford to be picky. The lure of Internet riches dragged tons of people into the coding field. Now they're all getting laid off and competing for the same jobs.
posted by TiggleTaggleTiger at 3:56 PM on August 1, 2001
Folks, I've done a lot of interviewing, and there's a good reason why there's a high rejection rate: most of the people who apply aren't qualified. I did a screening interview one time with a guy who claimed to have ten years experience programming in the language "C". I don't generally pay much attention to resumes, as it turns out; I want to know what you are, not what you've done. I rely on various kinds of tests such as coding examples. One of the things I've done is to write some deliberately faulty code of my own to give to a candidate and ask them to tell me if it would do what it claims to do, and if not why it would fail and how it should be fixed.
I gave this guy a piece of code which traversed a linked list maybe 10 lines, all told -- which had a deliberate off-by-one error in it. And he looked at it for a moment and then confessed to me that he didn't now how to use pointers and didn't understand the code. If you're familiar with the C language, you'll realize right now that it's unlikely that anyone had used the language for ten years without using pointers; in a word, the guy's resume was a lie.
The reason there are so many applicants is that programming jobs pay extremely well. But there isn't any law that says that someone who applies actually knows anything. I had another candidate straight out of college who claimed to have taken a one-year sequence in compiler theory, but he couldn't tell me what the difference was between a compiler and an interpreter, let alone where you'd use each. (If you think you never use interpreters, you flunk.) He claimed to have taken an operating system course but couldn't tell me the difference between preemptive and cooperative multitasking. You'd be amazed how many people lie about their credentials; it's the main reason I don't bother reading resumes.
Yes, we only hired about 2% of applicants, and this in 1997 when qualified applicants were scarce. That's because the other 98% would have been liabilities rather than assets. We were better off with empty desks than we would have been with those people as employees.
By the way, that company hired me when I was 43. I was helping to develop cell phones, and when I interviewed I knew nothing whatever about cell phones. I had never even held one in my hand. They didn't mind that part; what they wanted from me was good general skills in embedded software, which I had. They put me in classes to teach me about cell phones, but they didn't have time to teach me to program.
The difference in productivity between a good person and a turkey can be 5:1, and in some cases it can be infinite, because there are some jobs a good person can do rapidly and a lousy person can't do at all. The kinds of companies I have worked for have gotten cautious about hiring because they've been burned in the past. Engineering isn't like staffing a production line; in a production line pretty much anyone can learn any job if they demonstrate minimum capabilities. But in engineering you really have to have a lot of knowledge and also a lot of talent, and these things can't be taught in a reasonable period of time. If I have a problem to solve, I need staff which is effective right now. I can't afford to hire someone and spend two years training them before they become productive. (And if I did invest that two years, getting nothing out of that employee, there's a fair chance at the end that he'd leave and take the skills I paid for to my competitor.)
It's a truism that everyone thinks that they are good, so people who are rejected don't understand why. But their perception as applicants and mine as an interviewer are not the same, and I have to have my employer's interests in mind. If I don't think an applicant will work out, I'll say so and recommend against hiring.
"Give them a chance" isn't in the rules. We're not here to be fair; we're not here to help people out or give them chances. This isn't a charity. We're working on developing products and trying to get them out the door, and our responsibility to our stockholders is to do that the best way we know how. That means hiring only the best people. "Giving chances" and "helping people out" will put you out of business. We'd rather be short handed than to be stuck with an incompetent staff. (Do you have any idea how hard it is to terminate someone for incompetence?)
posted by Steven Den Beste at 5:15 PM on August 1, 2001
I gave this guy a piece of code which traversed a linked list maybe 10 lines, all told -- which had a deliberate off-by-one error in it. And he looked at it for a moment and then confessed to me that he didn't now how to use pointers and didn't understand the code. If you're familiar with the C language, you'll realize right now that it's unlikely that anyone had used the language for ten years without using pointers; in a word, the guy's resume was a lie.
The reason there are so many applicants is that programming jobs pay extremely well. But there isn't any law that says that someone who applies actually knows anything. I had another candidate straight out of college who claimed to have taken a one-year sequence in compiler theory, but he couldn't tell me what the difference was between a compiler and an interpreter, let alone where you'd use each. (If you think you never use interpreters, you flunk.) He claimed to have taken an operating system course but couldn't tell me the difference between preemptive and cooperative multitasking. You'd be amazed how many people lie about their credentials; it's the main reason I don't bother reading resumes.
Yes, we only hired about 2% of applicants, and this in 1997 when qualified applicants were scarce. That's because the other 98% would have been liabilities rather than assets. We were better off with empty desks than we would have been with those people as employees.
By the way, that company hired me when I was 43. I was helping to develop cell phones, and when I interviewed I knew nothing whatever about cell phones. I had never even held one in my hand. They didn't mind that part; what they wanted from me was good general skills in embedded software, which I had. They put me in classes to teach me about cell phones, but they didn't have time to teach me to program.
The difference in productivity between a good person and a turkey can be 5:1, and in some cases it can be infinite, because there are some jobs a good person can do rapidly and a lousy person can't do at all. The kinds of companies I have worked for have gotten cautious about hiring because they've been burned in the past. Engineering isn't like staffing a production line; in a production line pretty much anyone can learn any job if they demonstrate minimum capabilities. But in engineering you really have to have a lot of knowledge and also a lot of talent, and these things can't be taught in a reasonable period of time. If I have a problem to solve, I need staff which is effective right now. I can't afford to hire someone and spend two years training them before they become productive. (And if I did invest that two years, getting nothing out of that employee, there's a fair chance at the end that he'd leave and take the skills I paid for to my competitor.)
It's a truism that everyone thinks that they are good, so people who are rejected don't understand why. But their perception as applicants and mine as an interviewer are not the same, and I have to have my employer's interests in mind. If I don't think an applicant will work out, I'll say so and recommend against hiring.
"Give them a chance" isn't in the rules. We're not here to be fair; we're not here to help people out or give them chances. This isn't a charity. We're working on developing products and trying to get them out the door, and our responsibility to our stockholders is to do that the best way we know how. That means hiring only the best people. "Giving chances" and "helping people out" will put you out of business. We'd rather be short handed than to be stuck with an incompetent staff. (Do you have any idea how hard it is to terminate someone for incompetence?)
posted by Steven Den Beste at 5:15 PM on August 1, 2001
By the way, I'd like to mention that professors in universities are notorious in the industry for being out of touch. It's pretty astounding how many of them think that they're actually creating the state of the art. In actuality, most of what's being taught in universities now is at least five years behind what we're doing in the field. I've talked to professors before and it's amazing how little they actually understand about the real world. There's a reason we make snide jokes about the "ivory tower". I would take this report a lot more seriously if it had been written by someone who had actually been holding a real job for ten years.
posted by Steven Den Beste at 5:23 PM on August 1, 2001
posted by Steven Den Beste at 5:23 PM on August 1, 2001
As I am currently looking for employment right now, I would like to ask the employers of the world a simple request: please be realistic and knowledgeable in your ads.
Javascript is not Java.
Are there really that many people out there that can program C/C++, Java, Flash, Photoshop, SQL and have sales/project management experience (in one person)?
Or is it just me?
I know the economy is slow and you're not hiring much, but still...
posted by owillis at 5:50 PM on August 1, 2001
Javascript is not Java.
Are there really that many people out there that can program C/C++, Java, Flash, Photoshop, SQL and have sales/project management experience (in one person)?
Or is it just me?
I know the economy is slow and you're not hiring much, but still...
posted by owillis at 5:50 PM on August 1, 2001
owillis: of course there is. that person would be the boss's nephew/niece.
(s)he can do all that, and for only 9 bucks an hour.
posted by lescour at 8:34 PM on August 1, 2001
(s)he can do all that, and for only 9 bucks an hour.
posted by lescour at 8:34 PM on August 1, 2001
« Older | World's numbers may peek by 2100 Newer »
This thread has been archived and is closed to new comments
this company expected to find someone who did not need to be trained in powerbuilder. which is all well and good, except -- why are you interviewing college kids if that's the case?
another interview of mine was with a consulting firm that expected to find someone who required no training whatsoever. is there such a creature as the entry level programmer who requires ZERO training? i wonder if this is just HR's way of trying to justify their hours: interviewing people who have absolutely no shot of getting a job.
posted by moz at 3:08 PM on August 1, 2001