Do computer science graduates get jobs in the end?

Administrative units of the United Kingdom
Image via Wikipedia

Computer science has the worst employment rate for any broad class of (bachelor) degrees in the UK – and seems to have been in that place for some time.

But what happens a few years after they graduate … well the news is not so great, though the figures do indicate that the vast majority of graduates get jobs … some sort of job anyway … four years or so after they graduate.

The UK’s Higher Education Statistics Authority (HESA) has just published its longditudinal survey into the destinations and expectations of graduates from 2006/07.

By the winter of 2010/11, of all graduates from full-time courses in that year, just 3.8% were unemployed. But for computer science graduates the figure was a much higher 5.1%. Interestingly, the survey also shows just 3.7% of computer scientists remained (or were back) in full-time further study – compared to an average of 8.2% for all graduates.

It’s not all bad news, 81.5% of computer science graduates were in full time employment four years on from their degree, compared to just 73.2% of all graduates. For maths graduates the figure is 73.1% and for physical science graduates it is  just 66.0% – though a whopping 19.8% of them are in full-time education.

What is not clear is whether the physical scientists are in education because they love their subject and want to do more, or because they have nothing else to do.

The high employment rate of computer scientists may also be because they (we assume)are more likely to be male and so not breaking their career to have children – there is not enough data here to say.

21 thoughts on “Do computer science graduates get jobs in the end?

  1. I think it would be interesting to find out what the shortcomings of the British computer science curriculum is. I have, for years been seriously concerned at the serious discrepancy created since computer science and information technologies have been seen as a single group. Computer scientists are generally much more attractive since they are people who can fit into many different job types ranging from IT to mathematics. IT grads on the other hand shouldn’t even really be in the university for the most part. These are programs which should be learned in trade schools or certification programs.

    Additionally, it’s about time that web technologies jockies were spun off of the computer science track. Making web pages and coding business systems IS NOT a computer science. A computer scientist designs, implements, improves upon etc… algorithms generally mathematically. A web technologies geek on the other hand makes pretty web pages and hooks them up to databases… sometimes hooking them up to algorithms, but generally they use algorithms instead of making them.

    Computer scientists are the guys who create the technologies that are used by web tech jockies and IT jockies. They are in demand at all times and have very few problems finding jobs anywhere in the world. IT and Web Tech guys are a dime a dozen and unless they are exceptional, are generally easily replaceable and therefore of much lower market value.

    The major discrepancy comes in video game developers which are generally among the best educated computer scientists as they also have strong classical math education as well as strong physical science educations. Yet they have a hard time finding jobs in their field because their jobs are generally in very high demand. Though I happen to know for example that AMD/ATI’s drivers development department is among one of the most professionally organized and run development groups in the PC industry. They have a clear differentiation between junior and senior engineers etc… and their workers have to legitimately work their way up the ranks. Sadly, like with most teams, their task is too great for the resources available to them and as a result the quality of their work suffers.

    1. That’s not true, you think the people in Amazon, Google, Facebook are not software engineers? Just because they do web development doesn’t make them any less computer scientists than AMD’s engineers.

      1. Yes, I am saying that there is a difference between an engineer and a scientist. Problem is as I described in my earlier post that the problem is, people such as the author of this article don’t realize that. Facebook, Amazon and Google employ computer scientists and software engineers. Often people are members of both disciplines. However a software engineer is not necessarily a computer scientist. And therefore my answer is yes.

    2. Darren,

      Thanks for the reply. But I think you are being absolutist in your argument.

      Coding is probably not a higher level skill but rather a technical one. But good coding and design rather go hand in hand and computer science ceratinly has a role to play in that.

      Incidentally, you seem to have a rather odd view of what an algorithm is. It’s a method to solve a problem. All code – that produces a useful output – will use some algorithm to solve a problem.

      Nobody “hooks up” to an algorithm, but everybody hooks up to a program that implements one.

      (Some problems are not solvable, and for some we do not know what the algorithm is in advance.)

      1. Ok… now we’re obviously playing semantics. I see a software engineer as a person that may code a project which makes use of an H.264 encoder. There are literally hundreds and sometimes thousands of algorithms that are involved with a H.264 encoder. There are algorithms for identifying the most efficient and “proper” balance between applied motion vectors and quantization (there are more factors, but I’m limiting it here for this purpose) to achieve an optimal signal to noise effect. A software engineer may spend some time researching the different products from different vendors, they may even make minor changes to the code to integrate it in their projects. However, that engineer will have a limited if any actual understanding of the algorithms that are involved in how the H.264 encoder actually works.

        That majority of software engineers today do in fact simply “hook up” algorithms to accomplish their goals. In fact, the vast majority of developers today randomly choose which algorithm to use based on the API provided to them without bothering to understand the underlying logic behind what would make a List more or less optimal than an AVL tree or even a simple Vector. They often implement suboptimal code for retrieving data from these collections due to their lack of understanding of the algorithms. They may attempt to do things like insert items in a sorted fashion to a vector when it would substantially more efficient to append the item and sort the vector once the items have been added.

        A software engineer often thinks on this level since even the better ones are often far more focused on the full picture of a project as opposed to an individual module which could be coded far more effectively if science were properly applied to the problem.

        Also, I didn’t say being a software engineer excludes you from being a computer scientist, I just say the two things are two separate disciplines. You don’t need to be a computer scientist in order to be a software engineer. An engineer is generally focused more on creating a specific product. A computer scientist will often be more focused in creating optimal underlying technology to apply to the problem. Some people do both.

        To make an association which would clear things up a bit, a bridge designer/engineer will design a fantastic bridge for moving millions of people across a river over the next 200 years. Once he has designed the structure and asthetics, he’ll pass the design of the bridge to a group of scientists who will focus on things like helping to choose the right materials to build the bridge from, sometimes requiring the engineer to redesign components of the bridge in order to decrease the cost of the materials needed as the original design would require materials that cost too much in the given dimensions. Additionally, sometimes other scientists will be hired in order to attempt to produce materials that meet the needs of the cost as well as asthetics. While the bridge builder has a general understanding of material sciences and the material scientist has an pretty good idea of how bridges are engineered, they have two different disciplines. One is assembling components while the other is researching the optimal component for the job.

        The difference between a software engineer and a computer scientist should be similar. The fact that we cludge together database developers with digital signal processing specialists is the issue I addressed initially. We don’t group together materials scientists with bridge designers even though they might sit next to each other for the next 50 years in an office working on the same project. Even if over time the two will form a much more solid understanding of each others disciplines from the experience of working together. We do do it to the different disciplines of computers and that’s what is really wrong here.

      2. quite the old joke is “engineers” are “scientists” with thumbs – there is of course a rider to that “but the technician actually knows how it realy works🙂 Like when i pointed out to a senior engineer at the 4th largest consulting engineers in the world that he had used the wrong equation in his bridge design.

        oh i mean “real” chartered engineers here not some developer with a CS BSc and a sense of self importance..

  2. ” IT and Web Tech guys are a dime a dozen and unless they are exceptional,” — goes to show how little the commenter knows about the industry. We’re based in central london and have yet to find someone suitable in 9months, there is a huge shortage of even junior staff.

    1. I am not familiar with the IT web page jockey industry directly. But, It strikes me that there are some obvious issues here. Is you perceiving a shortage because you can’t somehow manage to attract the people you’re looking for? Are you taking into account that programmers are generally not as adept at finding job opportunities? Are you writing your job listings in a way that scares off the junior guys because they fear they are underqualified due to lack of formal experience or lack of self confidence? Are you making the biggest mistake which is to not publish the expected pay scale, an instant way of causing those guys to distrust you or fear they’ll have to negotiate or be screwed? Are you expecting them to come to you? Are you trying to find them at networking parties which is a great way to find a sales person, but a crappy way to find an IT guy? These are just a handful of things I can think of that can cause a company to misperceive a shortage. From the article, it sounds like the author and you have entire different perspectives of the problem.

      1. I think your first sentence says it all really; your inexperience with this particular industry and job market means you have actually got completely the wrong idea. Working for a web development company we’re definitely finding that there is a shortage of talented software engineers in the market, especially with regards to skills in open source languages as opposed to the glut of .NET and Java engineers. I’d say the reason for this is primarily because Computer Science degrees let them down, not because CompSci teaches better skills overall. My experience of CompSci at university I find to now be very out-of-touch with how the web industry operates, which is bleeding edge and often requires a wide-skill set. Semesters in low-level C and Java chucking together bad looking GUIs for finance algorithms don’t seem to satisfy what there is true demand for in software engineering anymore.

  3. But you’re not looking for computer scientists… you’re looking for software engineers with an education in the skills you’re looking for. If you hire a university grad with a degree in computer science, you’d expect that graduate to be able to learn the skills of interest to you. If you’re looking for a web developer who knows LAMP stuff (don’t know if that acronym is still in use or if the finance/marketing guys have come up with a fancy new one) then you should be looking into trade schools which teach those skills. A comp-sci grad should have no specific dependence on a language or an environment and as a result might actually completely lack any skills what-so-ever in the field of software engineering.

    I genuinely believe that if the schools did a better job of separating computer science from software engineering, you’d have a much better time finding the right people for the job. Computer science is about the science of computing. Software engineering is about making something out of those sciences. There is a HUGE demand for computer scientists. My company hires them like crazy. We have a tremendous problem finding them because we get hundreds or thousands of applicants from software engineers who got a degree in “computer science” at the university, it takes forever to sort through them. Often times, we simply toss out CVs that list the languages they know. We want to know what kinds of problems they have researched and solved.

    The problem as I see it is, you’re looking for coders and we’re looking for scientists. You’re getting scientists and I’m getting coders. But because some idiot at the university just gobbed this crap together, we’re both screwed.

    P.S. – if you hire a computer scientist who’s actually written something useful in ANY language, then you should start there. They’ll adapt to your environment…. eventually.

  4. The issue lies with the survey, I’m a CS grad from 2010 who went back into education.
    A lot of my friends from the course are now employed.
    I lived in a house with 3 other computer science grads and when that survey arrived it had common destination, the bin.

    CS grads have a unique attitude toward efficiency that prevents them from wasting time responding to surveys that do not benefit them, they usually have mire interesting thing’s to do.

    1. As a CS lecturer, I know exactly what you mean about a “unique attitude toward efficiency” as something that comes with the territory, and which can be a virtue. However, as a CS lecturer whose job is on the line because of indifferent student responses to surveys, I wonder if this “unique attitude toward efficiency” sounds more like youthful arrogance and irresponsibility. Please let us know somehow if there is some way we can encourage you to reply responsibly to surveys. You don’t realise how much this matters.

      1. Starting by telling people they’re being arrogant might he the wrong direction🙂 sorry couldn’t resist as with most of the rest of the thread semantics was in the spirit. The poster above has a valid point. Engineers and scientists with jobs are far less likely to spend time during their work day filling out surveys unless they are dissatisfied with their position. Offering to enter people with a mathematical education without clearly defining the parameters of the contest in such a manor that there would be less than a 1 in 100 chance they’d win something if value to themselves would be ineffective as well. Gambling is for people who don’t understand odds. Additionally, many developers are paranoid by nature as they have been taught to be paranoid with other people’s data. So they are less interested in having further information about themselves stored in databases which are run by groups that appear to be suspicious. And nothing seems more suspicious than surveys asking about your means to feed yourself. Frankly, I feel for your value is being assessed based on the input of people who have too much time on their hands, lack the problem solving ability required to get jobs, or people who are generally dissatisfied with their position in life. Another has mentioned that the survey might take some of these variables into account, but I would be shocked if they can clearly understand how extreme the case is with engineers and scientists.

        In short, I genuinely believe that there is little hope it would be possible to establish any meaningful findings from these statistics. So, you might need to take it upon yourself to establish a database of contact information for your students and contact them yourself and through your personal relationship with them, formulate statistics relative to your own personal success rates.

    1. This is a longditudinal survey by professional statisticians, not a piece of voodoo polling. I cannot be sure, but I very much doubt that the “we throw the survey in the bin” or “we’re all abroad” factors are not discounted.

  5. But how many are doing CS work 1 year, 5 years, 10 years, 20 years, 30 years, 40 years, 50 years after they start? (Yah, I know; it’s a government report and they either didn’t think to investigate that, or they think it would be too difficult.)

Comments are closed.