Needed: A New IT Employment Model

By Kurt Cagle
December 24, 2008 | Comments: 7

It's Christmas Eve as I write this, but after having put the children to bed and turning off the tree lights, I find that my thoughts are not on Santa Claus tonight ... at least not in a very positive way.

The news on the web is not likely to give a lot of holiday cheer to anyone - Google announcing that they were ending the contracts of some 10,000 part time and contract employees, Sun reducing their force by nearly 6,000, Yahoo "trimming" 1,000. Yet these are small potatoes compared to some 24,000 laid off at Bank of America and a mind-boggling 60,000 people - the size of a smallish city - being let go by Citibank.

Notice the euphemisms here: laying off, trimming, let go, reducing forces. These people are discovering, just after trying to put a brave face on Christmas this year, that they will be dumped into one of the most hostile job markets to emerge in the last forty years, and one that looks like it will get far worse than better for the foreseeable future. Yet the media (and I will readily admit to being a part of that media) have developed a kind of doublespeak about this reality, one that is as insidious and in its own ways as disturbing as the sales pitch of a funeral director selling caskets to the surviving heirs of "the recently deceased".

Some of those people, some of the luckier ones, will be eligible for severance packages, though in this era of tight cash flow those severance packages will likely be meager and stingy at best. Severance is essentially a bit of legal protection - companies eager for new talent in a tight labor market when business is booming make all kinds of interesting promises in a contract, knowing, especially in IT, that the better talent are often now employing lawyers of their own to make sure that those promises have substance.

Yet when times get tight and companies are forced to "lay-off" people, those contracts can prove problematic in our litigious society. Rather than risk going to court, businesses have begun the process of proactively paying out severance packages to keep people from suing them from wrongful dismissal. It's a convenient out, of course, especially as the people receiving the pink slip are often in a position of not knowing whether they will need the funds and whether, if they don't take the severance, they will discover that eventually the company will be unable to make any payments.

Sadly, severance packages are themselves something of a luxury. If you are an architect or senior manager with several years of experience, the likelihood of receiving such a package is relatively good. If, on the other hand, you are a relatively junior programmer working for a small start-up firm, then severances are rare and usually sparse.

The Challenges of Being Contingent

Unfortunately, especially in the IT sector, the vast majority of those people who get laid off likely won't get such severance packages, because technically speaking, they are in fact still employed. The case of Google illustrates one of the unspoken dirty secrets of employment in high tech. For a significant number of companies, as much as a half of the IT staff are contract workers, meaning that rather than being employed by the companies, they are employed by contracting firms that specialize in technical workers.

Contract workers occupy a weird nether realm of employment. They usually make a little bit more than the industry average for fully employed company workers, but the companies that hire them typically charge 15% to 30% above these wages to the companies that they are sourcing the workers to ... nominally in order to provide some of the same services - health care, 401k retirement funds and so forth - that the companies would provide if they were full time workers, though in practice the health care packages are meager (and typically apply only while the worker is actually working) and benefit packages non-existent.

Larger companies like this arrangement, however, precisely because such workers are contingency workers - they can be terminated without notice for a fairly small kill fee, which goes to the employment agencies, not the workers. When job markets are fairly strong, this usually just means that the person will find themselves with a couple weeks "breather", usually without pay, before starting to work for another client.

When job markets are weak, however, contract workers get hurt - badly. They cannot apply for unemployment insurance in many states, because they are technically still employed. They often can't work for a competing agency without violating their contracts with the first agency, which means that if a given agency has a lock on a particularly large company that dominates the IT industry in a given area, the worker loses any chance to go to work for that company again.

Finally, such contingency workers do not have the luxury of a pad that freelance workers are sometimes able to build up. A freelancer may charge that same 15-30% overage fee that the employment agencies do, but that money goes into savings as protection against those periods when they're not working (a point that most freelancers learn early in their business careers). In the case of contingency workers, however, most of that pad is taken by the agency as profit, and very little of it makes its way back to the contractors.

The challenge that those hiring IT workers face, however, is the typically short term requirements that they hire people for. If you are managing a software project, such a project typically has distinct milestones of development toward completion, at which point the project is done, save for an ongoing maintenance phase (more on that last point in a bit, because that caveat has interesting ramifications for employment).

In general, you do not want to keep a Java developer on after the development of a specific Java project if you're next IT requirements are for a Ruby developer, and while it behooves software developers to become proficient across technologies, most programmers are by nature specialists who manage to make the wages they do by being experts in a very narrow domain.

This differs dramatically from almost any other type of employment. Managers are generalists who can be moved from one project to another of a similar type without needing to be a specialist in the technology. Accountants have work to do year round, sales people typically work on commission (or bonuses, which amounts to the same thing) and aren't project constrained.

Programmers, on the other hand, fit into a weird niche world similar to that of a couple of other types of professions that run into similar issues. On one hand, you have the trades model, in which contract workers are typically assembled by a foreman to complete a building, while on the other you have the creative model, in which creatives (writers, illustrators, graphic designers, musicians, etc.) are paid on a per project basis, potentially with some form of profit sharing royalty. A royalty in this case is compensation to a creative that cannot be determined up front because sales are determined on a per-use basis.

The tech worker model can be described as having characteristics of both. Typically, programmers work on a project basis until the project is completed, at which point the programmer may periodically be called in to update the code, but likely will not be working on that code. Such a worker has specialized knowledge, skills and talents, which means that they can be used for certain projects but not for others, and usually once they've started a project, they hold internally a certain knowledge of the state of the project that is difficult to convey to others until the project ends -- in other words, when you hire a programmer, you will likely wish them to stay until the end of the project - though unless you have a lot of similar projects in the works they become superfluous past that point.

Unfortunately for programmers, our society doesn't really know how to handle professional contingent knowledge workers. Forty years ago, programmers tended to be lumped in with stenographers and secretaries, primarily because they worked on keyboards; older managers even today share that same bias about their IT department, seeing it as being a (very expensive) secretarial pool ... and as such, a significant liability.

An independent IT contractor (as with many creatives) tends to face a fairly punitive tax structure, as they usually have to pay self-employment taxes that cover social security, FICA and related "job security" programs, despite the fact that most larger enterprises are able to offset this by securing significant tax breaks in other areas for job creation, and typically because such contractors are services providers rather than resellers, the major benefits of business incorporation, such as being able to purchase goods (and be taxed for selling them) at wholesale rates isn't really relative.

Moreover, the agency model, which does have the potential to be a good fit for programmers, by and large has been subverted by large corporations who see it as a convenient way of creating a truly dispensable work force. The agencies, dependent upon the largess of these corporations for their bottom line, cease being effective advocates for their workers and instead simply become another employment service, albeit one with a disproportionate negative impact upon their workers' livelihoods.

We're going to see the effects of this model upon the IT community over the next few years. IT workers, like any other professional, will only work in a field, even one that they feel passionate about, until such time as it is no longer cost effective for them to be in that field. Because most are highly intelligent they can easily train for positions outside of IT. There has been a significant dearth of people entering IT through college, in part because many are warned off by those already in the field, in part because it is no longer seen as a highly renumerative field or even a secure and stable profession that can support people throughout their lives.

Outsourcing, which evolved in great part to reduce the perceived expense of that technical secretarial pool, will also slow down dramatically, primarily because many countries who initially benefited from outsourcing are now dealing with their own need for IT personnel, and as such are providing greater benefits to work on local projects than to be available as "cheap" contingency labor.

IT will be cut back dramatically in the next year, but the intrinsic evolution of IT technologies will likely continue and eventually (and likely sooner than later) the demand for IT professionals will rise again, especially since most companies simply don't have that much IT staff left to cut - many companies were only just beginning to recover from the tech recession of 2001-2003, and few had built up their IT staffs to the extent of the late 1990s. What that means is that these contingency workers can remain contingency for only so long, then they will move on into other fields or otherwise leave IT, and the already critical IT shortage will become a significant barrier to innovation at a time when innovation is itself a necessity.

Changing the System

Consequently, what exists today is an employment model for IT workers that benefits a few well-heeled interests in the short term but that is already failing in the long term. It is driving talented and experienced people out of IT even as the influx of new programmers is falling dramatically. It forces programmers with families to give up on their profession because of the inherent instability of being employed in IT (not a big problem for single code jockeys, but a major issue for those of use with mortgages and kids to put through college).

From the business side, it means that the quality of the development pool is decreasing significantly, which means more money spent on trying to train up new workers that may be gone in eighteen months. It means taking on the more talented IT workers, even if they don't actually have work to do, for fear that they won't be available later when they are needed, and ultimately it translates into higher attrition, more failed projects and much higher waste.

There are a few things that can be done that will help ameliorate the situation, however. Some are relatively simple, others require a more significant thinking about employment and labor in general.

1) Amend the tax code - not just the income tax, but the highly regressive FICA and social security taxes, as well as state business tax code - to provide a mechanism that doesn't penalize creatives and development professionals at the beginning of their careers, perhaps with a more stringent set of codes for those that have fully established themselves nearer to the end of their careers. This should make it at least possible for independent contractors to survive until they become established.

Similarly, reorganize the incorporation laws to recognize the independent contractor and the workers cooperative as distinct entities that fall somewhere between being workers and being businesses.

2) Make it illegal for a company over a specific number of employees to set up a single employment agency as a gate - this forces a local monopoly that can be extremely hostile to contract developers - requiring instead that no more than 40% of contingency workers come from any one agency. Moreover, make it illegal for an employment agency to put terms of exemption into their contracts that establish a window of time they cannot work for a competitor.

3) Establish a base-level health care system that is independent of one's employer. This does not need to be a single payer system, but it does need to be a system where people are fearful of leaving a job because of the loss of health care coverage. This may be as simple as revamping COBRA, the current system (such a lovely moniker for a health care system - bleurgh), though it is likely that a more comprehensive solution will need to happen.

4) Provide tax and business incentives to the establishment of IT worker cooperatives. Traditional unions evolved around the notion that union workers were typically long term employees of companies, and as such makes comparatively little sense for project-oriented contingency workers.

Workers cooperatives, on the other hand, are businesses that are owned by the workers themselves, that would provide both financial support during times when workers are "underemployed" and would extract a fee from workers when they are currently working. They would also establish competency levels (certification) that make such workers more attractive to employers than what most agencies provide now. These differ from current agencies in that the workers in such agencies typically have no say in the ownership or business practices of the organization.

5) Revamp the unemployment system in this country to recognize that contingency workers make up a large and growing segment of the work force, and that contingency workers tend to be bursty - they will typically receive a few large spikes of renumeration (much like most businesses) but they may also have several months in which they see little to no revenue whatsoever.

6) Make the government bid process in the US (and Canada) more amenable to independent contractors. There are typically a large number of contracts that can be effectively handled easily by a single independent contractor, but because of the organizational focus of the government, such people are often shut out of the process in favor of large corporations.

7) Universities and colleges with IT programs should include IT Business and Contract Law courses as part of any formal certification program. Most beginning programmers are woefully unprepared for surviving as independent contractors from a business standpoint, and as such are easy prey for employment agencies.

8) Revert software patents so that they reside solely with the individual and are non-transferable. This is a huge change, one that frankly will probably not come about in the US although it still has a chance elsewhere. Typically, innovations come not from a corporation, but from an individual or a team within that organization, yet a common provision in most employment contracts is that any such innovation reverts to the corporation. This also means that such patents end up being tradable warrants, often ending up in the hands of patent trolls or used as weapons for litigation. By keeping patents the property of an individual or team, you protect the innovators rather than the corporation, which would essentially become just a primarily licensee of the innovation.

9) Similarly, limit the extent of work for hire agreements concerning intellectual property. The work of most creatives (and there is a class of work produced by developers, especially in the entertainment industry) typically cannot be assessed effectively at the time of creation - this is part of the reason for royalties, but also accounts for various forms of serialization rights. Many companies take broad ownership of intellectual property, usually at the expense of the creator.

In the past, reprint rights in particular were considered an ancillary function, yet in today's media-rich environment, electronic reprint becomes a major factor

This is especially significant since creativity tends to be an age related function - the number of novel and creative innovations that any one inventor or artist can create tends to go down as they age, even though quality in general improves. This means that the bulk of creative work tends to be done nearer to the start of a person's career, and as such is typically owned by someone else.

My personal feeling is that one major change that would help would be to scrap the Digital Millenium Copyright Act (DMCA) and start over with it. This act has done more to destroy the livelihoods of creatives than just about anything else, and has additionally become obsolete as technology raced ahead faster than its provisions could cover.

10) Close tax loopholes that make outsourcing attractive, and place limits on pay differentials allowed for outsourcing. I'm not against outsourcing per se - in the global economy, the chances are pretty good that the person you need might very well be a Russian expert on quantum computing algorithms, but most outsourcing is done not because of need for technical expertise but because outsourced workers may often be paid at a much lower wage without the need for even the minimal protections given to contingency workers.

While many people tend to scream about the needs for free market forces to work, what they often fail to recognize is that in the US especially, this is not a free market. Instead it is a market that has given a huge amount of authority and autonomy to large corporations and the managers of those corporations, usually at the expense of both the worker and, surprisingly, all but a tiny minority of shareholders.

This has created a situation in which we have a consumption driven economy that at the same time reduces the spending capability of those same consumers (most of whom are workers), that is driven by innovation but penalizes the individual innovators in favor of the companies that they work for, and that relies on the entrepreneurial spirit of the independent freelancer and small business owner while at the same time placing major roadblocks in their way to becoming successful.

Not surprisingly, given this, the system is breaking down. Hopefully, we can use the opportunity of this downturn to rethink employment, especially for the newest class of workers in our society, and its relationship to the success of business and the economy overall.

Kurt Cagle is an author, developer, and online editor for O'Reilly Media, living in Victoria, BC, Canada. You can subscribe to his published articles here or follow him on Twitter.

You might also be interested in:


Thanks for such an eye-opening article.Neatly explained the life of contractor in IT industry and the struggles they go through and the exploitation of the Contracting agency. Not only just explaining the problem and providing a good solution to the problem.

Hope the IT contracting agency and the companies they send their employees play a fair game.

Reading through this article, I have to just wonder why the author of this article thinks this situation might be a surprise to contingent workers. All of those contingent workers know exactly the situation they are in and can look to find permanent employment. If they didn't want to be a contingent worker during an economic downturn, they should have been doing something about it when the economy was better.


My experience has generally been that most people who are working contingency generally are not people who would prefer to do so, but who took the jobs hoping that it would turn into full time work. Many companies, especially larger ones, do occasionally hire from contingent ranks, though in my experience they generally prefer to go direct hire, especially with college students on one hand or senior programmers on the other. After all, why pay all the benefits for full times if you can get them on a contingency basis.

-- Kurt

While I completely agree with everything in your article... I find myself EXTREMELY sceptical that the government will do anything to alleviate the situation as it focuses severely on an individual. While I would love to believe that our system is as much for the individual as it is for the group as a whole, time and time again we see that those who seem to effect change typically have large corporate budgets behind them and enough capital to lobby and essentially "buy off" those politicians who would otherwise be getting letters from the citizens of their respective state asking for change ...

Ironically, as much as I hate what the Union environment has turned into, I think I see a need for IT workers to "unionize", this at least would create a system of assurance for hiring agencies that a specific tech is of a certain proficiency and would likewise create a major body of people (I'd be willing to bet nearly larger than many of the other unions).

I can tell you as a blooming IT professional I found myself needing to accept anything that was available that would pay bills, which lead me out of IT related work. So I completely relate to everything you've touched on... however, one issue that you came very close to, but didn't directly comment on was the fac that a large majority of IT skilled people are either self taught or self certified individuals. While certifications do provide a large credit, another hurdle that large corporations have created to "protect" themselves, presumably to assure a certain proficiency level for quality assurance.

I do certainly wish that the awareness and perspective that you've provided would find itself deeply burned into the minds those who benefit from those contingency workers ... hmm... nearly EVERY person who uses a computer as part of their organizational or operational regime.

Thank you.

Hi Kurt,

I wrote a lengthy response to your fascinating article. When you first posted it, I was dying to respond but didn't have time. Well I have now.

My response goes into some creative differences about whether contractors are all faring as well based on seniority; I also talk about your workers collectives idea and the importance of telecommuting. Thanks for the many great ideas.



I've responded in turn to your comments on your site, but will repeat it here because I think it has some relevance to explaining the distinctions I see between consultants and contractors, though I agree for the most part with your assertions:

Some of the points that you bring up were indicative of my own experiences contracting during the 1990s in Washington state, including a couple of stints where I was out of work long enough that I needed to file for unemployment, only to discover the catch-twenty two situation that applied then - if you worked for more than one contracting agency, serially in a given year, your unemployment eligibility was based upon your most previous contract. I had a three month short term contract, and as a consequence, was deemed, by the rather byzantine rules of the time, to effectively have almost no seniority for gaining unemployment, despite having earned more than $75,000 that year. That may very well have changed (I hope so), but it still pointed out the problems in the system.

Health care varies by state (and again by the time) - at the time I was contracting in the mid-90s, there were few options available; that’s changed considerably since (COBRA didn’t come about until nearly the end of my contracting period in the early part of this decade).

Additionally, there are now a number of technical organizations that provide health care packages, but they aren’t that cheap. I was paying nearly $900 a month in premiums with a still high deductible when the bottom fell out of the market (this was about eight years ago) - the health care, as much as it was critical for a family of four, eventually went by the wayside when work faded away.

I remember seeing two-year provisions, which, when you were dealing with a market that had only a couple of dominant players (Microsoft, primarily), could have effectively killed your ability to work. I had a family with two small kids and a wife who didn’t drive, which strongly limited my ability to go to the work. As it was, I spent the time writing books, but it was touch and go for a bit.

I’d agree fairly strongly with your assertions about contracting being something best done by older workers … indeed, I think that by the time you hit forty, you’re either going to be in tech management, in academia or will be contracting, usually with a small, reputable firm. Most companies are simply not willing to pay senior journeyman and master level programmers for full time work (especially if they didn’t come through the ranks at that company), but would definitely seek to gain that experience on a contingent basis.

However, I would also contend that there is a profound qualitative difference between these consultants (and I think the term is apt in this case) and the bulk of contract/contingency IT labor, which does consist of younger programmers who either can’t get full time work yet or are in markets that are agency-locked.

I’ve learned over the years, how to tell when an agency is basically a body shop. I look at the age of the agents, the quality of their tailoring, and the location of their headquarters. One agency which I worked with (regretfully) for about a year had about a dozen agents, the most senior of which could easily have been a student in one of my classes when I taught college. They were usually impeccably dressed, wearing $400 dollar suits, and they had recently moved into a twentieth story floor office in one of the most expensive buildings in Bellevue, Washington.

As I walked into the office, my first thought was that 30% of my potential salary was going into the lease of this office … it wasn’t a comforting thought.

It’s hard to get estimates on the size of the contingent work force, especially since there is almost no reporting by the agencies themselves. What this means is that almost all of the evidence for arguments on industry reform require anecdotal evidence. I suspect that while the contingent workforce has aged somewhat, there’s still a small tier of senior consultants that are doing quite well (and may actually do better in this market, given that critical IT personnel may end up being cut in general RIFs) and a much larger tier of apprentice level programmers who are in a far more fragile position … and these are the ones that are most susceptible to being cut as economic positions deteriorate.

Thanks for taking the time to write this up, by the way - it was very well thought out and overall reflects my experience as a senior consultant as well. I’ll have to address that in a follow-up article.


IT workers occupy an interesting position in terms of economics. Most IT workers make a fairly decent wage out of college and average between $80K and $100K a year after ten years in the field, which is on the high end of the typical middle class spectrum. However, as professionals, they tend to be on the low end compared to doctors and lawyers and senior managers, most of whom have at least a Master's degree or the equivalent (i.e., law school or an MBA).

The problem with this is that it makes protecting programmers a much harder sell politically than protecting automobile line workers or Walmart employees making less than minimum wage (factoring in the purchase of uniforms and other "required" pay-outs). Do I expect a government bailout of IT workers? No, nor would I see it being welcome if it occurred, given the fairly broad libertarian drift of most programmers.

However, I think that instilling a collectivist ethic in programmers is not only possible, but necessary. As I mentioned in the original article, I don't really think that unionization is the answer - its a different model for a different kind of worker - but we are reaching a stage of maturity as a profession where we need to formally recognize that in order to be taken seriously either at bargaining time or in Washington we need to establish basic quality standards. I know that the IEEE has experimented with this somewhat, but haven't seen it really gain traction in the workforce overall.

I also think that technologically setting a standard is becoming doable. The big problem with most certifications is that they lag the technology - by the time you get a certification process up and running and widely accepted, the technology environment has changed sufficiently that the skills involved are becoming archaic.

Perhaps one solution to this is to make such certification contingent upon doing the virtual equivalent of defending a thesis - you need to go before a board (likely over the Internet) of three or certificate arbiters and prove to their satisfaction that you understand the material.

In essence, such certification would be the equivalent of standing a master's degree. After three years of being certified, you would in turn be asked to be an arbiter on someone else's certification, or asked to become a sponsor to someone else seeking that certification.

This process would do several things:

* it would establish a threshold for determining the point where someone becomes a formal "master" programmer,
* it provides a mechanism that keeps the skill levels appropriate to the existing environment (and assures a fairly broad knowledge base for a master programmer, something that should be essential),
* it provides a legitimacy that the many certification programs out there don't give(you've shown proficiency not in a given skill, but in the mastery of being a programmer) that may serve to put programmers on a professional level with doctors and lawyers,
* it establishes an equivalency level with a college masters program that makes it possible for senior programmers to teach at the university level without taking out valuable time to get an advanced degree, and
* it provides a means for master-level programmers to mentor a journey-level programmer through his or her career, providing a way for experienced programmers to give back something to the community and insure a certain level of continuity in the educational process throughout the industry.

Overall, the goal here isn't improving wages for programmers, but in establishing a consistent level of competency, set up a formal ethical structure, and in giving greater legitimacy to programmers as a profession.

-- Kurt

News Topics

Recommended for You

Got a Question?