That model seems to assume:
- every employee who leaves will go to close competitors
- their work for that close competitor will be perfectly zero-sum to their work for your company
- the next best replacement employee your competitor would otherwise hire would be worth 0 to them
It also assumes that you don't pay your employees well enough to retire immediately after leaving ;)
Some of that extra business will be taken from other companies. In fact, unless it's a duopoly or you're by far the largest player in the market, probably most of that extra value will be taken from other companies.
If you pay engineers $300K for every $150K in value they add to your product, your cost will exceed value and will accelerate as you add more engineers. It's therefore eminently clear that there is no meaningful sense in which the true value of an engineer's work adding $150K to your product value is $300K. (It's $150K less the necessary support costs that allowing them to work imposes on the company.)
The true value of the employee ==/== the total compensation of the employee. As I said above, the total compensation would be as close to $150k as possible - because at $149,999.99 the firm still makes a profit (of $.01) and gets product features shipped.
If you find yourself in that unusual situation, your concern probably isn't with paying your engineers properly, it's with reorienting your business so it's not locked in that kind of market.
Secondly, the salary of an employee isn't the entire cost to the employer. I've seen some different ranges, but a random Boston Business Journal article [0] cited the total cost as 2.7x base salary. A CNN Money article cites 1.18x to 1.26x [1]. It's obviously really different on a case-by-case basis, but whatever that multiplier is has to be taken into account.
Lastly, for many companies, engineers are worth far more than 150% of their salary as value to the company, on average [2]. Apple makes 1.87 million per employee, Microsoft at 732k, IBM at 244k. These are obviously huge tech giants, but they are employing quite a lot of the highly skilled employees that you think are being underpaid. I'm not interpreting your comment to mean that your idea should be adopted for everybody in every situation, but certainly in a lot of (most?) places it can't work. Many tech workers are not paid what their true value to their organizations are, but I'm not sure that's a problem in of itself.
[0]: http://web.mit.edu/e-club/hadzima/how-much-does-an-employee-...
[1]: http://money.cnn.com/2013/02/28/smallbusiness/salary-benefit...
[2]: http://www.businessinsider.com/see-how-much-tech-giants-like...
> Also, how can a company give bonuses or raises to employees when all employees are already being paid what the value they are generating for the company is?
is a terrible argument. In the event that my boss determines that I deserve a bonus, I consider that an adjustment for mis-estimating my value when determining my original salary. If you pay me enough that you can't afford to pay out a bonus, then I don't need a bonus.
If you plan to pay out bonuses, given that the practical value of future money is less than current money, I prefer to simply have that bonus money added to my base salary from the beginning.
I realize that for many people there is a psychological factor involved; getting a bonus feels good, in a way that getting the same amount of money spread out throughout the year does not, and there is some value in that. But the ability to choose one payment option over the other should be open to a potential employee during negotiations. You should be able to offer the option of a higher salary in exchange for no possibility of bonuses.
It did make me realize though, that there is a risk of having a downraise, if the value of what you produced in the past cycle is less than the one before that. Even with a hypothetical way to accurately estimate the value (unless we're averaging), having a downraise or even the risk of it, is a big hit against this system because of the stress it would create for some, imo. Additionally, for some people, there may not enough incentive to continue working in the current cycle, if they feel they have already delivered as much value as they feel they need. Some people would be content delivering 100k of value for instance, wherever that is in the cycle, while some would be trying to reach new heights.
Most see a salary, benefits, fringe costs (like hiring HR and finance teams), lunches, etc. Some of the better firms add in the cost of hiring employees with intangibles, like communication, poise, and professionalism, but the truly great see the largest potential cost - that they do better work at a competitor.
If you're paying an engineer $100k to add 150k worth of value to your product, the true value of their work is $300k: $150k worth of value to you, and $150k worth of value that isn't added to your next closest competitor.
It's the same reason football teams say "defense wins championships" - because Defense is the only position where you can simultaneously score points and stop your opponent from scoring. Offenses can only score.
Now, that employee shouldn't be paid $300k because that's the true value for the organization, rather, they should be paid as close to the marginal benefit they directly apply to the organization - in this case, $150k. Doing so may be a breakeven point from a cash perspective, but a billion dollar company that breaks even is worth more than a lemonade stand making $10 profit.
From an accounting perspective, this can be observed as employee expenses (temporary equity accounts) providing value to products (permanent asset). When the employee is terminated, the assets still remain, so for anyone with any equity stake in the firm, it is beneficial to pay as much as possible up to the point that the company a) doesn't run out of runway, b) retains talent to further grow the assets and c) doesn't grow competitor assets.
The CFO that cut soda costs by $10k also boosted the equity of all the competitors that hired the disgruntled employees - and most likely by a value greater than $10k, and multiplied by each employee that left.
Just my $.02
Edit: Cut soda costs by $10k total, not per employee