For the majority of STEM graduates who end up working outside their chosen fields of study, the seeds of the Engineer’s Dilemma are sown in their university experience. But they are nurtured and come to full fruition as graduates begin their post-university employment.

Employers exist to create profit, and with good reason. Profitable companies have time and money to spend on a whole host of good things. Unprofitable and marginally-profitable businesses are miserable places to work. We can’t fault the pursuit of profit alone for the course that some managers choose to take along the way.

There are two (legitimate) ways to increase profit: Growing revenue against a relatively fixed cost basis, or reducing the cost basis against a relatively fixed revenue stream. The former seems to be an order of magnitude more difficult than the latter, and so risk-averse managers tend to “focus on the bottom line” by “taking cost out of the business”. Sometimes this means finding more efficient ways to do things, like changing the way a factory runs to produce less waste, or improving the design of a product to require fewer parts.

Other times, and particularly in businesses dealing with intellectual property where the majority of the costs are people rather than things, the only way to take cost out of a business is to get the job done with fewer people or with people who are willing to work for less money. And STEM graduates are disproportionately likely to work in these types of businesses.

It should not surprise us, then, that the average compensation for all STEM jobs has been flat for the past 15 years on an inflation-adjusted basis. This doesn’t mean that STEM workers haven’t received a raise, rather that their employers have become very adept at taking cost out of their businesses.

There are two effects buried within this trend that each merit our attention:

1. STEM workers are still getting merit and promotion raises. Recent graduates start their careers well below average compensation and receive annual pay increases commensurate with performance. We can safely assume that there are about as many STEM workers who make more than the average salary as there are who make less than average.

2. The use of lower-cost outsourced labor has increased significantly in the last 15 years. This phenomenon effectively puts a lid on STEM salaries. Once a worker’s salary hits that ceiling, it makes financial sense to outsource or eliminate that person’s job even if doing so requires back-filling with more than one lower-cost worker.

The ugly truth: There is planned obsolescence built into the STEM career path. The only question is how soon a given worker will hit the salary ceiling, and paradoxically the first casualties are the highest performers.

Among engineers, high performers whose annual salary increases average 7-10% above the rate of inflation will hit the salary ceiling in as few as 10 years after graduation. The middle third of workers who realize annual salary increases of 4-5% greater than inflation will top out after about 15 years. And the bottom third of workers whose pay increases average 3% or less above the rate of inflation will hit the ceiling after 20-25 years.

For employers, the implications of this phenomenon are profound. The first workers to hit the ceiling are the best and brightest. They make big contributions and grow accustomed to receiving rewards commensurate with the impact of their work. As soon as they get the sense that they might hit a salary ceiling they will move on to greener pastures, either to another function with more lucrative rewards or another company that has figured out how to keep its superstars from topping out. These high performers have no financial incentive to go back to school for a STEM graduate degree, and they are often viewed as too young for leadership positions or don’t aspire to management. The optimal thing for them to do once they near the top of the pay scale is to move on, but why would a company be so foolish as to let its top performers walk out the door?

Because most managers are not high-potential talent.

The middle third of workers hit the pay ceiling at the right time in their careers to join the ranks of management. They are generally the sharpest people on staff at all but the most competitive companies, so they appear to be relatively distinguished in both their talent and their experience. The problem with this group occupying management positions is that they tend to make life difficult — sometimes unbearable — for the younger high-potential workers below them. A middling manager will try to guide a high-potential employee down the same career path that worked for the manager, and to the high-potential this is tantamount to career suicide. High-potentials recognize and can’t endure the relative incompetence they see in these managers. They run for the exits, which may be a short-term win for both the manager and the employee, but this kind of attrition is tremendously costly for the company in the long run.

The bottom third of engineers hit the pay ceiling too late in their careers to make a substantive career change, and they may not have seen planned obsolescence coming at all. They usually end up unemployed with few options.

Companies that don’t understand this phenomenon generally end up with teams of below-average talent led by average managers. High-potential talent will avoid teams like these, and young superstars who do find their way into these teams will generally move on to greener pastures before long.

Employers can make simple and substantive changes to the way they reward their employees, particularly their best and brightest, that will end the vicious cycle of talent flight and eventually increase the competency of managers. Better talent and better management will attract high-potentials and create a sustainable competitive advantage for any company.

But not all companies will understand how to create a system of rewards that allows talented STEM graduates to thrive. In such cases graduates can take charge of their own destiny and navigate through the risk of planned obsolescence.

Read on to learn more.