THE GREAT SUCKING SOUND IN ENGINEERING • Part 3 of 4

Outsourcing Our Work: Recipes Without a Kitchen

Mar 10, 2026

Outsourcing Our Work: Recipes Without a Kitchen

Herbert Roberts, PE

The Borescope Technicians

At a major engine OEM where I spent years designing compressor and turbine components, the service technicians who inspected engines with borescopes made more money than the engineers who designed those engines. This was not a secret. It was not a temporary anomaly caused by overtime or shift differentials. It was a structural fact of the compensation system—the predictable result of two different labor models operating under the same corporate roof.

The borescope technicians had a union. That union negotiated wages based on the irreplaceable nature of the work: field inspection of operating engines requires years of pattern recognition training, physical access to aircraft on the ramp, and FAA-regulated certification that cannot be performed remotely. The union understood that scarcity of qualified personnel justified premium compensation, and it had the collective bargaining power to enforce that understanding.

The structural design engineers had no such protection. We had bean counters controlling our value.

Ten Thousand Engineers

Consider the arithmetic. A major gas turbine OEM employs tens of thousands of people globally—manufacturing technicians, service representatives, supply chain managers, sales teams, administrative staff. Of those, the engineers who perform structural analysis and design of critical rotating and static components—the disks, blades, vanes, cases, and combustor liners whose failure modes determine whether an engine stays on the wing—number perhaps two thousand per company. By generous count, there are five OEMs in the United States that design and certify large gas turbine engines. That produces a national talent pool of roughly ten thousand engineers.

Ten thousand. Out of the millions of working engineers in the United States. These are professionals with specialized knowledge in fatigue analysis, vibratory response, high-cycle fatigue, fracture mechanics, creep life prediction, and the certification processes mandated by the FAA. Many hold advanced degrees. Most require five to ten years of mentored experience before they can independently sign off on a component design that will operate inside an engine producing tens of thousands of pounds of thrust at forty thousand feet.

If the basic economic principle of supply and demand applies—if rarity of supply drives procurement cost—the compensation for these ten thousand engineers should place their salaries closer to the executive level than to the middle of the corporate pay scale. The knowledge they carry is as rare as it is consequential. An engine that fails in flight does not produce a warranty claim. It produces a catastrophe. The engineers who prevent that outcome should be compensated accordingly.

They are not. And the reason they are not is the subject of this post.

The Portability Illusion

The documentation effort described in the first post of this series—the ISO 9000-driven mandate to formalize tribal knowledge into step-by-step procedures—was never really about quality. Quality was the label. The objective was portability. If engineering methodology could be captured in documents, then the work could be performed by anyone who could read the documents. And if the work could be performed by anyone, it could be performed by someone in a country where engineers earned a fraction of American salaries.

The economics were irresistible to executives managing quarterly earnings. An experienced structural engineer in the United States, fully burdened with benefits, might cost the company one hundred fifty to two hundred thousand dollars per year. An engineer in India, Eastern Europe, or Southeast Asia with comparable academic credentials might cost twenty to forty thousand. Even after accounting for the overhead of managing distributed teams—coordination costs, communication delays, quality review cycles—the savings appeared enormous on a spreadsheet.

What the spreadsheet did not capture was what the documentation did not contain. As we established in the first two posts of this series, the procedures formalized from lab notebooks and engineering practice captured the “how” but not the “why.” They documented step sequences but not the judgment behind boundary condition selections, the conservatism embedded in material property choices, or the operational context that connected an analyst’s assumptions to real engine behavior. When those procedures were shipped overseas, the receiving engineers got recipes. They did not get the kitchen, the ingredients, or the taste memory of the engineers who wrote them.

Engineers Without a Feedback Loop

The most damaging consequence of outsourcing was not the loss of analytical accuracy in any single calculation. It was the severing of the feedback loop that had sustained engineering judgment for generations. In the traditional model, a structural engineer who designed a compressor disk would, over the course of a career, see that disk manufactured, assembled into an engine, run on a test stand, installed in an aircraft, inspected at overhaul intervals, and—when things went wrong—returned from the field with cracks, wear patterns, or failures that told the engineer what the analysis had missed.

That feedback loop was the mechanism by which tribal knowledge was created. Every empirical correction in the old notebooks, every conservative assumption, every handwritten margin note existed because an engineer had once seen the gap between prediction and reality and adjusted the methodology accordingly. The loop connected design to manufacturing to test to service and back to design. It was slow, informal, and impossible to document—but it was the immune system of the engineering process.

When the work moved offshore, the engineers who received it had no access to manufacturing floors where they could watch their parts being forged, machined, and inspected. They had no access to engine test data showing how their analysis predictions correlated with measured stresses and temperatures. They had no access to teardown reports from overhaul shops showing where cracks initiated and how they propagated. They had procedures and computer tools and academic training, and they followed the steps with diligence and good faith. But they were designing in the dark.

The small cracks that had begun appearing when FEA replaced hand calculations—parts not meeting design life, vibratory issues driven by proximity effects and operational variability—were now being generated by a process that had no mechanism for self-correction. The engineers producing the analysis could not see the results of their work. The engineers who could see the results were increasingly removed from the analysis. The feedback loop did not merely weaken. It broke. We were producing people with procedural skills in following directions and losing people that had cognitive skills that could collect a number of inputs from past results and backfill the weaknesses. There was no continous improvement, there was continous creep, and nobody to see it until there was - failure.

Wage Suppression by Design

The standard corporate narrative framed outsourcing as a response to global competition—a necessary adaptation to market realities. The standard economic narrative framed it as comparative advantage—work flowing to where it could be performed most efficiently. Both narratives obscured a simpler mechanism: outsourcing was, among other things, a tool for suppressing domestic engineering wages.

The logic was circular and self-reinforcing. Once a corporation demonstrated that structural engineering work could be performed overseas at a fraction of domestic cost, every domestic engineer’s compensation negotiation occurred under the implicit threat of replacement. The engineer did not need to be told that his job could move to Bangalore or Warsaw. The existence of the offshore capability was itself the message. Salary demands above market became evidence that the domestic workforce was “overpriced,” and market was being defined by the very offshore alternative that the corporation had created.

This dynamic took advantage of two populations simultaneously. Non-U.S. engineers received wages that were excellent by local standards but a fraction of what the work would command in the domestic market—effectively capturing the value of their education and effort at a discount. Domestic engineers saw their compensation compressed into a mediocre middle range that bore no relationship to the rarity or consequence of their skills. The ten thousand engineers who kept engines on wings were paid as if they were interchangeable, because the corporation had constructed a system designed to make them appear so.

The borescope technicians, meanwhile, continued to earn their premium. Their work could not be digitized, documented, or transmitted overseas. It required a human being, physically present, looking into an engine with a specialized instrument and decades of pattern recognition. Their union understood this. The structural engineers, who possessed an equally irreplaceable form of judgment, had no equivalent protection—and no collective mechanism to articulate that the documentation of their methodology was not the same as the transfer of their capability.

What the Spreadsheet Could Not See

The outsourcing of gas turbine structural engineering was, by any conventional financial metric, a success. Labor costs declined. Headcount shifted to lower-cost regions. Quarterly earnings improved. The spreadsheets confirmed what the executives had promised.

What the spreadsheets could not see was the accumulating risk. Every analysis performed by an engineer without access to test data was a small bet that the procedures were complete enough to compensate for the absence of judgment. Every design released without the feedback loop of manufacturing and service data was a small bet that the documented steps contained all the conservatism the hardware would need. Each individual bet was small. But the bets compounded, and the engineers who could have identified the growing risk—the domestic experts whose careers had been built on closing the gap between analysis and reality—were being systematically devalued, displaced, and retired without replacement.

The great sucking sound that Ross Perot warned about did reach the engineering offices. It did not sound like a factory closing. It sounded like the quiet departure of institutional knowledge that no document could capture and no spreadsheet could price.

In the final post of this series, we’ll examine the domestic side of the same coin—how H-1B visa programs reshaped the engineering workforce inside American OEMs, and what it means when credential is confused with capability.

Thank you for following me on this journey. I would be very interested in reading about your experiences, opinions and feedback, good and bad. Please leave a comment, especially if you are outside the US or work in a different engineering field.

The Inventor's Mind Blog's Substack

Discussion about this post

Ready for more?