C&E Fabrication: Pay Compression and a Quality-Destroying Incentive

This is classic pay compression turning into inversion. The wage range was never re-priced during the hiring freeze, so the market moved underneath it. New hires were correctly brought in at market ($27.50), but they now earn roughly $1.40 to $2.70 per hour more than colleagues with six to twelve years of tenure. The structure no longer pays for skill or seniority; it pays for the accident of a hire date.

Consequences to name:

• Turnover concentrates among the most employable incumbents. The two who left are the leading indicator; experienced welders with external options are exactly the people whose equity perception is most damaged.
• Range collapse destroys progression. If an experienced welder cannot move up without exceeding the new hire's rate, the internal ladder stops functioning and promotions become a pay cut in relative terms.
• Recruiting gets harder, not easier. Once the market hears that C&E pays new hires more than veterans, offers are discounted.
• Supervisory credibility erodes. Foremen cannot explain the rates, which is why the spreadsheet spread.
• Potential legal exposure. Check whether the compression falls disproportionately along protected lines; an unexplained inversion inside the same job can support a pay-equity claim under federal or state equal pay law.

Fix the structure, not the symptom: refresh market data, re-price the range midpoint and maximum, then move incumbent rates toward the new midpoint using a defined adjustment (for example a seniority and skill matrix), rather than issuing selective raises to whoever complains. Cost the adjustment openly: lifting 100 welders by $2.00 per hour is roughly $416,000 a year in base wages, which is comparable to the replacement cost of a handful of experienced welders given recruiting, training, and rework.

Diagnosis. The plan pays for one variable, volume, with a single hard threshold at 900 units. Four design faults follow from that:

1. No quality gate. Units above the threshold earn $0.40 regardless of whether they are good, so the marginal unit is worth producing even if it fails inspection.
2. The threshold is gameable. A step function invites crews to sprint past 900 and then stop, and to skip any step that slows the count. The inspection step is exactly the step with no output of its own.
3. The standard may be stale. A 900-unit standard set before process changes may be loose enough that crews hit it without efficiency gains, or tight enough that the only route past it is cutting corners.
4. Social pressure compounds it. At crew level, one person slowing to inspect costs the whole crew money, so peers enforce the shortcut.

Redesign. (a) Add a quality measure with real weight: first-pass yield or defects per thousand, either as a multiplier on the bonus or as a gate below which no bonus is paid. A gate is simpler to administer than a multiplier and is easier to explain. (b) Re-time the standard with industrial engineering or a formal time study and set a graduated curve rather than a cliff, so effort is rewarded continuously and there is no incentive to stop at 901. (c) Extend the measurement period to a month or quarter so a bad week is not hidden. (d) Include a crew-level quality component alongside individual output, so peers police defects instead of policing inspection. (e) Model the payout at historical defect rates before launch and pilot in one plant.

One thing compensation cannot fix: if inspection was deliberately skipped, that is an integrity and internal-control problem. Audit the inspection records regardless of how the plan is redesigned.

The confidentiality rule is the immediate problem. A blanket rule forbidding employees from discussing wages is the kind of restriction the NLRB has repeatedly found unlawful because it chills Section 7 rights, and those rights protect nonunion employees as well as unionized ones. The rule sits in the handbook as an unfair labor practice waiting to be charged. Do not discipline the welder who sent the spreadsheet; discipline would turn a morale problem into a legal one and would be difficult to defend.

Rewrite the handbook to a narrowly drawn confidentiality clause covering genuine proprietary information (customer lists, formulas, pricing to customers, personnel records) and explicitly carve out discussion of wages, hours, and working conditions and any other protected activity. Have counsel review the rest of the handbook for similar overbreadth while you are in it, since these rules tend to travel in packs.

Then treat the disclosure as a symptom. Employees circulate pay data when they believe the system is unfair and unexplained. A town hall that explains the pay structure, the market data behind it, and the adjustment timetable does more for retention than any policy.

Transparency law readiness. With eight months: refresh the market data now, build defensible ranges rather than posting the current midpoints, audit internal equity within each job family before the ranges become public, decide how to handle anyone who falls outside a posted range, and train managers to answer range questions consistently. The law is a forcing function for a pay structure C&E needed to rebuild anyway.