Past lost earnings: the mechanics of back pay
Back pay is named simply and built carefully. The loss period, compensation inputs, interim earnings, and jurisdiction-specific adjustments each determine a defensible number.
Past lost earnings, often called back pay, is the component of an employment damages model that sounds the most straightforward and contains the most quiet complexity. The loss period begins on a specific date. It ends on a specific date. Compensation would have flowed during that period. Interim earnings offset. Arithmetic follows. Every step is subject to attack if the inputs are not defended.
The loss period
Back pay runs from the date of the adverse employment action to the date of trial or judgment. For wrongful termination, the adverse action is the termination date. For failure-to-hire matters, it is the date on which the plaintiff would otherwise have started. For constructive discharge, it is the last day of employment, with the economic expert working from the legal determination of when the conditions that caused the plaintiff to resign became actionable.
Jurisdictions treat the end date differently. Some fix it at the jury verdict. Others extend to entry of judgment. The damages model uses the rule of the forum where the matter is pending and notes the chosen end date explicitly.
What counts as earnings
Back pay is not base salary. It is the full compensation package the plaintiff would have received during the loss period, reconstructed from documented pre-loss earnings, documented raises and bonuses for comparable employees, and plan documents for bonus, equity, and deferred-cash components.
For each line item, the model uses actual plaintiff history where available and employer pay practices where not. A plaintiff with three years of documented earnings at the same employer before termination provides a strong base. A plaintiff who was newly hired or whose compensation was restructured close to the adverse action requires more inference.
Benefits that carry cash value (health insurance premiums paid by the employer, 401(k) match, pension accrual) enter the back-pay calculation. Benefits that do not convert cleanly to cash (paid time off that would have accrued, sabbatical eligibility) are disclosed separately and valued where the jurisdiction requires.
Interim earnings
Back pay is reduced by what the plaintiff actually earned during the loss period in substitute employment. This is “actual mitigation.” The plaintiff’s W-2s, 1099s, and pay stubs during the loss period are pulled in discovery and subtracted line by line.
Benefits received from a substitute employer reduce the corresponding benefits calculation. Unemployment insurance received during the loss period is treated differently by jurisdiction: some treat it as a collateral source and do not offset, others subtract it directly.
Prejudgment interest
Some jurisdictions allow prejudgment interest on back pay, compensating the plaintiff for the time-value of the wages that should have been paid each pay period during the loss period but were not. The rate and the compounding convention are set by statute or decisional law in the forum. The economic expert’s model either computes prejudgment interest as a line item or provides the inputs the court needs to do so.
Tax gross-up
A lump-sum award of multi-year back pay is taxed in the year received, at a higher marginal rate than the annual rate that would have applied had the earnings been received in each year of the loss period. Some jurisdictions and some federal statutes permit a tax gross-up to offset this effect. Where permitted, the model computes the gross-up based on the plaintiff’s marginal rate and the size of the award. The calculation is straightforward but needs to be shown explicitly in the report.
What the back-pay section of the report should show
A defensible back-pay section shows the year-by-year build: gross compensation by year, benefit values by year, interim earnings by year, net shortfall by year. Aggregating only to a total obscures the math. The year-by-year table is also what makes sensitivity analysis possible: removing or adjusting one year’s input lets the trier of fact see what drives the number.