How to Prepare a Union Remittance Report from Payroll

A union remittance report converts payroll-period data into the contributions a contractor owes a union or trust fund — typically dues, employer fringe benefits, employee deductions, and any negotiated add-ons such as industry advancement or training contributions. Most funds expect it monthly, with the standard pattern being submission and payment by the 15th of the following month. The exact cadence and due date are set in the collective bargaining agreement or trust agreement that governs each fund.

The fringe rule that sits underneath the entire report is unforgiving on overtime. According to Wage and Hour Division Fact Sheet #66E on Davis-Bacon fringe benefit compliance, under the Davis-Bacon and Related Acts, prevailing wages including fringe benefits must be paid on all hours worked at the site of the work, and applicable fringe benefits must be paid for all hours worked, including overtime hours. A report that computes fringe from gross-times-rate, or from hours paid rather than hours worked, has already drifted from what the fund expects.

Late contributions carry teeth. Most CBAs and trust agreements specify interest or penalty assessments in the range of 1.5 to 2.5 percent per month, often compounded, with some funds also permitting recovery of audit costs and attorney's fees on collection. The exact figure for any given fund lives in the agreement that governs it; treat published ranges as a calibration point, not a quotable rate. The same documents set the deadlines that govern timely remittance of employee contributions, and those deadlines — payroll-date relative, deposit-date relative, or fixed-calendar — are what an auditor or fund administrator measures against later.

The report is a reconciliation document, not a software output. Preparing the union remittance report from payroll without the reconciliation breaking is what separates a clean filing from a rejection. Most rejections happen at reconciliation — overtime fringe miscalculated, retro adjustments landing in the wrong period, mid-period classification changes hidden inside a single report row, multi-local hours combined when each fund wants only its share — not at form filling. The rest of this article walks the payroll-register-to-remittance-report mapping field by field, names the recurring mismatch patterns that cause submissions to fail, addresses honestly why teams still drop their payroll exports into a spreadsheet before submitting, and specifies the audit-ready support pack that has to sit behind every filing.

---

From Payroll Register to Remittance Report: The Field-by-Field Map

The payroll register for the period is the source of truth. Every figure that appears on the remittance report should trace back to a row in the register, and any figure that doesn't is the first signal something is being computed off-register and needs to be documented. That principle is the foundation of the payroll reconciliation process between the register and downstream reports, and it is what carries the remittance report through a fund audit later.

The map below walks the columns a typical payroll register exposes and shows where each one lands in a union remittance report. Some map directly. Others have to be derived. A few are contingent on what the governing CBA or trust agreement says.

Worker identifier. The register usually carries an SSN or internal employee ID. The fund's report layout often demands a fund-issued member number alongside or instead of the payroll identifier — pension funds in particular maintain their own member rolls that do not always sync with HR records. This is a translation step, not a copy. If the register does not carry the fund member number as a stored field, the cross-reference has to live somewhere reproducible.

Classification or craft code. This is the rate-and-fringe driver: the contribution rate, the fringe schedule, and often the fund itself are all keyed off classification. Mapping is direct only when no mid-period change occurred. When a worker is reclassified mid-period — promoted from apprentice to journeyman, moved from one craft to another, or shifted off covered work — the single column on the register hides what the report has to expose as two partial obligations.

Hours worked versus hours paid. This is the most common point of drift, and it deserves to be treated as a distinct line of the map. Many payroll systems track hours compensated as the operational default, summing regular, overtime, holiday, vacation, and sick into one column. Davis-Bacon and most CBAs measure fringe on hours worked at the covered site — a narrower definition. If the register does not separate hours worked from hours compensated, the difference has to be derived from time data before the report is built. Some funds do contribute on certain compensated-but-not-worked hours per the agreement (vacation hours funded through a vacation fund, for example), so the rule is fund-specific even when the column on the register is uniform.

Gross wages, base wage, and overtime premium. The remittance report rarely needs gross wages as a single figure; it needs the base wage and the overtime premium broken out, because the fringe base typically applies to hours rather than dollars but dues are sometimes percent-of-gross. A register that only stores gross forces an OT-premium derivation downstream.

Employee dues. Usually a direct map, often a percentage of gross or a flat per-hour rate per the CBA. The pitfall here is not arithmetic; it is making sure dues are computed on the same base the agreement specifies, not whatever base the payroll system defaulted to when the deduction code was set up.

Employer fringe by fund. Fringe is rarely a single sum on the report. Each fund — health and welfare, pension, annuity, vacation, supplemental — appears on its own line, often with its own rate and its own base. The register typically stores fringe as a per-employee total or per-pay-code total; the report needs it allocated per fund. Where the register doesn't allocate, the allocation has to be derived from the rate sheet in force for the period.

Employee deductions for vacation, annuity, or supplemental funds. These appear on the register as deduction lines and on the report as employee-side contributions to specific funds. Mapping is direct when the deduction codes correspond one-to-one with funds. When multiple deduction codes feed a single fund, or one code splits across multiple funds, the report needs the consolidated or split figure.

Employer add-ons. Industry advancement, market recovery, training contributions, and similar negotiated add-ons are the columns most likely to be miscalculated, because they often apply different bases than dues or fringe — per hour worked here, percentage of gross there, flat per check elsewhere. Applying the dues base or the fringe base to all of them is among the most common errors at this layer of the map.

Even where every column above maps cleanly, a residue of derived fields remains — multi-local splits, retro adjustments, mid-period classification changes — and those are where reconciliations actually break.

Where the Reconciliation Breaks: A Catalogue of Mismatch Patterns

Vendor pages mention these issues only as features their software handles. Naming the patterns is more useful — once a mismatch has a name, it can be detected in the data, documented in an exception log, and held against the register the next time the fund flags a return. Each pattern below has the same shape: a register column does not pass straight through to the report, the difference has to be derived, and the derivation has to be defensible to a fund auditor.

Overtime fringe applied only to base hours. The overtime premium attaches to the base wage; fringe attaches to the hour. A report that computes fringe by multiplying gross wages by a fringe rate, or that uses an "overtime is base-only" rule for fringe as well as for premium, will understate. The corrected calculation pays fringe on every hour worked, including overtime hours, at the fund's per-hour fringe rate. Funds that hour-reconcile catch this immediately when remitted hours fall short of register hours worked.

Retro adjustments processed in the current period for prior weeks. A payroll correction posted this period that belongs to last period — a missed time entry, a late classification change, a back-pay award — distorts the current remittance in two directions. Including the retro hours in the current report inflates current-period figures and matches them to a period in which they were not earned. Excluding them without filing an amendment to the prior period leaves the fund short for the period that actually owns the work. The clean fix is an exception log entry that names the period the adjustment belongs to, a corrected prior-period filing where the dollars are material, and current-period figures that exclude the retro.

Mid-period classification or craft changes. A worker reclassified mid-period generates two partial obligations within a single pay period — different rates, sometimes different fringe schedules, occasionally different funds. A single report row keyed off the worker's end-of-period classification hides the split. The reconciliation treats each segment as its own contribution event: hours worked under classification A at rate sheet A, hours worked under classification B at rate sheet B, with the date of change documented.

Multi-local employees split across funds. A worker performing covered work in more than one local's jurisdiction generates per-local hour shares. Each fund expects only its share. Reports that combine the totals into a single line for the worker, or that duplicate the totals across both fund submissions, get rejected. The split has to be tracked back to the time data — the geographic or jurisdictional time stamps on hours worked — and remitted to each fund separately, often on each fund's own form.

Terminations within the period. The final paycheck still carries fringe-bearing hours, employee deductions, and any accrued vacation or PTO contribution. Terminations late in the cycle are a frequent source of missed contributions because the worker drops off the active register before the report is built. The control here is running the remittance against everyone who had hours in the period, not against the active roster on the day the report is prepared.

Hours worked versus hours compensated. This is the named pattern that drives most overstatement and understatement at once on a union report, and the resolution is fund-specific rather than uniform. Reports built from hours compensated treat PTO, holiday, and sick hours as fringe-bearing — overstating against funds that contribute on hours worked only. Reports built strictly from hours worked, without checking the agreement, can understate against funds whose CBA does fund certain compensated-but-not-worked hours through a vacation fund or supplemental contribution. The reconciliation has to know which hours each specific fund considers fringe-bearing, not assume one rule fits all funds.

Add-ons calculated on the wrong base. Industry advancement, market recovery, vacation accrual, and training contributions often each apply a different base. One may be per hour worked; another may be a percentage of gross; another may be a flat per-check or per-pay-period amount; another may be capped at a weekly hour ceiling. Applying the dues base or the fringe base to all of them produces a report that looks consistent and is wrong on most lines. The reconciliation here reads each add-on's base directly from the rate sheet and computes against that base, not against a single reused figure.

The shared diagnostic across every pattern: a register column cannot be passed straight through to the report, and the difference has to be derived, documented, and held somewhere a fund auditor can see. That is what every fund audit ultimately tests for.

---

Why Payroll Exports Still End Up in a Spreadsheet

Most payroll systems used by union contractors ship a union or fringe-benefit report. Most teams using those systems still drop the data into a spreadsheet before submitting. That is not a workflow failure — it is a description of what happens when one system's report has to feed a constellation of fund forms and prior-period corrections that the system was not designed to produce in one pass.

The structural reasons the export pattern persists are worth naming, because once a team understands them, the spreadsheet step can be designed as a control rather than tolerated as a workaround.

Fund-specific form layouts differ. A payroll system's report is one layout. Each fund typically wants its own — sometimes a fund-portal upload with prescribed column order, sometimes a fund-specific Excel template with proprietary cell positions, sometimes a paper form with member-number formats that do not match the payroll IDs. Re-keying the same hours into three or four different layouts is often what the spreadsheet is actually doing.

Multi-local splits are rarely native. When a worker spans locals, system-level reports allocate by classification or pay code rather than by the local jurisdiction the fund cares about. The split has to be reconstructed from time data — geographic or job-coded — and the spreadsheet is where that reconstruction lives.

Prior-period corrections need a side pass. A retro adjustment posted in the current period shows up as a current-period figure in the built-in report. Producing both an amended prior filing and a clean current filing usually requires manual intervention outside the system, because the system's report logic does not separate "adjustment" from "current activity" in the way the funds want them separated.

Recordkeeper feedback loops. Third-party fund administrators return reconciliation files in their own formats — fund-portal PDFs, CSVs whose member numbers do not match the payroll register, paper acknowledgments. Matching those returns to the original submission is a reconciliation in its own right, and it usually happens in the same spreadsheet.

Add-on bases that vary by fund. A single payroll system computes one fringe base per code. Reports that need different bases per fund — per hour worked here, percentage of gross there, flat per check elsewhere — push teams to recompute downstream, and the recomputation tends to land in the spreadsheet alongside the rest of the side passes.

What every one of these reasons points to is the same underlying control. Before the remittance file is assembled, payroll registers, fund forms, and recordkeeper outputs need to be in one structured layer the team can hold against the register. Mixed payroll output — PDF registers from the payroll system, fund-specific paper or Excel templates, third-party recordkeeper PDFs — is the input to that layer; structured rows with worker, fund, hours worked, hours paid, and contribution figures are the output. That extraction-and-normalization step is where AI payroll data extraction fits into a remittance workflow: it turns mixed payroll documents into structured data the team can reconcile against the register, before the per-fund remittance files are built.

The product is not a remittance tool — funds do not want a generic structured file, they want their specific form — and treating it as one is the wrong framing. Its job earlier in the workflow is concrete: a payroll administrator can upload the period's payroll register PDF, the fund-specific reconciliation files, and the recordkeeper acknowledgments together, and prompt for the worker, classification, hours worked, hours paid, gross, dues, and per-fund fringe figures to be returned as a single Excel sheet. Each row carries a reference to its source file and page so the figure on the remittance report can be traced back to where it came from. What the spreadsheet does after that — fund-by-fund hour reconciliation, exception log, per-fund form fill — is where the operator's judgment lives, and that judgment is easier when the input data is already structured and traceable rather than re-keyed.

---

The Audit-Ready Support Pack Behind the Filing

A remittance filing is one artifact. The support pack behind it is what determines whether a fund audit, recordkeeper inquiry, or DOL Davis-Bacon investigation six months later finds a defensible record or a reconstruction job. The principle is simple: every figure on the report should be reproducible from a self-contained packet that ties back to the underlying payroll register without scramble.

The pack should contain the following:

• The payroll register for the period, in its native format — Excel, CSV, or PDF — plus any structured extract used to build the report. Hours worked, hours paid, gross wages, dues, and employer fringe should all be visible at the worker-and-classification level the report was built from.
• A fund-by-fund hour reconciliation showing how register hours map to remitted hours for each fund. Where the mapping required derivation — multi-local splits, hours-worked-versus-compensated adjustments, mid-period classification changes — the derivation method and the data behind it should be documented in the same workbook, not held in someone's memory.
• An exception log naming every adjustment that was not a clean register pass-through: each retro adjustment with the period it belongs to, each mid-period classification change with the date of change, each multi-local split with the jurisdictional basis, and each termination with the final-paycheck contribution detail. The exception log is the audit trail for any figure that is not a direct register-to-report mapping.
• A copy of the report as filed, in the layout the fund accepted — fund-portal export, fund-specific Excel template, or scanned paper form. The "as filed" version is the one auditors compare back to the register, not whatever working copy the spreadsheet was built from.
• Evidence of submission and payment: portal confirmation receipts, ACH or wire references, check numbers, or recordkeeper acknowledgments. Submission and payment evidence have to be paired — a confirmed submission without a confirmed payment leaves timeliness exposure even when the report itself was correct.
• Reference to the governing CBA or trust agreement section that drives the calculation: the rate sheet in force for the period, the fringe schedule, the cadence and due-date language, and the late-payment terms. Auditors verify the report against the agreement, not the report against itself, and the period of work matters because rate sheets are versioned.

What auditors typically reconcile narrows to three comparisons: hours on the payroll register against hours on the remitted report, payroll dates against deposit dates, and the contribution rates applied against the rate sheet in force for the period. A pack designed around those three comparisons covers most of the audit surface a fund or DOL inquiry will examine. The same approach lines up with the payroll compliance audit records auditors typically request when an employer's union contributions become part of a broader payroll audit. Where the union remittance is also reviewed inside an employee benefit plan engagement — common for multiemployer plan filings — the pack overlaps substantially with the employer-side audit package checklist for benefit plan filings, and building one pack that serves both keeps the two reviews from each demanding their own reconstruction effort.

Whatever system or spreadsheet builds the remittance, the support pack is what survives. Designing the workflow so the pack is a byproduct of the filing rather than a reconstruction job after the fact is the difference between a clean audit and a six-month recovery exercise across funds, payroll periods, and rate sheets that none of the people involved remember in detail anymore.