Reviewing the whole file
PA payroll CSV review: reconcile every row before using the results
Check that every expected record is accounted for, then check what each result means. A complete file can still contain unresolved addresses.
This guide is a file-review workflow, not a withholding calculation or an approval to import into payroll software.
Start with the scope accepted for processing
The service has input and row limits. An oversized, rejected or truncated file is not a complete result for the original roster. Keep any correction request or processing-limit notice and identify the accepted batch before comparing counts. Do not assume that a file of any size was processed in full.
Count data records, excluding the header. A CSV field can contain a line break, so counting text lines is not the same as counting records. Preserve the input as received; do not remove repeated references just to make the totals agree.
Four steps from file to review record
Step 1
Keep the original
Save the input and returned file separately. Keep the accepted scope and any processing-limit notice with them.
Step 2
Count and pair records
Compare expected and returned records, then trace each one to its original row. Equal totals alone can hide a missing row and a repeated row.
Step 3
Keep exceptions visible
Read Status and Note alongside the given and matched addresses. A blank code or rate is not a zero tax result.
Step 4
Confirm and record
Use the official address source for current information. Keep the evidence, date and next action with your review.
Two fictional files, two different checks
All references and results below are invented for this walkthrough. They are not customer records or live address determinations. No addresses, PSD codes or rates are supplied.
Case A · All records accounted for
6 in, 6 out — with 4 exceptions
Assume this six-record batch was accepted and each returned row has been paired with the correct input record.
| Row | Reference | Status |
|---|---|---|
| 1 | DEMO-A | ok |
| 2 | DEMO-B | ok |
| 3 | DEMO-C | confirm |
| 4 | DEMO-D | not_found |
| 5 | DEMO-E | no_match |
| 6 | DEMO-F | error |
6 = 2 matched records + 4 exceptions.
The file is accounted for: six expected records, six paired results. Only two have an ok software match; the other four remain on the review list. Even ok is not an official confirmation of current rates or a payroll instruction.
Case B · Deliberately incomplete example
6 expected, 5 returned — investigate the gap
Imagine a different file containing only five of the six expected results. This is a teaching counterexample, not a reported failure of HireTaxNexus.
6 − 5 = 1 record not yet accounted for.
- Keep both versions and confirm which batch was accepted.
- Trace each returned record using its original row position, reference and input address.
- Find the unmatched input record and check any limit or correction notice.
- Leave the reconciliation open until the difference is explained. Do not fill a missing result with a nearby code or a zero rate.
A repeated reference needs tracing, too
If two input records share the same employee reference, keep both positions visible during review. One repeated label does not prove a duplicate, and equal input/output totals do not prove that the rows correspond.
Keep evidence beside each check
| Check | Evidence to keep | What it does not prove |
|---|---|---|
| Record count | Expected count, returned count and the scope accepted for processing. | That each expected record is present, or that every address is resolved. |
| Repeated reference | Original row position, reference and address; trace both occurrences separately. | That two rows are duplicates that can be deleted. |
| Given and matched address | The original address beside the returned match, with any difference noted. | That a similar-looking match belongs to the intended location. |
| Status and Note | Every unresolved row and its next action, including blank code or rate fields. | That an empty field means a zero rate or no tax obligation. |
| Source and date | The official confirmation used, when it was checked and who reviewed the discrepancy. | That a dated directory match is a current withholding instruction. |
The statuses are HireTaxNexus processing labels. no_match means the address was not located; not_found means no matching register row was found; error reports a lookup failure. Read the Note rather than treating these as the same outcome. The existing employee CSV preview explains the output fields and statuses.
For not_pa, no resident Pennsylvania code is returned. That does not determine obligations connected to a work address or Ohio municipal taxes. Keep the row visible rather than turning it into a tax-exempt conclusion.
Before handing the review to your payroll team
- The original file, accepted batch and returned file are preserved.
- The record counts agree, or every difference is explicitly left open.
- Each result is traceable to an input record, including repeated references.
- Unresolved rows, blank fields and required next actions remain visible.
- Official checks carry their source and date; the person reviewing the file can see what still needs confirmation.
This checklist records a review. It does not approve withholding, certify compliance or guarantee that a payroll system can import the file.
Official address check and scope
Pennsylvania DCED: PSD Codes and EIT Rates, consulted , links to the official address search. If that search returns no result, DCED recommends identifying the counties for the home and work addresses and contacting the appropriate tax collector or officer for the codes and rates.
The file-reconciliation workflow and fictional examples are our operational guidance, not a CSV procedure mandated by DCED. Confirm current address-level information with the official source and review tax decisions with a qualified advisor.