
Costs
Part of Corporate responsibility: what you owe, not just what you give
Corporate Responsibility Checklist: 12 Pre-Board Audit Tests
A [corporate responsibility](/corporate-responsibility) checklist for finance or internal audit: 12 evidence tests to run before a board review or report.
A corporate responsibility checklist for finance or internal audit should test the records behind the program, not just the language in its report. These 12 pass-or-fail checks focus on traceability: whether an effect, commitment, payment, grant, or public claim can be followed to a dated document and an accountable owner.
Run them before a board review or publication, and keep the evidence file so another reviewer can repeat the checks. A pass means the record can be produced and checked; a policy statement alone is not evidence.
Set up the review
Ask for the working register, the records used to prepare the latest report, and the relevant finance or program documents. Record the version reviewed and the person who supplied each file.
For every test, mark pass or fail and note the document that supports the result; if the document is missing or cannot be reconciled, record a fail rather than filling the gap with an explanation.
The checklist is not a rating of whether the company is responsible. It is a review of whether the program's own records support what the company tracks and says. Do not combine the results into an average that conceals a missing register or an unsupported public claim. Use the failed item to identify which file, owner, or control needs follow-up.
Check the effects register
1. Can you identify the register's coverage? Pass if the file states the period and business activity it covers, identifies its scope, and records exclusions. Fail if a reviewer cannot tell which operations or groups the register was meant to include.
A list without a defined boundary is hard to check against the rest of the evidence.
2. Can you trace a register entry to its source?
3. Can you tell when and why the register changed?
4. Is each entry assigned to a named owner whose incentives do not control the answer?
Effects register: four tests
- List of effects on non-customers and non-staff
- Ranked by effect on them, not risk
- Every effect has a named person
- Owners whose pay does not depend on answer
Trace costs, missed commitments, and supplier payments
5. Can you trace the cost of a commitment through the records? For commitments that require spending or resources, follow the stated commitment to its approval and finance records.
Pass if the records show where it was authorized and how it is tracked; fail if the program describes a commitment but cannot identify the corresponding record or accountable owner. This test checks the audit trail, not whether a particular decision was morally or commercially correct.
6. Are slipped commitments recorded as exceptions?
7. Can you calculate the gap between agreed and actual supplier payment dates?
Cost tests: three items
- Last decision choosing the more expensive option
- What happened to commitments that slipped
- Payment terms actually met for smallest suppliers
Reconcile giving and communications records
8. Are giving records kept distinct from records of operational effects? Check the accounts, program lists, and responsibility-report drafts. Pass if giving is separately identified and the report does not use those records as evidence for effects caused by operations.
Fail if donations or volunteering are blended into a response about operational performance, or if the same item appears in both categories without an explanation.
9. Do grant records show the terms and stated purpose?
10. Is the communications spend-to-program spend ratio known and defensible?
Giving: three tests
- Giving kept separate from effects
- Shape of the givingmulti-year, unrestricted
- Communications spend against program spend
Verify claims and place-decision files
11. Can a reviewer substantiate a sample of published claims? Select three statements at random from the prior year's published material. For each, check that the basis is identified as measured, estimated, modeled, or reported by someone else, and that a supporting document exists.
Fail if supplier self-reporting is presented as independently established fact, or if the methodology was selected after the result was known. The Federal Trade Commission's policy statement on advertising substantiation expects advertisers to have a reasonable basis for claims before they are disseminated.
For environmental claims, the same evidence discipline applies: identify the basis of the statement and retain the supporting record. The related discussion of what can be claimed about environmental performance is in sustainability ethics. The FTC's Green Guides explain how marketers can substantiate environmental marketing claims.
12. Is there a traceable file for the last decision affecting a place? For a closure, relocation, or large redundancy, check whether the file records the decision chronology, the stated reason, the approval trail, and what was documented as left behind. Pass if the record can be assembled from dated documents and the people responsible are identifiable. Fail if the file consists only of a final announcement or an account written after the decision, with no supporting chronology.
This is a records test, not a legal review of consultation or notice requirements. Those requirements vary by jurisdiction and circumstance; the checklist does not determine what applies. A closure can also raise ethical dilemmas that a records test does not settle.
Keep any legal review separate from the audit finding about whether the company's own decision file is complete. For a worked example of a program, see the corporate responsibility case study.
Claims and places: two tests
- Published claims you could evidence if asked
- Last decision that affected a place
Read the results and assign follow-up
| Failed items | What the review establishes |
|---|---|
| 1–4 | The register's scope, sources, update history, or ownership cannot be reliably checked. |
| 5–7 | Cost records, missed commitments, or supplier payment timing lack a reproducible trail or accountable owner. |
| 8–10 | Giving, grant terms, or communications spending cannot be clearly reconciled to the program records. |
| 11–12 | The basis for published claims or the file for a place decision is incomplete. |
Reading the failures
Failing items
- 1 to 4
- No register, no program to assess
- 5 or 6
- Program has never constrained a decision
- 7
- Financing yourself with others' working capital
- 8 to 10
- Giving answers a question it does not answer
- 11 to 12
- External account and internal reality separated
Reading
- 1 to 4
- 5 or 6
- 7
- 8 to 10
- 11 to 12
For each failure, note the missing document or reconciliation, the person responsible for supplying it, and the next review point.
Keep the original fail on the record if a document arrives later; note the follow-up rather than rewriting the result as though the evidence had been available during the review. A repeated failure may indicate that the recordkeeping process needs attention, not simply that one file was misplaced.
Finance or internal audit should run the review rather than the team whose report is being tested. That reviewer should be able to request source documents and compare them with the program's own records.
The checklist can then support a more specific board discussion: which statements have evidence behind them, which records need correction, and which items remain unresolved. Keep the working file available to the next reviewer so the checks can be repeated. Where a failure points to a design gap rather than a missing file, the corporate responsibility framework is the place to look.







