Library audit: five findings taxpayers should see

Nearly $704,000 in the capital account. A wish list that includes a bookmobile, office redesign and lobby study pods. Financial reports that cannot keep their reporting months straight.

Berkeley Heights taxpayers deserve answers before the Library makes another big commitment.

Our November 5 review of available Library records covered 98 primary files and assessed all 23 capital-plan components. It found concrete reporting defects, unresolved payment questions and spending proposals that need a much stronger case.

This was a review of available records, not a new independent CPA audit or bank-confirmed fraud examination. It did not establish fraud or a dollar amount that can immediately be returned to taxpayers. Here are the five findings that matter most.

1. Capital keeps growing while expensive ideas await a convincing case

The capital account's October 31 book balance was $703,678.28. The Board moved $550,000 from operating funds into capital in 2024 and 2025.

Meanwhile, the budget presentation estimates $150,000-$200,000 for a bookmobile, $50,000-$100,000 for offsite lending depots, $100,000-$150,000 for an office redesign and $70,000-$100,000 for lobby study pods.

Together, that is $370,000-$550,000 in proposed concepts. These are estimates, not verified construction contracts or guaranteed savings.

The Library opened in the municipal complex in 2021. A broad office makeover roughly five years later needs a clear explanation. Before buying pods, show that existing rooms are regularly full. Before buying a bookmobile, test a modest delivery service for residents who cannot reach the Library.

The fix: Defer these four concepts until each has documented demand, cheaper alternatives and a full operating-cost estimate. Protect necessary repairs, accessibility and reliable core systems. Review legally releasable reserves and pursue approved transfers for municipal property-tax relief rather than automatically funding another project.

2. Budgets repeatedly exceed actual operating costs by large amounts

The audited 2025 expenditure budget was $1,713,525.31. Actual operating expense was $1,388,204.19. The gap was $325,321.12, or 19%.

For 2024, the comparable gap was $515,881.20, or 28.3%. These comparisons separate transfers to capital from the cost of operating the Library.

Vacancies, delayed work and other changes can explain underspending. November 2024 minutes specifically describe cash accumulating when positions were vacant or filled for only part of a year. But recurring gaps this large call for a more realistic budget.

The fix: Build the next budget from recent actual costs, documented staffing changes and separately priced projects. Explain large variances publicly. An unused allowance should not become a reason to find something else to buy.

These budget gaps are not another pot of cash to add to the capital balance. A tax-relief calculation must avoid counting the same money twice.

3. The monthly financial reports contain reproducible errors

The September financial workbook's salary total omitted September's $59,064.03. Its operating-expense formula omitted September while including October. The combined total mixed nine months of salaries with ten months of operating costs.

That makes the displayed totals unreliable for judging spending through a single date. A zero in the Township-paid expense line also does not mean the Library incurred no benefit or shared-service costs.

These are reporting defects, not evidence that those amounts disappeared.

The fix: Use one reporting cutoff. Reconcile the packet to the ledger, payroll, benefits and bank records. Have a second person check the formulas before the Board relies on the numbers, and publish dated corrections.

Taxpayers should be able to read a financial report without reconstructing it.

4. Payment records raise specific questions that need bank and vendor answers

The review found four check references reused for different payees within the same main-bank records, across 2025 and 2026.

That weakens the payment trail. It could reflect numbering or void-and-reissue errors. It does not establish that duplicate checks cleared.

There is also a narrow subscription-account question. Creativebug's stated three-year contract totals $2,640, payable in three $880 installments. Recorded payments total $3,520, leaving $880 above the stated whole contract to explain.

A credit, extension or other account history could resolve it. The records do not yet prove an overpayment.

The fix: Obtain cleared-check images, bank reconciliations and the issue/void register. Get the complete Creativebug account statement, payment applications and any credits or amendments. Recover any confirmed overpayment and report the result.

5. Professional-service spending needs a clearer price, purpose and paper trail

October trial-balance postings show $65,572.08 in professional services against a $45,000 category. Budget amendments, classifications, outstanding bills and insurance responsibility must be reconciled before calling this an unauthorized overrun.

There is a concrete paperwork defect, too: the June legal-services resolution names Wiley Malehorn Sirota & Raynes in its heading and recitals, but names Bibliotheca in the actual authorization paragraph. June minutes separately approve Wiley. The Board should correct or clarify the record.

The fix: Publish a simple contract schedule showing each engagement's purpose, rates, cap, amount paid, remaining obligations and delivered work. Separate necessary legal and purchasing-compliance services from optional consulting, and explain any overlap with paid staff.

The same value test belongs on subscriptions and promotional purchases. The new LinkedIn Learning commitment is $7,125 for three years, not one. Yet comparable April-September reports show course views falling from 116 in 2025 to six in 2026. Check access and reporting definitions, then require a credible renewal case.

Better service, with fewer expensive experiments

Put the budget to work where residents can see the benefit. Use waiting lists to guide book purchases. Help patrons use digital services already purchased. Compare optional subscriptions with state and MAIN resources before renewing them.

Protect successful children's and literacy programs. Review repeatedly weak-attendance events individually. Track use of the self-checkout equipment already acquired and determine whether it frees staff for more personal help or reduces overlapping coverage.

Staffing deserves a workload review before adding positions or automatically filling vacancies. But the reported 43,804 front-door events are not verified Library-only attendance; the counter's physical location remains unconfirmed. And the 24 employees in the 2024 survey included four full-time and 20 part-time workers. The 2025 staffing measure was close to its population group's average. Those figures do not justify a specific job-cut target.

Measure demand over 60-90 days, including desk work, reference help, program preparation and opening-hour coverage. Where excess coverage is demonstrated, reduce it through scheduling changes, consolidation or attrition.

The rent question deserves legal review as well. The signed agreement provides rent-free space. The Township should determine whether any lawful rent arrangement is possible, respecting restrictions on Library funding. First, publish and enforce the actual utility, cleaning and shared-expense obligations. Moving money between two public accounts only helps taxpayers if combined costs or levies fall.

New Jersey's excess-funds process offers a route to municipal property-tax relief, subject to operations, restrictions, commitments and State Librarian approval. It does not authorize an automatic refund of the entire capital balance. The statutory minimum Library levy also remains.

Before approving another six-figure purchase, show residents the problem it solves, the cheaper alternatives and the bill for keeping it running.

Records and corrections

Read the complete 28-page records review. It identifies source documents, page and cell references, limitations and the records needed to close unresolved findings. It also explains the Library's existing independent audit opinion and why accounting errors, poor-value decisions and fraud are different claims.

The June 26 Library article has been corrected to withdraw mis-mapped marketing, postage, summer-reading and Reimagining spending figures and misleading visit-based comparisons. The later records support specific oversight questions; they do not support those earlier claims.