Census Testing Audits: What Auditors Should Review Under GASB Standards

Bottom Line Up Front

• When auditing pension and OPEB valuations under GASB standards, census testing is one of the most effective and practical ways to evaluate the accuracy of a valuation without becoming an actuary yourself. 
• Auditors should focus their review on high-impact census data fields such as birth dates, hire dates, Medicare eligibility, benefit eligibility, compensation (for pension plans) and special employee populations like public safety personnel and teachers. 
• A risk-based, materiality-driven approach to census testing can help auditors identify potential issues, perform meaningful audit procedures, and gain confidence in pension and OPEB valuation results.

AUDITING THE ACTUARY: WHERE SHOULD YOU FOCUS?

GASB standards for pension and OPEB valuations have placed an increased burden on the audit community to “audit the actuary.” What exactly that means remains somewhat ambiguous. It seems obvious that auditors should not need to become actuaries themselves to understand every aspect of an actuarial valuation. Yet auditors must perform procedures sufficient to demonstrate that the valuation has been appropriately reviewed. 

Two of the most productive and reasonable areas to focus on are: 

• Census testing
• Reviewing actuarial assumptions

This article focuses on census testing and the practical steps auditors can take to perform an effective review. 

Prefer video? Watch our Ask an Actuary discussion on census testing and what auditors should focus on when reviewing pension and OPEB valuations under GASB standards.

Step 1: Understand the Census Fields

Actuaries often use specialized software that requires specific data fields, and they may create additional fields for valuation purposes. While many fields are self-explanatory, such as birth date or higher date, others might not be. 

If there are fields on the census file that you don’t understand, it is entirely reasonable to ask the actuary what they represent and how they are used in the valuation process. Understanding the purpose of each field is the first step toward performing an effective review. 

Step 2: Verify that the population makes sense 

Before reviewing individual records, step back and consider the census population as a whole. 

A lot can happen during the process of gathering and consolidating data. For example: 

• A department’s census file may have been omitted. 
• Employees who are not benefit-eligible may have been included. 
• Certain groups or departments, such as retirees or teachers, may have been unintentionally excluded.

Reviewing overall employee and retiree counts can help identify these types of issues before they materially affect valuation results. 

STEP 3: PAY SPECIAL ATTENTION TO PUBLIC SAFETY AND TEACHER POPULATIONS

Many retirement systems provide different retirement eligibility provisions for designated public safety personnel (pro tip: it’s not just police & fire. As an example, lineman at Light Departments are easy to miss). As a result, failure to properly identify public safety status may materially understate liabilities and service cost. Verifying that these populations have been properly identified can be an important audit procedure. 

Teachers also warrant special attention. Although the impact is generally less pronounced than for public safety employees, teachers often have different mortality, termination, and retirement assumptions than other employees, which can materially affect liabilities. 

For employers with municipal utility districts or similar organizations, it may also be worth reviewing whether employees eligible for public safety retirement benefits have been properly coded. 

Looking for an OPEB-specific checklist? We got you! 9 Things to Review for Your OPEB Census Audit

THE MOST IMPORTANT CENSUS ELEMENTS TO REVIEW

Birth Date

Birth date is one of the most important components of any pension or OPEB valuation because age directly impacts liabilities. 

For active employees, incorrect birth dates may understate liabilities if employees are older than reported. For retirees and covered spouses, younger ages generally produce higher liabilities. 

It is often worthwhile to select five to ten active and retired participants and verify that the dates on the census file agree with supporting records. 

Spouse Birth Date

For retirees with covered spouses, spouse age can materially affect liabilities. 

When actual spouse dates of birth are unavailable, actuaries frequently rely on assumptions such as assuming male spouses are three years older than female spouses or that same-sex spouses are the same age. While these assumptions are reasonable, actual spouse dates of birth are preferable and can materially affect results when spouses are significantly younger. 

Hire Date

Hire date is another critical census element because many pension and OPEB plans require minimum years of service before employees become eligible for benefits. 

If prior credited service is omitted, liabilities may be understated. Conversely, if an employer uses a more recent date due to promotions, transfers, or position changes, employees may incorrectly appear ineligible for benefits. 

A common issue occurs when employers update hire dates whenever an employee changes positions or departments. From an actuarial perspective, the relevant date is generally the employee’s original hire date as benefits are earned/accrued over the entire working lifetime. This normally should align with the date recognized under the pension system. 

Medicare Eligibility

Medicare eligibility status has a significant impact on OPEB liabilities. 

Retirees who remain enrolled in active healthcare plans after age 65 often have liabilities that are three to four times higher than retirees enrolled in Medicare supplement plans. Therefore, if retirees over age 65 remain coded as participants in active plans, it is worth verifying that the information is correct. 

Employers may also identify employees who are expected to be ineligible for Medicare at age 65, such as certain teachers hired before March 1986. However, these cases can be difficult to verify because individuals may have earned Medicare credits through other employment. 

Medical Coverage Level for OPEB Plans

For retirees, the level of medical coverage elected can materially affect liabilities. 

In particular, whether a retiree has covered dependents or a covered spouse can significantly change projected healthcare costs and should be verified when possible. 

Compensation (mainly for Pension Plans) 

While not generally important for OPEB plans (there are some pay based life insurance benefits), the large majority of public sector pension plans have pay related benefits (often a final three or five year average). So, it’s important to validate the pay is correct and also that the pay provided matches the plan’s compensation definition (e.g., should bonuses, overtime, and certain other pay types be included or excluded). 

Form of Payment for Pension Plans 

While we care about “coverage level” for OPEB plans, a key data element for retirees in pension plans is the form of payment (e.g., life annuity, joint & survivor, certain & life, cash refund, etc.). This along with the spouse birth date noted earlier can have a material impact on the pension liability for a retiree. 

Benefit Eligibility 

Some employers provide census files that include all employees, including those who are not eligible for retiree benefits. 

Failure to identify employees who are not benefit eligible can result in overstated liabilities. Therefore, auditors should confirm that only employees eligible for benefits are included in the valuation population. 

REMEMBER: IT’S ABOUT MATERIALITY

As with any audit procedure, materiality should guide your review. 

Birth dates that differ by a few days, months, or even years may not materially impact liabilities. However, if errors appear to be systemic, they should be investigated further. 

Similarly, the size of the employer and the census population will influence whether a particular anomaly is material. 

If questions arise regarding the impact of a census issue, auditors should consult with the actuary to determine the size of the issue, its materiality, and whether a revised valuation report may be necessary. 

FINAL THOUGHTS

Auditors should not be expected to become actuaries in order to review pension and OPEB valuations. However, performing thoughtful census testing is one of the most effective ways to gain confidence in valuation results and satisfy audit requirements under GASB standards. 

By understanding the census data, verifying key demographic information, reviewing special employee populations, and focusing on material items, auditors can perform meaningful procedures without becoming valuation experts themselves. 

If you have questions regarding a particular census issue or would like to discuss best practices for pension and OPEB census testing, the actuarial team at Odyssey Advisors is always happy to help. 

Frequently Asked Questions

What is census testing in a pension or OPEB audit?
Census testing is the process of verifying the accuracy and completeness of participant data used in an actuarial valuation. Auditors review key demographic and eligibility information to gain confidence that pension and OPEB liabilities are reasonably stated under GASB standards. 

Why is census testing important when auditing an actuarial valuation?
Actuarial valuations rely heavily on participant data. Errors in birth dates, hire dates, benefit eligibility, Medicare status, or employee classifications can materially impact reported pension and OPEB liabilities, making census testing a critical audit procedure. 

What census data fields should auditors review?
The most important fields typically include: 

• Birth date
• Hire date
• Benefit eligible status 
• Medicare eligibility
• Spouse date of birth
• Medical coverage level 
• Compensation (mainly for pension plans)
• Form of payment (the retiree’s benefit options for pension plans)
• Employee classification (such as public safety or teacher status) 

How many records should auditors test during census testing? There is no universal number. Sample sizes should be determined based on the size of the population, audit materiality, risk assessment, and applicable audit standards. 

What should auditors do if they identify census errors? Auditors should discuss the issue with the actuary to determine the impact on liabilities, evaluate materiality, and determine whether additional testing or a revised actuarial valuation may be necessary.

What documentation should auditors use to verify census data?
Auditors should compare census data to reliable source documentation, such as payroll records, personnel files, benefit enrollment records, retirement system records, or other supporting documentation maintained by the employer. The appropriate source will depend on the data field being tested.