The Importance of Good Data for DB Plan PRT Activities

While clean data is important for identifying deaths and locating participants and beneficiaries, it can also have a great cost benefit for plan sponsors implementing pension risk transfer activities.

It’s important for plan sponsors to have clean data on defined benefit (DB) plan participants to help them identify and validate deaths, locate participants and beneficiaries, and manage uncashed checks, according to speakers at a webinar, “The Data Dilemma: The Impact Bad Data Has on a Pension Plan,” presented by Pension Benefit Information (PBI) Research Services and DIETRICH.

Mike Irey, director of operations at PBI Research, told webinar attendees that plans can have inaccurate or missing personal identifiable information (PII) when data in older administrative platforms does not get updated or get carried onto a new system, or when data gets input incorrectly. A PBI Research study found 5% of all participant or beneficiary data is likely inaccurate or missing.

He said missing or inaccurate Social Security numbers can have the biggest and most severe impact, because they are often used to verify other information about a person.

“Plan sponsors can more easily clean up other data if they have the right Social Security number,” Irey said. “And to correct a Social Security number, the plan sponsor needs an accurate first and last name, date of birth and location—usually city and state.”

According to Irey, PBI Research found that having an inaccurate or missing Social Security number resulted in the highest likelihood of missing a participant’s death. “A study we did found a missing or inaccurate Social Security number resulted in 90% fewer found deaths,” he said.

Bad data can cause overpayments, added Geoff Dietrich, executive vice president at DIETRICH, especially if a plan sponsor doesn’t know a pensioner has died. He said doing regular death checks is a best practice.

It is easier to correct first and last names than other information, Irey said. He noted that the most common reason last names are wrong is an unreported marriage. “People are more likely to open communications if it has the right name on it,” Irey said.

Further explaining the importance of accurate PII, he noted that having a correct date of birth is not as big an issue when trying to locate a participant or beneficiary, but it makes verifying a death more difficult. A missing or inaccurate address is the least impactful dilemma, and the easiest to correct, he said, but a plan sponsor might have to spend extra money if an erroneous address results in multiple mailings.

Irey suggested plan sponsors use commercial databases to clean up data. Responding to an attendee question, he said about 90% of participant or beneficiary deaths are found using the Social Security Administration’s Death Master File (DMF)—and they are usually found within the first four days of death. However, Irey said the highest percentage of found deaths are through obituary searches. “We found a 0.08% error rate using the DMF, but our obituary search has a 0.05% error rate,” he said.

Clean Data for PRT Transactions

While clean data is important for identifying deaths and locating participants and beneficiaries, Dietrich told webinar attendees that his firm sees the application of data every day with pension risk transfer (PRT) activities—i.e., the offering of a lump-sum distribution window or the purchase of an annuity. “More benefits are settled via lump sums than annuities, and, with annuities, the transition to an insurance company is for the life of the benefit, so having good data is important,” he said.

When it comes to PRT activities, bad data can mean plan sponsor staff members have to spend more time cleaning up data or trying to communicate with participants and beneficiaries. It can also result in extra cost for mailings, project delays and even contract delays, Dietrich explained. “Time is money. A delay of communicating with participants can delay the contract with an insurer,” he said.

Bad data can also affect liability calculations, which are based on a participant’s date of birth and life expectancy, Dietrich added. “If you think someone is a certain age and they’re 10 years younger, a correction will come at a cost,” he said.  

Having correct participant and beneficiary addresses is important to ensure the receipt of required communications, Dietrich noted. He said a lump-sum distribution can be 10% to 40% less costly for a plan sponsor than an annuity purchase, so sponsors want to get the highest take-up rate they can because it will save the plan meaningful dollars in an annuity purchase. “If you don’t have the right address, the participant can’t take a lump sum, so you will either have to include that person in the annuitization of liabilities or hand that person over to the PBGC [Pension Benefit Guaranty Corporation],” he said.

In addition, missing or unresponsive participants will limit the number of insurance companies that will bid on a plan sponsor’s liability, resulting in less choice and potentially higher costs, Dietrich said.

Knowing participants’ locations can save plan sponsors on PRT costs another way, Dietrich explained. “Over time, insurance companies and other data miners have gotten better at estimating how long participants will live. That comes into play with the price to purchase an annuity—insurers will estimate how long they will have to insure liabilities,” he said. “Today, insurers look at life expectancy on a more granular basis using [a participant’s] ZIP code or ZIP+4. Having that data will improve the costs for annuitizing liabilities. The less information insurers have, the more conservative they will be, which will cost more for plan sponsors.”