The board of Pennsylvania’s biggest pension fund adopted an inflated number for its investment performance even after the state treasurer raised skeptical questions about the calculation last summer, newly obtained documents show.

That decision by the PSERS board has emerged as a costly and disruptive mistake, raising the possibility that the $64 billion pension fund for teachers may soon have to hike their payments to support the mammoth but underfunded plan. The panel is to meet Monday to consider doing that.

In his August 2020 letter, then-Treasurer Joe Torsella raised doubts about a decision by the fund’s professional staff to go back almost a decade to revise — and improve — figures for past investment performance.

» READ MORE: PSERS and its troubles: A guide to the woes facing Pa.’s biggest pension plan

The staff’s move was no academic exercise. Under so-called risk-sharing reforms that govern state-funded pensions, PSERS — the Public School Employees’ Retirement System — must collect more from teachers in pension payments if returns fail to meet goals. So the calculation is closely watched.

However, the plan’s executive director, Glen Grell, immediately rejected Torsella’s worries as unfounded.

“The adjustments are not errors,” Grell wrote back to the treasurer. “Nor were there any errors by PSERS staff or consultants.”

The Inquirer obtained copies of their exchange last week by invoking the state’s right-to-know law. The letters shed new light on what was plainly a long-simmering internal controversy over the crucial matter of what official figure to adopt for investments.

Torsella, a persistent critic of PSERS’s management while a board member, noted in his letter that the fund’s staff was about to toss out its previous official performance figures over a multiyear period and replace them with new ones proposed by a consultant. In polite words, he said he found this “surprising.”

In reply, Grell said that the adviser, Chicago-based Aon Investment Consulting Inc., was merely updating the numbers with late-arriving financial data from some of the fund’s “private market” investments, money put into firms not traded on the stock market.

In December, the debates seemed to be resolved when the PSERS board overwhelmingly endorsed an investments figure based on the revised numbers — despite abstemptions from Torsella and two board allies.

The figure was official for three months. In late March, the board admitted making an “error” despite the earlier assurances by Grell and other senior PSERS staff that all was right.

» READ MORE: FBI probes Pa. school pensions amid furor over false investment report, sources say

PSERS staff have refused to explain the error or detail its import. But an Inquirer review of data in Grell’s memo suggests the revisions in December were just enough to spare teachers from having to pay more into the plan.

The board has hired two sets of outside lawyers to address the issue. Separately, the panel in recent weeks has engaged other attorneys to help it cope with an FBI investigation.

As debate continues to roil within the agency, people familiar with the issue say the board’s new outside lawyers, as well as the staff of Treasurer Stacy Garrity, Torsella’s successor, now say PSERS has little choice but to recalculate the rate, even if that should trigger higher payroll deductions.

But the fund’s management has dug in. Grell’s staff has prepared its own analysis to make a case against recalculation, the sources said.

PSERS spokesperson Evelyn Williams declined to comment, saying that the issues were under investigation.

» READ MORE: Pa.’s largest pension plan admits ‘error’ that may have benefited teachers at taxpayers’ expense

In their often technical exchange, Torsella asked Grell to explain why Aon’s figures for overall fund growth differed from those in the agency’s Comprehensive Annual Financial Reports. Torsella cited the “importance” of those annual reports and “the system’s reliance on their accuracy.”

At issues were the years between 2010 and 2019. While the differences were often tiny, the upgrade was sharp for one year, 2015. For that year, Aon increased the reported return for the massive fund from 3.04% to 3.41%.

That made a significant difference. Based on Aon’s new numbers, the board in December approved a report saying that the average return during all those years was 6.38%. This slightly exceeded the mandated 6.36% “hurdle” that would have forced a hike in payroll deductions.

According to the data in Grell’s memo, the earlier, official annual returns averaged 0.04% lower — enough to drop the average to 6.34%, below the magic number.