The CalPERS board has decided to fight the inevitable by backing CEO Marcie Frost even though her days are clearly numbered. The board’s Performance, Compensation & Talent Management authorized a 4% salary increase and a $85,000 bonus payment yesterday which is sure to be ratified by the full board today.1
But as our story today shows more and more people are disturbed that Frost continues to get away with misrepresentations and performance exaggerations that go back much earlier than her time at CalPERS. People who have worked closely with Frost for many years in Washington are coming forward with first hand accounts of her poor performance and cover-ups.
Two employees of the Department of Retirement Systems (DRS), the back office/call center portion of the Washington state pension system where Frost worked for over a decade, contacted me to confirm the accuracy of our reporting on Marcie Frost and voice their concerns about her conduct and performance in Washington and its implications for CalPERS. As one put it, “The taxpayers of California need to know the truth.”
Both individuals spent many years at DRS and their tenures overlapped with that of Frost as Executive Director. Each at different times had substantial direct interaction with her and went to meetings and conferences with her. Each of them had also worked outside DRS. Through either direct experience at other agencies or having worked in interface roles with other agencies, they were also confident of the information one provided and the other was able to confirm about Frost’s claims regarding her work at her employer prior to DRS, the Department of Labor & Industries.2
The bottom line is that Frost is a poor manager who is nevertheless, as one put it, “cunning” and very skilled at self promotion. Each source described multiple instances of how Frost took credit that was not warranted for her work early in career and grossly misrepresented her initiatives at DRS, falsely depicting efforts that were either failures or never completed as successes.
As one stated:
Marcie ran the agency like a high school student body – unprofessional, catty, and a con game. She was an incompetent and dysfunctional manager.
As the other wrote:
Everything you say in the CalPERS articles is spot on about Frost. She highly embellishes everything she touches and is untrustworthy to her employees and her organization for her own personal gain, and regularly throws people under the bus. Not unexpectedly, she schmoozes well with upper management. Many people I know would not recommend her for the CalPERS job and were actually surprised to hear she got the job.
The specific issues we will cover are Frost’s poor management decisions, her misrepresentation and exaggeration of her supposed accomplishments, and her toxic management style.
Poor management decisions. After Frost returned to DRS in 2010, she decided to reorganize its call center. As you can imagine, the call center is the key interface between beneficiaries and DRS. It is not just their first but often their only channel for getting questions answered and problems resolved.
As readers may recall, DRS handles roughly 15 separate defined benefit pension plans, including those of teachers, judges, and firefighters, along with a defined contribution plan. Each of the defined benefit plans has its own terms and they are often complex. Historically, each plan had its own dedicated call center reps who were familiar with the idiosyncrasies of the plan they handled. The system had worked well for 20 years.
Frost decided to end the specialized call center teams and have every call center representative handle all plans. This radical change was implemented without giving the staff much in the way of additional training or upgrading the caliber of the staff. The apparent aim was to improve efficiency.
The initiative backfired. Not only did wait times rise and beneficiaries increasingly wind up having their calls go into voicemail, but even worse, the overloaded call center workers were making significant errors more often. One of several examples provided was of a retiring employee who had over $10,000 of accumulated vacation which would by default be paid out as a lump sum. Instead, he asked (as was permissible) to have it treated as deferred compensation and rolled over into his defined contribution account at DRS. Even though the call center agent said she was going to implement his request, the funds were paid to him, with 20% withheld, resulting in him paying more in taxes than necessary. He was not able to get DRS to reverse this transaction.
The misguided combining of call center operations also hurt other beneficiary service efforts. DRS also had a walk-in customer service center, where wait times had seldom been longer than ten minutes before Frost’s reorganization. The queues increased greatly.
What did Frost do when confronted with these problems? She did not attempt to fix them, say by increasing training of the call center staff or undoing or partly undoing her reorganization (one option would be to re-create dedicated teams for the 3 or 4 biggest plans and have the rest handled by another group). She instead resorted to scapegoating and external image management.
Frost pressured the managers of the call center, getting them to write up the front line staff for their supposed performance failures, based on new metrics Frost had implemented, when in fact the new system imposed unreasonable demands on them. As one put it:
People in the call center were crying. They even left when they had no new job lined up. But Frost cherry picked data to present to the Governor [Jay Inslee], even though it was an utter failure.
Needless to say, high turnover in an operation where staff performance problems result in large measure from them lacking sufficient knowledge to execute a greatly expanded job scope is a recipe for continued poor results and higher expenses. New employees are guaranteed to need time to learn the required information and are likely to make mistakes. Hiring staffers in and of itself involves costs, plus new employees require training, yet another cost.
And the charge of cherry-picking data is credible since Frost has been doing that at CalPERS.
At the end of this post, we’ve embedded the slides from a presentation at the July 2018 offsite from the presumably most persuasive, two case studies Frost could show to demonstrate that her implementation of “lean” management techniques was paying off.
If you don’t know any better, the slides look compelling. The problem is the underlying work is bogus.
We’ve embedded the slides from the first of the two examples at the end of the post. They show an egregious case of data distortion.
CalPERS implemented alleged “lean” changes in its death benefits processing and the slides depict CalPERS as having achieved marked improvements, such as reducing the number of cases that took more than 45 days to process.
Why are the touted results clearly the result of cherry-picking?
CalPERS used an obviously cooked-up basis of comparison. Rather than take the same time period in successive years, CalPERS instead chose a set number of cases to examine (300 each) before and after a suspiciously-artbitrary-looking cutoff date, February 12. Under questioning, the presenters admitted the “before” cases included ones submitted in November and December.
Why does this matter? The beginning of November through end of January is certain to be the worst time of year in terms of efficiency for a government agency. First, you have a high density of holidays compared to the rest of the year (Veterans’ Day, Thanksgiving, Christmas, Martin Luther King Day). Output suffers due to distractions like holiday shopping, more interaction with family members, and even getting out of the mood to work. Second, many employees also take vacation days around these holidays (and recall that CalPERS employees have generous vacation allowances). So there was also almost certainly reduced manpower to process claims during this period.
Finally, even if this comparison had been constructed properly, a single iteration proves nothing. Any apparent improvement could be the result of sheer happenstance. You would need to have multiple properly-contructed comparisons before you could conclude anything.
The second case study, on purported improvements in the processing time of disability benefits, suffers from similar deficiencies. Again there is only one performance comparison. And there is no control for quality, such as proportion of claims rejected. Any improvement in turnaround could be due to rejecting more difficult cases, the same way that the fastest way to reduce average time per call in a call center is to hang up more often.
Even third party evaluators who have reason to be as nice as possible to CalPERS can’t do enough in the way of porcine maquillage to hide what is going on. The normally very CalPERS-friendly Chief Investment Officer Magazine reported CalPERS Gets a D for Fixed Income Management, But B For Overall Program:Consultant Wilshire Associates also gives overall CalPERS leadership a C grade. If you click through and read the article, the reason for giving the giant pension fund a “D” in fixed income (vacancies in too many key roles) applies even more so to private equity, which accounts for more than half the investment risks and fees CalPERS incurs. So how exactly does that add up to an overall “B” grade, with leadership also getting a “C”?
Misrepresentation and exaggeration of accomplishments. The call center fiasco shows Frost misrepresenting a terrible outcome as a success. This is not an isolated case. Frost has a pattern of exaggerating results that goes well beyond acceptable apple-polishing.
Frost has represented herself repeatedly at conferences in Washington, and presumably also to CalPERS, as an expert in migrating systems off mainframes to desktops. That is flat out false:
DRS’ reliance on mainframes was not seen as a good thing. Marcie was going to all these IT conferences telling the audience how we’d gotten off them. The IT staff from DRS at them would be poking each other in the ribs about her statements.
We did have a huge number of projects to get off the mainframe, but we haven’t made any substantial progress. 90% to 95% of the IT processing power at DRS is still on the mainframe.
Another clear misrepresentation was one we cited on our website, that of Frost claiming she had implemented the first human resources IT system in the State of Washington:
Washington was one of the first states to adopt computer processing in the 1960s and 1970s, as you wrote earlier. It was her embellishment to claim that she had anything to do with a first implementation. And Labor & Industries [the department she worked in] was not the one that developed human resource systems. That was the Department of Personnel.
The most she would have been responsible for was training for the rollout of a Department of Personnel systems program.
The only other thing she might have referred is Labor & Industries working on their own version of a leave system (sick leave, vacation leave, etc) using the enterprise Department of Personnel data. But the creation of that interface would be considered a worst IT practice, not a best IT practice. It would be like a vendor modifying a Microsoft product.
So predictably, that Labor & Industries system has had ongoing problems, even though the people at Labor & Industries would try to deny that. Employee leave balances were many times out of sync with the official Department of Personnel leave balances. This would result in people taking vacation leave for example, when they did not have the official balance to take the time off. It became challenging to manage at L&I. On top of that, when the Department of Personel made a change to their enterprise system, L&I had to scramble to change their system. It was an ongoing maintenance challenge.
Toxic management style. Frost was an abusive, scapegoating boss:
Marcie said she believed in six sigma type practices, such as the the “leader-servant” style of management, where low-level employees had a lot of authority, just like factory workers at Toyota could stop the production line if they saw something wrong happening.
But what she would say and what she would do were two different things.
She would regularly discredit and embarrass junior people before senior management. Morale plunged. Frost also is politically well connected and vengeful, so her victims and people who saw this sort of thing happening and were unhappy about it felt powerless to do anything about it.
If you look at LinkedIn, you can see a significant number of bios on it with short times of employment at DRS, which is not what you’d expect from a government agency, since staffers get good benefits and have strong incentives to stay in order to get a pension. These insiders confirmed that DRS suffered from a high level of departures under Frost:
Turnover was much higher under Marcie since Marcie returned to DRS from WSECU around 2010. That is when Marcie reorganized the call center to “everyone does everything,” and people started getting written up. People were stressed out and leaving at higher rates than in the past. I am pretty confident you could confirm this from information you could get by FOIA, but Marcie admitted to it by her aggressive attacks on managers at DRS.
Marcie blamed the turnover on the the human resources management and secondarily on the call center managers, accusing them of not creating a supportive environment. In actuality, the front-line phone reps were overwhelmed by the new demands and there was nothing their supervisors could do to change that.
The higher turnover numbers were numbers that were reported to Governor Inslee on his performance measures so Marcie refused to take responsibility for the higher turnover and blamed the HR management. The head of HR eventually quit, and most people I know at DRS believe that it was directly due to Marcie’s unfair criticism and pressure.
The reports also suggest that the level of CalPERS employees contacting me about Frost’s misrule is certain to rise. California civil servants are well protected from dismissal, and it would not be difficult to provide evidence of problems at CalPERS without exposing confidential information. So keeping Frost at the helm will guarantee a continuing flow of bad news about CalPERS.
1Frost’s salary level last year was $318,000 and her bonus was $80,000.
2 The sources insisted on anonymity out of a belief that Frost would seek to retaliate and still had the connections in Washington to be able to do so. A third potential source backed out due to a stated fear of Frost’s vindictiveness.CalPERS data doctored lean case study