Retirement Committee

DRAFT Burlington Retirement Committee Meeting of the Committee Members Agenda August 19, 2014 5:00pm – 7:00pm BCA Art Gallery 2nd Floor Conference Room 135 Church Street, Burlington Vermont 05401 Present: Mayor Weinberger, Councilor Bushor, Councilor Mason, Joe Keenan, John Federico, Susan Leonard, Jim Strouse, Bob Rusten (facilitator), Mike Flora, Eileen Blackwood, Susan Leonard Absent: Councilor Paul, Councilor Knodell, Bob Hooper, Jeffrey Wimette, Bill Rasch 5:00pm – 5:05 pm Approve Agenda Councilor Bushor moves to approve, Jim Strouse seconds Review and Approval of August 5, 2014 meeting minutes Councilor Bushor moves to approve, Councilor Mason seconds 5:05pm – 5:10pm Public Comment Linda Blanchard – questions about Committee work related to current retirees or near retirees Ron Ruloff – concerns about the Committee’s work 5:10pm – 6:30pm Review of key problem areas and ideas identified by the Committee on August 5 All, with Keith Brainard Bob Rusten: What are the parameters for what we are trying to achieve – we have talked to about a number of parameters, including national norms, predictability, attracting employees, the definition of a healthy system, and risk-sharing, in addition to the original five goals the Committee discussed.1 My understanding of today is that we are going to focus on the “what” – what the Committee hopes to achieve – and in the next meeting the “how,” as in the way we get there. Keith Brainard: That fits my understanding. I have a number of questions the Committee should contemplate as we discuss the “what”:  Who will be affected? Three categories to consider. First, new hires only? Second, new hires and currently active (and if so how close to retirement)? And third, retired members in addition to new employees and current members.  The goal: Trying to reduce cost, risk level, volatility (cost to the City), etc. 1 For reference, those original five goals are: Taxpayer contribution level, ability of the City to recruitment and retain good employees, growing unfunded liability, voter support for the system, and complexity of the system. 1  DRAFT At this point, I am going to focus my comments on the problems identified by the Committee on August 5, excluding for the moment the first five on the sheet I received [please see footnote #1 below] that have been the subject of our discussion this far. Starting then with #6 – 6. The City currently bears a disproportionate share of the risk (specifically mortality, market, and inflation risk): After reviewing your retirement system, I agree that in the system’s current arrangement puts a disproportionate share of risk on the City. This seems to be a primary concern that the Committee is trying to address. 7. Troubling some retirees earn more than when working – a. the concern here is specific to individuals retiring normally but receiving compensation in excess of their final average salary. b. This is an issue that can be addressed in plan design, for example through a cap on the salary (I believe this is done in the State of Vermont, though I do not know if this is the best approach – a cap of 60 or 70 percent of final salary can encourage an employee to leave…there may be better ways to approach that problem) 8. Instability of the plan – the constant discussion of retirement issues strains relationships a. A problem around the country, and it appears to be a real issue in Burlington and my hope is this is something the Committee can address and resolve 9. Uncertainty about the system’s future creates fears for employees a. When considering New Brunswick example, the community felt the old system was precarious and needed to take action to preserve solvency. There are solid reasons for everyone to have a system that is, and is perceived to be, financially sound – recruitment, employees wanting a stable retirement b. Bob Rusten, interjecting: #9 also identified by the Committee in the context of #8, where there is uncertainty created by continuing discussions about possible changes to the retirement system. c. Keith Brainard, continuing: Yes, that makes sense. 10. Time at which employee is asked to choose multiplier a. The Burlington feature that permits/requires employees to choose multiplier at retirement is unique in all the different retirement systems that I know of – I’m not certain what the purpose of this is, and I’m unsure what the utility of it is. It adds uncertainty to the system, and the City Council may want to create a system where the multiplier is known earlier and this reduces some of the uncertainty. b. John Federico, interjecting: I believe changes made post 2006/2007 have already made this adjustment (Susan Leonard agrees it was that timeframe) 11. Burlington deviates from a national norms on a number of dimensions (COLA, etc) a. The concern here is not just or not primarily that you have strayed from national norms (which is not always a bad thing on its own), but that Burlington has strayed from the national norm on a number of dimensions and at the same time has seen substantial growth in its unfunded liability 12. Lack of clarity on the goal of the retirement system – what should it actually do? a. For the employer to attract and retain, for taxpayers to have public service delivered in an effective and professional manner, and in a cost-effective manner. This is an issue that the City Council and BERS should spend some time discussing 13. Impact of past decisions a. The present state of the plan is the product of past decision – plan design, asset allocation, etc. b. The extent to which BERS or the City Council want to make changes to those who are currently in the plan will be a factor determining system design. 14. How longer vesting affects recruitment a. Existential, starting point question – do you want to attract and reward people who come to work for the City for a relatively short period of time? On the Class B side, 2  DRAFT people who stay less than 7 years leave with relatively little retirement savings (and the same on the Class A side). Some believe longevity should be rewarded, and others think that no matter when you walk away you should have some retirement benefit accrued b. You should know that nationally this is a discussion among those who tend to favor cash-balance plans over DB plans. Particularly focused on school teachers and school retirement systems, but it is a question that ought to be considered by a broader audience. c. Of note, your system probably generates an actuarial gain from turnover that reduces employer’s costs by the cost of the benefit. 15. Lack of predictability in the system a. An issue in Burlington, and an issue in many retirement systems across the country b. Collective defined contribution rate, at the heart of the New Brunswick system, has a fixed employer contribution rate, with other components that can be adjusted i. There is always some risk in the system, the question is who will bear it Bob Rusten: Keith, thank you. How should we proceed to get you the info you need to come back with options for us to consider? Keith Brainard: The City is on the path to pay off its UAL over 30 years. Is this something the Committee wants to address? And if so, it gets into questions of how. I am assuming that the Committee would like to make changes to the system, but I do not want to be presumptuous. Councilor Mason: Changes are necessary. Speaking on behalf of my constituents, no – we are not on a sustainable path forward. Mike Flora: I would advocate on behalf of current employees – to avoid creating animosity among the workforce, I think if we consider structural changes we should start with new employees coming in. New employees will know what they are getting into. Councilor Bushor: This plan is not sustainable. People want change, and we need a dependable plan. I believe it would be a hard sell to change the system for current employees. I want to protect those close to retirement, and would like the changes to effect new employees only – my concern is that change to new employees would not be sufficient to make the necessary corrections to the system. Joe Keenan: My constituents in my union would certainly not want to see the status quo changed. To affect current members would be a hard sell. There are some gains to be made by reviewing our actuarial assumptions. We should also understand that the effect over time of the benefit tiers we’ve instituted alone may be able to keep a level contribution on the City’s taxpayers / ratepayers. That is yet to be determined. If we can change the curve on funding level and taxpayer contributions, do we really need to look at large systemic changes in the design? Councilor Paul: If you look at the first five problems we have identified – taxpayer contribution, UAL, support for the system among voters, to reference three as examples – it seems as though we do need change to the system. In terms of complexity, when you talk about no changes to current employees, I understand that – we have 21 plans already. But, these small adjustments haven’t worked. I agree with Councilor Bushor - voters have seen the cost of retirement increase. Having Keith here has illuminated the fact that we are not within the national norm on several categories. I don’t know that it makes sense to go back and determine why, but I do think in fairness to taxpayers we need to be within national norms. And, I believe we should have a better sharing of the risk burden. 3  DRAFT John Federico: First, solve the riddle of what we want the system should be doing. I think the Committee is well-equipped to do that, but less so to go out and search for other kinds of systems that will have impacts that are difficult to predict in the context of our system. People will be more disappointed in the future if we offer the prospect of improved shared risk, but do not actually deliver. This could worsen animosity toward the system. I don’t know how we jump off a DB plan – and when I hear discussion about changing risk I hear a shift away from a DB plan. Mayor Weinberger: John, I think you raise an interesting point worth discussing further here, and it gives me some hope there is some common ground. Keith’s work has provided examples of adjustments in risk sharing in other municipalities that are not changes out of a DB plan. I think the steps other municipalities have taken include measures to stop adding layers of complexity by agreeing to solutions now for how to keep the plan at an acceptable funding level, or in other words adjusting risk while keeping a defined benefit plan. Mike Flora: Have some of the changes you proposed (Keith) consistent with the Mayor’s point worked? Keith Brainard: You mean changes in plan design around the country? (yes) As a rule, it is too early to tell – most occur around 2010. Between 2010 and 2014, virtually every state made changes of unprecedented magnitude. Changes in plan design do work – and I’m not trying to sell them – but reducing a benefit level lowers an employer’s cost, for example. Certain changes in plan design will have an outcome within an expected range. Bob Rusten: So here are the different issues I have heard so far – I am summarizing only here, and choosing my words carefully: Changes that are designed to stop the increase in UAL. Changes designed to stop the increase in the City contribution. No impact on current retirees. No impact on those close to retirement. Minimal impact on current employees. Adjustments within the framework of a defined benefit plan. Measures of success. Agreement on future adjustments. Thoughts? Councilor Mason: In principle, I agree with not impacting current retirees or those close to retirement. But without numbers it is hard to know if that is sufficient to make the changes needed. And this is in the context of keeping the investment return projection at 8 percent – that is not guaranteed. Bob Rusten: BERS is examining the rate of return, and the impact such a changes would have on the system. Councilor Bushor: I’ve heard that since we made changes there are many new employees. Maybe you look at only those who have five years or less in the system. I like the Mayor’s idea of putting in place changes that we agree to in advance, if those changes we make now do not work out. I don’t want another group to be in this room struggling this same issue years down the road. Eileen Blackwood: I have an issue with the characterization of stopping the increase in the UAL (Bob: I am only summarizing here). John Federico: Are we asking Keith to come up with a plan design for us? I failed to see changes in plans that are similar to ours, and I don’t want to make changes without understanding the impact. Keith Brainard: I’d like to elaborate on the answer on how we know changes have an impact. In CO and MN, a number of changes were instituted that affected all members. Contributions for 4  DRAFT employers and employees went up, years or age of service requirements increased, COLA changes, etc. I should not have said it is too early to tell – in those cases and many others, the reforms have the intended effect of reducing the UAL or the cost of the plan or both. It is clear mathematically that if you reduce the benefit for retirement or current employees, you get more impact on the plans cost than when you make changes affecting new employees. Mike Flora: 13 changes in Class A, and 9 changes in Class B over the recent past, along with some benefit changes – maybe it’s too early to tell, or maybe that continued trajectory of growing unfunded liability is proof these changes haven’t worked. And, Bob, I think we need to define what “close to retirement” means. John Federico: I don’t think this Committee needs to target who we are going to effect. The committee should focus on what kind of retirement system we should have and leave it to the natural bargaining process with the City administration, unions, and City Council. We didn’t see immediate impact on the plan despite the number of changes (Mike’s reference to 13 and 9 changes), and that could be tied to actuarial assumptions that we’re reviewing. Councilor Bushor: Is the intent to have a dialogue here about active or retired? I need better guidance from you on the question of who is impacted. Keith Brainard: The question is what categories the Committee is considering. First, new hires only? Second, new hires and currently active (and if so how close to retirement)? And third, retired members in addition to new employees and current members. Councilor Bushor: My line in the sand is active members with up to five years of service and new hires – not those close to retirement. Those coming into the system and those relatively new in the system, if we can negotiate that. I am an employee too and understand what I’m saying – to make changes that give us necessary stability. Mike Flora: To clarify – I don’t think we have legal authority to impact what someone has already earned in the system. Eileen Blackwood: People have the right to benefits, but not necessarily getting benefits in the same way they always received them. Joe Keenan: Do you mean with current retirees? Eileen Blackwood: More on active employees – it may be possible to affect current retirees legally, but those options are limited. What I’m saying is the benefit could be provided in different ways, but not a decrease in benefit. Bob Rusten: Cannot eliminate benefit already earned, but you could eliminate benefit not earned. Bob Rusten: So the question is what the system could accomplish, and then what is the design for the system based on the outcome we want – I don’t think we are actually this far apart. Other thoughts or comments? Councilor Paul: I think I want a system that honors the notion of shared risk, and more certainty and predictability to avoid what we’ve been doing – continually going back and rejiggering this plan ineffectively & created uncertainty for all parties. DB plan with shared risk by all parties and more in line with all parties and more predictable for all parties. 5  DRAFT John Federico: Is a shared risk plan a DB plan? Keith Brainard: Yes – strictly speaking, a DB plan is an employer sponsored plan that pays a certain amount known in advance and funded by employer. The prevailing DB plan in the public sector varies on that concept by, typically, requiring some employee contribution (and is thus not a DB plan in the perfect sense, but in the form we understand it generally). Think about it on a spectrum – at one end is a DB plan and on the other end is a DC plan. Most public sector retirement plans fall somewhere along that spectrum, and that’s where you get into the concept of shared risk. In the last 5 – 10 years, significant increase in elements that shift to transfer risk – high contributions by employee, or employee bears more mortality risk, or will receive a benefit at an age based on retirement (and thus choose to retire earlier), or inflation risk so COLA is reduced or eliminated, actuarially reduced benefit at retirement coupled with a COLA. There are many potential variations around this idea. Bob Rusten: John, I’d offer a concrete example in Burlington as well: Buck Consultants has estimated City had to pay the $5.9m in the FY2013 valuation in past service. Looking to next year, that figure has jumped to $6.1m. The City bears that difference – funded by ratepayers and taxpayers solely. Other systems would share that cost increase when we don’t hit, for example, our actuarial assumptions. Bob Rusten, continuing on a separate topic: I think we want to make the next meeting as productive as possible – we need agreement about what we are asking Keith. What additional info can we give him? John Federico: I haven’t seen any comparable plans to Burlington -- Mayor Weinberger, interjecting: Keith has provided a number of examples – July 12 paper looks at a large number of comparable plans. Councilor Mason: And John, right now, whatever the comparison, if an unexpected event happens, the taxpayer is liable for everything. It is not tied directly to the questions of the UAL. John Federico: I think having a less abstract model to deal with would make it easier to understand – something more grounded could make this easier to understand and transition. Keith Brainard: I have a database of 45 states and 60 / 70 municipalities and the changes that have been made – retirement ages, vesting periods, changes to COLAs, etc. You don’t need a comparably sized City to understand the concept in front of you – money comes in, benefits go out, and within those parameters there are lots of ways to do it. Bob Rusten: John’s question has me thinking – I’m assuming that next meeting, Keith, you’ll be coming in with ideas specific to Burlington, rather than East Oshkosh, for example, and that these ideas would have options for different types of approaches. No changes to retirees / active members, fundamental reform to include active members, etc. Councilor Bushor: Many ideas put forth – DB/DC, the Boston College model, New Brunswick, etc. I want to know whether a hybrid plan would work for Burlington, for the taxpayer. Stability and reduced risk distribution would be the goal in my opinion. Eileen Blackwood: To build on Councilor Bushor’s ideas – As an example, I’d like to know if we don’t make changes for retirees, what is the impact on other measures we have to take? Or, if you 6  DRAFT want to stop the growth of the UAL or instead focus on decreasing the UAL, what would be the impact on the plan? Bob Rusten: Let me go to an answer we have danced around a bit – shared risk. With the 5.9m to 6.1m example, do we want that to all be on the City? What are the parameters that we are talking about here? John Federico: Keith’s answer was interesting – we already have a shared risk system by his definition through our contribution or benefit reductions. People have different ideas about how to share the risk. And, I would also add I see the automatic adjustment agreement as a separate question. Mayor Weinberger: It’s an interesting point – we do have shared risk, but the arrangement is currently unsatisfactory to everyone at the table. The system leaves too much uncertain – potentially dramatic changes in the future that drive people to retire earlier than they might otherwise, and it puts a great deal of instability on the City side during these negotiations. Further, making changes to the system adds to its complexity and makes administering the system more difficult. Other systems, where there is an agreement about how to make changes in the event of a shortfall before the fact, could allow you to make changes to bring the system into alignment. To shift points slightly – I have been impressed by Burlington-specific analysis Keith has provided. I think we should not miss the opportunity here to have him put forth ideas following his review – I have felt consensus on that point until tonight. As John rightly continually reminds us, this Committee’s role is limited, but it is part of our purpose to get Keith to provide some instruction on our situation. We should not miss this opportunity, when we are on the cusp of taking what we have learned into recommendations. Mike Flora: 8 months at this Retirement Committee, going on 9 months. I agree with Miro. What do we want to put forth as recommendations? DB plan with a shared risk portion in that pool, and with new members look at the hybrid plan? That’s one idea – we need to give Keith the ingredients to come up with some recommendations. After 8 months, we are at the point of coming up with our own model. Keith Brainard: What I am hearing is that I should develop options for plans only for new hires, adjustments to benefits should Committee wish to go that way instead, come up with models that you could adjust as you see fit. John Federico: We are still missing the boat on what we want. We do have consensus on the problems and issues, but not a clear sense of an ideal system. Where could we find some agreement, savings, and stability? Ask Keith to consider from what is the ideal way. Mayor Weinberger: To just ignore current system is to talk about a fantasy world we do not live in, though. Keith should go through the exercise assuming that we are sticking to a DB plan, but that it is in trouble, and that we need to make significant progress relatively quickly (within 3 years) to transition from shared risk only in the way John described it to a better system that was more predictable than our current system and that got there in part by making adjustments to areas where we know Burlington is an outlier nationally – what would that look like? Completing that process – maintaining a DB system but sharing risk better, more in control of UAL, more in line with national norms in how we fund it – means exploring what that system looks like. Bob Rusten: Keith, have you heard enough in the last 30 minutes to get guidance? 7  DRAFT Keith Brainard: Yes. Bob Rusten: If I think about the first meeting and what we are now discussing, we are essentially looking at continuing the DB – not where we started out. Better way of sharing the risk to be discussed in contract negotiations. A more predictable system. I think there has been movement on a number of issues and it is easy to lose sight of that. Councilor Paul: I would add that through Keith’s input, we know that our system has a significant number of outliers when compared to national norms – and we should move more in line with national norms. Keith Brainard: Scheduled to meet two weeks from today – September 2 – and I’m hopeful to be back to the group with some solid information in time to consider before the meeting. Bob Rusten: Who do we want to facilitate the next meeting? Joe Keenan: Eileen? Eileen: I am happy to do it. Meeting ends at 6:52pm. Next Meeting Time: 5pm – 7pm, Tuesday, September 2, 2014 (TBD) 8  Burlington Retirement Committee Meeting of the Committee Members Agenda August 19, 2014 5:00pm – 7:00pm BCA Art Gallery 2nd Floor Conference Room 135 Church Street, Burlington Vermont 05401 5:00pm – 5:05 pm Approve Agenda Review and Approval of August 5, 2014 meeting minutes 5:05pm – 5:10pm Public Comment 5:10pm – 5:30pm Review of key problem areas and ideas identified by the Committee on August 5 All, with Keith Brainard 5:30pm – 6:30pm Discussion of shared risk and retirement plan design components with Keith Brainard in light of the problem areas and ideas identified by the Committee All, with Keith Brainard 6:30pm – 6:35pm Input for Next Meeting Agenda and selection of facilitator All Next Meeting Time: Tuesday, September 2, 2014 (TBD)