The Wisconsin Retirement System (WRS) is one of the few public pension systems nationally to avoid the funding gaps that have beset other plans like those in the city and county of Milwaukee. Adopting elements of the WRS approach would be challenging in the short term but could aid the Milwaukee plans in the long run.

In recent years, Wisconsin’s state pension system has won acclaim while its two counterparts in Milwaukee have faced stark challenges. The fiscal impacts of those challenges have contributed to the elimination of city police positions, a huge county infrastructure repair backlog, and a push for a new local sales tax.

The contrasts between the three public pension systems in the state are striking. The Wisconsin Retirement System (WRS) outperforms most of its peers nationally on a range of criteria, including its annual cost to taxpayers and its ability to pre-fund the benefits promised to state and local workers and retirees. The pension plans for the city and county in Milwaukee are healthier than many public plans, but still closed 2018 with a combined unfunded liability of $2 billion.

The WRS, the pension plan for the state and most local governments in Wisconsin, was more than fully funded at 102.9% at the end of 2017, a testament to the disciplined contributions by those employers and WRS’ unusual approach of linking retiree payments to investment performance. At that point, state employee retirement funds nationally had funded only 69% of the $4.1 trillion in obligations promised to beneficiaries, according to a Pew Charitable Trusts report from June. The state pension system in neighboring Illinois had only 38.4% of the pension assets needed to cover its liabilities.

The Milwaukee area plans are better than the average state system in terms of funding levels. And, like WRS, both plans will benefit from strong 2019 investment returns that won’t show up on their financial statements until this summer. But they still have sizable unfunded liabilities, which require growing annual taxpayer contributions that are placing severe pressure on city and county finances. In the 2020 budget, for example, Milwaukee Mayor Tom Barrett and the city council cut 60 police positions to help prepare for a huge increase in the city’s employer contribution to its pension plan in 2023.

Meanwhile, Milwaukee County’s employer contribution to support its pension obligations (including pension obligation bond debt) grew by nearly 50% in the past decade to nearly $98 million in the 2020 budget. The increased costs contributed significantly to difficult decisions in recent years to reduce staff and service levels and implement a $30 wheel tax.

These contrasts raise the question: why has Wisconsin’s system weathered financial challenges so well and could the two public pension plans in Milwaukee benefit from adopting elements from the WRS approach? In this issue of The Wisconsin Taxpayer, we consider those questions.

OVERVIEW OF THE THREE PENSION SYSTEMS

Wisconsin Retirement System (WRS)

The state is fortunate that the WRS is so well-positioned given its massive size – by one ranking it is the eighth-largest state or local pension system in the nation and 25th-largest in the world. The WRS provides retirement, disability, and death payments to survivors and covers nearly 642,000 retirees, workers, and other participants who were or are employed by over 1,500 state agencies and local governments. Financial statements show WRS benefit payments provided to retirees and beneficiaries rose 5.9% in 2018 to $5.5 billion. The growth reflects increases in the number of beneficiaries and in their benefits because of rising employee salaries and earnings on pension investments.

Among the participants are teachers, police and firefighters, university professors, and some elected officials as well as a host of other state and local employees from all school districts in Wisconsin, all technical colleges, all counties except Milwaukee County, essentially all cities except Milwaukee, two-thirds of villages, roughly one-fifth of towns, and more than 200 special districts. All told, more than 1,430 local governments participate. Since 2001, the WRS has gained employers, helping to boost the number of state and local participants by more than 30%, while the city of Milwaukee plan’s membership has been roughly flat and the county’s is down 8%.

After World War II, Wisconsin officials brought together a number of smaller retirement funds, according to the state Legislative Fiscal Bureau (LFB). In 1951, Wisconsin became the first state in the country to extend participation in the federal Social Security program to some state and local employees and nearly all active WRS workers qualify for it. Most participants in the two Milwaukee pension plans also qualify for Social Security benefits but the city’s police and firefighters do not – a notable difference that means the city funnels more money into its pension plan for those workers but does not have to pay federal Social Security payroll taxes. A 1976 law consolidated funds for teachers, state workers, and local employees, including the Milwaukee Teachers Retirement Fund, but did not extend to the plans for the city and county of Milwaukee, a decision with far-reaching implications.

Milwaukee pension plans

Like the WRS, both Milwaukee plans provide retirement, disability, and death benefits. The city of Milwaukee pension plan has twice as many participants and roughly three times the assets of the Milwaukee County retirement system, though both plans are considerably smaller than WRS (see Table 1).

The city of Milwaukee Employees’ Retirement System is overseen by an eight-member board and serves police officers, firefighters, and other city workers as well as employees in the Metropolitan Milwaukee Sewerage District, the Wisconsin Center District, the city housing and redevelopment authorities, and certain Milwaukee Public Schools staff who are not teachers or administrators. The county plan is run by a 10-member board and serves general employees and sheriff’s deputies. It does not include county circuit court judges (who are part of the WRS) or Milwaukee County Transit System workers (who have their own pension plan).

Like the state, the city and county pension plans provide a guaranteed benefit to retirees based on factors such as salary and years of service. The defined benefit approach provides a high degree of certainty for employees regarding the pension payments they will receive upon retirement. (See box on calculating this benefit.) However, both Milwaukee plans differ from the state in providing automatic cost of living adjustments to retirees’ pension payments regardless of investment performance.

Defined benefit plans differ from the defined contribution approach used by most private employers, who often favor 401k plans. Under that approach, the employer and employee make defined contributions to individual retirement accounts, but the benefits available upon retirement are uncertain and based on the investment performance of individual accounts.

The state’s practice of providing the necessary funding, sharing risk with participants, and relying on more conservative assumptions have helped the WRS maintain funding levels at or near 100% of liabilities since 2001, a period which includes two recessions. Investment losses in 2018 reduced the plan’s net assets, or the resources available to fund benefits, and as of Dec. 31, 2018 (see Figure 1) resulted in a drop in WRS’s funded ratio to 96.5% of the $100.3 billion in benefits promised (we previously cited the 2017 ratio of 102.9% because that is the most recent year for which national comparison data were available). However, strong investment returns in 2019 should help to offset the previous year’s losses.

In contrast, the city of Milwaukee retirement system reports it was 78.7% funded at the end of 2018 and the Milwaukee County system was 70.6% funded. Like the state, both systems saw a decrease in 2018 and will likely see an increase at the end of 2019 because of better investment returns. These figures – like most others in this report – are based on the fluctuating fair market value of the investments for each pension system, making their funded ratios more up to date but also more volatile.

Funding ratios also can be calculated by actuaries using methods to smooth out short-term differences in asset values and these data are available going back further in time. Figure 2 charts these ratios using smoothed asset values, showing how funding levels for the two Milwaukee plans have fallen since 2008 while the state has kept steady at or near 100%.

HOW DEEP ARE MILWAUKEE’S CHALLENGES?

For both the city and county of Milwaukee plans, taxpayer-financed contributions have been rising. Government pension systems facing unfunded liabilities seek to fill those gaps through annual employer and employee contributions recommended by plan actuaries. The recommendations reflect the amounts needed to offset the cost of pension benefits paid out during that year and to eliminate any unfunded liabilities within a specific timeframe known as the “amortization period.”

The calculation of these payments rests on a series of assumptions approved by the pension system’s governing board. One of the most important is the assumed rate of annual investment return on the fund’s assets, but there are many other important assumptions, including mortality rates of plan members, the rate of inflation, and projected changes in the salaries and number of active workers.

Any unfunded liabilities – as well as corresponding employer and employee contributions – rise or fall depending on how closely actual investment returns and other outcomes mirror the actuarial assumptions. The challenging investment environment has led many plans nationally – including all three in Wisconsin – to lower their assumed investment returns. Other factors that can impact annual contribution amounts include decisions to modify key assumptions and to contribute more or less than what the pension fund actuary recommends.

Milwaukee County

The 70.6% funded status of the Milwaukee County pension system at the end of 2018 meant that the plan was facing an unfunded liability of about $675 million at that time. The county has a policy of paying down the unfunded liability completely over 20 years but in practice it has grown in four of the last five years.

In 2018, the county paid an employer contribution of $59.5 million, while its employees contributed $12.7 million. The county also paid $33.2 million for principal and interest owed on $400 million in pension obligation bonds (POBs) issued in 2009 to reduce the unfunded liability, for a total taxpayer pension contribution of $92.7 million.

To put that number into context, the county’s overall property tax levy for 2018 was $292.8 million. In other words, the county’s pension payments in 2018 equated to nearly a third of its total levy, though the county has other revenues it also uses to make pension payments.

Unfortunately, the county plan’s unfunded liability and employer contribution have been trending upward in recent years. The 2018 contribution was $10.6 million (11%) higher than just two years earlier (see Figure 3), and the county is budgeting $97.7 million for 2020.

One past contributor to the county’s unfunded liability was the package of pension enhancements approved in 2000, which included the lump sums paid to retirees and known as “backdrop payments.” Though the county has not recently calculated the long-term net impact of these changes, they clearly added to the plan’s net liability. In 2017 and 2018, the county plan made lump sum backdrop payments of $19.9 and $20.7 million respectively, which amounted to about one-tenth of the total benefits paid out in each year.

One of the most significant recent contributors to the rise in taxpayer funding was a decision by the pension board to reduce the assumed rate of investment return from 8.0% to 7.75% in 2018 and again to 7.5% in 2020; and a 2015 decision to reduce the amortization period from 30 years to 20 years. In addition, the required contribution has grown in part because investment returns have fallen short of assumptions, though they were strong in 2019.

In the summer of 2017, the county executive formed a Retirement Sustainability Task Force (which was facilitated by the Forum) to assess potential solutions to rising pension payments. The task force received technical assistance from staff of the Pew Charitable Trusts, who projected the county’s annual employer contribution (including POB debt service) would grow to more than $130 million by the mid-2030s.

Pew also found that because the size of the contribution was driven primarily by legally binding promises already made to retired and active workers, any newly enacted plan design changes aimed at reducing benefits for new employees would have little impact on reducing the size of the contribution for at least a decade. One possible change with more immediate impact would be to modify the annual cost of living adjustments made to retirees’ benefits and we discuss that potential change below.

Pew’s work included another concerning projection: contributions to the county Employees’ Retirement System (ERS) from general employees could increase from 6.5% of salary in 2017 to more than 9.7% within two decades. Today, that increase would mean an additional contribution of $1,795 a year for an employee with the average plan salary of $55,779 a year. Such an increase could pose a significant threat to employee retention and recruitment.

City of Milwaukee

At the end of 2018, the 78.7% funded status of the city of Milwaukee’s pension plan meant that the city and other units of government in the plan were facing an unfunded liability of what was then projected to be $1.34 billion. Like the county, the city’s system has seen its unfunded liability grow in four of the last five years (see Figure 4). However, the magnitude of the increase for the city is particularly notable, with the system’s financial report for the end of 2018 showing an increase of nearly $1 billion in the net liability from the previous year, when the city pension system was considered to be 93.7% funded.

The financial statement shows two notable developments that produced the shift. First, after a series of years with positive investment returns, the pension plan’s investments yielded a loss of $160.2 million in 2018 – a stark contrast to the gain of $787.8 million they produced in 2017.

Second, in both 2017 and 2018 the city’s estimated pension liability grew substantially because of changes in the underlying assumptions used to calculate it. As we will discuss in greater detail below, the Annuity and Pension Board in April lowered the assumed rate of return used for its 2018 financial statements to 7.5% from the previous rates of 8% for 2018 through 2022 and 8.24% from 2023 into all future years.

In all, the changes in assumptions caused the liability to grow by $245 million in 2017 and $475.8 million in 2018 for a total of $720.8 million over the two-year period. In 2017, the negative impact of the change in assumptions was easily absorbed by the positive investment returns, but in 2018 there were no investment gains to offset the even larger effect that year of the changes in assumptions.

To cover the much larger liability, the city will need to substantially increase its pension contribution, although the boost will not need to occur until 2023. The delay results from the city’s “stable contribution policy,” which resets the employer contribution rate every five years, as opposed to annually.

In 2018, the city and its other employer participants paid a contribution to the pension system of $83.2 million. As of last fall, the contribution was expected to rise sharply to $158 million by 2023 – or roughly double current levels. Whereas in 2022 the projected employer contribution would account for about 23% of the city property tax levy for that year (based on annual inflationary increases), the much larger contribution in 2023 would require a whopping 51% of that year’s levy.

To prepare for that impact, city leaders contributed $8 million to a pension reserve fund in the 2020 budget and hope to make larger contributions to the reserve in the following two years. They will then withdraw reserve funds beginning in 2023 to soften the blow somewhat.

City employees contributed $32.1 million to the pension system in 2018. The contribution rate for general employees enrolled in the plan prior to Jan. 1, 2014 is 5.5% of salary and for those enrolled after that date it’s 4.0% of salary. These are lower amounts than those paid by their counterparts within WRS or the county plan, a factor that we will discuss in greater detail later in this report. Still, the city’s employer contribution rate for its general employees (7.5% of payroll) is not much higher for now at least than the employer rate for WRS general employees (6.75%), though the city’s contribution is expected to rise sharply in 2023.

Public safety workers represent the main driver of the city’s higher pension cost. After excluding other units of government like Milwaukee Public Schools with some workers in the city plan, 80.8% of the city’s required employer contribution is for police and firefighters despite the fact that they make up 44.3% of the city’s active workforce.

In addition to their earlier retirement age, the larger impact for the city’s police and firefighters partly reflects the fact that they do not receive Social Security benefits. It is worth noting that the public safety workers themselves do pay a higher employee contribution rate (7.0%) than the city’s general employees. However, the contribution rate for those workers as a share of payroll (25.7%) is much higher for the city than it is for the small number of public safety workers in the WRS (local firefighters) who are not covered by Social Security (14.95%). Reasons for this could include the city’s unfunded pension liability, the larger guaranteed annual benefit increases for these city employees, and their earlier potential retirement age compared to their public safety counterparts in WRS.

WHAT CAN THE CITY AND COUNTY LEARN FROM WRS?

Several elements stand out in helping to explain the stronger position of the WRS, including the practice of making the required contributions, generally conservative assumptions regarding investment returns and other key factors, and the policy of tying benefit payments to those returns and other factors.

These practices have paid off. Largely because of WRS, government contributions to state and local pension systems in Wisconsin accounted for just over 2% of all state and local general spending in fiscal year 2017, according to U.S. Census Bureau figures. That was the third-lowest share in the country, behind only Wyoming and South Dakota.

Shared Risk

The WRS follows a “shared-risk” model which helps spread the potential rewards of investing – as well as the risks – between government employers and pension recipients through increases or decreases in annual payments to retirees. Like many other pension systems, the Milwaukee plans offer automatic annual cost-of-living increases (COLAs) to retirees’ pension payments regardless of the plan’s investment performance and the size of its unfunded liability. Milwaukee County retirees are entitled to annual increases of 2% of their original pension benefit, general employees within the city of Milwaukee plan receive 2% increases in their pension payments starting on the fifth anniversary of their retirement, and city public safety workers receive up to 3% increases.

Instead of automatic COLAs, the WRS provides higher payments to beneficiaries only if the fund outperforms its target rate of return that applies to those recipients. The reverse is also true – in the case of many employees previous benefit increases can be taken back due to poor investment performance. Figure 5 shows how payments to WRS retirees fell during and after the Great Recession instead of rising as they did for both Milwaukee plans.

To manage their risk, WRS participants can choose between two trust funds with investments managed by the State of Wisconsin Investment Board (SWIB). The diversified Core Trust Fund is by far the larger of the two, accounting for more than 90% of WRS assets, and holds stocks, bonds, real estate, and other investments. At least 50% of the contributions for each employee must go into the Core Fund. Participants can choose whether to place the other half in the smaller Variable Trust Fund, which invests only in stocks and therefore has greater potential risks and rewards.

Gains and losses within the Core Fund are distributed to employee accounts over rolling five-year periods, smoothing out increases and decreases in pension payments. In addition, a retiree’s pension benefit from the Core Fund cannot fall below the floor of his or her initial payment. Gains and losses in the Variable Fund are distributed every year and there is no limit to decreases to benefit payments, making them more volatile.

This approach also applies if other factors turn out better or worse than projected, including whether participants retire earlier or live longer. In addition, public employers and active employees paying into the system may have their contributions lowered or raised when plan investments and other factors beat expectations or fall short of them.

Any WRS benefit changes occur automatically, freeing elected officials from taking difficult votes. In addition, decreases are unlikely to face legal challenge from pensioners because they and their employers have essentially agreed in advance to the shared risk approach and because these provisions have also been written into state law.

Prudent Projections

The WRS has also avoided banking on a high rate of return that could lead to higher employer and employee contributions if the investment earnings fail to materialize. At the end of 2018, the WRS lowered its investment return assumption from 7.2% to 7%, below the national median of 7.25% reported by the National Association of State Retirement Administrators and the 7.5% currently used by both Milwaukee plans.

As discussed above, the city’s near-term return assumption was 8% (and slightly higher when blended over 30 years) before it was decreased to the lower rate in 2019, while the county lowered its rate from 7.75% to 7.5% on January 1, 2020. Figure 6 shows that while all three pension plans have lowered their assumed rate of investment return in recent years, the state’s has remained the most conservative.

In addition, the WRS is actually being more cautious than the numbers above might imply. Once workers in the WRS retire, the state only needs to earn a 5% rate of return on the assets associated with those retirees to make the minimum necessary payments to them. If the state’s returns exceed the 5% target, then the extra returns on those assets can be used to make additional payments to the retirees, as discussed above.

Making WRS Contributions More Affordable

Another key to the current funding level of the WRS has been the steps taken to share the cost of employer contributions with workers. With the passage by lawmakers and then Gov. Scott Walker of 2011 Wisconsin Act 10, collective bargaining was eliminated for most public employees and the share of pension contributions being made by state and local government employees increased dramatically. While we take no position on the overall merits of Act 10, there is no question that it lowered the cost of pension contributions for the state and other participating employers. It did so as well for the two Milwaukee governments, though arguably not to the same extent because of the way employee contribution requirements have been applied.

Prior to Act 10, WRS employers made more than 99% of all contributions in that plan, but since its passage the share contributed by employers in the system has fallen to roughly 52%. Act 10 also increased contribution rates for city and county workers within the two Milwaukee area pension plans, though the law did not apply to most law enforcement officers and firefighters. For the city that is particularly significant given that police and fire personnel make up 44% of the active workforce. Figure 7 shows how taxpayer contributions to WRS fell as employees’ amount rose.

Using LFB figures on WRS contributions, we estimate that having employees pick up a greater share of those pension contributions reduced state and local government costs – and before-tax wages for public employees – nearly $5.2 billion over the seven-year period from 2011 to 2017. The analysis looked at only the increase in required employee pension contributions and not other Act 10 changes, such as increases in employee health premiums or a decrease in the pension multiplier for elected officials and appointed executives. The seven-year pension savings was nearly $1.7 billion for the state and almost $3.5 billion for the local governments participating in WRS – not counting additional pension savings for the Milwaukee plans.

State and local WRS employers now pay less into the system even as contributions have risen overall. In 2008, the contributions in WRS for general employees such as teachers amounted to 11.2% of their gross earnings, with almost all of that paid in by employers such as state agencies and school districts. As of January 2020, the overall contribution rate had risen to 13.5% of the wages of general employees but employers were only paying half of that, or 6.75%, with workers paying the rest.

Act 10 also contained provisions requiring city of Milwaukee and Milwaukee County employees to pay more toward their pensions, although the two governments implemented those provisions differently. The county calculates the annual employee contribution as 50% of the portion of the actuarially required contribution corresponding to active employees. For 2020, that percentage is 6.2% for general employees (compared to 6.75% for WRS participants).

The city simply ascribes a longstanding employee contribution rate of 5.5% that general employees started paying prior to Act 10 in January 2010. While the employee contribution rate had existed on paper before that, the city had previously agreed to pay it for its general employees when they were covered by collective bargaining agreements. For general employees hired after 2014 the rate is only 4%, as pension benefits are lower for that group of employees (for example, the multiplier is 1.6% as opposed to the 2% provided to pre-2014 workers).

The Act 10 provisions requiring employee contributions for general city and county employees have helped both governments weather the impact of growing employer contributions much as they have for the state and other local governments and school districts covered by WRS. However, there may be latitude for both governments (and in particular the city) to consider higher employee contributions, although officials would have to weigh potential impacts on employee retention and recruitment. We discuss this issue in greater detail later.

For added context, Figure 8 shows the employer’s required contribution as a percentage of payroll for the three Wisconsin pension plans in the wake of Act 10. According to 2018 financial statements, the 7.2% contribution rate as a percentage of payroll for all WRS employers and types of employees in 2018 was half the 14.5% rate for the city of Milwaukee pension plan before its even higher rate expected for 2023; and less than one-fifth of Milwaukee County’s worrisome 32.3% rate. As noted, the city’s higher payment reflects in part a larger share of public safety workers who retire earlier and who do not receive Social Security benefits.

Factors contributing to the higher contribution rates for the Milwaukee plans include their greater unfunded liabilities and recent decreases in assumed investment returns. Milwaukee County also is coping with the decline in its workforce resulting from the transfer of certain social services to the state, the 1995 closure of the county’s John L. Doyne Hospital, and, more recently, personnel reductions caused by outsourcing initiatives and budget challenges.

A few WRS provisions may lead to higher liabilities within the state plan compared to at least one of the Milwaukee plans. For example, the city of Milwaukee plan does not include employee overtime in its covered pay or final earnings calculation but the WRS and county plan do (the state plan also monitors workers’ overtime for spikes). As mentioned in the “Calculating Benefits” box, the alternate WRS money purchase formula also can benefit shorter-term workers.

WRS has also faced some challenges. In 2003, the state approved issuing bonds to help fund accumulated liabilities for state employers within the system. As of December 2018 the state still owed $1.5 billion in principal on these bonds, according to LFB. Milwaukee County took a similar step in 2009 when it issued about $400 million in pension obligation bonds to pay down a portion of the retirement system’s unfunded liability; as of December 2019 about $262 million of that debt was still outstanding. The city has not taken such a step so far.

OPTIONS FOR MILWAUKEE’S PENSION PLANS

The city and county’s rising pension contributions largely reflect liabilities for benefits already earned by retirees and active employees. Consequently, while WRS policies and practices may offer insights for policy changes that could be applied to future workers or years of service, in most cases the financial benefits associated with those changes would not be realized for quite some time. Figure 9 helps illustrate why – both Milwaukee plans have more retirees and other beneficiaries with fixed benefits than they have active workers whose benefits for future service could be reduced.

Nevertheless, to help address their challenges the two governments could consider three categories of options: accounting and financing changes; design modifications to their existing plans; and closure of the existing plans with a transfer to WRS. We view each through the lens of the policies and practices that have worked well for WRS.

Given the complexity of the analysis, we do not consider eliminating the city and county defined benefit plans and prospectively switching to a defined contribution or hybrid approach. Also, the county Retirement Sustainability Taskforce rejected this approach in favor of a shift to WRS.

Accounting and financing changes

As discussed above, required employer contributions reflect the calculations made by plan actuaries using dozens of assumptions about future events and trends. The actuaries recommend changes to the assumptions when actual experience diverges too greatly from them. While changes to most assumptions only modestly affect required contributions, modifications to others – like the assumed rate of investment return on plan assets – can produce annual swings in the tens of millions of dollars for plans the size of Milwaukee’s two systems.

Contributions also are impacted by each system’s accounting and financing policies. For example, the amount of time over which any unfunded liabilities are amortized can have a substantial effect on contributions. A longer amortization period spreads required contributions over a greater number of years, reducing annual contributions but delaying the plan’s arrival at full funding. Plans also adopt different approaches for how their actuary calculates their required contribution; the city’s stable contribution policy uses a five-year time frame to calculate the required contribution and resets it every five years, while the county does so on an annual basis.

To hold down employer contributions in the short term, the city and county could:

• Extend the amortization period. A longer amortization period would push out rising employer contributions over a longer period. Officials would have to consider each plan’s financial outlook and the impacts on current and future taxpayers and budgets. Generally speaking, the trade-offs are similar to those a home buyer faces when considering a 15-year vs. a 30-year mortgage; paying off the mortgage more quickly reduces interest costs (and total payments) and generally promotes better financial health, but the larger annual payments required under a shorter repayment schedule can produce other untenable impacts.
  Analysis conducted by Pew as part of the county’s Retirement Sustainability Taskforce (RST) estimated that extending the county’s amortization period to 25 or 30 years from the current 20 would reduce the annual employer contribution by $4 million to $9 million per year until 2036 but would leave contributions substantially higher after that. The RST did not recommend pursuing this option because of the greater financial benefits of a shorter amortization period.
• Make changes to contribution policies. The city’s stable contribution policy is meant to avoid large annual swings by estimating required employer contributions over a five-year period and keeping them steady over that time but it can lead to greater shocks when a reset does occur. The city could consider a shorter or a lengthier period between resets depending on whether the goal was a steady contribution in most years or a smaller change in the years when one does occur. The county could consider a similar policy. Still, these policy changes likely would have little impact on the total amount of employer contributions required. In addition, the sharp projected increase in the city’s contribution for 2023 shows a multi-year window does not guarantee stability.
• Issue pension obligation bonds. As discussed above, both the state and county have issued pension obligation debt as a strategy for stabilizing annual contributions and shoring up funding levels. Under this approach, some or all of a projected unfunded liability is paid off through a deposit of bond proceeds into the pension fund. The government pays off the bonds (with interest) over a lengthy period of time and the debt schedule provides predictability with regard to annual payments. POBs can be financially advantageous if earnings on the investments purchased with bond proceeds generate a higher rate of return than the interest rate paid on the bonds. However, there is considerable risk involved, as the opposite can also occur in cases of a downturn in the financial markets. Also, the additional debt could limit the city or county’s ability to borrow for other needs and potentially impact bond ratings.

Design changes to existing plans

Rising employer contributions may prompt policymakers to consider reducing pension benefits to hold down costs, though such strategies may have little immediate effect. In their deliberations, city and county officials should weigh the impact of any potential changes on employee recruitment and retention as well as any effects on retirees and other beneficiaries. Also, as noted before and shown in Figure 10, the great majority of the city’s contribution is for police and firefighters. Any changes to their pensions would have to be collectively bargained but changes that do not include them would only have limited impacts on the city’s long-term costs. Any benefit changes for county public safety workers would also require collective bargaining, but they make up far less of the county workforce.

The options include:

• Implement risk sharing. As discussed above, WRS links annual changes in its payments to beneficiaries to factors such as investment performance while the city and county offer automatic annual 2% COLAs to retirees (up to 3% for retired city police and firefighters) regardless of investment performance and funding levels.
  The RST recommended the county analyze a variety of options for adjusting its COLA policy, including “an approach that mirrors that used by the WRS, maintaining the COLA at a sustainable amount, and revisiting the COLA during each annual budget.” So far, no action has been taken on this recommendation.
  Pew’s analysis revealed reducing the guaranteed COLA from 2% to 1% could reduce the county’s annual employer contribution by about $10 million annually, while eliminating it entirely could yield annual savings in the $20 million range. Similar estimates for the city are not available but could be substantial. However, the Pew projections assume the changes would be applied not only to new workers, but also to existing retirees and active employees prospectively. Whether such a move would be legally permissible is unknown; such changes have withstood legal scrutiny in other states but it is impossible to know how Wisconsin’s courts would rule and also whether a previous legal agreement affecting the city known as the Global Pension Settlement might preclude such an option for that plan.
  Depending on how it was implemented, adopting the WRS approach might have the practical impact of reducing COLAs to zero for city and county retirees for the foreseeable future, given the unfunded liabilities in the two plans. The resulting impact on retirees with more modest pensions would have to be considered. Yet, if full funding is achieved, city and county retirees might benefit from COLAs in some years that exceed the current 2% depending on investment performance.
• Reduce the defined pension benefit. Policymakers could consider whether pension benefits in Milwaukee are out of line with other public plans and whether any reasonable and legal changes would yield significant savings. Indeed, the city already lowered its pension multiplier from 2.0% to 1.6% for newly hired general employees in 2014, and the county also lowered its multiplier from 2.0% to 1.6% for both new general employees and future years of service for most active general employees earlier in the decade.
  For the future service of new or existing workers, local leaders could further reduce the multiplier, push back the retirement age, or change the final average salary calculation (such changes would need to be collectively bargained for public safety employees). As shown in Table 1 on page 2, city and county benefits are generally on par with WRS. There are some differences, however, such as the minimum age at which recently hired police with 25 years of service can retire and still receive full benefits – 50 for city of Milwaukee officers and 53 for those in the WRS.
• Increase employee contributions. Under Act 10, the employer and employees covered by WRS essentially share each year’s required contribution equally, which for employees is translated into a percentage of their annual salary. As discussed above, for general employees that percentage was 6.75% in 2019.
  For the city and county, the situation is more complicated. Instead of dictating a 50/50 split of the required contribution, Act 10 simply prohibits both governments from paying the employee share. Prior to Act 10, all three of Wisconsin’s pension plans calculated an employee contribution, but except for the city’s general employees the employer typically paid most or all of that contribution on behalf of its workers. Current law gives each of the Milwaukee plans the leeway to calculate the employee contribution as it sees fit and simply requires the employees to actually pay it, though any change is still subject to collective bargaining in the case of law enforcement officers and firefighters.
  Prior to Act 10, the city had established its employee contribution rate at 5.5% of salary, and it continues to apply that percentage to general employees hired before 2014. For those hired after 2014, a rate of 4% is used, which reflects the lower pension multiplier for those workers (though it is now the same as the one used by WRS). Public safety workers not covered by Act 10 have their contribution rate established through collective bargaining. The county established a new methodology for calculating the employee contribution after Act 10’s passage that has the employer and employees share equally the “normal cost” (i.e. the cost of benefits earned by workers each year), but applies only 20% of the cost of the unfunded liability to the employee share (the rest is for workers who are no longer with the county). For 2020, that produces a 6.2% contribution rate for general employees.
  The fact that city and county general employees are contributing lower percentages of their salaries to support their pensions than WRS participants may be reason to consider potential changes. Of course, as with other plan design changes, potential impacts on employee retention and recruitment as well as sound actuarial principles would need to be taken into account.

Transfer to WRS

Given WRS’ fiscal health and its high regard nationally, it is logical to ask whether the city and county in Milwaukee should simply close their plans and have their participants become part of WRS. Indeed, that approach has been seriously considered by Milwaukee County and its retirement task force, and it may receive similar consideration by the city.

For their part, WRS leaders have repeatedly expressed willingness to consider that option, but only for new employees and future years of service for active employees, and not for existing retirees and the benefits already earned by active employees.

That stipulation – which is understandable given the myriad legal and administrative issues involved – does create formidable challenges. For example:

• • Under a “soft freeze” scenario in which only new employees of the city and county would become members of WRS, those employees would contribute to that system, thus depriving the two existing plans of some revenue to offset the annual required contribution. To address that issue, either the active workers remaining in the county plan, the employer, or most likely both would have to contribute much more. In addition, as noted earlier, WRS has both more conservative assumptions and in some ways higher benefit levels for workers – both factors could increase employer contributions.
  • In a so-called “hard freeze,” both new city or county employees and the future service of active employees would be covered by WRS. While this scenario would leave the legacy plans responsible only for administering benefits already earned and would shorten their existence, it would essentially eliminate all employee contributions, as both active and new employees would contribute to WRS. That, in turn, would drive up the employer contribution to a much greater extent than the soft freeze scenario.

• Under either scenario, the city and county would be expected to contribute to WRS as well as pay off the unfunded liabilities in the legacy plans, thus producing a higher combined contribution for the foreseeable future. In reviewing a soft freeze scenario for the county, Pew projected that the need for two employer contributions would increase overall employer costs by $200 million over a 30-year period, or about 2% of payroll (see Figure 11). That estimate grew to $380 million, or 3.9% of payroll, for a form of hard freeze considered by the county task force because of the elimination of employee contributions from active employees.
• Finally, under a scenario in which the existing plans are closed or partially closed, national standards for actuaries might dictate that the amortization period for paying off the unfunded liability would need to be shortened. That requirement also would drive up the annual employer contribution over the relatively near term.

Despite the potential for higher short-term employer contributions, there are several benefits associated with a shift to WRS that merit consideration. One is the opportunity to ease and eventually eliminate the need for both governments to administer their pension plans, which for the county in particular would be a substantial relief because of its extremely complicated system. Perhaps more compelling is the opportunity to reduce risk by virtue of being part of a much larger and more stable pension system with greater economies of scale and more predictable employer contributions, as well as risk sharing with employees. The ability to provide parity in pension benefits for virtually all city, county, and state workers in Wisconsin also would be a sound policy outcome.

Consequently, additional analysis for both the city and county would seem warranted, including consideration of possible state assistance or other strategies to reduce or eliminate the short-term financial impacts for both governments so that they might realize the potential long-term benefits. Such an analysis might also explore whether and how much each Milwaukee system could save in plan administration and investment management costs through a merger.

CONCLUSION

It is somewhat ironic that while growing pension challenges are creating severe fiscal stress in Wisconsin’s largest city and county, those governments are located in a state that has seemingly developed a blueprint for effectively managing such challenges. This report has highlighted key elements of that blueprint that might be adopted by the city of Milwaukee and Milwaukee County and has considered whether the two governments should simply close their systems to new employees and join WRS.

As we have discussed, there are both pros and cons – as well as potential legal and logistical barriers – associated with any of the changes the city and county might contemplate to more closely mirror WRS or to have their employees be covered by it. As noted, the county’s task force recently spent more than a year examining and weighing the advantages and obstacles associated with several of those options. While it did not propose a clear-cut solution, it did recommend further consideration of both greater risk sharing and a shift of new county employees to WRS.

The alarming projected growth in the city’s employer contribution and the county’s huge infrastructure repair backlog should now create a new sense of urgency for leaders of both governments to consider those options. Those leaders recently have focused on a 1% county sales tax proposal to address the legitimate weaknesses in their revenue streams that are significant contributors to their ongoing budget challenges. We would suggest that equal emphasis needs to be placed on their growing pension costs, which are perhaps the foremost driver on the expenditure side of the ledger.

It is critical to note that as with the sales tax proposal, any efforts to comprehensively address Milwaukee’s city and county pension challenges likely will require assistance from state government. Specifically, the transfer of new city and county employees to WRS not only would require state approval to carry out, but also would seemingly call for consideration of some form of state assistance or additional flexibility for local revenues to help both governments manage the near-term financial hit and reap the long-term benefits.

Given the stakes involved – which include appropriate city police staffing, the needs of county parks and cultural institutions, and the area’s vulnerable citizens who depend on public services – a partnership between the state and its two largest local governments to identify a comprehensive solution would appear to be in order. We hope this report helps generate that needed discussion and partnership.