HOW WILL LONGER LIFESPANS AFFECT - PDF document

 S  L P P HOW WILL LONGER LIFESPANS AFFECT STATE AND LOCAL PENSION FUNDING? By Alicia H. Munnell, Jean-Pierre Aubry, and Mark Cafarelli* * Alicia H. Munnell is director of the Center for Retirement Research at Boston College (CRR) and the Peter F. Drucker Professor of Management Sciences at Boston College’s Carroll School of Management. Jean-Pierre Aubry is assistant director of state and local research at the CRR. Mark Cafarelli is a re - search associate at the CRR. The authors wish to thank David Blitzstein, Keith Brainard, Gene Kalwarski and his colleagues at Cheiron, Steven Kreisberg, and Joseph Silvestri for helpful comments. I The fact that people are living longer is good news from a human perspective. But longer lifespans also make dened benet pension plans more expensive because sponsors must pay benets to retirees for a longer period of time. The question is the extent to which state and local plans have already incorporated this pattern of continued longevity improvement into their cost estimates. For example, CalPERS – one of the nation’s largest plans – revised its longevity assumptions in , signicantly increasing its liabilities and reducing its funded ratio by  percent - age points. This change raises the question whether more cost increases due to longevity improvements are on the horizon. To answer the question, this brief explores what public plan liabilities and funded ratios would look like under two alternative scenarios: ) if public plans were required to use the new mortality table designed for private sector plans; and ) if public plans were required to go one step further and fully incorporate expected future mortality improvements. The discussion proceeds as follows. The rst section describes how public and private plans cur - rently incorporate longevity improvements into their cost estimates. The second section presents a simple model that relates the impact of improved longev - ity to liabilities, showing that, if beneciaries live an additional year, liabilities increase by . percent. The third section estimates the impact of changing the longevity assumptions to: ) the new standard stringent standard that incorporates future mortality improvements. The results suggest that, under the rst standard, public plans underestimate life expec - tancy by only . year. Adopting the second standard would increase life expectancy by . years and reduce the funded ratio of public plans from  percent to  percent. Of course, public plans vary signicantly, so the impacts would be much larger for some and smaller for others. LEARN MORE Search for other publications on this topic at: crr.bc.edu Center for Retirement Research  Note: Alabama Teachers Retirement System (), DC Teachers Retirement System (), and North Dakota Teachers Retirement System () use dierent mortality tables for male and female retirees. For these plans, the gure reects the male mortality tables. For a description of the various methods, see endnote . Source : Various pension plan actuarial valuations. E\n L\n\t\b I\t The private sector is under much more specic guid - ance than the public sector in terms of how to calcu - late expected mortality. The Pension Protection Act of  directed the Internal Revenue Service (IRS) to publish mortality tables for private sector funding calculations. Currently, these IRS tables are based on the RP- mortality table, which was constructed by the Society of Actuaries (SOA) with data from over  private pension plans for the period - . These mortality rates are then updated using the mortality improvement Scale AA.  In an eort to approximate future mortality improvements, the  IRS table actually uses estimated retiree mortality rates for . In , the SOA initiated a new study of mortal - ity trends, focusing on death rates of participants in private pension plans in . They then applied an updated mortality improvement scale, MP-, to create RP-. It is uncertain when these new tables will be adopted.  Unlike the private sector, public sector plans are not required to use a specic mortality table and, at the turn of the century, state and local plans used a F\r\f\n . D\t\b\f\b  M\b\b
T\n\t U\t\n 
L\r\n S\b\b\n  L P\n\t P\t,    wide variety of approaches (see Figure ). By , however,  percent of plans in the Public Plans Data - base (PPD) used the RP- as their base.  But the base table is only the starting point; public plan actu - aries make a variety of adjustments to align the tables with the expected mortality of their plan members.  The common adjustments are mortality improvement scales, setbacks, or some combination of the two. A mortality improvement scale species the pace at which mortality rates will decline each year. A set - back involves applying mortality rates at younger ages to older ages. For example, a -year setback would use age- mortality rates for a -year old. In developing mortality tables, actuaries use two dierent approaches: “static” and “generational.”  The static method is a snapshot of mortality rates at a given point in time. As noted, the IRS tables used by private plans choose a point in time that is seven years in the future in an eort to partially reect fu - ture mortality improvements. The generational meth - od goes further, fully incorporating anticipated future improvements in longevity. Interestingly, while state and local plans primarily use a static approach, they have been gradually moving toward an explicit gen - erational method (see Figure  on the next page).   11% RP-2000 GAM 1994 Other Proprietary GAM 1983 UP-1994 73%10%4%10%2%1% Issue in Brief  Figure  compares life expectancies for men and women at age  from old, scaled, new, and gen - erational tables. The rst comparison shows that – between the original RP- and the current  IRS table – life expectancy for healthy annuitants in - creased by . years for men and . years for women. The second comparison – between the  IRS table and life expectancy implied by RP- – suggests that the IRS tables (which, as noted, actually project mortality rates in ) do not fully account for all the gains in life expectancy that occurred from - .  The third comparison – between RP- life expectancy calculated on a static basis and on a gen - erational basis – shows that the latter adds . years for men and . years for women. The comparisons suggest that even though the IRS tables are intended to be up to date, they still show lower mortality improvements than the RP-. In addition, the ap - plication of generational tables to the RP-, which incorporate future improvements, would result in a further increase. H D D  L E\b A L? The overall goal of this analysis is to calculate how much applying private sector life expectancy assump - tions would aect public sector liabilities and funded status.  The rst step is to establish a relationship be - tween life expectancy and liabilities. To this end, we estimate a model where the present value of pension liabilities (L) is approximated as follows: L = p b -(+r) -n r This relationship can be transformed into a linear equation as follows: Log(L) = a+ ß  log(p) + ß  log(b) + ß  log(r) + ß  n + ß  n log(r) + , where p is the number of participants; b is the aver - age annual benet; r is the discount rate; and n – our life expectancy variable – is the average length of expected future payouts. The linear equation can then be estimated using data from the  state and local pension plans in the PPD over the period -.  The variable of interest is life expectancy, which reects the specic mortality assumptions for men and women for each year for each plan.  The PPD data suggest that the average age for current annuitants is  in police and re plans and  in plans for teachers and general employees, so the life expectancy is calculated at those ages for each type of plan. The male-female ratio is assumed to be - for police and re plans, - for teacher plans, and - for plans for general em - ployees; the life expectancies for men and women are weighted to reect these aggregate ratios. Source : Various pension plan actuarial valuations. F\r\f\n . N\f\n  S\b\b\n  L P\n\t P\t U\t\r G\n\n\b S\r, - Source : Authors’ calculations from Society of Actuaries (); and Internal Revenue Service (). F\r\f\n . L\n E\n\b
\b A\r\n    H\n\b­
A\f\b\b\t U\n V\f\t M\b\b
T\n\t 000023324591315220510152025 Women RP- IRS  RP- RP- generational ( ) Center for Retirement Research  The results in Table  show that state and local pension plans would see their liabilities increase by . percent for each additional year of life expec - tancy.  These results are consistent with previous research on private sector plans and hypothetical arrangements.  impact on liabilities. Finally, we recalculate the liabili - ties and reestimate the funded ratio.  The results of the exercise show that, on average, public plan life expectancies were . year lower than that implied by static RP- tables. This dierence means that liabilities would increase by . percent if plans adopted RP-, which would reduce the  funded status of state and local plans from  percent to  percent (see Figure ). If plans were required to adopt a generational, rather than a static, version of RP-, their assumptions would fall short by . years, implying an -percent increase in liabilities and a funded ratio of  percent. Notes: The data for liabilities and participants are for retirees only. The coe€cients report eects from an OLS estimation and are signicant at the -percent level. The model includes plan and year xed eects. Source : Authors’ calculations. T\n . F\b\t A\n\b\r P\n\t L\b\n\t Variables Coe€cients Log (p)Number of participants .‚ Log (b)Average benet level . ƒ Log (r)Discount rate -.„  Life expectancy . Constant . R-squared .„  Number of observations 1,750 U\n P P A  S  G RP-\r\f The results from the equation are then used to calcu - late what pension liabilities and funded ratios of state and local plans would be if liabilities were calculated based on the new RP- mortality table and then on a generational version of RP-. (We are not advocating that state and local plans adopt RP-, since their mortality experience is quite dierent from private plans. Rather, RP- is simply used as a benchmark.)  The exercise starts with each of the  plans’ cur - rent male-female weighted life expectancies at  or  and  liabilities and assets to get a base funded ratio. Public plans show enormous variation in their life expectancies (see Figure ). Life expectancy at  for men ranges from - years and for women from - years, which means that some of the high projections far exceed those implied by RP-. The next step is to compare each plan’s assumed life expectancy with that implied by RP- and multiply that dierence by . percent to estimate the Source : Authors’ calculations. F\r\f\n ƒ. D\t\b\f\b  S\b\b\n  L P\t 
A…\n\r\n L\n E\n\b
\b A\r\n   Source : Authors’ calculations. F\r\f\n  . A…\n\r\n F\f\n R\b 
M\b\b
T\n 793739251852100009234532238301020304050 Men Women 20%40%60%80% 100% Current mortalitytablesRP-2014 staticRP-2014generational Issue in Brief  The results for each of the  plans are shown in the Appendix. Three conclusions emerge from examining the individual plan data. First, the biggest decline in funded ratios occurs among the smallest plans; large plans appear to keep their life expectancy assumptions more up to date (see Figure a). Second, the decline in funded status appears to be inversely related to the initial funded level – that is, the worst funded plans tend have the most outdated mortality assumptions (see Figure b). Finally, adopting mor - tality assumptions designed for private plans appears to have a roughly equal impact on the funded ratio of plans for teachers (.-percent decline in funded ratio), general employees (.-percent decline), and police and re personnel (.-percent decline). C The question underlying this analysis is whether outdated mortality assumptions are a serious problem among state and local plans. The answer appears to be “no.” It’s true that if plans were to adopt the gen - erational version of RP-, the aggregate funded ratio would drop from an estimated  percent to  percent; but even the private sector is not considering using such low mortality rates. Simply adopting the static RP- would only reduce the funded ratio from  percent to  percent. In short, public sec - tor plans seem to be making a serious eort to keep their life expectancy assumptions up to date. The big increase in  of CalPERS’ liability and decline in funding was reective of an eort to better incorpo - rate future mortality improvements when estimating mortality, not a sign of a serious problem.  Source : Authors’ calculations. F\r\f\n\t   . P S†\n  F\f\n R\b  P\t ‡\b­ B\r\r\n\t\b  S\n\t\b D\n\n\t  F\f\n R\b  A\b\r G\n\n\b RP-ƒ   . Plan Size (Average Liability)   . Current Funded Ratio $11.7 63.6% 81.9% 0% 20% 60% 100% Biggest decline Smallest decline Billions Center for Retirement Research   For example, the mortality rate for a -year-old man in the RP- is . percent and the annual per - centage decline in mortality based on the Scale AA is . percent, so to calculate the mortality rate in  requires reducing the initial rate by . percent for  years – producing a  mortality rate of . percent.  Some critics suggest that, because of the sample used, RP- may be biased toward faster rates of longevity improvement. See American Academy of Actuaries Pension Committee ().  The PPD is developed and maintained through a collaboration of the Center for Retirement Research at Boston College, the Center for State and Local Government Excellence, and the National Associa - tion of State Retirement Administrators. It contains data for  large state and local plans –  state and  local – and accounts for  percent of assets and workers relative to the totals reported by the U.S. Census Bureau.  Plan actuaries perform periodic experience stud - ies (every three to ve years for most large plans) to ensure that assumptions used by the plan align with the plan’s actual mortality experience.  Alternative terms for “static” and “generational” projections of life expectancy are, respectively, “pe - riod” and “cohort.” An example of how the two ap - proaches dier may be helpful. Under the basic static method, for a -year-old in  the mortality rates at , ,  etc. are the rates applicable to individuals currently at those ages in . In contrast, a “genera - tional” approach would take into account that mortal - ity rates for individuals would likely decline in the future. Thus, for a -year-old in , the mortality rate at  would be that for a -year-old in ; at  that for a -year-old in , etc. Since death rates are projected to decline in the future, a static calcula - tion signicantly understates how long someone is actually likely to live.  Each mortality table is based on dierent sources of actual mortality experience. The RP- is described in the text. The UP- (Uninsured Pensioner) tables are based on group annuitant experience from -, the federal Civil Service Retirement Sys - tem experience, and Social Security’s Actuarial Study No. . The  GAM (Group Annuity Mortality) tables are based on the same experience as UP- except that the GAM- tables include a -percent margin designed for insurance reserves. The  GAM tables are based on insured group annuity ex - perience submitted by Prudential and by the Bankers Life, U.S. white population statistics for the period from -, Canadian population statistics from -, and mortality rates for persons covered under Medicare during -.  To test for consistency between the RP- and the RP- rates, SOA actuaries applied both the Scale AA and the Scale MP- to the RP- rates and concluded that the Scale MP- was more ac - curate.  The following analysis builds on a similar study by Kisser et al. () for private dened benet plans over the period -.  Complete historical data are not available for every plan, so the total number of observations is ,.  Life expectancy can be derived from mortality rates in three steps: ) compute survival rates from mortality rates – that is, a -percent chance of dy - ing turns into a -percent chance of surviving; ) calculate the probability of, say, a -year-old living to , to , to  and so on, where each year’s rate is the product of the previous survival rates; and ) sum the conditional survival rates to determine the number of years the -year-old is expected to live.  The dependent variable is the liability for annui - tants – that is, for those already retired. The percent - age increase in active worker liability will be of a similar order of magnitude.  Antolin () computes pension liabilities for a hypothetical pension fund that is closed to new en - trants and nds that an unexpected improvement in life expectancy of one year per decade could increase pension liabilities by - percent. Dushi, Friedberg, and Webb () nd that updating the mortality tables used to estimate the pension liabilities reported on Forms -K, which typically reect mortality rates in the early s, would increase life expectancy at age  by about three years and increase liabilities by  percent for the average male plan participant. Kisser et al. () estimate the above equation for E Issue in Brief  R American Academy of Actuaries Pension Committee. . RP- and MP- Comments . Washing - ton, DC. Antolin, Pablo. . “Longevity Risk and Private Pen - sions.” OECD Working Papers on Insurance and Private Pensions :-. Paris, France. CalPERS Actuarial O€ce. . CalPERS Experience Study and Review of Actuarial Assumptions . Sacra - mento, CA. Dushi, Irena, Leora Friedberg, and Anthony Webb. . “The Impact of Aggregate Mortality Risk on Dened Benet Pension Plans.” Journal of Pension Economics and Finance : -. Internal Revenue Service. . Updated Static Mor - tality Tables for the Years  and  . Notice -. Washington, DC. Kisser, Michael, John Ki, Eric Oppers, and Mauricio Soto. . “The Impact of Longevity Improve - ments on U.S. Corporate Dened Benet Plans.” Working Paper. Washington, DC: International Monetary Fund. Public Plans Database . -. Center for Retire - ment Research at Boston College, Center for State and Local Government Excellence, and National Association of State Retirement Administrators. Society of Actuaries. . Mortality and Other Rate Tables . Chicago, IL. Database available at http:// mort.soa.org/ private dened benet plans and nd that an addi - tional year of life expectancy increases liabilities by about  percent.  Public plans were excluded from the mortality data used to create RP- because their mortality experience diered signicantly from those of private plans for which the RP- table was devised. In response to comments, the SOA recommended a separate study of public plan mortality experience, with the expectation that the study would include separate tables for public safety workers, teachers, and other public entities.  The variation in assumptions and methodology means that some rules are required to determine how plans would respond to the imposition of RP-. First – for plans that currently use the static method – if a plan’s current life expectancy exceeds that implied by RP-, we assume that the plan maintains its current life expectancy under the RP- static sce - nario. In these cases, to project life expectancy under the generational approach, we add the dierence between the RP- static and generational assump - tions to the plan’s own static assumption. Second – for plans that currently use the generational method – we calculate a new life expectancy only under the RP- generational scenario and do not include any estimate of life expectancy under the RP- static scenario.  Specically, CalPERS shifted from virtually no projection of future mortality improvement to a - year static projection (the approximate duration of CalPERS’ benet payments). APPENDIX Issue in Brief  Total ˆ.„ ‚.ƒ . ˆ ˆ ˆ Alabama ERS  . ‚. „.     Alabama Teachers ‚. ‚. .ƒ    Alameda County Employee's Retirement Association ˆ. ‚. „. ˆ ˆƒ ˆ Alaska PERS ˆ. ‚. „.     ƒ   Alaska Teachers . . . ƒ‚ ƒ‚ ƒ  Arizona Public Safety Personnel . .  .ƒ  „     Arizona SRS ‚.„ ‚.„ .  ˆ  ˆ  ˆ Arizona State Corrections O€cers ‚.‚ .  .ƒ ˆ   ˆ Arkansas PERS .„ ‚. „. ˆƒ ˆ ‚ Arkansas Teachers ‚.„ ‚.„ . ˆ ˆ ‚ Boston Retirement Board .„ ‚. „.     California PERF ‚. ‚. „. ‚ ‚ ˆ„ California Teachers ‚. ‚. .ƒ ˆ ˆ  Chicago Municipal Employees . ‚. „. ˆ    Chicago Police „. .  .ƒ  ‚  Chicago Teachers . n/a . ƒ„ ƒ„ ƒ City of Austin ERS ‚.ˆ n/a . ˆ ˆ ˆ Colorado Municipal ‚.  ‚.  . ˆ ˆ „ Colorado School ‚.  ‚.  .     Colorado State ‚.  ‚.  .  ˆ  ˆ  ƒ Connecticut Municipal . ‚. „. ‚‚ ‚ ˆ‚ Connecticut SERS ‚.ƒ ‚.ƒ . ƒ ƒ „ Connecticut Teachers „. „. .ƒ  ˆ  ˆ   Contra Costa County ‚. ‚. . ˆ ˆ ˆ Cook County Employees ‚.ˆ n/a .  ˆ  ˆ  ƒ Dallas Police and Fire „. .  .ƒ ˆ ˆ  DC Police & Fire . .  .ƒ    „„ DC Teachers „.ƒ „.ƒ .  „ „ ‚ƒ Delaware State Employees ˆ. n/a „. „ „ ‚ Denver Schools .„ ‚. .ƒ ‚ ˆ‚ ˆ Duluth Teachers „.„ n/a .  ƒ  ƒ   Fairfax County Schools „. „. .ƒ ˆ  ˆ  ˆ Florida RS ˆ.  ‚. „. ‚  ‚ƒ ˆ„ A\n T\n . L\n E\n\b
 F\f\n R\b  S\b\b\n  L P\t \f\n C\f\n\b  RP-ƒ M\b\b
A\t\t\f\b\t Plan name Life expectancy Funded ratio RP-ƒ RP-ƒ Current Current Static Static Generational Generational % % % Georgia ERS  . ‚. „. ˆ    Georgia Teachers „. „. . ‚ ‚ ˆ  Hawaii ERS ‚. ‚. „.    ˆ Houston Fireghters .„ .  .ƒ ‚ˆ ‚  ‚ Idaho PERS .‚ ‚. „. ‚  ‚ ˆˆ Illinois Municipal ˆ. ‚. „. ‚‚ ‚  ‚ Illinois SERS ˆ. ‚. „. ƒ   Illinois Teachers „. „. . ƒ ƒ ‚ Illinois Universities ‚. ‚. .ˆ ƒ ƒ „ Indiana PERF .ƒ ‚. „. ‚ ˆ ˆ Indiana Teachers ‚. ‚. .ƒ ƒ ƒ ƒ Iowa Municipal Fire and Police .  .  .ƒ ˆƒ ˆƒ „ Iowa PERS . ‚. „. ‚ ˆ ˆ Kansas PERS  .  n/a „.     Kentucky County  . ‚. „.  „      Kentucky ERS  . ‚. „.  ƒ  Kentucky Teachers ‚.ƒ ‚.ƒ .      ƒ‚ Kern County Employees Retirement Association ˆ.ƒ ‚. „.    ˆ LA County ERS ‚. ‚. „.„ ˆ  ˆ  ˆ Los Angeles City Employees Retirement System ˆ. ‚. „. „ ˆ  Los Angeles Fire and Police .„ .  .ƒ ‚ ‚ ˆ Los Angeles Water and Power ˆ. ‚. „. ˆ„ ˆˆ ˆ Louisiana Municipal Police . .  .ƒ ƒ   ˆ Louisiana Schools . ‚. „.   ‚    Louisiana SERS . ‚. „.      Louisiana State Parochial Employees .  ‚. „. „ ‚‚ ‚ Louisiana Teachers ‚. ‚. .ƒ       Maine Local ˆ. ‚. „. ‚‚ ‚  ‚ Maine State and Teacher ‚. ‚. .ƒ ˆ‚ ˆ‚ ˆ Maryland PERS  . ‚. „.   ‚    Maryland Teachers ‚.‚ ‚.‚ .„ ˆ ˆ  Massachusetts SRS ˆ. ‚. „. „ ˆ  Massachusetts Teachers ˆ.  ‚. .ƒ    ƒ   Michigan Municipal .„ ‚. „. ˆ „   Center for Retirement Research  Plan name Life expectancy Funded ratio RP-ƒ RP-ƒ Current Current Static Static Generational Generational % % % Michigan Public Schools ˆ. ‚. .ƒ   ˆ   Michigan SERS . ‚. „.   ˆ  ƒ Milwaukee City ERS ‚. n/a „. „  „  „ Minneapolis ERF ˆ.ƒ ‚. „. ˆƒ ˆ „ Minnesota GERF „. n/a . ˆ ˆ „ Minnesota Police and Fire Retirement Fund . n/a ƒ. ‚ ‚ ˆ Minnesota State Employees ‚.‚ n/a .ƒ ‚ ‚ ˆ‚ Minnesota Teachers .ƒ n/a .  ˆ ˆ ˆ Mississippi PERS .‚ ‚. „.  ‚      Missouri DOT and Highway Patrol .„ .  .ƒ ƒ ƒ ˆ Missouri Local .  ‚. „. ‚ˆ ‚ ˆ‚ Missouri PEERS ˆ. ‚. „. ‚ ‚ ˆ  Missouri State Employees ˆ. ‚. „. ˆ ˆ ˆ Missouri Teachers ‚. ‚. .  ‚ ‚ ˆƒ Montana PERS ˆ. ‚. „. ‚ ˆˆ ˆ Montana Teachers ‚.„ ‚.„ . ˆ ˆ  Nebraska Schools ‚.„ ‚.„ . ˆˆ ˆˆ ˆ Nevada Police O€cer and Fireghter „.ƒ .  .ƒ ˆ   Nevada Regular Employees ˆ. ‚. „. „ ˆ  New Hampshire Retirement System ‚. ‚. .  ˆ  ˆ  ƒ New Jersey PERS .„ ‚. „.    ˆ New Jersey Police & Fire . n/a .ƒ ˆ ˆ ˆ New Jersey Teachers ‚. n/a .ƒ  ˆ  ˆ   New Mexico PERA ˆ. ‚. „. ˆ ˆ ˆ New Mexico Teachers ‚.ˆ ‚.ˆ .„     New York City ERS ‚. ‚. „. ‚ ‚   New York City Fire .  .  .ƒ  ƒ  ƒ   New York City Police .  .  .ƒ ˆ ˆ  New York City Teachers ‚. ‚. .ƒ  ‚  ‚  ƒ New York State Teachers ‚. ‚. .ƒ ‚‚ ‚‚ ‚ North Carolina Local Government  .ƒ ‚. „.  „ ‚ North Carolina Teachers and State Employees ‚. ‚. „. „ƒ „ƒ ‚„ North Dakota PERS ‚.  ‚.  .    „ Issue in Brief  Plan name Life expectancy Funded ratio RP-ƒ RP-ƒ Current Current Static Static Generational Generational % % % North Dakota Teachers ‚.‚ ‚.‚ .„  „  „    NY State & Local ERS ‚. n/a „. ‚„ ‚„ ‚ƒ NY State & Local Police & Fire .  n/a .ƒ „ „ ‚ƒ Ohio PERS ˆ. ‚. „. ‚ ‚ ˆ Ohio Police & Fire „. n/a .ƒ ˆ ˆ  ‚ Ohio School Employees ˆ.ˆ ‚. „.      Ohio Teachers ‚.ˆ ‚.ˆ .„    Oklahoma PERS .ˆ ‚. „. ‚ ˆ‚ ˆƒ Oklahoma Police Pension and Retirement System . .  .ƒ ‚„ ‚  ‚ Oklahoma Teachers „. „. .  ˆ  ˆ   Orange County ERS ‚.  ‚.  .    Oregon PERS ‚. n/a „. „ „ ‚ Pennsylvania Municipal Retirement System . ‚. „. „„ „ ‚‚ Pennsylvania School Employees ‚.  ‚.  . ƒ ƒ  „ Pennsylvania State ERS . ‚. „.  „     Philadelphia Municipal Retirement System ƒ.ˆ ‚. „. ƒˆ ƒ ƒ Phoenix ERS . ‚. „. ƒ   ˆ Rhode Island ERS ˆ. ‚. „.  ˆ      Rhode Island Municipal ˆ. ‚. „. ‚ ˆ„ ˆ  Sacramento County ERS ˆ.ˆ ‚. „. ‚ ‚ ˆ‚ San Diego City ERS . ‚. „. ˆ   San Diego County ˆ.ˆ ‚. „. ˆ„ ˆ‚ ˆƒ San Francisco City & County ˆ. ‚. „. ‚ ˆ„ ˆƒ South Carolina Police ‚. .  .ƒ „   ‚ South Carolina RS .‚ ‚. „.    ˆ South Dakota RS ˆ. ‚. „.  „ˆ „ St. Louis School Employees . ‚. .ƒ ‚ƒ ˆ„ ˆ St. Paul Teachers „.‚ „.‚ .     Texas County & District ˆ. n/a „. ‚„ ‚„ ‚ Texas ERS ‚. n/a „.‚ ‚ ‚ ˆ  Texas LECOS ‚. n/a .ƒ ˆ ˆ  Texas Municipal ‚. n/a „.„ ‚ƒ ‚ƒ ‚ Texas Teachers „. „. .ˆ ‚ ‚ ˆ  TN Political Subdivisions ‚. ‚. „. „  „  „ TN State and Teachers ‚. ‚. „. „ „ ‚‚ Center for Retirement Research  Plan name Life expectancy Funded ratio RP-ƒ RP-ƒ Current Current Static Static Generational Generational % % % University of California „. „. . ˆ ˆ ˆ Utah Noncontributory ˆ. ‚. „. ‚ ‚ ˆ Utah Public Safety .  .  .ƒ ˆ ˆ ‚ Vermont State Employees .ˆ ‚. „. ˆˆ ˆ ˆ Vermont Teachers „. „. .ƒ     Virginia Retirement System ˆ.ˆ ‚. „.     Washington LEOFF Plan  . n/a .ƒ      Washington PERS / . n/a .‚   „ˆ Washington School Employees Plan / . n/a .‚   „ Washington Teachers Plan / .„ n/a ƒ.     „‚ West Virginia PERS . ‚. „. ‚ ˆƒ ˆ West Virginia Teachers .‚ ‚. .ƒ  ‚      Wisconsin Retirement System ‚. ‚. „.   „  Issue in Brief  Plan name Life expectancy Funded ratio RP-ƒ RP-ƒ Current Current Static Static Generational Generational Source : Authors’ calculations based on various actuarial valuations. % % % A  C The mission of the Center for Retirement Research at Boston College is to produce rst-class research and educational tools and forge a strong link between the academic community and decision-makers in the public and private sectors around an issue of criti - cal importance to the nation’s future. To achieve this mission, the Center sponsors a wide variety of research projects, transmits new ndings to a broad audience, trains new scholars, and broadens access to valuable data sources. Since its inception in , the Center has established a reputation as an authorita - tive source of information on all major aspects of the retirement income debate. A I The Brookings Institution Massachusetts Institute of Technology Syracuse University Urban Institute C I Center for Retirement Research Boston College Hovey House  Commonwealth Avenue Chestnut Hill, MA - Phone: () - Fax: () - E-mail: crr@bc.edu Website: http://crr.bc.edu © \r\f, by Trustees of Boston College, Center for Retirement Research. All rights reserved. Short sections of text, not to exceed two paragraphs, may be quoted without explicit permission provided that the authors are identied and full credit, including copyright notice, is given to Trustees of Boston College, Center for Retirement Research. The CRR gratefully acknowledges the Center for State and Local Government Excellence for its support of this research. The Center for State and Local Government Excellence ( http://www.slge.org ) is a proud partner in seeking retirement security for public sector employees, part of its mission to attract and retain talented individuals to public service. The opinions an conclusions expressed in this brief are solely those of the authors and do not represent the opinions or policy of the CRR or the Center for State and Local Government Excellence. pubplans.bc.edu Visit the: Center for Retirement Research