Report t,o Congressional Committees GAO -- Au#rst IWO PRIVATE PENSIONS Impact of New Vesting Rules Similar for Women and Men Th IllIIIII 142052 ----- 1 . United States GAO General Accounting Office Washington, D.C. 20648 Human Resources Division B-239063 August 21,199O The Honorable Lloyd Bentsen Chairman, Joint Committee on Taxation Congress of the United States The Honorable Lloyd Bentsen Chairman, Committee on Finance United States Senate The Honorable Edward M. Kennedy Chairman, Committee on Labor and Human Resources United States Senate The Honorable Dan Rostenkowski Chairman, Committee on Ways and Means House of Representatives The Honorable Augustus F. Hawkins Chairman, Committee on Education and Labor House of Representatives For workers who change jobs, vesting in pension benefits can add to retirement income. Vesting-gaining the nonforfeitable right or entitle- ment to employer-provided pension benefits-is largely dependent on years of employment with the company sponsoring the pension plan.’ Federal rules limit how long a participant in a qualified plan must wait to vest in pension benefits.” The Tax Reform Act of 1986 (TRA) cut the maximum allowable vesting period in half for most workers in qualified private pension plans. TRA targeted plans that were not “top-heavy.” Top-heavy plans are those in which over 60 percent of the benefits or contributions go to company owners or other key employees.” ‘A worker is always fully vested in any benefits derived from his or her own contributions to the plan. This report only deals with participants’ vesting in employer-provided benefits, ‘Employer-sponsored plans that qualify for preferential tax treatment must comply with a variety of federal rules, including vesting rules, designed to improve the equity and security of benefits. “Top-heavy plans must vest workers more quickly than those under TFL4and observe other special rules. Generally, the smaller the plan, the more likely it is top-heavy. Page 1 GAO/HRD-90-101 Impact of TRA Vesting Rules on Workers Prior vesting rules do not meet the needs of many workers who change jobs frequently and so do not vest in their pension plans, the Joint Com- mittee on Taxation stated in describing the need for TRA.Women com- prise one group singled out aa being disadvantaged by these rules; they tend to be more mobile employees who are less likely to vest in pension benefits. TRA’S more rapid vesting would enhance the retirement income security of shorter-tenured workers, the Committee concluded, by enti- tling them to some pension benefits. Under the old rules, these workers, some of whom changed jobs voluntarily or lost jobs due to layoffs or plant closings, probably would not have been entitled to pension bene- fits. Longer-tenured workers would be entitled to pension benefits at retirement under either set of rules. This report estimates the impact of TRA vesting rules on participants in plans not considered top-heavy. We prepared it pursuant to a require- ment in the Retirement Equity Act of 1984 (REA) that we study the effect of federal pension rules, including vesting rules, on women.4 To do so, we examined a sample of pension plans and compared . the proportion of women and men vested under the old and the new rules and l the change in vested benefits due to TRAfor women and men. In addition, to gauge the impact that further reductions in vesting stan- dards might have, we estimated the effect on both women and men of applying rules for shorter vesting than under TRA. The vesting changes in the Tax Reform Act of 1986 will improve the Results in Brief vesting status of shorter-tenured workers with a similar effect on women and men. Judging by our sample, about three-fourths of all pen- sion participants in 1986 would have been vested in their pension bene- fits had the TRA rules been in effect then, In comparison, about half the participants were vested under the rules then in effect. In defined ben- efit plans, which contained most participants in our analysis, about 4 in 10 participants would be affected by TRA.” Among participants not vested under TRA,an estimated 1 in 3 would be vested if plans were 4!3eeRelated GAO Products for other reports prepared under the REA requirement. “In a defined benefit plan, the employer promises a specific retirement benefit that is generally based on a worker’s years of service, earn@s, or both. The employer is responsible for fundii the plan sufficiently to pay promised benefits. Page 2 GAO/HRD-90-101 Impact of TRA Vesting Rules on Workere R-239063 required to use the faster top-heavy rules instead of the TM rules. (See wp. I.1 If the defined benefit participants with increased vested benefits under TRA (4 in 10 in our analysis) left their plans tomorrow, they would be entitled to additional vested benefits equal to an average of 5 percent of their compensation. The median annual gain in the dollar value of vested benefits for those affected by TRA is about $1,240 (in 1990 dollars). Vesting standards first were established for all qualified private pension Background plans by the Employee Retirement Income Security Act of 1974 (ERISA). These standards limit the waiting periods for participants to vest in employer-provided pension benefits. Plans may use any vesting method, as long as the waiting period does not exceed ERISA’S standards. Two common methods are “cliff” and “graded.” l Under cliff vesting, a participant does not secure the right to future ben- efits until employed a specified number of years, at which time the par- ticipant becomes fully vested. For a cliff schedule, ERISA required that participants not be made to wait more than 10 years to move from nonvested to fully vested status. About 20 percent of the plans we ana- lyzed used the cliff method. They tended to be defined benefit plans sponsored by employers with 100 or more employees (large employers) that contained over 800 participants on average. . Under graded vesting, a participant gains partial vesting rights after a specified length of service and the percentage periodically increases until the participant is fully vested. ERISA’S standard for a graded schedule required that, at a minimum, a participant be partially vested after 5 years of service and his or her vesting rights increase by a fixed percentage each year until full vesting is reached after 16 years. About 70 percent of the plans in our analysis used the graded method of vesting. Most were defined contribution plans sponsored by employers with less than 100 employees (small employers) that contained fewer than 20 participants on average.” T&4provisions accelerated vesting for many pension plans, effective in 1989. Specifically, TRA reduced the maximum years workers must wait “In a defined contribution plan, a formula specifies the rate at which the employer makes contribu- tions to each participant’s account. The retirement benefit will depend on the amount of contributions and the investment experience of the account. Page 3 GAO/HRD-90-101 Impact of TR.A Vesting Rules on Workers IL233333 L for full vesting from 10 to 5 for cliff vesting.7 For graded vesting, TRA lowered the 5 to E-year standard to 3-to-7 years. Nine of 10 plans sponsored by large employers will need to provide more rapid vesting to comply with TRA, compared with about half the plans sponsored by small employers. The remainder would not have to change because they already used shorter time limits for vesting than TRA requires. To estimate the effect of TRA vesting changes on participants in private Scopeand pension plans that were not top-heavy, we used nationally representa- Methodology tive data already gathered in response to REA. These data were from surveys of two samples of private pension plans operating in 1984 and 1986. (See app. II.) The plan and participant data covered the most recently completed plan year for which information was available, usu- ally ending in 1986, before TRAbecame effective. Our estimates of the effect of the TRA vesting changes on the proportion of women and men vested are representative of approximately 6.1 mil- lion participants in about 26,100 pension plans (see app. II).* We used participant tenure data and plan vesting schedules to simulate the TRA vesting changes for women and men separate1y.OWe also simulated vesting changes assuming that the faster vesting rules that apply to top- heavy plans-3-year cliff vesting and 2- to 6-year graded vesting- applied to this universe of plans. Our estimates about the effect of TRA on the dollar value of vested bene- fits were limited to defined benefit plans in our universe.lO These plans comprised about 30 percent of our universe but contained about 60 per- cent of the participants, an estimated 3.8 million people. Where we have reported differences between women and men or types of plans, these differences are significant at the 96-percent confidence ‘Multiemployer plans satisfy TRA’s vesting requirements if employees covered by collective bar- gaining agreements are fully vested after 10 years. HOfthe 26,100 plans represented, small employers sponsored about 16,700 with about 270,000 par- ticipants, and large employers sponsored about 9,400 with about 6.8 million participants. “Evaluating whether women’s and men’s tenure and mobility patterns would be affected differently by TRA was outside the scope of our work. loWe did not have the information required for these calculations for participants in defined contri- bution plans. Page 4 GAO/HRDQO-101 Impact of TRA Vesting Rules on Workem R-239063 level. This means that there is less than a S-percent chance that we would have identified differences from our sample that do not exist in the universe (see app. III). We were unsure if any employer-sponsors would change their plans’ vesting methods (that is, from graded to cliff or cliff to graded). Hence, we assumed that plans would use the same method after TRA as before. This assumption was reasonable because employer-sponsors’ choice of particular vesting methods under the old rules was based on considera- tions specific to the plan and company. These include workforce demographics, administrative burden, and plan costs. We had no reason to expect the sponsors’ rationale to change under TM. But to the extent that companies change methods to comply with TRA, actual vesting changes might differ from our estimates. More participants (an estimated 76 percent) will be vested in their pen- More Participants sion benefits under TRA than under the old rules (63 percent). TRA’S Vested Under TRA effect will be similar for women and men (see fig. 1). This is because Rules similar proportions of women and men were in the range of service where they were not vested under the old rules but were vested under the TFU rules, for example, from 6 to 9 years of service in a lo-year cliff vesting plan.” If the plans are required to use the faster vesting sched- ules that apply to top-heavy plans, about 86 percent of participants will be vested (see app. I). ’ ‘Women and men in our sample had similar average years of tenure --the mean tenure for both was 8, the median, 6. Page IS GAO/HRD9O-101 Impact of TRA Vesting Rules on Workers IS239063 , Flgure 1: Impact of TRA on Vesting game for Women and Men 100 Psrcsnt ot Participants Vsstod All Womon Man Paftlcipant oroupa Old Rules TRA Rules NP6.1 million participants in total. N=2.Q million women. N&3.2 million men. TRA will have a greater impact on participants in plans using the “all-or- nothing” cliff vesting method than on those in plans using graded vesting (see fig. 2). Under the old rules, many participants in cliff plans did not work long enough to become vested but would be fully vested (100 percent) under TRA.12 Participants in graded plans are more likely to be at least partially vested under the old rules. The TRA rules tend to increase these participants’ level of vesting incrementally rather than from 0- to loo-percent. In either case, participants who were fully vested under the old rules would remain fully vested under TRA. “Almost 90 percent of cliff vesting plans used 10 years as the waiting period under the old rules. Page 6 GAO/HRD-90-101 Impact of TR.A Vesting Rules on Workers Figure 2: Impact of TRA on Vesting &eater in Plans Using Cliff Method loo Pemenfof Particlpnt~~ Vastod cliff Graded Vostlng Method Uaod I Old Rules TRA Rules N9.1 million participants in cliff vesting plans. N~2.7 million participants in graded vesting plans. For 4 in 10 participants in defined benefit plans, TRA will increase the Vested Benefit value of vested benefits (see fig. 3). If these participants left their plans Increases Under TIXA tomorrow, they would be entitled to more benefits every year in retire- Large for Some, ment than they were entitled to under the old rules. For 1 in 10 partici- pants, the increase would be over $2,208.13 Increases would range from Nonexistent for Many $2 to $76,217. The median gain would be $1,242 or 5 percent of compen- sation (see table 1). ‘“Plans usually reported dollar benefit data for the plan year ending in 1986. We adjusted these data for inflation to represent 1990 dollars, using the Consumer Price Index. Page 7 GAO/HRD-@MO1 Impact of TRA Vesting Rules on Workers B288063 Figure 3: TRA Will Increase Verted Benefits for 4 in 10 Participant8 In Deflned Benefit Plans Increased vested benefits under TRA Fully vested under old rules and TRA Not vested under old rules and TRA 3.8 million participants in defined benefit plans. Table 1: Impact of TRA on Dollars of Vested Benefits Median increase No. of participants Annual amount Percent of Type of plan (in millions) (1990 dollars) compensation All 1.5 $1,242 5.4 Women 0.8 980 5.1 ~- Men 0.7 1,987 5.9 Cliff vesting 1.2 1,546 5.9 Women 0.6 1,118 5.6 Men 0.6 2,263 7.0 Graded vesting 0.3 223 1.0 Women 0.16 160 1.1 Men 0.16 297 1.0 The stream of additional benefits that participants would receive in retirement (each year from normal retirement age until death) as a result of TRA can be expressed as a present value. This “present value” represents the value to participants in 1990 of the stream of additional benefits in retirement. It is the amount that, if invested today at a given rate of compound interest, would generate annual benefits during retire- ment equal to the additional vested benefits under TRA (see app. II). Page 8 GAO/HRD-90-101 Impact of TRA Vesting Rules on Workers R-239033 The gain in annual benefits under TRA for women is about half that for men, though the gain is a similar percentage of compensation. The pre- sent value of the increase for women is less than half that for men and is a smaller percentage of compensation.14 Under the TRA rules, men would receive about $2,000 more annually in retirement on average, based on our analysis. The present value of this increase in benefits during retire- ment is $3,011, or about 9.6 percent of compensation. Women would receive about $1,000 more annually in retirement as a result of TM. The present value of this increase is about $1,169, or 6.1 percent of compensation. The difference in present values for women and men is greater than the difference in the annual increase in vested benefits because women in our analysis are younger than men and women usually start receiving retirement benefits later than men. If the stream of benefits in retire- ment for men and women were the same, the present value for women still would be less than for men because of these differences in ages and retirement times. Participants in plans with cliff vesting schedules, where the participant is either not vested or is fully vested, will have substantially larger increases in vested benefits than participants under graded schedules (see table 1). This is true for both dollar increases and increases relative to compensation. For employers, the effect of increased vesting under TRA on annual pen- sion plan costs probably will be relatively small. It will vary with the type of plan and its actuarial profile. An employer who experiences little turnover in employees will have few additional pension costs because plan participants eventually would have the same amounts vested under the old rules, Conversely, an employer with a higher turn- over rate could have added costs, but only to the extent that the new rules add to the amount of vested benefits terminated employees would have under the old rules. The additional cost of 5-year cliff vesting (TRA’S standard) was estimated by the Employee Benefit Research Insti- tute (EBRI) at 2 to 7 percent of private pension plan contributions to the system as a whole.‘” ‘“See app. III for sampling errors associated with these estimates. “EBRI, “Pension Vesting Standards: ERISA and Beyond,” Issue Brief, Feb. 1986. Page 9 GAO/HRD@O-101 Impact of TRA Vesting Rules on Workers B233003 L The effect on the federal treasury in terms of additional tax expendi- tures for qualified pension plans due to increased vested benefits prob- ably will be small also. The Joint Committee on Taxation estimated in 1987 that the TRA vesting changes would have a “negligible effect” on tax receipts because of their effect on employers’ costs and employees’ tax-deferred inc0me.l” These are the main sources of the revenue for- gone through the preferential tax treatment granted qualified pension plans. Nothing in our analysis leads us to expect otherwise. We did not obtain written comments on this report because we were not reviewing specific agency functions or programs. However, we discussed the contents of the report with representatives of the Department of Labor and the Internal Revenue Service and incorporated their com- ments where appropriate. Copies of this report are being sent to other interested congressional committees, the Secretary of Labor, and the Commissioner of the Internal Revenue Service, and will be available to others upon request. If you have questions about information in the report, please call me at (202) 2756193. Other major contributors are listed in appendix IV. Joseph F. Delfico v Director, Income Security Issues ‘“Staff of the Joint Committee on Taxation, General Explanation of the Tax Reform Act of 1986,OQth Gong., 2nd Sess., 1987. Page 10 GAO/HRD9@1O1 Impact of TRA Vesting Rules on Workers Y Page 11 GAO/HRDBO-101 Impact of TRA Vesting Rules on Workers --. 7 Contents Letter 1 Appendix I 14 Top-Heavy Rules Result in Some Additional Increases in Vesting Appendix II 17 Further Details on Plans Sponsored by Small Employers 17 Plans Sponsored by Large Employers 19 GAO’s Scopeand Scope of GAO’s Analysis of TRA’s Vesting Changes 21 Methodology Calculating the Present Value of the Dollar Change in 21 Vested Benefits Appendix III 23 Sampling Errors for Estimates Appendix IV 26 Major Contributors to This Report Related GAO Products 28 Tables Table 1: Impact of TRA on Dollars of Vested Benefits 8 Table II. 1: GAO’s Universe and Sample of Plans 18 Sponsored by Small Employers Table 11.2:GAO’s Universe and Sample of Plans 20 Sponsored by Large Employers Table 111.1:Sampling Errors for Figure 1 and Figure I. 1 23 Table 111.2:Sampling Errors for Figure 2 24 Table 111.3:Sampling Errors for Figure 3 24 Table 111.4:Sampling Errors for Figures I.2 and I.3 24 Table 111.6:Sampling Errors for Table 1 25 Page 12 GAO/HRD-MI-101 Impact of TRA Vesting Rules on Workera _ _ ,_ . __ .._.__.. Contents Figures Figure 1: Impact of TRA on Vesting Same for Women and 6 Men Figure 2: Impact of TRA on Vesting Greater in Plans 7 Using Cliff Method Figure 3: TRA Will Increase Vested Benefits for 4 in 10 8 Participants in Defined Benefit Plans Figure I. 1: Top-Heavy Rules Improve Vesting of Some 14 Participants Figure 1.2: Many Participants Not Affected by Top-Heavy 16 Vesting Rules Figure 1.3: Impact of Top-Heavy Rules on Vesting Greater 16 in Defined Benefit Plans Using Graded Method Abbreviations EBRI Employee Benefit Research Institute ERISA Employee Retirement Income Security Act of 1974 REX Retirement Equity Act of 1984 TRA Tax Reform Act of 1986 Page 13 GAO/HRD9O-101 Impact of TRA Vesting Rules on Workers , Appendix I Top-Heavy RulesResult in SomeAdditional * Increasesin Vesting We applied the top-heavy standards to gauge the impact that these faster vesting standards might have compared to the TRA rules. This appendix contains the results of the top-heavy vesting simulation. Some participants who were not vested under the TRA rules would be vested under the standards that apply to top-heavy plans. Under the top-heavy rules, 85 percent of participants would be vested compared with 76 percent under TRA (see fig. 1.1). The top-heavy rules would improve vesting for about the same proportion of men and women, as was the case with the TRA rules. Figure 1.1:Top-Heavy Rules Improve Verting of Some Participants 100 Percent of Participants Vested All Women Men Participant Groups Old Rules TRA Rules Top-HeavyRules N-6.1 million participants in total. N-G.9 million women. Nd.2 million men. Page 14 GAO/HRD-90-101 Impact of TRA Vesting Rules on Workers Appendix I Top-Heavy Rules Result in Some Additional Increases in vesttng Most participants in defined benefit plans would be fully vested under the TRA rules and would remain so under the top-heavy rules (see fig. 1.2). An estimated 18 percent of participants in defined benefit plans will have increased vested benefits under the top-heavy rules. Figure 1.2:Many Pnrticlpants Not Affected by Top-Heavy Vesting Rules Increased Vested Benefits Under Top-Heavy Rules Keested Under TRA and Top-Heavy Fu;,lisVested Under TRA and Top-Heavy Nd.8 million participants in defined benefit plans. The effect of the top-heavy rules would be similar for women and men, but participants in graded vesting plans would be more likely to be vested under the top-heavy rules but not the TRA rules (see fig. 1.3). This is because participants in graded vesting plans tend to have shorter tenure than participants in cliff vesting plans. In fact, participants in defined benefit plans with graded vesting had an average tenure of about 6 years, compared to almost 9 years for participants in defined benefit plans with cliff vesting. Page 15 GAO/HRD-90-101 Impact of TRA Vesting Rules on Workers Appendix I Top-Heavy Rules Result in Some Additional Increases In Vesting Figure 1.3: Impact of Top-Heavy Rules on Vesting Qreater in Defined Benefit Plans Using Qraded Method Porcont ot Pattlclpants 100 Qmded Cliff Typa of Vntlng VestedBenefit8IncreasedUnderTop-HeavyRules Not Vested Under TRA and Top-Heavy Rules Fully Vested Under TRA and Top-Heavy Rules N10.7 million participants in defined benefit plans with gradad vesting. N4.1 million participants in defined benefit plans with cliff vesting. Among the almost 20 percent of participants who would have increased vested benefits using the top-heavy rules, the median annual increase for women and men is similar, both in terms of the dollar amount and as a percent of compensation. The median increase for women is an esti- mated $306, about 2 percent of compensation. The present value of this increase is $341, or 1.8 percent of compensation. The median increase for men is about $290, or 1 percent of compensation. The present value of this increase is about $344, or 1.3 percent of compensation. Page 16 GAO/HR.D-@O-101Impact of TRA Vesting Rules on Workers Appendix II Fbrther Details on GAO’sScope and Methodology We analyzed the effects of TRA vesting rule changes using data we had already gathered in response to a requirement in the Retirement Equity Act of 1984 that we study the effect of federal pension rules on women. This appendix provides details about our samples of plans and our methodology. From ERISA reports for employee benefit plans filed for the plan year beginning during 1984,’ we drew two samples of private pension plans operating in both 1984 and 1985. One sample contained plans sponsored by employers with fewer than 100 employees (small employers); the other contained plans sponsored by employers with 100 or more employees (large employers). The reports maintained by IRS were the most up-to-date information available on pension plans operating in 1984 and 1985 at the time we drew our samples. The reports did not, however, include plans that began operating in 1986. Consequently, both samples include only plans that started before 1986. We estimated that about 202,300 plans sponsored by small employers Plans Sponsored by met our sampling criteria (see table II. 1). The plans met all of the fol- Small Employers lowing criteria: 1. They were ongoing plans of the four most prevalent types-fixed benefit and unit benefit defined benefit plans, and profit-sharing and money purchase defined contribution plans.’ 2. They were in one of the five industry groups with the most of these types of plans: wholesale trade; retail trade; finance, insurance, and real estate; legal, medical, and health services; and other services. 3. They were sponsored by a single employer with fewer than 100 employees. ‘Form 6600 for plans with 100 or more participants and Form 6600-C for plans with fewer than 100 participants. “A fixed benefit plan provides a retirement benefit that is not related to the years of service of the plan participant. An example is a specified percentage of compensation, such as 60 percent of the participant’s final pay. A unit benefit plan uses a formula that provides an explicit unit of benefit for each recognized year of service with the employer; for example, 1 percent of compensation per year of service. In contrast, rather than fixing benefits by a formula, profit-sharing and money purchase plans fix the amount of the employer’s contribution to each participant’s account. In a profit-sharing plan, the total employer contribution is a function of profits and the amount contributed to each participant is generally in proportion to the participant’s share of total compensation paid to all par- ticipants. In a money purchase plan, the employer is committed to periodic contributions according to a specific formula, usually a percentage of salary. Page 17 GAO/HRD-90-101 Impact of TRA Vesting Rules on Workers .- .- Appendix II * Further Details on GAO’s Scope and Methodology 4. They had more than one participant. 6. They were not Keogh plans for self-employed individuals. Table 11.1:GAO’s Universe and Sample of Plans Sponsored by Small Employers Response Original Original Eligiblea Adjusted rate Population Type of,.._ ._..._..- plan/ industry ~-. .._.__.-..--_- group universe sample sample universe (percent) estimate Fixed benefit - plans..- - - __... trade Wholesale .-____.__. __.___.-. __._.-_-.._...-_.- 3,855 31 20 2,487 85 2,114 Retarl trade 3,356 17 10 1,974 80 1,579 Finance, insurance, and--.~-_--~.- _ ___-....-. real estate 4,416 25 10 1,766 60 1,060 Legal, medical, _.... and health ._.- -~_ services --. 17,641 119 78 11,566 59 6,821 Other servrces . “._ .._._. .._-.-.-- ._.._- ...~ 11,054 -?I 39 6,072 54 3,270 Unit benefit plans ~--- Wholesale trade - 478 34 27 380 78 296 Retail trade ----- 430 28 24 369 71 261 Finance: insurance, and real estate 984 53 39 724 72 520 Legal, medical, and health services 1,659 -- 82 51 1,032 61 627 Other services -- ..-- _...-. ...- - -.- ...-.. -- 936 56 34 568 65 368 Profit-sharing plans Wholesale-trade 10,942 33 23 7,626 61 4,642 Retarl trade 11,254 20 15 8,441 80 6,753 Finance, insurance, -and .~ ~_real estate ~---.-.. .-.- .___. -.-~~--- 9,902 -- 21 9 4,244 78 3,301 Legal, medical, and health services 44,633 94 61 28,964 70 20,417 Other services 25,605 81 37 11,696 41 4,742 Money purchase .,.....I .” _...plans _.____-..-- Wholesale trade 3,431 16 11 2,359 64 1,501 Retail trade 3,254 15 10 2,169 100 2,169 Frnance, insurance, and real estate 4,881 24 12 2,441 67 1,627 Legal,~medical;and _ -.. health - ._--- services .__-_._.- --- --3ij%- --- 153 98 20,303 65 13,112 Other services --___.. 11,885 50 22 5,229 55 2,852 Totals 202,299 1,023 630 120,410 6!jb 70,031= “Origrnally sampled plans were ineligible if they were (1) Keogh plans for self-employed persons, (2) plans with only one participant, (3) sponsored by employers with 100 or more employees, or (4) terminated during the 1984 plan year. “The response rate is weighted to represent industry and plan types in proportion to their representa- tion in the universe. ‘Population estimate has total precision of + 5,471 plans (It 7 percent). Page 18 GAO/HRD-90-101 Impact of TRA Vesting Rules on Workers . Appendix II Further Details on GAO’s Scope and Methodology Our original stratified sample included a total of 1,023 plans selected from each of the four plan types. Within each plan type, we allocated the sample across selected industry groups generally in proportion to each group’s representation in the universe. We determined the final sample size of 630 and adjusted our universe estimates after we identi- fied 393 cases in the original sample that did not meet our sampling cri- teria. The adjusted universe included an estimated 120,440 plans (& 7,400). Among these 630 sampled plans, 66 percent (407) responded across all the sampled plan types and industries. We compared respondents and nonrespondents on several characteristics-plan size, top-heavy status, integration with social security, vesting method, industry, and plan type-and found some significant differences. For example, defined contribution plans that did not respond tended to be smaller than those that did respond. Because of these differences, our estimates apply only to that proportion of the adjusted universe that responded to our survey. As indicated in the final column of table I. 1, our respondents represent an estimated 78,000 plans (f. 6,600). These plans contained an estimated 700,000 participants (-& 100,000). In our sample, only small employers’ plans that were not top-heavy plans were eligible for inclusion in our analysis of the effects of TRA. About one-fourth of the plans (18,900 f. 3,600) in our universe of small employers’ plans were not top-heavy. We estimated that 19,600 plans sponsored by large employers met our Plans Sponsored by sampling criteria (see table 11.2). These were ongoing plans in one of the Large Employers three most prevalent plan types- fixed benefit, unit benefit, or profit- sharing-in one of six industry groups containing most of these types of plans. The six were nondurable manufacturing; durable manufacturing; wholesale trade; retail trade; finance, insurance, and real estate; and legal, medical, and health services. In addition, sampled plans were sponsored by a single employer or a controlled group (in which all the business entities are under common control) with 100 or more employees and contained more than one participant. Page 19 GAO/IiRD-90-101 Impact of TRA Vesting Rules on Worker6 Appendix II Purther Detalis on GAO’s Scope and Methodology Table 11.2:QAO’r Unlverre and Sample of Plans Sponsored by Large Employers Response Original Original Eligible’ Adjusted rate Population Type -~ of plan/ Industry group universe sample sample universe (percent) estimate Fixed benefit plans -__ iondurable manufacturing 526 4 4 526 25 132 --l_l_ Durable manufacturing 587 10 8 470 50 235 .--_--.l-- ---..--____-- Wholesale trade 187 3 1 62 0 0 _--_--- ~__ Retail trade 151 2 1 76 0 0 Finance. insurance. and real estate 295 4 4 295 50 148 .-_. Legal, medical, and health services 235 4 3 176 33 59 Unlt ----~ beneflt plans _- Nondurable manufacturing 2,796 31 29 2,616 83 2,165 --.-~ Durable manufacturing 4,251 50 39 3,316 46 1,530 I-- Wholesale trade 429 5 4 343 50 172 -.--.. Retail trade 426 5 3 256 100 256 Finance, insurance, and real estate 1,169 13 11 989 73 719 ---_._I_- Legal, medical, and health services 1,278 15 14 1,193 79 937 __--.._.-“_.~-..- -----_- Protlt-sharlna -- Iplans --.-. ~_____ ~-- Nondurable manufacturing 1,735 28 25 1,549 76 1,177 --_-.--~ Durable manufacturing - ~-_____-- 2,244 29 25 1,934 64 1,238 --_-.-- Wholesale trade 824 11 11 824 45 375 -..---._- Retail trade 4 13 921 69 638 Finance, insurance, and real estate 1,056 14 12 905 67 603 ---_-. Legal, medical, and health services 372 6 4 248 50 124 -.z- --_- Totals 19.553 248 211 16,099 63b 10,507c aOriginally sampled plans were ineligible if they were (1) sponsored by employers with less than 100 employees or (2) terminated during the 1984 plan year. hThe response rate is weighted to represent industry and plan types in proportion to their representa- tion in the universe. CPopulation estimate has total precision of + 1,019 plans (f 9.7 percent). The original sample included 248 plans allocated across the selected plan types and industry groups generally in proportion to each plan type’s and group’s representation in the universe. We determined the final sample size of 211 plans and adjusted the universe estimates after identifying 37 cases in the original sample that did not meet our sam- pling criteria. The adjusted universe included an estimated 16,700 plans (& 800). Page 20 GAO/HRD9O-101 Impact of TRA Vesting Rules on Workers Appendix II Further Detab on GAO’s Scope and Methodology Among these 211 sampled plans, 63 percent responded across all the included plan types and industry groups. We compared respondents and nonrespondents on several characteristics-plan size, age of the plan, integration with social security, industry, and plan type. We found one significant difference. As with the sample of plans sponsored by small employers, defined contribution plans that did not respond tended to be smaller than those that did. Consequently, the estimates in this report apply only to that proportion of the adjusted universe that responded to our survey. As indicated in the final column of table 11.2,our respon- dents represent an estimated 10,500 plans (+ 1,000). These plans include an estimated 6.2 million participants (f. 1.9 million). None of the sampled plans sponsored by large employers was top-heavy. Consequently, all these large employers’ plans were eligible for our anal- ysis of the effects of TR.A vesting changes. Combining plans sponsored by small employers and plans sponsored by Scopeof GAO’s large employers, our analysis of the effect of TRA’S vesting changes rep- Analysis of TRA’s resents an estimated 6.1 million participants in about 25,100 plans. Our Vesting Changes analysis of TM’S effect on dollars of vested benefits focused on the esti- mated 3.8 million participants in defined benefit plans. We excluded par- ticipants in defined contribution plans because we did not know the account balances for participants who were not vested. Consequently, we could not determine the dollar changes under the new rules for these participants. The estimated changes in the value of vested benefits under TM and the Calculating the top-heavy simulation are expressed as present values. Present values Present Value of the are sums of money that, if invested now at a given rate of compound Dollar Change in interest, will accumulate to specified amounts at specified future dates. To convert the additional vested benefits under TRA to present values we Vested Benefits chose a 7-percent interest rate.” Using Department of Labor data on retirement ages, we assumed benefit payments would begin at age 61 for men and 63 for women4 “We performed a sensitivity analysis using alternative interest rate assumptions. A one-percent change in the rate changed the present value by about 6 percent, though the choice of interest rates did not affect the comparisons discussed in the report. 4Department of Labor, Office of Pension and Welfare Benefit Programs, Findings from the Survey of Private Pension Plan Benefit Amounts, 1985. Page 21 GAO/HRD-90-101 Impact of TRA Vesting Rules on Workers Appendix II Further Details on GAO’sScope and Methodology We also made certain life expectancy assumptions for men and women. At age 34, the median age of men with increased vested benefits under TRA in our analysis, men had a life expectancy of 41 years. At age 31, the median age for women with increased vested benefits under TRA in our analysis, women had a life expectancy of 49 years.” For the subset of participants with increased vested benefits in our top-heavy simula- tion, men had a median age of 31 and a life expectancy of 43 years; women had a median age of 30 and a life expectancy of 50 years. Were alternate assumptions of interest rates, life expectancies, or benefit com- mencement used, a different present value would result. “Data for 1986 reported in Department of Commerce, Bureau of the Census, Statistical Abstract of the United States, 1989, based on data from U.S. National Center for Health Statistics, Vital Statistics of the United States. Page 22 GAO/HRD-00-101 Impact of TRA Vesting Rules on Workers f&p&%x III SamplingErrors for Estimaks Because our estimates about participants are based on a sample of pen- sion plans and their participants rather than the universe of plan par- ticipants, each reported estimate has a sampling error associated with it. The size of the error reflects the precisionsf the estimate-the smaller the error, the more precise the estimate. We calculated sampling errors for estimates in this report at the 95-percent confidence level. This means that the chances are 19 out of 20 that the actual number or per- centage being estimated falls within the range of our estimate, plus or minus the sampling error. For example, if we have estimated that 30 percent of a group has a characteristic and the sampling error is 6 percentage points, there is a 95-percent chance that the actual per- centage is between 24 and 36. Tables III. 1,111.2, and III.3 include sampling errors for participant esti- mates in figures 1, 2, and 3, respectivel.y. Sampling errors for the esti- mates from the top-heavy rules simulation appear in tables III.1 and 111.4.Sampling errors for the change in dollars of vested benefits under TRA are included in table 111.5. Table 111.1:Sampling Errors for Figure 1 and Flguro I.1 No. of participants Sampling (in millions) Estimate error All oarticloants 6.1 Old rules 53 10 TRA rules --____-. 76 7 Top-heavy rules 85 5 Women 2.9 Old rules -______- - 52 11 TRA rules _-- 76 -__ 10 Top-heavy rules 86 .__-- 6 Men .___ 3.2 Old rules 54 6 TRA rules 75 6 Top-heavy rules 85 5 Figures are percentages of participants vested Page 23 GAO/HRD90-101 Impact of TRA Vesting Rules on Workers Appendix Ill Sampling Errors for Estlmatem fable 111.2:Sampling Errors for Figure 2 No. of participants Sampling (in millions) Estimate error Cliff vestina Plans 3.1 Old rules 32 2 TRA rules 69 3 Waded vesting plans 2.7 Old rules 73 6 TRA rules 80 5 Figures are percentages of participants vested Table 111.3:Sampling Errors for Figure 3 Sampling Estimate error Fully vested under old rules and TRA 30 5 Not vested under old rules and TRA 31 5 Increased vested benefits under TRA 39 5 Figures are percentages of participants vested. Total participants, 3.8 million Table 111.4:Sampling Errors for Figures 1.2and 1.3 No. of participants Sampling (in millions) Estimate error All dens 3.8 Fu;:yl;;sted under TRA and top-heavy 63 7 No;u;zssted under TRA and top-heavy ----___ 19 4 Increased vested benefits under top- 18 4 -. heavy rules Cliff vesting plans 3.1 Fully vested under TRA and top-heavy rules 70 4 -______--- Nc$;e;ted under TRA and top-heavy 19 2 i&eased vested benefits under top- 11 2 ---.-heavy rules Graded vesting plans 0.7 Fully vested under TRA and top-heavy rules -.----..-- 36 8 Nort$.e;ted under TRA and top-heavy 19 5 Increased vested benefits under top- heavy rules 45 6 Figures are percentages of participants vested. Page 24 GAO/HRD-99-101 Impact of TRA Vesting Rules on Workers . Appendix To Sampling Errora for Estimates Table 111.5:Sampling Error8 for Table 1 No. of participants Estimate (1988 Sampling (in milliona) dollarsa) error All plans 1.5 $1,080 $372 Women 0.7 852 227 Men 0.8 1,728 470 Cliff vesting plans 1.2 1,344 282 Women 0.6 972 195 Men 0.6 1,968 280 Qraded vesting plans 0.3 194 170 Women 0.16 139 161 Men 0.16 258 289 aDollars in body of report are adjusted to 1990 dollars Y Page 26 GAO/HRDM-101 Impact of TRA Vesting Rules on Workers Appendix IV Major Contributors to This Report Human Resources Sharon A. Ward, Assignment Manager Division, Washington, DC. Kenneth Wachner, Evaluator-in-Charge Detroit Re@ona1 Office EdnaMeSaltzma,nEvaluator Edmund 0, Price, bomputer Programmer/Analyst Page 26 GA0/IiRD.90.101 Impact of TRA Vestine Rules on WOrkem Page 27 GAO/HRLMN-101 Impact of TRA Vesting Rules on Workera RelatedGAO products Private Pensions: Impact of Vesting and Minimum Benefits and Contri- bution Rules in Top-Heavy Plans (GAO/HRD-90-4BR, Oct. 23, 1989) Private Pensions: Plan Provisions Differ Between Large and Small Employers (GAOIHRD-89-105BR,Sept. 26, 1989) 401(k) Plans: Participation and Deferral Rates by Plan Features and Other Information (GAOIPEMD-88-ZOFS, Apr. 29, 1988). 401(k) Plans: Incidence, Provisions and Benefits (GAO/PEMD-88-15BR, Mar. 29, 1988). Pension Plans: Vesting Status of Participants in Selected Small Plans (GAO/HRD-88-31,&t. 30, 1987). Pension Integration: How Large Defined Benefits Plans Coordinate Bene- fits With Social Security (GAO/HRD-86-118BR, July 21, 1986). Y (106617) Page 28 GAO/HRD-90-101 Impact of TRA Vesting Rules on Workers E -- The first. five copies of t~.ch GAO report, are free. Additional copies art* $2 each. Or&m shoulti be sent. to the following address, acconi- panitd by a check or rnout~y order made out. t,o the Superinteudent of Documents, wlrtm necessary. Orders for 100 or more copies to be ruailt~d t.0 a single actdress are distvnmt.ed 25 jwrw~~t.. Gaithtmburg, MI) 20877 Orders may also be placed by calling (202) 275624 1. i Permit No. GlOO 1 Official Ihasiwss Ii hmllt y for I’rivatt~ IJse $300
Private Pensions: Impact of New Vesting Rules Similar for Women and Men
Published by the Government Accountability Office on 1990-08-21.
Below is a raw (and likely hideous) rendition of the original report. (PDF)