4 3 Thank you for your invitation to submit written testimony for the record T-53 2 of the joint hearing held by the Subcommittee on Aging and the Subcommittee on of the files has been sytematically denied. If this valuable research file is Clearly, it is possible for the federal government to be penny-wise and nationwide basis and to project those trends to the future. Energy, Nuclear Proliferation, and Government Processes on the subject of population since the 1960s, when a similar attempt was funded by the Federal pound-foolish eventually released in datatocollection. the public Public when DOL policy now must anticipates continueittowill be grounded be, the Statistical Policy in an Aging America. The Employee Benefit Research Reserve Board. It provides information on the distribution of personal assets information on benefits will be eight years out of date. on a firm understanding of the facts; the facts are only available with a Institute (EBRI) is a nonprofit, nonpartisan policy research organization IV. 5500 Annual Report Data including employer-sponsored pensions. Information on pension entitlement good statistical foundation and reliable research. We applaud the committed to accurate and timely research and education on a wide range of from employees was matched to employer data, providing the most complete data The release of these data also has been further encumbered because of Subcommittee for this inquiry into this important subject and remain willing publicSinceandthe employer-sponsored Employee Retirement benefits, Income Security such asAct retirement (ERISA) of and 1974,heamost lth available. An ongoing survey of this type, in which savings can be tracked as another to providetypefurther of access assistance problem.if you The should Survey desire of Benefit it. Amounts was matched to insurance, and their role within the income security system of the nation. pension and welfare plans are required to file an annual report with the with the aging of the population, would be invaluable to the research and earnings records from the Social Security Adminstration so researchers could Since InternalEBRI's Revenue founding Servicein 19 which 78, weincludes have relied considerable upon qualit finyancidaalta provided informationby policy community. It would also provide the Congress with a more reliable track earnings throughout a retiree's entire career to investigate how about the federthese al government plans. through No other its madny ata stasource tistical provides programs. comparable Over the yeasset ars, picture of the real economic condition of the elderly, in the event that pensions are related to lifetime earnings. Social Security earnings records we inform have ation. providedThe suggestions Department of on Labor continuing receivesdatacopies needs ofto these help forms researchers under athe nd reforms are considered in the various public programs. have long been linked to survey data and released as public use files after policy reporting analand ysts disclosure study retirement provisions andofheaERISA. lth issues. Since the For first-year this reason,annual we ensuring that the information is not so specific that individuals can be welcome the opportunity to share our thoughts with the Subcommittees. reports were filed (for plan year 1975), the research and policy community has II. Supplements to the Current Population Survey identified from the data. This usually has involved the deletion of Statement for the Record hoped that the Department of Labor (or the Internal Revenue Service) would identifiers and removal of other unique information to ensure that that any provide Despite a stsome atistically very positive reliable steps sample in dof ata these collection annual over reportthe forms. past deca Early de, In 1972, the Bureau of Labor Statistics, the Social Security one individual could not be pointed out. Based on intempretations of the 1976 the availability and timeliness of federal data has not entirely met the needs years of dat8 were available with a significant by time lag and, despite a recent Administration, and the Department of Treasury funded a survey conducted by Tax Reform Act, however, which, in essence, specifies that tax-return of researchers. We also fear reductions in data availability may occur in the flurry of activity, the last plan year available for research use is for 1981 the Census Bureau as a supplement to the Current Population Survey collecting information cannot be released in microdata form without extreme safeguards, future, when information on benefits and benefit recipiency will be even more (although preliminary 1982 taEmily bulations S. Andrews, are available). Ph.D. These data constitute information on pension plan participation and benefit entitlement among the continued public-use release of survey data matched with earnings records crucial as the nation ages. a potentially valuable resource to study the financial aspects of pension full-time private-sector workers. In 1979 the Department of Labor's Office of has virtually been precluded. This anomalous situation occurred because Submitted to the plans. Although the Federal Reserve Board is benchmarking the pension asset Pension and Welfare Benefit Programs conducted a similar survey on pension Social Security earnings are reported on Internal Revenue Service forms. The data Ininaddition, the Flow given of Funds the onlikely the Department passage ofoftaLabor's x reform 5500 legisl forms, ation, an we ongoing will coverage and individual retirement account (IRA) utilization among all upshot of a series of incomprehensible negotiations on the part of the Social need to know its impact on the pension U.S. Senate and health plans upon which millions of funded statistical program is needed. Ad hoc efforts since 1974 are the workers. In 1983 EBRI joined with the Department of Health and Human Services Security Administration and the Internal Revenue Service has been that, with workers result of and resource retirees limitations depend. In which the event forestalled of as faan r-reaching ongoing astatistical change as to conduct a similar survey including information on universal IRAs and 401(k) very few exceptions, any survey that attaches earnings records has virtually basic tax reform, analysts and researchers cannot accurately foresee the program in this Jointarea. HearingThese on Statistical 5500 data are Policy gathered in an by Aging imposing America considerable plans. The 1983 data have been used widely to analyze the effect of been embargoed--whether there is any technical privacy reason or not. future and reach reliable conclusions. Follow-up evaluation with up-to-date costs and effort on the private sector. To allow these valuable data to retirement income provisions found in current tax reform and retirement income Researchers have begged both the agencies and the Congress to return authority rem inform aination underutilized will be necess is ary a serious to evaluaand te unnecessary its effect andwaste correcof t for the undesired nation's Committee on Labor and Human Resources legislative proposals. EBRI has proposed to fund a similar survey in 1988 in over the earnings records to Social Security and to the pre-1976 situation, to resources. impacts. concert with the Social Security Adminstration. While as a private-sector no avail. Also, a major pension forecasting model may soon become outdated nonprofit institution we are happy Subcommittee to sponsor on Aging a CPS supplement in conjunction without additional earnings record data. V. Access Our testimony Issues discusses six specific areas of concern to the statistical with the federal government, we regret that the survey is not conducted on a community interested in issues related to aging in America. Subcommittee on Energy, Nuclear Proliferation, yearly or biannual basis as an ongoing federal statistical program. Because VI. New Data Sources The first three statistical concerns we discussed revolve around problems and Government Processes Congress relies on private-sector programs to supplement benefits provided I. Recent Initiatives of irregular data collection. Our fourth concern, the 5500-form issue, is one through the public sector, an ongoing federal statistical program of this Many fear that in today's federal budget environment few new data sources of access and processing; the data are collected but are not available to nature would provide vital information on the extent and adequacy of such will be developed. While the Social Security Administration conducted a major researchers. Two salutoryResenew archers initiatives would ahlso ave been benefittaken fromin more recenttimely years:or accessible the Survey supplementation. June 3, 1986 study of retirees with the New Beneficiary Survey, there is no assurance that of information Income andfrom Progrother am Parti administrative cipation (SIPP)sources, sponsoredincluding by the filings Census Bure with au athe nd another study of that type will be funded. Their longitudinal Retirement the InternSurvey al Reven ofueConsumer Service Finfor ances plans (SCF)thatsponsored have terminated by the Feder andal plans ReservethatBoaare rd III. Level of Benefits Survey History Study provided researchers with one of the major databases to study and other federal agencies. starting up. Better access to Pension Benfit Guaranty Corporation data, the retirement decision and the income of the elderly. Other longitudinal including Since 1979 information the Bureau on plan of Labor terminations, Statistics would has be collected valuable detailed as well. annual Once data such as the National Longitudinal Survey funded by the Department of SIPP is a large, complex survey providing detailed information on many again, neither of these agencies has an ongoing statistical program in many data from medium and large firms on the prevalence and provisions of pension, Labor have also ended. While we do not argue that either of these surveys areas of concern. health, types of and ca other sh aemployer-sponsored nd in-kind income.benefits. As thatThissurvey effort,continues, often called expanded the should be resuscitated (many technical factors would go into that decision), technical support for the survey may also be needed. Without technical Level of Benefits Survey, in part replaced the Employer Expenditure we do argue for the need for longitudinal data to respond to many of the Our final concern is that other data of an ongoing or one-time nature may Compensation support, thisSurvey complexwhich data was set discontinued is inaccessible. in the mid 1970s. The new effort policy research questions about an aging labor force and retirement--questions be collected and processed adequately but then are not released in a timely is the only nationally representative source of benefits information. But it that more and more congressional committees are asking. In a changing While SIPP can serve many purposes, other special-purpose surveys should manner. Problems of timely release have plagued the 1977 National Health Cave does not provide data on benefits offered by small employers. Although the society, information about individuals in the 1970s may provide little not be refused clearance by the Office of Management and Budget on the theory Expenditures Study (NHCES) and the 1982 Nursing Home Survey. Public access to Bureau of Labor Statistics is considering the addition of a small-employer information on the behavior of workers and retirees in the 1990s. that SIPP can do everything. Some issues, such as the numbers and status of these data is slowed when they are only provided to certain researchers under sample in the future, this effort is not under active consideration within the contract low-income toelderl the yagency , the orelderl wheny information in nursing homes, is held orfortheresearchers circumstanceswithinof the the next Emilycouple S. Andr ofewsyearsis (although director aofsurvey research of state for the and Employee local governments Benefit Res may earch be Conclusion very old elderly, over age 85, may require separate surveys of special agency to use on a monopoly basis. onInstitute. a faster The track). views Currently expressed more in than this half statement of all are American solely those workers of are the populations to achieve meaningful sample sizes. employed author, in and small do notestablishments. necessarily represent These workers the views are potentially of the t_ustees the most and We feel that all the data issues we have raised are vital to understanding The Department of Labor Survey of Private Pension Benefit Amounts is vulnerable sponsors oftothechanges Employeein Benefit public policy Research andInstitute, benefits or costs. its staff. With substantial an aging society. We understand data gathering is costly and that difficult The 1982 SCF is the first comprehensive survey of the assets of the another example of a data collection effort that has suffered from these current interest in the benefits provided by small employers, the decisions must be made. We have tried to be specific in our suggestions and small-employer problems. Thesegment survey ought has been to be available conducted only on an to accelerated a limited schedule. number of have not presented a "wish list" of data needs without regard to importance. researchers under contract to the Labor Department. Outside requests for use These data would also enable us to track trends in plan participation on a