• Mature Women cohort
• Young Women cohort

Mature Women cohort

This section reviews information on receipt of pension benefits and on pension plans for which the respondent is eligible. It also describes the special pension plan matching project conducted in conjunction with the 1989 survey. For details on income from Social Security/Railroad Retirement or disability insurance, see the Social Security & Disability section.

Pension income

One focus of the Mature Women pension data collection was receipt of various entitlements by the respondent and/or other members of the respondent's household over a 12-month period. The 1967-72 and 1977 surveys gathered information on whether income had been received in the past calendar year from any local, state, or federal governmental pensions or from any other retirement pensions and, if so, the pension benefit amount received from each. The universe for the 1967-72 survey years was any (unspecified) family member; during 1977, receipt and amount information was collected separately for the respondent, her husband, and other family members. The 1974, 1976, and 1981 surveys fielded questions which did not distinguish the source of the pension benefit, combining Social Security and other pension income.

The 1982 and subsequent surveys included a set of pension income questions that asked, in a separate series, whether the respondent or her husband had received any pension income in the past calendar year/last 12 months and, if so, the amount. Beginning in 1982, coding categories consistently identified plan providers as one of the following: private employer, the military, the federal government, state or local government, a union, a personal plan (e.g., IRA, Keogh, or 401k), or another source.

Pension plans

In the late 1970s, the Mature Women surveys began to ask respondents about their expected retirement income. A series fielded in 1979 included questions about working respondents' eligibility for a pension plan--other than Social Security/Railroad Retirement—from their employers, the age at which eligibility for full or reduced benefits would be attained should they remain with their employer, and the age at which benefits could be drawn should they leave their current job today. Also collected was information on the eligibility of the respondent's spouse for other pension benefits from one or more of the following sources: a personal plan, a private employer, a government employer, or the military.

The 1982 and 1986 questionnaires included similar sets of questions designed to explore current/expected pension coverage of the household from current or past employment. Questions included: (1) whether the respondent or her spouse was eligible for or already receiving pension benefits from a current employer or from another job held in the past; (2) whether the respondent was eligible for or already receiving survivor benefits or other benefits from a husband's or ex-husband's employment or military service; and (3) whether the respondent or husband had a personal retirement (IRA or Keogh) plan.

More detailed data were collected beginning in 1989. This survey included questions designed to identify the most important pension plan for which a respondent was eligible through her own employment, the most important pension plan of a spouse, and any other benefits for which a respondent was eligible through a spouse's or ex-spouse's employment or military service. During this fielding, information was gathered on:

• characteristics of the pension provider (whether the source of the pension was a private employer; the military; another federal, state, or local governmental unit; a union; etc.)
• the industry of the provider and corresponding occupation of the respondent or spouse
• characteristics of each plan (age/service year requirements and estimated pension benefit amounts for full versus reduced benefits; method of determining benefits; vesting rules; and for those already receiving benefits, actual monthly benefit amounts)

Information collected in 1989 about the respondent's pension plans was used to contact the plan providers and gather additional data. This special pension matching project is described below.

In 1992 and each subsequent survey, extensive data were collected regarding multiple employer-provided pension plans; coverage under pension plans obtained through self-employment was not included. Separate questionnaire sections collected information for both the respondent and her husband on: (1) future pensions from current employers, (2) current pensions from previous employers, and (3) future pensions from previous employers. Each series gathered details on participation in defined benefit and/or defined contribution pension plans offered by an employer. For those participating in a defined benefit plan, data were collected on the number of years included in the plan, the amount of money contributed, the age at which full or reduced benefits would be/were being received, and expected/actual benefit amounts at retirement. For those participating in a defined contribution plan, information was gathered on the type(s) of account plan (e.g., thrift or savings, 401k, 403b, Supplemental Retirement Account, profit sharing, stock purchase), amounts both employer and respondent contributed, the total dollar amount of contributions ever made, and how the dollars were invested. All respondents providing pension plan information were asked whether an early retirement option with incentives had been offered and, if so, the type(s) (credit for extra years of service, increased benefits, early benefits, lump sum settlement, etc.). Additional information was collected from those covered under a pension plan from a previous employer on type of employer, years worked for that employer, years included in the pension plan, and employment stop date. Those currently receiving a pension from a previous employer reported the year they started receiving this pension, the amount received, cost of living adjustments, and changes in the pension benefit amount over time.

Researchers should be aware of a shift in the way pension data were organized between 1995 and 1997. In 1995, the pension questions are included as part of the employer roster (a part of the questionnaire that collects information about any employers since the previous interview), so the pensions are organized by employer. This means that respondents reported all pensions from employer #01, then all pensions from employer #02, and so on. The employer number (#01, #02, etc.) is included in the variable title. In 1997-2003, pension data are located in a separate roster, so that plans are organized in the order they were reported by the respondent. The variable titles include only "PN #01," "PN #02," "PN #03," etc., for the first plan reported, second plan reported, and so on, regardless of which employer that plan is associated with. A set of ID variables then permits researchers to link the plans with the appropriate employer. The two systems are described further below.

The following example in table MW1 illustrates the implications of this change. Consider a respondent with four pensions, two from a current employer listed on line 2 in the employer roster, and two from a past employer listed on line 5 of the employer roster. In 1995, the respondent would start with question name RSP-108-ARR-02 and answer questions about the first plan for employer #02. She would then return to the same question, now named RSP-208-ARR-02, and provide information about the second plan with that employer. This pattern would repeat for plans three and four. In 1997, the respondent would answer a series of questions, beginning with question name RSP-102-ARR-01, about her first pension plan. She would next answer the same series of questions, now named RSP-102-ARR-02, about her second plan, and so on until all plans are reported. Researchers can then look at the R7PENS variables (R7PENS is the prefix of the question name) to determine which plan number a given plan is for a specific employer. Finally, researchers can use the R7EMPS variables to determine which employer matches with a given plan. Note that, in the example, the plans are listed by employer, but they would not necessarily be listed in that order.

Pension Matching Project (1989)

Address information collected during the regular 1989 survey permitted the Census Bureau to contact pension providers identified during the 1989 survey for the subsequent Pension Matching Project. Copies of the relevant Summary Plan Descriptions (SPD), actual pension plans, and Internal Revenue Forms-5500 were obtained. Details on each defined benefit or defined contribution plan were systematically coded by the Institute for Survey Research (ISR) at the University of Michigan using the protocol developed for the Survey of Consumer Finances (SCF). Respondents eligible for the pension-matching project were those who provided, during administration of the main questionnaire, pension plan and provider locator information for one or more of the pension sources listed in Table MW2. Of the approximately 1,900 respondent households reporting pension eligibility from one or more sources, plan information was linked to respondent information for 1,329 respondents. Use the series of 'ISR Pension Plan Matching' variables on the Mature Women main data file to locate the correct pension plan.

Of the 815 unique pension plans, 538 (66.0 percent) are defined benefit plans, 259 (31.8 percent) are defined contribution plans, and 18 (2.2 percent) are plans with combined characteristics. Users should note that coverage under multiple and different types of pension plans is possible. For example, a household may receive or be eligible to receive pension benefits from one defined benefit and three defined contribution plans. The source of the four plans could be solely from one household member's (e.g., the spouse's) employment, or from more than one member.

The SCF-based instruments recorded data on plan definitions, benefit formulas, and other provisions applicable to six different retirement conditions: early retirement, normal retirement, late retirement, employment termination prior to retirement, disability retirement, and death or survivor's benefits. Each record contains the identification, linkage, and plan characteristic variables listed below in Figure MW1.

Reference

Gustman, Alan L. and Steinmeier, Thomas L. "Retirement in a Family Context: A Structural Model for Husbands and Wives." NLS Discussion Paper 94-17. Washington, DC: Bureau of Labor Statistics, 1994.

Young Women cohort

This section reviews the pension coverage information from recent surveys and the pension plan data collections. For details on income from Social Security/Railroad Retirement or disability insurance, see the Social Security & Disability section.

Pension benefits

The 1993-2003 surveys collected information about pension benefits actually received as income by the respondent and her husband or partner. Pension sources included a personal plan (e.g., IRA/401k), a private employer, the military, the federal government, a state or local government, a union, or another source.

Pension plans

The 1991-2003 interviews included the collection of extensive information on employer pension plans for which the respondent and her spouse were eligible and participating.

1991 survey

The 1991 survey included questions on actual pension coverage and vesting rights of those respondents who indicated, during administration of the fringe benefit series, that a retirement pension program was available from their current or last employer. Respondents enrolled in an employer's pension program were asked for information on (1) the method used to determine their benefit amount, i.e., a defined benefit formula based on years of service/salary, the amount accumulated in their pension account, or both, and (2) whether they had worked under the plan long enough to be vested or entitled to some pension benefits. Those not vested were asked for information on the number of additional years needed until such rights would be obtained, what would happen to the money accumulated in each retirement account should the respondent leave her job, the dollar amount in the account now, and the amount of money that would be received in a partial or full cash settlement.

1995-2003 surveys

Information on employer-provided pension plans was collected in 1995-2003. The respondent provided information about the eligibility of herself and her spouse for current pensions from current employers. Details were gathered on participation in both defined benefit and defined contribution pension plans offered by an employer. Data were collected from those participating in a defined benefit plan on the number of years included in the plan, the amount of money contributed, age at which full or reduced benefits would be/were being received, and expected/actual benefit amounts at retirement. For those with defined contribution plans, information included the type(s) of account plan (e.g., thrift or savings, 401k, 403b, Supplemental Retirement Account, profit sharing, stock purchase), amounts both the employer and respondent contributed, total dollar amount of contributions ever made, and how the dollars were invested. All respondents who provided pension plan information were asked whether an early retirement option with incentives had been offered and, if so, the type(s) (e.g., extra service credit, increased benefits, early benefits, lump sum settlement, etc.).

Researchers should be aware of a shift in the way pension data were organized between 1995 and 1997. In 1995, the pension questions are included as part of the employer roster (a part of the questionnaire that collects information about any employers reported since the previous interview), so the pensions are organized by employer. This means that respondents report all pensions from employer #01, then all pensions from employer #02, and so on. The employer number (#01, #02, etc.) is included in the variable title. In 1997 and 1999, pension data are located in a separate roster, so that plans are organized in the order they were reported by the respondent. The variable titles include only "PN #01," "PN #02," "PN #03," etc., for the first plan reported, second plan reported, and so on, regardless of which employer that plan is associated with. A set of ID variables then permits researchers to link the plans with the appropriate employer. The two systems are described further below.

The following example in Table YW1 illustrates the implications of this change. Consider a respondent with four pensions, two from a current employer listed on line 2 in the employer roster, and two from a past employer listed on line 5 of the employer roster. In 1995, the respondent would start with question name RSP-108-ARR-02 and answer questions about the first plan for employer #02. She would then return to the same question, now named RSP-208-ARR-02, and provide information bout the second plan with that employer. This pattern would repeat for plans three and four. In 1997, the respondent would answer a series of questions, beginning with question name RSP-102-ARR-01, about her first pension plan. She would next answer the same series of questions, now named RSP-102-ARR-02, about her second plan, and so on until all plans are reported. Researchers can then look at the R7PENS variables (R7PENS is the prefix of the question name) to determine which plan number a given plan is for a specific employer. Finally, researchers can use the R7EMPS variables to determine which employer matches with a given plan. Note that, in the example, the plans are listed by employer, but they would not necessarily be listed in that order.