In conjunction with larger Social Security solvency plans, many policymakers have proposed introducing benefit increases for older beneficiaries. This brief analyzes the projected effects of two such policy options on beneficiaries aged 85 or older in 2030 using the Modeling Income in the Near Term model. Both options target older beneficiaries' primary insurance amounts for a 5 percent increase, but they differ in how the increase would be calculated. Both proposals would increase monthly benefits for nearly all older beneficiaries, and both would reduce poverty levels among the aged, relative to currently scheduled benefits. However, the options differ in how the benefit increases would be distributed among older beneficiaries across shared lifetime earnings quintiles.
Kevin Whitman and Dave Shoffner are with the Office of Retirement Policy, Office of Retirement and Disability Policy, Social Security Administration.
Acknowledgments: The authors thank Natalie Lu, Anya Olsen, Patrick Purcell, Kathleen Romig, and Mark Sarney for their helpful comments and suggestions.
The findings and conclusions presented in this brief are those of the authors and do not necessarily represent the views of the Social Security Administration.
The original version of this Policy Brief contained an error in Chart 1. The median changes in average PIAs for the lowest and second lowest shared lifetime earnings quintiles were incorrectly given as 6.7 and 4.9, respectively. The corrected percentages appear in this version of the Brief. [Posted: January 24, 2014.]
|MINT||Modeling Income in the Near Term|
|PIA||primary insurance amount|
Many reform plans designed to return the Social Security program to long-term solvency also include a benefit increase targeted toward older beneficiaries. Policymakers use two core rationales for such targeted increases. First, older beneficiaries tend to be more economically vulnerable and reliant on the income Social Security provides.1 Second, certain benefit reductions (such as changes to cost-of-living adjustments) can compound over time, and including a benefit increase for older retirees in a larger reform plan can ameliorate those reductions.
Generally, the benefit increase proposals provide slightly larger monthly benefits starting at around age 80, but can vary along multiple lines. This policy brief analyzes two options, and projects their respective effects. The two options vary in how the benefit increase is calculated:
- The Individual PIA plan provides an increase of 5 percent of the individual's primary insurance amount (PIA), which is the benefit an individual will receive based on lifetime earnings, if retirement benefits start at the normal retirement age.
- The Average PIA plan provides an increase of 5 percent of the average PIA for all retired workers, rather than an individual's own PIA.2
We analyze those two particular provisions because they appear in various publicly available reform plans.3, 4 Both 5 percent targets may appear identical when described only as “a 5 percent benefit increase,” but they can produce different outcomes for beneficiaries.
Our analysis focuses on beneficiaries aged 85 or older in 2030 and we express the results as percentage differences from scheduled benefits under current law. We do not project the relative costs of the proposals.5
The analysis is based on data from the Modeling Income in the Near Term model, version 6 (MINT6). MINT6 includes data for respondents to the 2001 and 2004 Surveys of Income and Program Participation (SIPP), matched to Social Security administrative records through 2009. For 2010 and later, MINT6 projects life events, earnings, and benefits for those respondents.6
- The Individual PIA plan would produce a larger median increase in monthly benefits. However, the Average PIA plan would provide more progressive benefit changes, with lower lifetime earners receiving higher proportional benefit increases.
- Poverty would decline by a small amount under both of the plans, but by a slightly larger percentage under the Average PIA plan.
- Although nearly all beneficiaries aged 85 or older would receive the targeted benefit increase provided by each plan, roughly 2 percent would not receive higher benefits. In general, that group would not receive higher benefits either because they are dually entitled beneficiaries, and the increase in their own worker benefit is offset by a decrease in their auxiliary benefit; or because they are auxiliary-only beneficiaries receiving benefits based on the earnings record of an individual who has not yet reached age 81.7, 8
Comparing the Policy Options
Box 1 differentiates the 5 percent targets in the plans we model.
|Plan||Benefit increase equals 5 percent of—|
|Individual PIA||The individual's PIA, which is the benefit an individual would receive based on his or her own earnings record if he or she started benefits at the normal retirement age. Dollar amounts vary, mirroring PIA variations among beneficiaries.|
|Average PIA||The average PIA for all retired-worker beneficiaries as of the year they reach age 80 (specifically, the weighted average age-80 retired-worker PIA for each year, based on MINT6 projections for scheduled current-law benefits). All affected beneficiaries receive the same dollar-amount increase.|
The other provisions of the two proposals are identical. The plans start in 2013, apply to beneficiaries starting at age 81, and phase up linearly to the full increase at age 85. We analyze effects for beneficiaries starting at age 85 rather than age 81 to provide a clearer picture of how the policies look when fully phased in. Table 1 presents the projected characteristics of the beneficiary population aged 85 or older in 2030.
|Less than 12 years||5.4|
|Completed high school or equivalent||39.3|
|Shared lifetime earnings quintile a|
|SOURCE: Authors' calculations using MINT6.|
|NOTES: Weighted population size for beneficiaries aged 85 or older in 2030 is 6,800,966.
Rounded components of percentage distributions do not necessarily sum to 100.0.
|a. Quintiles are not evenly distributed because quintile bounds are based on the beneficiary population aged 60 or older while the population shown reflects beneficiaries aged 85 or older.|
Both plans increase benefits by raising the retired worker's PIA. This approach means that the benefit increase an individual receives each month may differ significantly from the original PIA increase calculated using either of these methods (for example, the final benefit could be adjusted based on early or delayed claiming). An additional outcome of modifying the PIA as we do in this analysis is that auxiliary beneficiaries can receive the benefit increase even before they reach age 81.9 To maintain equity for auxiliary beneficiaries receiving benefits on the earnings record of someone who dies prior to age 81, the benefit increase is payable starting when that deceased beneficiary would have reached age 81.
These options do not apply to disability insurance beneficiaries who have been receiving benefits for a period comparable to that of a retired worker reaching age 81. Some elder benefit increase provisions specify that the increase will occur “X number of years post-eligibility,” allowing disability insurance recipients to obtain the same type of benefit increase as an older, retired-worker beneficiary. That approach addresses the compounding effect of benefit reductions for long-time beneficiaries. Our analysis does not incorporate that approach, as we focus on the beneficiary population aged 85 or older. Disability insurance beneficiaries who convert to retired-worker status at normal retirement age are also eligible for the 5 percent PIA increase at age 81, but they do not receive special consideration nor the ability to receive the increase before age 81 under the options analyzed here.
Both Options Affect Nearly All Older Beneficiaries
Because these elderly benefit increases do not include any means testing or qualification requirements other than age, we project that nearly all beneficiaries aged 85 and older would see their benefits increase by at least 1 percent above scheduled benefits under either policy option (see tabulation below).
|SOURCE: Authors' calculations using MINT6.|
|NOTE: Reflects beneficiaries whose benefit would be at least 1 percent higher than the benefit scheduled under current law.|
However, some older beneficiaries would not receive a benefit increase. In general, beneficiaries who did not receive higher benefits would belong to one of two groups:
- Dually entitled beneficiaries for whom the increase in their own worker benefit would be offset by a decrease in their auxiliary benefit because the spouse whose PIA provides the auxiliary benefit is younger than age 81.10
- Auxiliary-only beneficiaries with benefits based on the PIA of a spouse younger than age 81. This outcome is another result of applying the benefit increase to the PIA, rather than the actual benefit.
Median Benefit Increases are Higher Under the Individual PIA Plan
Overall, median monthly benefit increases would be higher under the Individual PIA plan, at 5.0 percent above scheduled benefits (see tabulation below). For the Average PIA plan, the median benefit increase would be 3.9 percent. However, the median figures do not capture the greater variety of benefit changes that appear under the Average PIA plan.
|Plan||Benefit increase (%)|
|SOURCE: Authors' calculations using MINT6.|
Benefit Increases are Proportionally Larger for Lower Income Households Under the Average PIA Plan
The average PIA is a single value that produces a benefit increase of equal dollar value for all elderly beneficiaries, which amounts to a larger percentage increase for those in the lower shared lifetime earnings quintiles (Chart 1).11 For example, among beneficiaries in the lowest shared lifetime earnings quintile, the median benefit increase would be 6.4 percent, compared with 3.4 percent for beneficiaries in the highest quintile. The Individual PIA approach provides an equal percentage increase for all older beneficiaries (5.0 percent), and thus a larger dollar increase for those in the higher shared lifetime earnings quintiles.12
Median percentage change in Social Security benefit under alternative plans to increase PIAs for beneficiaries aged 85 or older, by shared lifetime earnings quintile, projected 2030
Elderly Poverty Would Decline Under Both Options
Both of the options would decrease projected poverty among beneficiaries aged 85 and older. Under scheduled benefits, 1.5 percent of this group is projected to be in poverty in 2030 (see tabulation below). The relatively low poverty rate under scheduled benefits reflects projected differential mortality (that is, wealthier individuals are more likely to live to older ages; see Table 1) and the projected decline in poverty due to real wage growth. Under either elder benefit plan, the poverty rate would decline to between 1.1 and 1.2 percent. However, note that if these elderly benefit increases were coupled with other policy changes that reduce benefits, the number of beneficiaries aged 85 or older at or near poverty might not decline (or could rise), as any increase might be offset by reductions elsewhere.13
|Plan||Poverty rate (%)|
|SOURCE: Authors' calculations using MINT6.|
1 Johnson, Goldwyn, and Favreault (2004) project that median family income among individuals aged 85 or older in 2040 will be $9,000 lower than that for individuals aged 67–74, in part because a larger share of the older group will be widowed. Additionally, the older group will rely substantially more on Social Security benefits as a source of income. Reno and Lavery (2009) suggest two reasons for that greater reliance: (1) the oldest old cannot easily supplement their income through work, and (2) private pension benefits, if the individual has any, generally fail to keep pace with inflation.
2 For more information on the PIA, see http://www.socialsecurity.gov/OACT/COLA/piaformula.html.
3 The Social Security Administration's Office of the Chief Actuary presents a comprehensive listing of proposed reform plans on its “Provisions Affecting Level of Monthly Benefits” web page at http://www.socialsecurity.gov/oact /solvency/provisions/benefitlevel.html. The proposals analyzed in this policy brief are listed under section B6: Benefit Increases for Older Beneficiaries.
4 Policymakers have also proposed a benefit increase for older beneficiaries based on 5 percent of the PIA for a newly eligible average wage worker (as of the year the individual turns age 62). We modeled a version of that proposal and the results are nearly identical to those for the Average PIA plan. However, those results could differ if the benefit increase were part of a larger plan that modified the PIA formula in other ways.
5 For additional details on the projected costs of benefit increases for older beneficiaries, see the “Provisions Affecting Level of Monthly Benefits” web page (note 3).
6 For more information on MINT6 and the projection methodology, see SSA (n. d.).
7 Dually entitled beneficiaries are eligible for a benefit based on their own earnings record, and for an auxiliary (spousal or survivor) benefit based on the PIA of a current or previous spouse. Auxiliary-only beneficiaries do not have sufficient work history to be eligible for their own retirement benefit.
8 These results arise from applying the benefit increase for older workers to the PIA rather than to the final benefit itself.
9 However, not all proposed benefit increase options for older beneficiaries share this feature. Some do not apply to auxiliary beneficiaries under the specified age limit, and some apply to auxiliary beneficiaries if they are at the age threshold but do not apply to retired-worker beneficiaries who are below the age threshold. See the “Provisions Affecting Level of Monthly Benefits” web page (note 3).
10 In rare cases, a dually entitled beneficiary who receives a higher PIA under one of these options could still ultimately receive a lower final monthly benefit. That could happen if the portion of the benefit based on one's own PIA were reduced because of early claiming (a worker benefit is reduced by as much as 30 percent, if claimed at the earliest eligible retirement age). Then, if the proposed policy changes caused the beneficiary's own PIA to increase but not the spouse's (because the spouse has not yet reached age 81), the own-worker share of the benefit would increase at the expense of the auxiliary share. Thus, the 30 percent early-retirement reduction would apply to a greater proportion of the total benefit than it would apply to under current law, resulting in a lower total monthly benefit.
11 The shared earnings quintile “represents the sum of wage-indexed shared lifetime total earnings. While other quintiles account only for the current spouse's benefits or earnings, this quintile accounts for all previous marriages as well. We divide the sum by 40 for presentation purposes, but earnings across a whole lifetime are included in the sum. If a person is single in a given year, all of his or her earnings that year are counted as shared earnings. If a person is married in a given year, half of the married couple's combined earnings that year are counted as shared earnings. We calculate the quintiles for each year for all elderly beneficiaries aged 60 or older. Wage indexing controls for differences across generations. The dollar ranges shown are in annual real 2012 dollars” (SSA n. d.).
12 For the Average PIA plan, the overall median benefit change is 3.9 percent, while the median benefit change for beneficiaries in the middle shared lifetime earnings quintile is 4.1 percent. This discrepancy reflects the higher share of beneficiaries aged 85 or older who are in the higher quintiles. We define the quintile bounds based on the population aged 60 or older, and differential mortality leads to a greater share of higher lifetime earners at age 85 or older. See Table 1 for information on the percentage of beneficiaries aged 85 or older in each shared lifetime earnings quintile.
13 For widows, often targeted by policymakers for elderly benefit increase options, results are consistent with those for the total beneficiary population aged 85 or older (not shown in table).
Johnson, Richard W., Joshua H. Goldwyn, and Melissa M. Favreault. 2004. “Social Security COLA Reductions Would Weaken Financial Security for the Oldest and Poorest Retirees.” Retirement Project Brief No. 18. Washington, DC: Urban Institute. http://www.urban.org /UploadedPDF/311063_retirement_no18.pdf.
Reno, Virginia P., and Joni Lavery. 2009. Fixing Social Security: Adequate Benefits, Adequate Financing. Washington, DC: National Academy of Social Insurance. http://www.nasi.org/research/2009/fixing-social-security.
[SSA] Social Security Administration. No date. “Projection Methodology.” http://www.socialsecurity.gov/retirementpolicy/projection-methodology.html.