Testimony of Kathy A. Ruffing Senior Fellow, Center on Budget and Policy Priorities, Before the Subcommittee on Social Security Committee on Ways and Means U.S. House of Representatives
Mr. Chairman, Ranking Member Becerra, and members of the subcommittee, I appreciate the invitation to appear before you today.
The Social Security Disability Insurance (DI) program provides modest but vital benefits to workers who become unable to perform substantial work on account of a serious medical impairment. Although some critics charge that spending for the program is “out of control,” the bulk of the rise in federal disability rolls stems from demographic factors: the aging of the U.S. population, the growth in women’s employment, and Social Security’s rising retirement age. Other factors — including the economic downturn — also have contributed to the program’s growth, but its costs and caseloads are generally in step with past projections. There is little evidence that DI benefits are going to people who could support themselves by working.
The Social Security trustees project that the DI trust fund — which is legally separate from the Old-Age and Survivors Insurance (OASI) trust fund for the retirement and survivors’ programs — will become insolvent in 2016; the Congressional Budget Office concurs. If policymakers take no action to bolster the fund, beneficiaries’ checks will have to be cut by about one-fifth after that. But the fund’s anticipated insolvency should come as no surprise; when policymakers last changed the allocation of taxes between DI and OASI in 1994, they expected the DI fund to run dry in 2016.
Ideally, policymakers should address DI’s pending depletion in the context of overall Social Security solvency. Both DI and OASI face fairly similar long-run shortfalls; DI simply requires action sooner. Key features of Social Security — including the tax base, the benefit formula and cost-of-living adjustments, and insured-status requirements — are similar or identical for the two programs, and most DI recipients are near or even over OASI’s early-retirement age. Tackling DI in isolation would leave policymakers with few — and unduly harsh — options, and lead them to ignore the strong interactions between the disability and retirement programs. A balanced solvency package would also be an opportunity to make needed improvements in the Supplemental Security Income (SSI) program, which is distinct from Social Security but has important intersections.
There is no reason to restructure DI fundamentally, and many reasons not to do so. If policymakers are unable to agree in time on a sensible solvency package, they should reallocate taxes between the retirement and disability funds — a traditional and noncontroversial action that has occurred often in the past.
Demographics Explain Most of Growth in Disability Insurance Beneficiaries
Contrary to the impression conveyed by many recent critics, changes in the workforce explain most of the growth in the disability rolls.
In December 2012, 8.8 million people received disabled-worker benefits from Social Security. Payments also went to some of their family members: 160,000 spouses and 1.9 million children. The number of disabled workers has tripled since 1980, and doubled since 1995 (see the left panel of Figure 1).
- Baby boomers have aged into their high-disability years. Aging takes a toll on many workers’ bodies and minds long before retirement age. People are roughly twice as likely to be disabled at age 50 as at age 40, and twice as likely to be disabled at age 60 as at age 50. (See Figure 2.) As the baby boomers — the huge cohort of people born between 1946 and 1964 — have grown older, the number of disability cases has risen substantially.
- More women have qualified for disability benefits. In general, workers with severe impairments can get DI benefits only if they have worked for at least one-fourth of their adult life and for five of the last ten years. Until women joined the workforce in huge numbers in the 1970s and 1980s, relatively few women met those tests; as recently as 1990, male disabled workers outnumbered women by nearly 2 to 1. Now that more women have worked long enough to qualify for DI, the ratio has fallen to 1.1 to 1.
The Social Security actuaries express the number of people receiving DI using an “age- and sex-adjusted disability prevalence rate” that controls for these factors. That rate rose from 3.1 percent of the working-age population in 1980 to 3.5 percent in 1995 and an estimated 4.6 percent in 2012. (See the right panel of Figure 1.) Expressed another way, age- and sex-adjusted rates of receipt are 48 percent higher than in 1980 and about 30 percent higher than in 1995. That is a significant increase. It is not nearly as dramatic, however, as some alarmists have painted.
Other Factors Have Also Boosted Rates of Receipt
Yet rates of receipt have indisputably risen, even when adjusted for age and sex and for the rising retirement age. Why? The reasons are not fully understood, but include:
Legislative changes. In the early 1980s, the Reagan Administration used its influence over the process of determining eligibility, including new powers to conduct medical reviews granted in a 1980 law, to limit the number of people approved for DI and to terminate benefits for thousands of people already on the rolls. Disability caseloads fell even during a deep economic slump. A backlash ensued from governors, members of Congress, and the courts. Ultimately Congress unanimously enacted the Disability Benefits Reform Act of 1984 (DBRA) to clarify eligibility and to limit terminations to cases where the agency could show that the beneficiary’s medical condition had improved. Notably, DBRA required the agency to consider the impact of multiple impairments and to issue new regulations for evaluating mental impairments that “realistically evaluate the ability of a mentally impaired person to engage in [substantial work] in a competitive workplace.” Although some scholars disparage the new rules as “liberal” or “subjective,” they nevertheless reflected Congress’s determination to give fair weight to the full range of medical evidence in complex cases.
An unfortunate tactic of some program critics is to compare today’s receipt rates with those of the early- and mid-1980s. That amounts, however, to cherry-picking the data. Rates of receipt fell to record lows in 1982 through 1984 in the heyday of the Reagan Administration crackdown, and those thus are atypical years for DI receipt. By enacting DBRA on a bipartisan basis, lawmakers convincingly repudiated the practices of that time.
Workplace factors. Work is less physical than in the past, leading some analysts to expect a declining prevalence of disability. But a surprisingly large fraction of jobs — including those performed by older workers — still involves arduous physical demands or difficult working conditions. Even sedentary work carries its own set of health hazards, such as obesity.
The accelerating pace of globalization and technological change has been particularly unforgiving to older, less-educated workers and those with cognitive impairments. Whereas in the past such workers — even if they had serious health problems — might have been able to find jobs, the combination of poor health and poor labor market prospects has probably tipped many onto the disability rolls. A trio of researchers who generally argue that the shift to jobs that emphasize “mind over muscle” bodes well for the future employment of older workers nevertheless caution that “[c]ognitively demanding work may be better suited for older people than physically demanding work, but probably not for those with limited education.”
- Rising cost and declining availability of health insurance. DI beneficiaries qualify for Medicare after a two-year waiting period. With employer-sponsored health insurance eroding and the individual-policy market becoming costlier or outright unavailable, Medicare eligibility may loom larger and larger in some workers’ decisions to apply for DI. Researchers have found evidence that it is a significant factor for some applicants, and some suggest that implementation of the Affordable Care Act may diminish pressures on the DI program.
- Rising retirement age. Social Security’s rising retirement age has a very simple, direct effect on disability caseloads by delaying the conversion to retirement benefits. It also has an indirect effect by making disability benefits relatively more attractive. The basic benefit for a disabled worker is the same benefit paid to a worker who files at his or her full retirement age. Workers may file for reduced retirement benefits as early as age 62. As the full retirement age rises, that reduction becomes deeper. The full retirement age has already risen from 65 to 66 and will reach 67 in the next decade. Some researchers believe this growing wedge between reduced and full retirement benefits explains significant growth in the DI program, although others conclude it has boosted applications more than actual receipt.
- Economic downturn. Many observers — buttressed by press stories and academic studies — assume that the Great Recession and its aftermath account for rapid growth in the disability rolls. Yet economists generally find that while a sour economy significantly boosts applications to the program, it has a much smaller effect on awards. The implication is that economic downturns tend to attract more marginal, partially disabled applicants, but their applications are more likely to be denied. Therefore, while the economic downturn has surely contributed to the program’s growth, its influence should not be overstated.
One frequently overlooked facet of recent growth in the DI rolls is the fact that women have caught up with men. Until the mid-1990s, insured women of any age — that is, women who had worked enough to qualify for DI in the event of disability — were only about three-fourths as likely as insured men to receive DI benefits. Now they are equally likely to do so. (See Figure 3.) Because this comparison is limited to insured workers, this change is not simply explained by women’s rising labor force participation. Researchers — who have overwhelmingly focused on how DI affects males’ labor force participation — have rarely noted this trend and have studied it even less. Whatever the reasons for this trend, it does not seem valid to criticize as a deficiency, or a sign of recent laxness in the program, the growth in DI receipt that results from insured women reaching parity with insured men.
Trends in DI contrast with those in Supplemental Security Income (SSI), which is also run by the Social Security Administration but which — unlike DI (which receives a share of the Social Security payroll tax) — is means-tested and financed by general revenues. SSI pays subsistence benefits to people who are elderly or disabled and have little or no income and assets. People with severe disabilities who lack the work history for DI — as well as some who receive a very small DI benefit — can turn to SSI for help to meet their basic needs. Until the recent economic downturn, the number of working-age SSI recipients between 18 and 64 had been stable or declining as a percentage of the U.S. population since the mid-1990s. That trend is almost certainly related to the maturation of the DI program. As more people (especially women) qualify for DI on the basis of their prior work history and receive DI benefits that lift them over SSI’s meager income limits, fewer qualify for SSI — a fact that is often overlooked.
Eligibility Criteria Are Stringent
- Insured status. Applicants for DI benefits must be both fully insured and disability insured. In general that means they must have worked for at least one-fourth of their adult lives and in at least five of the last ten years. Applicants who cannot meet these requirements do not qualify for DI. (They may turn to SSI if their income and assets are very low.)
- Severe impairment. Applicants must show that they suffer from a “severe, medically determinable physical or mental impairment that is expected to last 12 months or result in death.” Acceptable medical sources are licensed physicians or (for certain conditions) licensed psychologists, optometrists, speech/language pathologists, or podiatrists. The agency generally gives greater weight to the applicant’s treating physician, but treats that provider’s opinion on the nature and severity of the applicant’s impairment as controlling only when it is well supported by clinical and laboratory diagnostic techniques and is consistent with the other substantial evidence in the case record. Other professionals — such as nurse practitioners or licensed clinical social workers — do not suffice, nor do statements from the applicant’s family, friends, teachers, or co-workers. The Social Security Administration (SSA) will order and pay for a consultative examination where merited.
- Inability to perform substantial work. Applicants must be unable to perform substantial gainful activity, which is currently defined as an inability to earn $1,040 per month ($1,740 for the blind). That threshold amounts to working less than full-time (about 35 hours a week) at the minimum wage of $7.25, or less than 40 percent of the median earnings of full-time workers with a high school diploma but no college. The law specifically requires that the applicant’s impairment must render him not just unable to do his past work, but unable — considering his age, education, and work experience — to do any other kind of work that exists in the national economy, regardless of whether that work exists in his geographic area or whether he would be hired if he applied. So-called vocational factors — experience and education — are considered for older applicants with limited skills and education.
Claimants apply to the SSA, which rejects people who are technically disqualified (chiefly because they lack insured status) and submits the remaining applications to each state’s disability determination service (DDS) for medical evaluation. If denied by the DDS, the applicant may appeal. Ultimately, of about 1,000 initial applications, about 410 are awarded benefits — more than one-third of them on appeal.
Typical processing times at the DDS level are three to four months, and processing times at the hearing level average about a year. The allowance rate at the Administrative Law Judge (ALJ) level (also known as the hearing level, generally the second level of appeal) is quite high, which has led to some valid concerns about inconsistency in decisions; yet it is important to remember that ALJs are often seeing claimants whose condition has deteriorated since their application was turned down and whose case file is better documented when it reaches the ALJ (often with the help of an attorney) than it was at the DDS stage.
Some critics imply that the geographic disparity in receipt of DI and SSI benefits is a sign of inconsistent standards. That is not correct. States that have a less-educated population (as evidenced by lower rates of high-school graduation), an older median age, fewer immigrants, and an industry-based economy (that is, with a greater-than-average concentration of mining, manufacturing, and forestry) also tend to have more disability recipients. (See Figure 4.) The program’s eligibility rules explicitly take into account applicants’ age, education, and ability to do past work or to transfer skills to another field of work. Geographic variation is a natural result.
If accepted, claimants are subject to periodic review to verify that they are still disabled. These continuing disability reviews (CDRs) are, by law, supposed to be conducted at least once every three years unless the beneficiary’s disability has been judged to be permanent. SSA estimates that CDRs result in eventual savings of nearly $10 in benefits (in Social Security, SSI, Medicare, and Medicaid) for each $1 they cost to conduct. Nevertheless, as discussed below, Congressional cost-cutting efforts have hampered SSA’s ability to conduct these reviews on schedule.
DI Provides Modest, but Critical, Benefits
DI recipients receive modest benefits, which are calculated by applying a progressive formula to their average earnings from early adulthood until the onset of disability (with up to five years of zero or low earnings dropped). Under the formula, higher earners receive larger dollar benefits but a lower percentage of their past earnings — a fraction known as the “replacement rate” — than do workers who received lower wages over their careers.
Most disabled workers collect benefits only for themselves. In a minority of cases, other family members may also be eligible to collect — most commonly, the minor children of the worker.
The economic circumstances of most disabled workers are modest, and in some cases, even precarious. The average monthly DI benefit in December 2012 was just $1,130 (or $13,564 on an annual basis). Only 7 percent of DI beneficiaries collected more than $2,000 a month. A careful comparison of disabled workers’ benefits to their past earnings found that their benefits replaced about 55 percent to 60 percent of average lifelong earnings for a median worker, and about 50 percent to 55 percent of final earnings prior to the disability. People who receive disability insurance benefits undergo a sharp drop in their standards of living.
Few Beneficiaries Could Support Themselves by Working
Practically since the DI program’s creation, economists and policymakers have debated whether it results in workers leaving the labor market. Evidence suggests, however, that few beneficiaries could earn more than very small amounts if they did not receive DI.
The typical DI beneficiary is in his or her late 50s — 70 percent are over age 50, and 30 percent are 60 or older — and suffers from a severe mental, musculoskeletal, circulatory, respiratory, or another debilitating impairment. (See Figure 5.) Mortality rates among DI beneficiaries are three to five times as high as for the general population. Nearly one-quarter of beneficiaries lack a high school diploma, and only 10 percent have a four-year college degree. Labor-market prospects for such applicants are poor.
It is important to note that DI beneficiaries are permitted to work. After all, the criterion for eligibility is not complete inability to work, but rather the inability to perform substantial gainful activity (SGA). There is no bar on recipients earning up to the SGA threshold — currently $1,040 per month — while collecting benefits. Recipients may earn unlimited amounts for a short period without jeopardizing their benefits, while they test their ability to return to work. DI benefits are low, and one would expect beneficiaries to take advantage of these rules by trying to supplement their benefits with earnings if they are able to do so.
If beneficiaries could readily work, we might expect a substantial number to make use of the program’s work incentive features to maximize their earnings without losing DI benefits. Yet there’s scant evidence of such behavior. Studies of beneficiaries who hold their earnings just under the SGA (a behavior known as “parking”), presumably to avoid triggering benefit suspension, and of DI beneficiaries who are converted to retirement benefits (when their earnings are no longer subject to any restrictions) indicate some work capacity but not enough to be economically meaningful.
Researchers note that even rejected applicants — who are, presumably, less disabled than successful claimants — fare poorly in the labor market, thus illustrating that the program’s eligibility criteria are indeed stringent. The latest and most exhaustive study finds that barely half of rejected applicants have any earnings; even fewer had significant earnings; and — for those with earnings — median amounts are very low. (See Table 1.)
|Table 1 |
Even Rejected Applicants Fare Poorly in Labor Market
|Accepted applicants||Rejected applicants||Non-applicants|
|With any earnings two years after application||20%||53%||82%|
|With significant earnings two years after application||13%||43%||79%|
|Median nonzero earnings||$3,500||$10,000||$35,000|
|Source: CBPP based on von Wachter, Song, and Manchester, 2011. Data are for men age 45 through 64. Nonapplicants were selected to mimic applicants in terms of age and previous earnings. For simplicity, figures for accepted applicants are a weighted average of those allowed at the DDS and ALJ levels. “Significant” earnings were defined as the equivalent of three months of full-time work at minimum wage, or about $2,700 in 2000. Median earnings are expressed in 2000 dollars.|
A widely cited recent study provocatively implies that in as many as one-quarter of cases, applicants’ fates might hinge on whether they are assigned to a relatively lenient or extremely tough DDS examiner. It nevertheless concludes that the effects on work and earnings are relatively small — on average, only $1,600 to $2,600 a year.
In short, there is little reason to think that many DI beneficiaries could support themselves by working. The program’s beneficiaries are people who worked in the past, lost their ability to work substantially, and only rarely recover. Its criteria are sufficiently stringent that it rejects many applicants who struggle mightily in the labor market thereafter.
Policymakers Should Address DI in Context of Overall Solvency
As Stephen Goss, the chief actuary of the Social Security Administration, testified last week, the DI program faces financial challenges. It is currently experiencing its peak demographic stress, and pressures will lessen as the economy recovers and as the baby boomers “age out” of the disability program and onto the retirement rolls. Nevertheless, the DI program faces sustained deficits under current policies, and its trust fund is expected to be exhausted in 2016.
Over the long run, DI and the much larger OASI program face similar funding gaps. For both programs, the 75-year imbalance is about one-fifth of income or one-sixth of costs. DI’s insolvency, however, looms much closer. The separate OASI trust fund would face depletion in 2035. Combined, the two trust funds would run out in 2033.
DI’s projected exhaustion should not come as a surprise. When lawmakers last redirected some payroll tax revenue from OASI to DI in 1994, the program’s actuaries projected that step would keep DI solvent until 2016. Despite fluctuations in the meantime, the current projection anticipates depletion in 2016.
When the trust funds are depleted, if policymakers took no action, benefits would be cut to whatever level could be covered by incoming tax receipts. In the case of DI, that means benefits would be cut by about one-fifth in 2016, though by slightly smaller fractions in later years.
Such a sudden and sharp cut in benefits — benefits that recipients depend on for most or all of their income — is unacceptable. Because the DI and OASI programs face similar shortfalls, and because their eligibility criteria and benefit calculations are so closely intertwined, it makes sense to address them together. Lawmakers should take steps reasonably soon to put the entire Social Security program on a sound footing for the long run and divide payroll tax revenues between the two programs as necessary.
Addressing the programs in tandem makes compelling sense:
- More options are available. Many leading options to improve solvency — such as raising the taxable maximum (currently $113,700) or using the “chained CPI” to compute cost-of-living adjustments in Social Security and other programs (as well as to adjust features of the tax code that are indexed for inflation) — would bolster both the OASI and DI programs. Limiting the menu to options that would affect only DI (other than a straightforward payroll-tax increase) would leave few options, most of which would be draconian.
Many features are common to both programs. Key features of the OASI and DI programs are similar or identical. The insured-status tests, the method of computing Average Indexed Monthly Earnings (AIME), and the formula for calculating the Primary Insurance Amount (PIA, or basic benefit) from AIME are seamless in the two programs, and provisions for spouses’ and children’s benefits are similar. Some proposals to achieve solvency treat these similarities as an afterthought, concentrating on the retirement program and offhandedly stating that the options in question would not apply to DI recipients until they’re converted to retirees. But that can lead to perverse and inequitable results.
Changes in the AIME and PIA calculations are powerful tools for affecting the future level of benefits and should be carefully coordinated across Social Security’s retirement, disability, and survivors’ programs.
|Table 2 |
Rising Retirement Age Widens the Wedge Between Disability and Early-Retirement Benefits
|Year attaining age 62||Full retirement age (FRA)||% of PIA paid to |
|% of PIA paid to disabled worker|
|2000-2005||FRA increases by 2 months per year||100%|
|2017-2022||FRA increases by 2 months per year||100%|
|2022 and beyond||67||70%||100%|
|PIA=Primary Insurance Amount, the basic amount on which all Social Security benefits are based. Source: Social Security Administration.|
Some Social Security retirement changes would have strong spillover effects onto disability benefits. Another potent tool for achieving savings in Social Security is to increase the retirement age. This option appears in many deficit-reduction plans — for example, in the Bowles-Simpson plan, in proposals advanced in 2008 and 2010 by Congressman Paul Ryan, in illustrative options developed for the National Academy of Sciences, and many others. While raising the retirement age would not directly affect disabled-worker beneficiaries, it would — in the absence of other changes — worsen pressures on the disability program, by widening the gap between disability and early-retirement benefits and by delaying the age at which DI beneficiaries are converted to Social Security retirement beneficiaries
The basic benefit for a disabled worker is 100 percent of PIA — the same benefit paid to a worker who files at his or her full retirement age. That equivalence dates back to the inception of the DI program, when the full retirement age was 65 and there was no early-retirement option for men. The introduction of early retirement for men created a differential between disability and early-retirement benefits, and the increase in the retirement age to 66 — and eventually to 67 — widen that differential. (See Table 2.) Options to raise the retirement age further would exacerbate the gap.
Policymakers should take a hard look at the wedge between early-retirement and disability benefits before the retirement age rises to 67 under current law, and they certainly must address the issue if they propose to raise the age further.
- Changes in related programs should be considered. Social Security retirement and disability insurance, Medicare, Medicaid, and SSI serve overlapping populations. A balanced solvency package would consider those interactions and make selected changes to non-Social Security programs, as appropriate.
Because Social Security benefits are so modest and make up the principal source of income for most recipients, legislators should use tax increases to generate at least half of the savings in a solvency package. Those could come from raising the maximum amount of wages subject to the payroll tax (which now encompasses only about 83 percent of covered earnings, well short of the 90 percent figure envisioned in the 1977 amendments); broadening the tax base by subjecting voluntary salary-reduction plans, such as cafeteria plans and health-care Flexible Spending Accounts, to the payroll tax (like 401(k) plans and similar retirement accounts); and raising the payroll tax rate. Future workers are expected to be more prosperous than today’s. Under the trustees’ assumptions, the average worker will be 50 percent better off — in real terms — in 2040 than in 2012, and twice as well off by 2070. It is appropriate to devote a small portion of those gains to the payroll tax, while still leaving future workers with much higher take-home pay. Social Security is a popular program, and poll respondents consistently express a willingness to support it through taxes.
If policymakers cannot agree on a well-rounded solvency package before 2016, they should — as a stopgap — reallocate payroll taxes between the two programs. Congress has done that on at least six occasions in the past, and in both directions, demonstrating that this is a traditional and uncontroversial step. The actuaries estimate that temporarily raising the DI share (currently 1.8 percentage points) of the 12.4 percent payroll tax by 0.8 percentage points through 2014, and then by amounts that gradually shrink to just 0.2 percentage points in 2021-2029, would enable both of the trust funds to pay scheduled benefits through 2033 — their combined exhaustion date.
Policymakers Should Provide Sufficient Administrative Funds
Appropriations for SSA’s operations (including the tasks performed for SSA by the state disability determination services) are part of discretionary spending, a category that faces a tight squeeze in the years ahead. SSA’s administrative funding has been frozen since 2010, despite growing caseloads in all three of its programs (OASI, DI, and SSI). Moreover, the Budget Control Act (BCA) of 2011 adopted aggregate caps on discretionary spending that will cut non-defense discretionary programs significantly, in real terms, by 2021; and that’s before the automatic cuts imposed by sequestration.
SSA needs adequate funding to perform its jobs. Those include not just processing applications and administering payments, but carrying out crucial program integrity activities. The BCA included a special “cap adjustment” for such activities — which in SSA’s case include continuing disability reviews (CDRs) and SSI redeterminations of financial eligibility. These limited funding increases do not require offsetting reductions in other non-defense appropriations; in effect, such increases are outside the statutory caps on annual non-defense appropriations that the BCA established.
As noted previously, CDRs are estimated to reduce eventual benefit payments by nearly $10 for every $1 in increased administrative funding, by removing from the rolls people who are no longer eligible. Congress nevertheless failed to take full advantage of the allowable cap increases adjustment in the 2012 appropriation, and is on track to do so again in 2013. While lawmakers debate how to restore long-run solvency to Social Security, funding disability reviews properly is a sensible, money-saving step that they should take right now.
DI’s trust-fund exhaustion comes as no surprise — it was anticipated in 1995 — and should not be considered evidence that the program is out of control. While researchers cannot fully dissect all of the reasons for the program’s growth, it’s clear that the bulk of it comes from demographic factors, women’s entry into the labor force in large numbers, and the increase in the Social Security retirement age, and that the DI program’s growth will taper off in the next decade.
DI is an integral part of the Social Security program, and legislators should address it in the context of overall Social Security solvency. The common features and interactions of DI and OASI would make efforts to fix the two programs separately a mistake.
Because Social Security’s finances are fairly predictable, it is not difficult to craft revenue and benefit proposals that would place the program on a sound long-term footing. The best proposals would protect vulnerable workers and beneficiaries and give all participants ample warning of future changes to this vital program. That will enable them to plan their work, savings, and retirement with confidence — while continuing to count on Social Security’s protection in the event of early death or disability.
If policymakers are unable to agree on a well-rounded solvency package before DI faces depletion, they should reallocate taxes between the two programs as a stopgap, as they have done multiple times in the past, while intensifying efforts to develop a long-term solvency package that restores the program’s financial health for decades to come.
 Most of the material in this testimony appears in Kathy A. Ruffing, Social Security Disability Insurance is Vital to Workers with Severe Impairments, Center on Budget and Policy Priorities, August 9, 2012, https://www.cbpp.org/cms/?fa=view&id=3818.
 Katharine P. Collins and Anne Erfle, “Social Security Disability Benefits Reform Act of 1984: Legislative History and Summary of Provisions,” Social Security Bulletin, Vol. 48, No. 4, April 1985.
 Hye Jin Rho, “Hard Work? Patterns in Physically Demanding Labor Among Older Workers,” Center for Economic and Policy Research, August 2010; Darius Lakdawalla, Dana Goldman, and Baoping Shang, “The Health and Cost Consequences of Obesity Among The Future Elderly,” Health Affairs 10 (2005); H. Stephen Kaye, “Improved Employment Opportunities for People with Disabilities,” Disability Statistics Report (17), U.S. Department of Education, National Institute on Disability and Rehabilitation Research, 2003.
 See, for example, David H. Autor, David Dorn, and Gordon H. Harrison, “The China Syndrome: Local Labor Market Effects of Import Competition in the United States,” National Bureau of Economic Research Working Paper 18054, May 2012.
 Richard W. Johnson, Gordon B.T. Mermin, and Matthew Resseger, “Employment at Older Ages and the Changing Nature of Work,” AARP Public Policy Institute, November 2007. Reductions in funding for sheltered work, and the difficulty faced by the mentally ill in obtaining job accommodations notwithstanding the Americans with Disabilities Act, may also have pushed some people who are unable to cope with a competitive, more technological workplace onto the disability rolls; see Virginia P. Reno and Lisa D. Ekman, “Social Security Disability Insurance: Essential Protection When Work Incapacity Strikes,” Journal of Policy Analysis and Management, Vol. 31, No. 2 (2012).
 Kajal Lahiri, Jae Song, and Bernard Wixon, “A Model of Social Security Disability Insurance Using Matched SIPP/Administrative Data,” Journal of Econometrics 145 (2008); Norma B. Coe, Kelly Haverstick, Alicia H. Munnell, Anthony Webb, “What Explains State Variation in SSDI Application Rates?” Center for Retirement Research at Boston College, Working Paper 2011-23, December 2011.
 Mark Duggan, Perry Singleton, and Jae Song, “Aching to Retire? The Rise in the Full Retirement Age and Its Impact on the Disability Rolls,” National Bureau of Economic Research Working Paper 11811 (December 2005); Xiaoyan Li and Nicole Maestas, “Does the Rise in the Full Retirement Age Encourage Disability Benefits Applications? Evidence from the Health and Retirement Study,” Michigan Retirement Research Center Working Paper 2008-198 (September 2008); Norma B. Coe and Kelly Haverstick, “Measuring the Spillover to Disability Insurance Due to the Rise in the Full Retirement Age,” Center for Retirement Research at Boston College, Working Paper 2010-21, December 2010.
 Stephan Lindner and Austin Nichols, “The Impact of Temporary Assistance Programs on Disability Rolls and Re-Employment,” Working Paper 2012-2, Center for Retirement Research at Boston College, January 2012; Matthew S. Rutledge, “The Impact of Unemployment-Insurance Extensions on Disability Insurance Application and Allowance Rates,” Working Paper 2011-17, Center for Retirement Research at Boston College, October 2011; Alan Krueger and Andreas Mueller, in progress, “Applications for Disability Insurance and the Exhaustion of Unemployment Insurance Benefits: New Evidence from a Survey of Unemployed Workers.”
 Kalman Rupp and David C. Stapleton, “Determinants of the Growth in the Social Security Administration’s Disability Programs—An Overview,” Social Security Bulletin, Vol. 58, No. 4, October 1995;. Till von Wachter, Jae Song, and Joyce Manchester, “Trends in Employment and Earnings of Allowed and Rejected Applicants to the Social Security Disability Insurance Program,” American Economic Review, December 2011; Kalman Rupp, “Factors Affecting Initial Disability Allowance Rates for the Disability Insurance and Supplemental Security Income Programs: The Role of the Demographic and Diagnostic Composition of Applicants and Local Labor Market Conditions,” Social Security Bulletin, Vol. 72, No. 4, 2012.
 Annual Report of the Supplemental Security Income Program, May 2012, Table IV.B7.
 For applicants who become disabled very young — before age 31 — the recency requirement stipulates that they must have had earnings in half of the years since attaining age 21 (rather than in five of the last ten years).
 These acceptable medical sources are lists in the Code of Federal Regulations, §404.1513.
 See Code of Federal Regulations, §404.1527.
 According to the Bureau of Labor Statistics, median weekly earnings for high school graduates 25 and older, employed full-time, were $652 in 2012.
 Table 3.5, “Key Performance Targets,” in Social Security Administration, Full Justification of Estimates For Appropriations Committees, February 2012, http://www.socialsecurity.gov/budget/2013FullJustification.pdf.
 Kathy A. Ruffing, “The Geography of Disability,” Center on Budget and Policy Priorities Off the Charts blog, November 28, 2012, http://www.offthechartsblog.org/the-geography-of-disability/.
 Social Security Administration, Annual Report of Continuing Disability Reviews, May 1, 2012, http://www.socialsecurity.gov/legislation/FY%202010%20CDR%20Report.pdf.
 Spouses are eligible for benefits only if they are either age 62 or older, or are caring for the worker’s eligible children.
 Computed from http://www.ssa.gov/OACT/ProgData/benefitlevel.html. Specifically, 11 percent of male beneficiaries and 3 percent of females received $2,000 per month or more.
 L. Scott Muller, “The Effects of Wage Indexing on Social Security Disability Benefits,” Social Security Bulletin, Vol. 68, No. 3, 2008.
 Bruce D. Meyer and Wallace K.C. Mok, “Disability, Earnings, Income and Consumption,” National Bureau of Economic Research Working Paper 18869, March 2013.
 Anne DeCesaro and Jeffrey Hemmeter, “Characteristics of Noninstitutionalized DI and SSI Program Participants,” Social Security Administration, Office of Research, Evaluation, and Statistics, Research and Statistics Note No. 2008-02, January 2008.
 Table 3.C6, Annual Statistical Supplement to the Social Security Bulletin, 2011.
 Age data for December 2011, from the Social Security Administration, Office of the Chief Actuary, http://www.ssa.gov/OACT/ProgData/beniesQuery.html; diagnostic data from Social Security Administration, Annual Statistical Report on the Social Security Disability Insurance Program, 2011.
 De Cesaro and Hemmeter, op. cit.
 Arif Mamun, Paul O’Leary, David C. Wittenburg, and Jesse Gregory, “Employment Among Social Security Disability Program Beneficiaries, 1996-2007;” Su Liu and David C. Stapleton, “Longitudinal Statistics on Work Activity and Use of Employment Supports for New Social Security Disability Insurance Beneficiaries,” Social Security Bulletin, Vol. 71, No. 3, 2011.
 Jody Schimmel, David C. Stapleton, and Jae Song, “How Common is ‘Parking’ Among Social Security Disability Insurance Beneficiaries? Evidence from the 1999 Change in the Earnings Level of Substantial Gainful Activity,” Social Security Bulletin, Vol. 71, No. 4, 2011; Nicole Maestas and Na Yin, “The Labor Supply Effects of Disability Insurance Work Disincentives: Evidence from the Automatic Conversion to Retirement Benefits at Full Retirement Age,” Michigan Retirement Research Center, Working Paper 2008-194, September 2008.
 The table focuses on applicants age 45-64, who dominate the DI rolls. The authors found somewhat higher rates of employment among younger applicants: 58 percent of rejected male applicants age 30 to 44 (versus about 19 percent of accepted applicants) had non-negligible earnings two years after application, compared with 85 percent of nonapplicants. Again, however, the amount of earnings for applicants was paltry. The median rejected applicant who worked made just $8,000; accepted applicants earned far less.
 Nicole Maestas, Kathleen Mullen, and Alexander Strand, “Does Disability Insurance Receipt Discourage Work? Using Examiner Assignment to Estimate Causal Effects of SSDI Receipt,” Michigan Retirement Research Center Working Paper 2010-241, 2010.
 Testimony by Stephen C. Goss, Chief Actuary, Social Security Administration, before the House Committee on Ways and Means, Subcommittee on Social Security, March 14, 2013.
 The 2012 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Federal Disability Insurance Trust Funds, April 2012, http://www.ssa.gov/OACT/TR/2012/index.html; Kathy Ruffing, What the 2012 Trustees' Report Shows About Social Security, Center on Budget and Policy Priorities, May 10, 2012, https://www.cbpp.org/cms/index.cfm?fa=view&id=3774.
 Kathy Ruffing, Paul N. Van de Water, and Robert Greenstein, “Chained CPI Can Be Part of a Balanced Deficit-Reduction Package, Under Certain Conditions,” Center on Budget and Policy Priorities, February 22, 2012, https://www.cbpp.org/cms/?fa=view&id=3690.
 See, for example, options C1.1 through C2.8 on the actuaries’ “menu” at http://www.ssa.gov/OACT/solvency/provisions/index.html.
 Early retirement for women was introduced in 1956, for men in 1961. Few women qualified for DI in those early years of the program — male beneficiaries outnumbered females by about 4 to 1.
 Jasmine V. Tucker, Virginia P. Reno, and Thomas N. Bethell, Strengthening Social Security: What Do Americans Want?, National Academy of Social Insurance, January 2013; Pew Research Center for the People and the Press, Public Wants Changes in Entitlements, Not Changes in Benefits, July 7, 2011; Colette Thayer, Social Security 75th Anniversary Survey Report: Public Opinion Trends, AARP, August 2010.
 Using a narrow definition of “reallocation” — one in which the total tax rate remained the same but the split between OASI and DI changed — there were six such instances (three from OASI to DI, three from DI to OASI). Using a broader definition — one in which the total tax rate changed and the OASI and DI rates changed in opposite directions — there were an additional five instances (three from OASI to DI, two from DI to OASI).
 Memorandum to Alice Wade, Deputy Chief Actuary, from Chris Chaplain and Jason Schultz (Supervisory Actuaries) and Daniel Nickerson (Actuary), April 23, 2012. Estimates are based on the assumptions of the 2012 trustees’ report.
 SSA also handles many administrative functions related to Medicare enrollment, but those costs are reimbursed by the Medicare trust funds.
 Richard Kogan, “Congress Has Cut Discretionary Funding By $1.5 Trillion Over Ten Years,” Center on Budget and Policy Priorities, November 8, 2012, https://www.cbpp.org/cms/index.cfm?fa=view&id=3840; Richard Kogan, “OMB Announces Amounts of Budget Cuts Under Sequestration,” Center on Budget and Policy Priorities Off the Charts blog, March 4, 2013, http://www.offthechartsblog.org/omb-announces-amounts-of-budget-cuts-under-sequestration/.
 Kathy A. Ruffing, “Failure to Fund Disability Reviews Is Penny Wise and Pound Foolish,” Center on Budget and Policy Priorities Off the Charts blog, March 11, 2013, http://www.offthechartsblog.org/failure-to-fund-disability-reviews-properly-is-penny-wise-and-pound-foolish/.
 The 1994 Trustees Report anticipated that DI would be depleted in 1995. The 1995 report anticipated depletion in 2016. In the meantime, the Social Security Domestic Employment Reform Act of 1994 (Public Law 103-387, signed on October 22, 1994) made several changes to the OASDI program, including directing a modestly larger portion of the 6.2 percent payroll tax to DI.