A T A G L A N C E. The Gap Between Expected and Actual Retirement: Evidence From Longitudinal Data, by Sudipto Banerjee, Ph.D.

November 2014 Vol. 35, No. 11 Views on the Value of Voluntary Workplace Benefits: Findings from the 2014 Health and Voluntary Workplace Benefits Survey, p. 2 The Gap Between Expected and Actual Retirement: Evidence From Longitudinal Data, p. 10 A T A G L A N C E Views on the Value of Voluntary Workplace Benefits: Findings from the 2014 Health and Voluntary Workplace Benefits Survey, by Paul Fronstin, Ph.D., EBRI, and Ruth Helman, Greenwald & Associates Three-quarters of workers state that the benefits package an employer offers prospective workers is extremely (32 percent) or very (44 percent) important in their decision to accept or reject a job. Nevertheless, 34 percent are only somewhat satisfied with the benefits offered by their current employer, and 22 percent are not satisfied. Eighty-six percent of workers report that employment-based health insurance is extremely or very important, far more than for any other work place benefit. Workers identify lower cost (compared with purchasing benefits on their own) and choice as strong advantages of voluntary benefits. However, they are split with respect to their comfort in having their employer choose their benefits provider, and think the possibility that they may have to pay the full cost of any voluntary benefits is a strong or moderate disadvantage. The Gap Between Expected and Actual Retirement: Evidence From Longitudinal Data, by Sudipto Banerjee, Ph.D., EBRI The 2008 economic recession sharply increased the gap between expected and actual retirement. Pre- September 2008, before the investment markets crashed, 83.9 percent of workers retired either earlier or no later than three years after their expected retirement compared with only 59.3 percent who did so post- September 2008. Longitudinal findings (comparing same cohort at different points in time) show that more people (35.9 percent) actually retired after 65 than expected (18.9 percent), and among those who expected to retire after 65, more than half (56.6 percent) did so. It also shows that 38.0 percent retired before they planned, 48.0 percent retired after they planned, and 14.0 percent retired the year they planned. Longitudinal data also show that people who have a retirement plan tend to retire closer to when they planned, compared with those without a plan. It also found that the gap between expected and actual retirement is generally very small between those with defined benefit plans and defined contribution plans. A monthly newsletter from the EBRI Education and Research Fund 2014 Employee Benefit Research Institute

Views on the Value of Voluntary Workplace Benefits: Findings from the 2014 Health and Voluntary Workplace Benefits Survey By Paul Fronstin, Ph.D., EBRI, and Ruth Helman, Greenwald & Associates Introduction The Employee Benefit Research Institute (EBRI) has been conducting value of benefits surveys for 20 years to determine the relative importance of different benefits to workers and to assess the role played by benefits in job choice and job change over time. The surveys show consistency in the value of some benefits and substantial change on others. Workers continue to rank health insurance as the first- or second-most important benefit provided by employers. Between 1999 and 2014, the percentage of workers ranking health insurance as the first- or second-most important benefit varied between 74 percent and 82 percent (Figure 1). While the ranking of a retirement savings plan fell from 2001 to 2014, this may be due to the introduction of additional benefits in the survey, such as paid time off. This report examines public opinion surrounding voluntary workplace benefits. Data come from the 2014 EBRI/Greenwald & Associates Health and Voluntary Workplace Benefits Survey (WBS). Among other topics, the survey examines a broad spectrum of workplace benefits issues, with a particular focus on voluntary workplace benefits. The Importance of Employee Benefits The benefits package that an employer offers prospective workers is an important factor in their decision to accept or reject a job. One-third (32 percent) of workers say the benefits package is extremely important, while 44 percent say it is very important (Figure 2). In fact, 21 percent of workers report they have accepted, quit, or changed jobs because of the benefits, other than salary or wage level, that an employer offered or failed to offer. Nevertheless, many workers are not especially satisfied with the benefits package offered by their employer. While 11 percent report being extremely satisfied and 33 percent are very satisfied, another 34 percent are only somewhat satisfied, and more than 2 in 10 are not too satisfied (11 percent) or not at all satisfied (11 percent) (Figure 3). Furthermore, job satisfaction and worker morale are strongly correlated with benefits satisfaction. For example, more than one-half (55 percent) of those who are extremely satisfied with their benefits are also extremely satisfied with their current job, compared with just 25 percent of those who are very satisfied. Less than 10 percent of those who are at most somewhat satisfied with their benefits say they are extremely satisfied with their job. Workers overwhelmingly consider health insurance to be the most important workplace benefit. Nearly two-thirds (65 percent) say this benefit is extremely important, while an additional 21 percent consider it to be very important (Figure 4). Indeed, having access to health insurance through their employer is considered so important that almost 6 in 10 (58 percent) report they are planning to work longer than they would like in order to continue receiving health insurance through their employer. When asked why continuing to receive health insurance through their employer was important enough to delay retirement, the plurality (43 percent) responded with a comment regarding the cost of insurance if they had to purchase it on their own, while another 15 percent cited the cost of medical care. Other reasons mentioned include the importance of having health insurance (14 percent), having medical problems or preexisting conditions (7 percent), and the quality of the employer s plan (6 percent). ebri.org Notes November 2014 Vol. 35, No. 11 2

Figure 1 Percentage of Employees Identifying Benefit as Firstor Second-Most Important, Selected Years, 1999 2014 Health Insurance Retirement Savings Plan Paid Time Off Traditional Pension or DB* Plan 9 8 82% 76% 76% 75% 78% 74% 7 67% 65% 6 55% 5 5 4 37% 41% 3 23% 26% 29% 2 21% 21% 16% 13% 13% 13% 13% 1999 2001 2004 2012 2013 2014 Source: EBRI/Greenwald & Associates 1999 and 2001 Value of Benefits Survey, 2004 and 2012 Health Confidence Survey, 2013 and 2014 Health and Voluntary Workplace Benefits Survey. * Defined benefit plan. 5 Figure 2 Importance of Benefits Package in Decision to Accept Job, 2013 2014 45% 45% 44% 4 35% 33% 32% 2013 2014 3 25% 2 18% 19% 15% 5% 4% 2% 1% 1% Extremely Important Very Important Somewhat Important Not Too Important Not at All Important Source: EBRI/Greenwald & Associates 2013 and 2014 Health and Voluntary Workplace Benefits Survey. ebri.org Notes November 2014 Vol. 35, No. 11 3

4 Figure 3 Satisifaction With Benefits Package, 2013 2014 2013 2014 35% 3 33% 31% 31% 34% 25% 2 15% 14% 12% 11% 12% 11% 11% 5% Extremely Satisfied Very Satisfied Somewhat Satisfied Not Too Satisfied Not at All Satisfied Source: EBRI/Greenwald & Associates 2013 and 2014 Health and Voluntary Workplace Benefits Survey. Figure 4 Importance of Various Employee Benefits, 2013 2014 Extremely Important Very Important Somewhat Important Not too Important Not at all Important Health insurance 2014 65% 21% 8% 3% 2 2013 63 25 9 2 1 Retirement savings plan 2014 40 35 19 4 2 2013 30 40 23 6 1 Dental or vision insurance 2014 30 36 23 8 2 2013 29 38 25 6 2 Traditional pension or defined benefit plan 2014 23 27 31 14 6 2013 17 30 34 16 3 Life insurance 2014 17 26 34 17 6 2013 18 29 31 16 5 Retiree health insurance 2014 17 25 32 19 6 2013 16 24 37 19 5 Disability insurance 2014 15 29 36 15 5 2013 16 32 35 14 3 Other health-related insurance 2014 11 24 38 20 6 2013 NA Long-term care insurance 2014 11 23 38 22 6 2013 12 23 40 20 5 Other benefits 2014 5 10 28 37 20 2013 4 9 27 37 23 Source: EBRI/Greenwald & Associates 2013 and 2014 Health and Voluntary Workplace Benefits Survey. ebri.org Notes November 2014 Vol. 35, No. 11 4

A retirement savings plan (rated extremely or very important by 75 percent of workers) and dental or vision insurance (66 percent) are also among the highest-rated benefits. One-half (50 percent) of workers say a traditional pension or defined benefit plan is extremely or very important, while at least 4 in 10 indicate disability insurance (44 percent), life insurance (43 percent), and retiree health insurance (42 percent) are important. Benefits Coverage in the Workplace Benefits coverage in the workplace, including health insurance, is far from universal. More than three-quarters of workers (78 percent) report their employer offers them health insurance (Figure 5). More than 7 in 10 each indicate they are offered dental insurance (72 percent, up from 67 percent in 2013) and a retirement savings plan (71 percent, up from 66 percent), and almost two-thirds each say they are offered vision insurance (64 percent) and life insurance (64 percent, up from 58 percent). At least one-half each report their employer offers them short-term disability insurance (57 percent), long-term disability insurance (53 percent), a health savings account (HSA) (50 percent), and accidental death and dismemberment insurance (50 percent). However, only one-third each say they are offered accident insurance (34 percent), a traditional pension or defined benefit plan (33 percent, down from 38 percent), and long-term care insurance (30 percent, up from 25 percent). Fewer report being offered supplemental health insurance for workers (23 percent) or other non-core ancillary benefits. Further, not all workers offered a benefit at the workplace take advantage of it. Approximately 8 in 10 who are offered health insurance (82 percent), dental insurance (81 percent), a retirement savings plan (80 percent), and life insurance (79 percent), each report they currently have these benefits through their employer (Figure 6). Between two-thirds and three-quarters each of those offered vision insurance (74 percent), a traditional pension or defined benefit plan (74 percent), and short-term disability insurance (66 percent, down from 71 percent in 2013) indicate they have this coverage through the workplace, while approximately 6 in 10 each have long-term disability insurance (63 percent) and accidental death and dismemberment insurance (60 percent, down from 70 percent). Fewer report taking up other benefits offered by their employer. However, a substantial minority of workers may be confused about some of the benefits their employer offers them and how those benefits are funded. Roughly 3 in 10 each state they do not know whether their employer offers them health insurance for early retirees (36 percent), supplemental health insurance for retirees on Medicare (34 percent), home health insurance (33 percent), critical illness insurance (32 percent), cancer insurance (31 percent), and supplemental health insurance for workers (28 percent). One-quarter each do not know if they are offered long-term care insurance (25 percent) and accident insurance (24 percent). With the exception of health, vision, and dental insurance and a retirement savings plan, at least 1 in 10 of those offered each benefit examined in the survey do not know whether their employer pays all, some, or none of the cost of the benefit, while many others report their employer picks up some or all of the cost of non-core ancillary benefits. For example, one-half of those offered supplemental health insurance for workers (50 percent), accident insurance (48 percent), and prepaid legal services (45 percent) each say their employer pays all or some of the cost. Attitudes Toward Voluntary Benefits Workers see a number of advantages to voluntary benefits. Foremost among these are cost and choice (Figure 7). Nearly 6 in 10 (56 percent) report that a strong advantage of voluntary benefits is that purchasing these benefits through an employer may cost less than purchasing them on their own, with another 28 percent saying this is a moderate advantage. In fact, one-half of workers are extremely (18 percent) or very (32 percent) confident that insurance and other benefit products are less expensive when purchased through the workplace (Figure 8). One-half (49 percent) report that the ability to choose which benefits they want to purchase is a strong advantage, and 36 percent say it is a moderate advantage. Other advantages workers cite are portability (80 percent say it is a strong or ebri.org Notes November 2014 Vol. 35, No. 11 5

Figure 5 Percentage of Employees Reporting Benefits Are Offered by Employer, 2013 2014 Offered Not Offered Don't Know 2013 2014 2013 2014 2013 2014 Health insurance 76% 78% 23% 2 1% 2% Dental insurance 67 72 30 23 3 5 Retirement savings plan 66 71 31 25 3 4 Vision insurance 60 64 36 30 5 6 Life insurance 58 64 36 28 6 9 Short-term disability insurance 55 57 34 30 10 13 Long-term disability insurance 49 53 38 31 12 15 Accidental death & dismemberment insurance 48 50 38 33 13 17 Accident insurance NA 34 NA 42 NA 24 Traditional pension or defined benefit plan 38 33 52 55 10 12 Long-term care insurance 25 30 56 45 19 25 Supplemental health insurance for workers NA 23 NA 49 NA 28 Critical illness insurance 17 19 60 49 23 32 Stock options 16 19 76 70 8 12 Pre-paid legal services 14 17 72 69 14 14 Cancer insurance 13 16 66 53 21 31 Home health insurance 13 15 65 52 22 33 Medicare NA 14 NA 52 NA 34 Health insurance for early retirees NA 13 NA 51 NA 36 Auto insurance 7 8 86 85 8 7 Homeowner's insurance 4 5 87 86 9 9 A health savings account (HSA) NA 50 NA 39 NA 11 Pet insurance 5 4 86 87 9 9 Source: EBRI/Greenwald & Associates 2013 and 2014 Health and Voluntary Workplace Benefits Survey. Figure 6 Reported Take-up of Workplace Benefits, 2013 2014 Among Employees Among All Employees Offered Benefit 2013 2014 2013 2014 Health insurance 63% 64% 83% 82% Dental insurance 54 59 80 81 Retirement savings plan 53 57 80 80 Life insurance 47 50 81 79 Vision insurance 44 47 73 74 Short-term disability insurance 39 38 71 66 Long-term disability insurance 32 34 66 63 Accidental death & dismemberment insurance 34 30 70 60 Traditional pension or defined benefit plan 29 25 76 74 Health savings account (HSA) NA 19 NA 38 Accident insurance NA 15 NA 45 Long-term care insurance 10 11 39 36 Stock options 9 9 57 49 Critical illness insurance 7 7 41 34 Supplemental health insurance for workers NA 6 NA 26 Prepaid legal services 6 6 39 33 Cancer insurance 4 5 34 33 Home health insurance 3 3 27 23 Supplemental health insurance for retirees on Medicare NA 3 NA 20 Health insurance for early retirees NA 3 NA 21 Auto insurance 3 3 47 34 Homeow ner's insurance 2 1 50 22 Pet insurance 1 <0.5 26 9 Source: EBRI/Geenw ald & Associates 2013 and 2014 Health and Voluntary Workplace Benefits Survey. ebri.org Notes November 2014 Vol. 35, No. 11 6

Figure 7 Advantages and Disadvantages of Voluntary Benefits, 2014 Strong Advantage Moderate Advantage Neutral Moderate Disadvantage Strong Disadvantage 1% Purchasing benefits through your employer may cost less than purchasing them on your own 56% 28% 15% 1% You can choose which benefits you want to purchase 49% 36% 14% 1% You may be able to take the benefits with you when you leave your employer 46% 34% 19% 3% 2% The benefits are paid through payroll deduction 34% 35% 27% Your employer chooses the companies that provide the benefits 6% 15% 48% 25% 7% You may need to pay the full cost of any voluntary benefits you chose 4% 9% 38% 28% 22% 2 4 6 8 10 Source: EBRI/Greenwald & Associates 2014 Health and Voluntary Workplace Benefits Survey. 4 35% 3 Figure 8 Employee Confidence That Benefits Purchased Through the Workplace are Less Expensive, 2013 2014 32% 32% 38% 36% 2013 2014 25% 2 19% 18% 15% 12% 5% Extremely Confident Very Confident Somewhat Confident Not Too Confident Not at All Confident 1% 3% Source: EBRI/Greenwald & Associates 2013 and 2014 Health and Voluntary Workplace Benefits Survey. ebri.org Notes November 2014 Vol. 35, No. 11 7

moderate advantage) and payments made through payroll deduction (69 percent say it is a strong or moderate advantage. However, workers also see some disadvantages. One-half (50 percent) identify the potential of having to pay the full cost of any voluntary benefits they choose as a strong or moderate disadvantage. In addition, workers are more likely to say that the employer choosing the companies that provide the benefits is a disadvantage (32 percent) rather than an advantage (21 percent). Moreover, workers are split with respect to how comfortable they feel having their employer pick their benefits providers. While 4 in 10 (40 percent) are extremely or very comfortable having their employer pick the companies that provide their health insurance benefits, another 4 in 10 (41 percent) are only somewhat comfortable, and 19 percent are not too or not at all comfortable (Figure 9). Similar splits are found for comfort with having their employer pick their retirement benefits provider (36 percent extremely or very comfortable; 19 percent not too or not at all comfortable) and voluntary benefits providers (33 percent and 18 percent). A majority of workers think it is important for their employer to offer them a choice of benefit plans, particularly when it comes to health plans (Figure 10). More than 8 in 10 say it is extremely (46 percent) or very (36 percent) important for their employer to offer them a choice of health plans. More than 7 in 10 feel it is extremely (32 percent) or very (40 percent) important to be offered a choice of retirement plans, while almost 6 in 10 indicate it is extremely (22 percent) or very (35 percent) important to have a choice of disability plans. As findings from the WBS clearly show, worker benefits continue to be important to workers. Even with the Patient Protection and Affordable Care Act, employers who offer a strong worker-benefits package should find themselves with a competitive advantage over other companies when it comes to attracting and retaining desirable workers. Appendix The 2014 WBS These findings are part of the 2014 EBRI/Greenwald & Associates Health and Voluntary Workplace Benefits Survey (WBS), which examines a broad spectrum of attitudes regarding workplace benefits, including voluntary benefits and health benefits. The survey was conducted online June 12 19, 2014, using the Research Now consumer panel. A total of 1,517 workers in the United States ages 21 64 participated in the survey. The data are weighted by gender, age, and education to reflect the actual proportions in the employed population. Previously published trend data from the EBRI/Greenwald & Associates Health Confidence Survey (HCS) may differ from those published in more recent reports, as the prior data have been recut from the total adult population to match the survey population of the WBS: workers ages 21 64. In addition, comparisons of 2014 data with data from years prior to 2013 should be viewed with caution due to the move from telephone to online methodology in 2013. No theoretical basis exists for judging the accuracy of estimates obtained from non-probability samples such as the one used for the WBS. However, there are possible sources of error in all surveys (both probability and nonprobability) that may affect the reliability of survey results. These include imperfect sampling frames, refusals to be interviewed and other forms of nonresponse, the effects of question wording and question order, interviewer bias, and screening. While attempts are made to minimize these factors, it is impossible to quantify the errors that may result from them. The WBS is co-sponsored by the Employee Benefit Research Institute (EBRI), a private, nonprofit, nonpartisan, public-policy research organization, and Greenwald & Associates, Inc., a Washington, D.C.-based market research firm. The 2014 WBS data collection was funded by grants from 11 private organizations. Staffing was donated by EBRI and Greenwald & Associates. WBS materials and a list of underwriters may be accessed at the EBRI website: www.ebri.org/surveys/hcs/ ebri.org Notes November 2014 Vol. 35, No. 11 8

Figure 9 Employee Comfort With Having Employer Pick Benefits Providers, 2014 6 Health Insurance Benefits Retirement Benefits Voluntary Benefits 5 49% 45% 4 41% 3 3 27% 26% 2 14% 15% 14% 9% 7% 5% 4% 4% Extremely Comfortable Very Comfortable Somewhat Comfortable Not Too Comfortable Not at All Comfortable Source: EBRI/Greenwald & Associates 2014 Health and Voluntary Workplace Benefits Survey. 5 45% 46% Figure 10 Importance of Employer Offering a Choice of Benefit Plans, 2014 Health Plans Retirement Plans Disability Plans 4 4 35% 32% 36% 35% 34% 3 25% 2 22% 23% 15% 15% 5% 8% 4% 2% 1% 1% 1% Extremely Important Very Important Somewhat Important Not Too Important Not at All Important Source: EBRI/Greenwald & Associates 2014 Health and Voluntary Workplace Benefits Survey. ebri.org Notes November 2014 Vol. 35, No. 11 9

The Gap Between Expected and Actual Retirement: Evidence From Longitudinal Data By Sudipto Banerjee, Ph.D., Employee Benefit Research Institute Introduction Some past studies have shown that expectations and realities about the timing of retirement can be very different. For example, according to the 2014 Retirement Confidence Survey (RCS), only 9 percent of workers planned to retire before age 60, but 35 percent of retirees reported retiring that early. Retiring earlier than planned can affect retirement security adversely especially, as it often is, for reasons beyond the worker s control and such gaps between expected and actual retirement should be a cause of concern. But while interpreting such findings, it should be kept in mind that such cross-sectional comparisons have a drawback specifically, comparing expectations and realities of two different cohorts, often many years apart in age. This report attempts to address this drawback by analyzing the gap between expected and actual retirement using longitudinal data. It also points out the differences between the two sets of results. Data The data for this study come from the University of Michigan s Health and Retirement Study (HRS), which is sponsored by the National Institute on Aging. HRS, the most comprehensive national survey of older Americans, is a biennial, longitudinal survey started in 1992 with primary respondents who are at least 50 years old, along with their spouses, irrespective of the spouses ages. Each (survey) year, HRS respondents report their self-assessed probability of working full time after age 65. This information, combined with the actual percentage of respondents above age 65 who report working full time, is used for the cross-sectional comparison. HRS respondents also report the age or calendar year at which they plan to retire. To calculate the gap between expected and actual retirement for the same group of people (the longitudinal comparison), the planned retirement year of those between ages 50 and 61 in the year 1992 is compared with their actual year of retirement reported in surveys between 1992 and 2012. Cross-Sectional Difference Between Expected and Actual Full-time Work After Age 65 For the cross-sectional comparisons, data from 2006, 2008, 2010, and 2012 survey years are used. Figure 1 shows the average self-reported probability of working full time after age 65 for those between ages 50 and 64. The results are shown separately for men and women as well as for full-time workers and the entire sample (everyone between ages 50 and 64). A number of facts can be noted from Figure 1: First, the expected probability of continuing full-time work after age 65 is higher for full-time workers than the population as a whole. Second, men have higher expected probability of continuing full-time work after age 65 than women. Third, between 2006 and 2012, the expected probability of full-time work after age 65 has increased slowly, with large increases between 2006 and 2008. For example, in 2006, it was 41.7 percent among men working full time, which increased to 48.3 percent in 2008 and 48.7 percent in 2012. Among full-time working women, it was 39.7 percent in 2006 and 46 percent in 2012. ebri.org Notes November 2014 Vol. 35, No. 11 10

However, the actual percentages of people working full time after age 65 were much lower, as shown in Figure 2. Among men between ages 65 and 70, in 2006 only 10.9 percent worked full time, which increased to 13.0 percent in 2008 and dropped slightly in 2012 to 12.7 percent. Among women of the same age, only 4.7 percent worked full time in 2006, which increased to 6.0 percent in 2012. Figure 3 shows the percentage of people still working between ages 65 and 70, and this includes both full-time and part-time workers. Although Figure 2 shows that the percentage of full-time workers has increased slightly between 2006 and 2012, the sum of full-time and part-time workers hasn t changed much during this period. In 2006, 23.9 percent of men in this age group were working (either full or part time), and this increased to 24.6 percent in 2012. For women, it was 17.3 percent in both 2006 and 2012. So, in 2012, the expected probability of working full time after age 65 was 37.5 percent for men, but even after including part-time workers, just 24.6 percent were actually working between ages 65 and 70. The RCS (2014) reports similar large differences in retirement after age 65. For example, in 2014, 33 percent of workers in the RCS reported that they expected to retire after age 65 while only 16 percent of retirees reported actually doing so. So, large gaps exist between expected and actual labor-force participation behavior after age 65, at least in cross-sectional survey results. Before moving on to longitudinal results, the expected probability of working full time after age 65 is correlated with two key job characteristics: retirement plan participation and wages. Figure 4 shows how these probabilities vary between retirement plan participants and non-participants. This sample includes part-time and full-time workers between ages 50 and 64. In all the survey years between 2006 and 2012, retirement plan participants reported lower expected probability of working full time after age 65. For example, in 2012, the expected probability of working full time after age 65 was 50.8 percent for those who did not participate in a retirement plan in their current job, as opposed to 44.0 percent among those who did. Also, these probabilities are much higher than the 2006 numbers for both groups. Figure 5 shows how the expected probability of working full time after age 65 differs by the retirement plan type. The plan types are self-reported, and they are broadly classified as defined benefit (DB) or defined contribution (DC) plans. This sample includes part-time and full-time workers between ages 50 and 64 who participated in a retirement plan and had at least five years of tenure. Overall, the expected probabilities are higher for those in DC plans. But, those in DB plans showed a much higher increase (between 2006 and 2012) in the expected probability of working full time after age 65. For them, the expected probability was 26.9 percent in 2006 and rose to 38.7 percent in 2012, an increase of 11.8 percentage points. In comparison, the increase for those in DC plans during the same period was 4.0 percentage points, from 41.0 percent to 45.0 percent. Finally, Figure 6 shows how the expected probability of working full time after age 65 varies across the wage distribution (quartiles). As in Figure 4, this sample contains all part-time and full-time workers between ages 50 and 64. People with higher wages have a higher opportunity cost of leaving the labor force, so they might want to continue working longer. On the other hand, higher-income people may be better prepared for retirement and therefore able to retire earlier, so it is not clear what type of trend to expect with respect to wage income. And it turns out that there is no clear pattern in the data: Except for 2012, the difference in expected probability of working after age 65 between the top and bottom (hourly) wage quartiles was 1 percentage point or less. However, in 2012, the top wage quartile had an expected probability of working after 65 of 43.7 percent, which was almost 4 percentage points lower than bottom wage quartile (47.6 percent). ebri.org Notes November 2014 Vol. 35, No. 11 11

6 Figure 1 Expected Probability of Working Full Time After Age 65 Among Workers Ages 50 64 5 4 3 2 Men Women Men Women Full Sample Working FT 2006 30.9% 23.7% 41.7% 39.7% 2008 34.8% 28.1% 48.3% 45.8% 2010 35.2% 27.4% 45.9% 41.8% 2012 37.5% 30.2% 48.7% 46. 14% Figure 2 Percentage of Men and Women Working Full Time Ages 65 70 12% 8% 6% 4% 2% Men Women 2006 10.9% 4.7% 2008 13. 5.3% 2010 11.1% 5.2% 2012 12.7% 6. ebri.org Notes November 2014 Vol. 35, No. 11 12

Figure 3 Percentage of Men and Women Working Full Time and Part Time Ages 65 70 3 25% 2 15% 5% Men Women 2006 23.9% 17.3% 2008 27.4% 17.9% 2010 22. 16.9% 2012 24.6% 17.3% Figure 4 Expected Probability of Working Full Time After Age 65 Among Workers Ages 50 64, by Retirement Plan Participation 6 5 Participate in a Plan? No Yes 4 3 2 2006 2008 2010 2012 No 44.6% 49.3% 47.3% 50.8% Yes 35.4% 43.4% 39.3% 44. ebri.org Notes November 2014 Vol. 35, No. 11 13

Longitudinal Difference Between Expected and Actual Retirement The expected and actual retirement decisions for the same set of people haven t been studied much, as very few surveys follow its respondents over long periods. However, HRS does. The original HRS cohort included people born between 1931 and 1941 who were first interviewed in 1992 and every two years thereafter. This section focuses on this group. They first reported their expected retirement year in 1992 and their actual retirement date has been recorded over the surveys between 1992 and 2012. Figure 7 shows the frequency distribution of expected retirement age and actual retirement age in this group. As expected, there are large spikes at ages 62 and 65, but the spikes in expected retirement at those ages are much larger than the spikes in actual retirement: 22.0 percent expected to retire at either age 62 or 65 but only 13.2 percent and 8.9 percent, respectively, actually did. From ages 67 and above, the percentages of actual retirement are higher than expected retirement. Figure 8 shows the cumulative frequency distribution of expected and actual retirement age in this group. Before age 62, actual retirement is higher than expected retirement. This changes at age 62 with a large spike in expected retirement, and a large gap opens up between expected and actual retirement by age 65. Eighty-one percent expected to retire by age 65, but only 64.1 percent actually did so. This is consistent with other recent research (Banerjee and Blau, 2015; Copeland, 2014; Moffitt, 2012) which has shown that both labor-force participation and full-time work have been increasing among those ages 65 and above post mid-1990s. 18.9 percent in this group expected to retire after age 65 (not including 65) and among those who expected to retire after age 65, 56.6 percent did so. Figure 9 and 10 shows the frequency distribution and the cumulative frequency distribution of the difference between their actual retirement year and the planned retirement year: Negative values indicate how many years earlier people actually retired than their planned retirement year. Similarly, positive values indicate later-than-planned retirement. The distribution is slightly positively skewed: 1 38.0 percent of people retired before their planned year while 48.0 percent retired after their planned retirement year, meaning the group of later-than-planned retirees is 10 percentage points larger than earlier-than-planned retirees. The median (mid-point) and modal value (highest frequency) are both zero (where the planned retirement year and the actual retirement year was the same). Fourteen percent in this group retired when planned, while 8.5 percent retired one year earlier than planned and 8.2 percent retired one year later than planned. A total of 30.7 percent retired within one year (plus or minus) of their planned retirement year and 55.2 percent in this group retired within three years (plus or minus) of their planned retirement year. Figure 10 also shows that in terms of working longer than planned, 19.1 percent worked five or more years than their planned retirement year, and 5.7 percent worked more than 10 years after their planned retirement. Among those who retired earlier, 13.3 percent did so five years or earlier than their planned retirement year. Figures 11 and 12 show the frequency and cumulative frequency distributions, respectively, of the differences between expected and actual retirement for this group by the type of pension they had in 1992 when they reported their retirement expectations. Respondents are divided into three groups: those who participated in only a DB plan, those who participated in only a DC plan, and those who did not participate in any pension plan at work. Almost 16 percent of DB participants, 18.0 percent of DC participants, and 10.1 percent of those with no pension retired the exact year they expected. The percentage of those with no pension plan increases relative to the two other groups toward the tail of the distribution on both sides: This means more respondents with no pension had the larger differences between expected and actual retirement. The cumulative distribution shows that the difference between DB and DC participants is largest in the first four years after planned retirement, with DB participants having a higher actual retirement percentage. But for the most part of the distribution, the difference between two types of pension plan participants is very small. ebri.org Notes November 2014 Vol. 35, No. 11 14

5 Figure 5 Expected Probability of Working Full Time After Age 65 Among Retirement Plan Participants Ages 50 64, by Retirement Plan Type 45% 4 35% 3 25% 2 15% 5% DB DC 2006 26.9% 41. 2008 39. 43.8% 2010 34.1% 41.1% 2012 38.7% 45. 6 Figure 6 Expected Probability of Working Full Time After Age 65 Among Workers Ages 50 64, by Hourly Wage Quartile 5 4 3 2 2006 2008 2010 2012 1 38.7% 43.5% 41.8% 47.6% 2 37.1% 44.2% 43.1% 46. 3 41.4% 47.9% 41.6% 48.1% 4 37.7% 44.3% 41.7% 43.7% ebri.org Notes November 2014 Vol. 35, No. 11 15

25% Figure 7 Frequency Distribution of Expected and Actual Retirement Age for Workers Ages 50 61 in 1992 Who Retired Between 1992 2012 2 15% 5% 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 Expected Retirement 0.1% 0.1% 0.2% 0.2% 0.2% 2.9% 2. 1.6% 2.3% 2.2% 6.3% 3.8% 22. 13. 2.1% 22. 10.9% 1.2% 0.7% 0.4% 2.9% 1.4% 0.3% 0.1% 0. 0.5% 0.2% 0. 0.1% 0. 0.2% Actual Retirement 0.5% 0.4% 0.3% 0.6% 1.2% 2.2% 1.9% 2.3% 3.1% 3.5% 5.4% 6.3% 13.2% 8.8% 5.6% 8.9% 5.8% 4. 4.5% 3.2% 4.4% 3.8% 2.3% 2.2% 1.6% 1.5% 0.9% 0.7% 0.6% 0.4% 0.2% 10 Figure 8 Cumulative Frequency Distribution of Expected and Actual Retirement Age for Workers Ages 50 61 in 1992 Who Retired Between 1992 2012 9 8 7 6 5 4 3 2 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 Expected Retirement 0.1% 0.2% 0.5% 0.7% 0.9% 3.8% 5.8% 7.5% 9.8% 12.18.2%22.1%44.57.59.1%81.1%91.9%93.1%93.8%94.2%97.98.4%98.7%98.8%98.8%99.4%99.6%99.6%99.7%99.7%99.9% Actual Retirement 0.5% 0.9% 1.2% 1.8% 2.9% 5.1% 7. 9.3% 12.4%16.21.3%27.6%40.8%49.6%55.2%64.1%69.9%73.9%78.4%81.6%86.89.7%92.94.3%95.8%97.3%98.2%98.9%99.5%99.8% 100.0 ebri.org Notes November 2014 Vol. 35, No. 11 16

16% Figure 9 Frequency Distribution of the Difference (in Years) Between Actual and Planned Retirement for Workers Ages 50 61 in 1992 Who Retired Between 1992 2012 14% 14. 12% 8% 8.5% 8.2% 6% 6.7% 5.9% 5.7% 6.2% 4% 2% 3.6% 3.7% 4.5% 4.2% 3.1% 3.1% 2.5% 2.6% 2.1% 1.8% 1.6% 1. 1.3% 0.4% -10-9 -8-7 -6-5 -4-3 -2-1 0 1 2 3 4 5 6 7 8 9 10 10 Figure 10 Cumulative Frequency Distribution of the Difference (in Years) Between Actual and Planned Retirement for Workers Ages 50 61 in 1992 Who Retired Between 1992 2012 9 8 7 6 5 4 3 2-10 -9-8 -7-6 -5-4 -3-2 -1 0 1 2 3 4 5 6 7 8 9 10 Cum. Fr. 4. 5. 6.3% 8.3% 10.2% 13.3% 16.9% 23.7% 29.6% 38. 52. 60.2% 66. 72.1% 76.7% 80.9% 84.5% 87. 90.1% 92.7% 94.3% ebri.org Notes November 2014 Vol. 35, No. 11 17

Differences Between the Cross-Sectional and Longitudinal Findings The cross-sectional findings emphasize that the error in expectations is mostly one-sided: While workers expect to work longer, retirees tend to retire earlier than what the workers expected. But this is a comparison across two different cohorts, and longitudinal results show that is not necessarily true. In fact, they show that many retire after their expected retirement date. For example, as Figure 1 and Figure 2 show, in 2012, the expected probability of working full time after age 65 was 48.7 percent (for men) and 46.0 percent (for women) among full-time workers. But only 12.7 percent of full-time male and 6.0 percent of full-time female individuals worked after age 65. This indicates that many retirees left the work force earlier than when current workers expect to retire. However, the longitudinal results in Figure 8 show that while 18.9 percent expected to retire after age 65, 35.9 percent actually did so. And among the 18.9 percent who expected to retire after age 65, 56.6 percent did so. Figure 10 shows that 38.0 percent retired before they planned and 48.0 percent retired after they planned. These longitudinal results indicate that more people retired later than planned, which is different than what is suggested by the crosssectional findings. It should be noted that the workers and retirees in the cross-sections belong to different cohorts, and some may be many years apart. They may face very different economic conditions during their lifetimes, including tax rates, labor market conditions, Social Security rules, etc. These factors should be kept in mind while discussing the gap between expected and actual retirement in cross-sectional data. Differences Between Pre- and Post-September 2008 Retirements Figures 13 and 14 show the frequency and the cumulative frequency distributions respectively of the difference in years between expected and actual retirement for those who retired before September 2008, when the investment markets crashed, and those who retired later. While 14.8 percent in the pre-september 2008 period retired the year they expected, only 9.2 percent in the post-september 2008 period did so. As the cumulative distribution (Figure 14) shows, the differences between the two periods really start to show up after zero (i.e., on the delayed-retirement side). In the pre-september 2008 period, 72.4 percent retired either earlier or no later than one year after their expected retirement, but post-september 2008, only 49.6 percent did so. Pre-September 2008, 83.9 percent retired either earlier or no more than three years after expected retirement, while only 59.3 percent did so post-september 2008. Conclusion Past studies have suggested that there is a large gap between expected and actual retirement. Most of these studies are based on cross-sectional data, which compare the retirement expectation of workers to the actual retirement decision of retirees. These findings suggest that workers are overly optimistic about working longer than what the retiree experiences suggest. But the workers and retirees are two separate cohorts in any cross-section. This study reports the expectation and actual retirement decision for the same set of individuals and also compares this to the cross-sectional findings. Major findings include: Cross-sectional findings suggest that workers expect to work later than shown by those who have already retired. In 2012, the expected probability of working full time after age 65 was 48.7 percent for full-time men and 46.0 percent for full-time women. But only 12.7 percent and 6.0 percent of men and women actually worked full time after age 65. Longitudinal findings show that more people actually retired (35.9 percent) after 65 than expected (18.9 percent), and actual retirement is more evenly distributed between pre- and post-years of expected retirement than what the cross-sectional findings suggest. Longitudinally, 38.0 percent retired before they planned, 48.0 percent retired after they planned, and 14.0 percent retired the year they planned. ebri.org Notes November 2014 Vol. 35, No. 11 18

The longitudinal findings also show that people without a retirement plan had larger differences between expected and actual retirement, compared with those who had either a DB or DC retirement plan. Those who participated in a retirement plan reported a lower-than-expected probability of working full time after age 65. For example, in 2012, the expected probability of working full time after age 65 was 50.8 percent for those who did not participate in a retirement plan in their current jobs, compared with 44.0 percent among those who did. The expected probabilities of working full time after age 65 are higher for those in DC plans compared with those in DB plans. But, those with DB plans showed a much higher increase in the expected probability of working full time after age 65 between 2006 and 2012. The expected probability of working full time after age 65 does not vary with wage earnings. A large difference exists in later-than-expected retirement between pre- and post- September 2008, when the markets crashed. Pre-September 2008, 83.9 percent retired either earlier or no later than three years after their expected retirement, but only 59.3 percent did so post-september 2008. Clearly, the economic recession delayed the retirement of many people. ebri.org Notes November 2014 Vol. 35, No. 11 19

Figure 11 Frequency Distribution of the Difference (in Years) Between Actual and Planned Retirement for Workers Ages 50 61 in 1992 Who Retired Between 1992 2012, by Pension Type of the Job Held in 1992 2 18% 16% 14% 12% 8% 6% 4% 2% -10-9 -8-7 -6-5 -4-3 -2-1 0 1 2 3 4 5 6 7 8 9 10 Defined Benefit 0.1% 0.7% 0.7% 2. 1.3% 2.7% 4.3% 7.2% 7. 11. 15.6% 10.4% 6.1% 6.5% 2.8% 4.2% 2.5% 2.3% 2.2% 2.8% 1.3% Defined Contribution 0.2% 1.3% 0.8% 1.5% 1.3% 4.6% 2.7% 6.5% 5.6% 7.5% 18. 7.3% 5.4% 6. 6. 4.6% 4.4% 3.1% 3.5% 2.5% 1.9% No Pension 0.7% 0.8% 2.1% 2.4% 2. 2.1% 3.3% 5.9% 4.3% 6.3% 10.1% 6.8% 5.6% 7.2% 6. 4.8% 5.3% 2.8% 4.4% 2.4% 2.3% 10 Figure 12 Cumulative Frequency Distribution of the Difference (in Years) Between Actual and Planned Retirement for Workers Ages 50 61 in 1992 Who Retired Between 1992 2012, by Pension Type of the Job Held in 1992 9 8 7 6 5 4 3 2-10 -9-8 -7-6 -5-4 -3-2 -1 0 1 2 3 4 5 6 7 8 9 10 Defined Benefit 1.8% 2.5% 3.2% 5.1% 6.5% 9.1% 13.5% 20.6% 27.6% 38.6% 54.2% 64.6% 70.6% 77.1% 79.9% 84.1% 86.6% 88.8% 91. 93.7% 95.1% Defined Contribution 1.3% 2.7% 3.5% 5. 6.3% 10.9% 13.6% 20.2% 25.7% 33.2% 51.3% 58.5% 63.9% 69.9% 75.8% 80.4% 84.8% 87.9% 91.4% 93.9% 95.8% No Pension 5.5% 6.3% 8.4% 10.8% 12.8% 14.9% 18.2% 24.1% 28.4% 34.6% 44.7% 51.5% 57.1% 64.3% 70.3% 75.1% 80.4% 83.2% 87.6% 90. 92.3% ebri.org Notes November 2014 Vol. 35, No. 11 20

Figure 13 Frequency Distribution of the Difference (in Years) Between Actual and Planned Retirement for People Who Retired Pre-September 2008 and Post- (Including) September 2008 16% 14% 12% 8% 6% 4% 2% -10-9 -8-7 -6-5 -4-3 -2-1 0 1 2 3 4 5 6 7 8 9 10 Pre-September 2008 0.9% 1.3% 2.2% 2.7% 2.8% 3.9% 4.4% 7.4% 6.7% 8.8% 14.8% 8.8% 5.9% 5.6% 3.9% 2.9% 2.6% 1.9% 1.8% 1.3% 0.5% Post-September 2008 0.8% 1.5% 2.3% 2.1% 2. 3.1% 2.5% 5.2% 5.5% 4.8% 9.2% 5.1% 5.2% 4.6% 4.1% 6. 4.1% 3.3% 3.8% 3.3% 3.6% Figure 14 Cumulative Frequency Distribution of the Difference (in Years) Between Actual and Planned Retirement for People Who Retired Pre-September 2008 and Post- (Including) September 2008 10 9 8 7 6 5 4 3 2-10 -9-8 -7-6 -5-4 -3-2 -1 0 1 2 3 4 5 6 7 8 9 10 Pre-September 2008 8.6% 10. 12.2% 14.9% 17.7% 21.6% 26. 33.3% 40. 48.8% 63.6% 72.4% 78.3% 83.9% 87.8% 90.6% 93.2% 95.1% 96.9% 98.2% 98.6% Post-September 2008 6.5% 8. 10.3% 12.4% 14.4% 17.4% 19.9% 25.1% 30.5% 35.3% 44.5% 49.6% 54.7% 59.3% 63.3% 69.3% 73.3% 76.6% 80.4% 83.7% 87.2% ebri.org Notes November 2014 Vol. 35, No. 11 21

References Banerjee, Sudipto, and David Blau. Employment Trends by Age in the United States: Why are Older Workers Different? Journal of Human Resources (Forthcoming), 2015. Copeland, Craig. Labor Force Participation Rates of Population 55 and Older, 2013 EBRI Notes, Vol. 35, no. 4 (Employee Benefit Research Institute, April 2014). Moffitt, Robert. The U.S. Employment-Population Reversal in the 2000s: Facts and Explanations, NBER Working Paper 18920 (National Bureau of Economic Research, November 2012). Helman, Ruth, Nevin Adams, Craig Copeland, and Jack VanDerhei, The 2014 Retirement Confidence Survey: Confidence Rebounds for Those With Retirement Plans, EBRI Issue Brief, no. 397 (Employee Benefit Research Institute, March 2014). Endnotes 1 The mean is 0.705 which is higher than the median 0, so, the distribution is slightly positively skewed. ebri.org Notes November 2014 Vol. 35, No. 11 22

Where the world turns for the facts on U.S. employee benefits. Retirement and health benefits are at the heart of workers, employers, and our nation s economic security. Founded in 1978, EBRI is the most authoritative and objective source of information on these critical, complex issues. EBRI focuses solely on employee benefits research no lobbying or advocacy. EBRI stands alone in employee benefits research as an independent, nonprofit, and nonpartisan organization. It analyzes and reports research data without spin or underlying agenda. All findings, whether on financial data, options, or trends, are revealing and reliable the reason EBRI information is the gold standard for private analysts and decision makers, government policymakers, the media, and the public. EBRI explores the breadth of employee benefits and related issues. EBRI studies the world of health and retirement benefits issues such as 401(k)s, IRAs, retirement income adequacy, consumer-driven benefits, Social Security, tax treatment of both retirement and health benefits, cost management, worker and employer attitudes, policy reform proposals, and pension assets and funding. There is widespread recognition that if employee benefits data exist, EBRI knows it. EBRI delivers a steady stream of invaluable research and analysis. EBRI publications include in-depth coverage of key issues and trends; summaries of research findings and policy developments; timely factsheets on hot topics; regular updates on legislative and regulatory developments; comprehensive reference resources on benefit programs and workforce issues; and major surveys of public attitudes. EBRI meetings present and explore issues with thought leaders from all sectors. EBRI regularly provides congressional testimony, and briefs policymakers, member organizations, and the media on employer benefits. EBRI issues press releases on newsworthy developments, and is among the most widely quoted sources on employee benefits by all media. EBRI directs members and other constituencies to the information they need and undertakes new research on an ongoing basis. EBRI maintains and analyzes the most comprehensive database of 401(k)-type programs in the world. Its computer simulation analyses on Social Security reform and retirement income adequacy are unique. EBRI makes information freely available to all. EBRI assumes a public service responsibility to make its findings completely accessible at www.ebri.org so that all decisions that relate to employee benefits, whether made in Congress or board rooms or families homes, are based on the highest quality, most dependable information. EBRI s Web site posts all research findings, publications, and news alerts. EBRI also extends its education and public service role to improving Americans financial knowledge through its award-winning public service campaign ChoosetoSave and the companion site www.choosetosave.org EBRI is supported by organizations from all industries and sectors that appreciate the value of unbiased, reliable information on employee benefits. Visit www.ebri.org/about/join/ for more. 1100 13 th Street NW Suite 878 Washington, DC 20005 (202) 659-0670 www.ebri.org www.choosetosave.org

EBRI Employee Benefit Research Institute Notes (ISSN 1085 4452) is published monthly by the Employee Benefit Research Institute, 1100 13 th St. NW, Suite 878, Washington, DC 20005-4051, at $300 per year or is included as part of a membership subscription. Periodicals postage rate paid in Washington, DC, and additional mailing offices. POSTMASTER: Send address changes to: EBRI Notes, 1100 13 th St. NW, Suite 878, Washington, DC 20005-4051. Copyright 2014 by Employee Benefit Research Institute. All rights reserved, Vol. 35, no. 11. Who we are What we do Our publications Orders/ Subscriptions The Employee Benefit Research Institute (EBRI) was founded in 1978. Its mission is to contribute to, to encourage, and to enhance the development of sound employee benefit programs and sound public policy through objective research and education. EBRI is the only private, nonprofit, nonpartisan, Washington, DC-based organization committed exclusively to public policy research and education on economic security and employee benefit issues. EBRI s membership includes a cross-section of pension funds; businesses; trade associations; labor unions; health care providers and insurers; government organizations; and service firms. EBRI s work advances knowledge and understanding of employee benefits and their importance to the nation s economy among policymakers, the news media, and the public. It does this by conducting and publishing policy research, analysis, and special reports on employee benefits issues; holding educational briefings for EBRI members, congressional and federal agency staff, and the news media; and sponsoring public opinion surveys on employee benefit issues. EBRI s Education and Research Fund (EBRI-ERF) performs the charitable, educational, and scientific functions of the Institute. EBRI-ERF is a tax-exempt organization supported by contributions and grants. EBRI Issue Briefs are periodicals providing expert evaluations of employee benefit issues and trends, as well as critical analyses of employee benefit policies and proposals. EBRI Notes is a monthly periodical providing current information on a variety of employee benefit topics. EBRIef is a weekly roundup of EBRI research and insights, as well as updates on surveys, studies, litigation, legislation and regulation affecting employee benefit plans, while EBRI s Blog supplements our regular publications, offering commentary on questions received from news reporters, policymakers, and others. The EBRI Databook on Employee Benefits is a statistical reference work on employee benefit programs and work force-related issues. Contact EBRI Publications, (202) 659-0670; fax publication orders to (202) 775-6312. Subscriptions to EBRI Issue Briefs are included as part of EBRI membership, or as part of a $199 annual subscription to EBRI Notes and EBRI Issue Briefs. Change of Address: EBRI, 1100 13th St. NW, Suite 878, Washington, DC, 20005-4051, (202) 659-0670; fax number, (202) 775-6312; e-mail: subscriptions@ebri.org Membership Information: Inquiries regarding EBRI membership and/or contributions to EBRI-ERF should be directed to EBRI President Dallas Salisbury at the above address, (202) 659-0670; e-mail: salisbury@ebri.org Editorial Board: Dallas L. Salisbury, publisher; Stephen Blakely, editor. Any views expressed in this publication and those of the authors should not be ascribed to the officers, trustees, members, or other sponsors of the Employee Benefit Research Institute, the EBRI Education and Research Fund, or their staffs. Nothing herein is to be construed as an attempt to aid or hinder the adoption of any pending legislation, regulation, or interpretative rule, or as legal, accounting, actuarial, or other such professional advice. EBRI Notes is registered in the U.S. Patent and Trademark Office. ISSN: 1085 4452 1085 4452/90 $.50+.50 2014, Employee Benefit Research Institute Education and Research Fund. All rights reserved.