Given the controversy about public employee benefits, it is important to get a clear picture of how public retirement benefits compare to private retirement benefits. This post updates and expands a previous post on this website.
Question: All-in, who makes more, public sector workers or private sector workers?
Answer: On average, public sector workers make more than private sector workers but they may make less when compared to private sector workers with similar education and skills.
The absolute numbers are clear and put public employees as much as 50% above private employees in total compensation. Fully adjusted comparisons, reflecting all factors depend on complex statistical analysis, but make compensation levels more equal. For local wage comparisons see this post; for a previous discussion of all-in compensation, see this post.
Question: Retired public employees get health insurance. What about private sector retirees?
Answer: Most don’t.
In 2009, only 11% of firms in the United States provided health insurance to retirees, whether over or under 65. The number is higher for the largest private employers; among those with 1000 or more employees, 34.5% insured retirees under 65 and 31.8 insured retirees over 65. See Table 1.A.2.e (2009) in the Medical Expenditure Panel Survey conducted by the federal Agency for Healthcare Research and Quality. For an older discussion of the trend towards reducing retiree health care benefits, see this EBRI issue brief from 2003.
Question: What kind of retirement allowance do public employees in Massachusetts receive?
Answer: 80% of their highest three years of compensation.
State and local government employees in Massachusetts are eligible for defined benefit retirement plans which will pay them a life annuity equal to 80% of their highest level of compensation (three year average) if they serve for 32 years. The minimum retirement age and other features vary across occupational categories, but the basic parameters of the benefits are standardized across Massachusetts governmental entities by Chapter 32 of the General Laws. This is an example of a “defined benefit plan” based on a percent of “terminal earnings”. Terminal earnings plans are typically more generous than plans that look at career earnings or pay some fixed dollar amount. A recent study found that state retirees had a median benefit of $50,130 (focusing only on non-public-safety retirees who had 30 or more years of service and who had retired in the past 5 years).
Question: Do private sector employees have access to defined benefit pensions like those given to public employees?
Answer: Not any more. Back in the 70s and 80s, defined benefit plans were standard for medium and larger employers. But the latest survey data show that only 20% of private sector workers have access to defined benefit plans. Only 7% have access to defined benefit plans based on a terminal earnings formula.
The annual National Compensation Survey by the Bureau of Labor Statistics is the primary source for benefits data. It is a rigorous probability sample of all employers — it included over 9000 respondents in 2010, achieving over a 2/3 response rate (See Appendix table 1). The NCS excludes agricultural, household workers and the self-employed, as well as federal government workers.
The Employment Benefits Research Institute has compiled the primary data from the NCS to show coverage in defined benefits programs decline from 84% in 1980 to 33% in 2009 for workers in private industry establishments in over 100 employees. The drop has occurred in part as a result of a shift from defined benefit plans to defined contribution plans — among all employees covered by pension plans in all sizes of employers in the private sector, 84% had defined benefits in 1980; by 2006, only 37% did.
The following statistics from the most recent NCS apply to private industry employers and workers:
- Only 10% of employers offer defined benefit pension plans. The number is higher for larger employers — 26% among employers with 100 to 499 workers, 50% among establishments with 500 or more workers. (Table 1)
- 20% of workers (averaged across all employer sizes) have access to defined benefit plans. The number is higher for full time workers — 24%. Among union members, 69% have access to defined benefit plans while only 15% of non-union workers have access to defined benefit plans. (Table 2) Unionization may explain the higher rates of plan participation among larger employers — among private sector industries, union membership rates are highest in utilities. Among workers participating in private sector retirement plans, only 11% are in defined-benefit only plans.
- Of the 20% of workers that are in defined benefit plans, 22% are in frozen plans — plans that are closed to new entrants or are not accruing benefits for all participants (Table 4) and 2/3 of these have frozen within the past five years (table 6), suggesting a trend that will further reduce the number of workers covered by defined benefit plans. See this additional BLS analysis.
- 45% of establishments offer defined contribution pension plans (Table 1) and these are available to 59% of workers (Table 2). It wasn’t until 1985 that the survey began covering defined contribution plans as a separate category.
Recent data on corporate retirement spending confirm a continuing trend towards more modest retirement plans with more risk carried by employees. For some consistent Massachusetts specific data see this survey from the Associated Industries of Massachusetts.
Question: Do private sector retirees derive much income from private retirement allowances — defined benefit plans, defined contribution plans, IRA’s, annuities, etc.?
Answer: Many don’t derive any. Focusing on men, more likely than women to have been in the work force for their whole life and so more likely to have accrued pension benefits, only a minority, perhaps as few as 1 in 3 among those who don’t have government pensions, have any private sector retirement benefit. The numbers vary depending on the survey instrument one looks at.
The Census Bureau’s Current Population Survey captures data about income sources. According to the Current Population Survey, Annual Social and Economic Supplement:
- Among men aged 65 to 74 in 2009 (in other words, likely workers retired within the past decade), only 2.44 (25.6%) of 9.53 million had union or corporate retirement income. Only 1/3 of those without a government retirement income source (federal, military or state/local — combined as if exclusive categories) had any form of private retirement income (corporate, union, railroad, annuities, IRA/Keogh or 401(k)). Among men 10 years older (over 75, earlier retirees) , the percentages are 33% and 42%, perhaps reflecting the slippage of private retirement security or perhaps reflecting higher retirement rates.
- Older publications in the Current Populations Report series group income categories differently, but EBRI has cross-tabulated much of the historical data. Chapter 7 of EBRI’s databook (see the 2010 update to Table 7.2) shows that for the last 20 years the share of persons over 65 receiving some form of private retirement income has remained relatively constant around 24%.
- EBRI notes some indications that the CPS data underreport the prevalence of pension income. See e.g., endnote 1 in this EBRI analysis. For in depth analysis of this issue, see this study from the Social Security Administration. It may be that the CPS understates the prevalence of income from private pensions by as much as 10 or 15 percentage points. To the same effect, see this EBRI note at pages 2 and 6.
For a couple of pieces that are still pessimistic, but suggest higher estimates of the prevalence of private savings, see the following. Neither of these actually distinguishes public from private workers, but the numbers are high enough to suggest somewhat higher prevalence of savings than the Current Population Survey.
- This article in the Wall Street Journal.refers to special-request work done by Center for Retirement Research at Boston College.
- This publication of the AARP is based on the Survey of Consumer Finances conducted by the Federal Reserve Bank. Figure 3 on page 6 suggests, backing out public workers, that roughly half of private workers over 65, have some form of retirement savings. The data in this study confirm the low prevalence of defined benefit plans. Figure 24 on page 18 suggests that for people nearing retirement age, the total amount saved in IRA/Keogh/401(k) type plans is a little over $100,000 (not at at all comparable to the value of a defined benefit retirement). The summary conclusion of the report is:
[H]owever pension coverage is measured, it remains low, just over 50 percent. Coverage rates vary markedly by age group and income level and with ethnicity, part-time versus full-time status, size of plan sponsor, and tenure. The low rates of coverage of the employer-provided pension system remain a matter of real concern. Workers without coverage through an employer are not making up the difference through contributions to IRAs or other personal saving.
Question: For the fraction of private sector retirees that do have retirement allowances, how do they compare to public sector allowances?
Answer: Private sector allowances appear to be considerably smaller.
EBRI estimates for 2008 put the mean annual income from pensions and annuities for males over 50 in the public sector at $29,374 as compared to $15,888 in the private sector (see figure 1 at page 12 at the link). One of the factors explaining this contrast has to be the relative instability of private sector employment — workers with longer tenure get higher benefits, but average tenure in the private sector is only half average tenure in the public sector. In addition, only 7% of private sector workers have access to defined benefit plans offering benefits based on terminal salary — the majority have defined contribution plans or defined benefit plans with less generous terms. Many private plans include terms that reduce benefits to reflect social security benefits received. As another measure, the pension assets accumulated per worker in the private sector are roughly twice what they are in the private sector — see figure 1 in this issue brief.
Among all men over 65 in 2009, those with corporate or union retirements received $14,393 from that source as compared to $26,255 for those with state and local government retirement income.
Here are some additional details on the plans available to the 20% of private employees who do have access to a defined benefit plan:
- Only 35% are based on terminal earnings. Table 19 from the 2009 BLS Survey.
- Among those based on terminal earnings, 72% use 5 year averaging as compared to the 3 year average in the Massachusetts pension system. Table 47 from the 2005 BLS Survey.
- 28% adjust downwards based on social security benefits. Table 41 from the 2005 BLS Survey. See also, Table 131, 2000 BLS Survey, showing 49% adjusting downwards.
- For those receiving a fixed dollar formula pension, the average is $41.65 per month per year of service. Table 48, 2005 BLS Survey.
- 55% allow use of earnings other than straight time earnings — in this respect they may be more favorable than Massachusetts pensions. Table 64, 2000 BLS Survey.
- For terminal earnings plans using a flat per-year-of-service percentage (only 35% of terminal earnings plans), the average percentage was 1.48, suggesting a pension of 50% of terminal earnings after a full career of service. Table 127, 2000 BLS Survey.
For more discussion on the comparison between public and private pension plans, see this issue brief from the Boston College Center for Retirement Research.
Question: But public employees in Massachusetts don’t get social security while private sector workers do . . .?
Answer: Yes. But social security benefits replace a relatively small fraction of income at retirement age. For teachers, judges and public safety workers, the public pensions are vastly more generous than social security. Even for a typical non-teacher in a non-hazardous job, the public pension is roughly twice as large as social security.
For a typical full-career state worker, who retires at age 65 at a salary of $63,000 after 35 years of service, the state pension might replace approximately 77% of the last year’s income. With the same earnings history, a private sector worker would be entitled to only 34% from social security.
Social security favors low wage earners and disfavors high wage earners so the comparison depends on income level. This spreadsheet enables exploration of pension-to-social-security comparisons at different income levels (and includes notes with links to sources on social security computations).
One benefit under social security is better protection for a non-working spouse — social security would provide a 50% add-on to the base benefit if the spouse had no benefits of his or her own. However, only about 1 in 15 social security recipients is able to add benefits for a spouse. In December 2008, there 32.2 million retired workers receiving social security old age benefits. There were only 2.3 million spouses receiving benefits as such. See the Annual Statistical Supplement, Table 5.C1.
Question: How does the benefit comparison add-up between public sector retirees without social security income and private sector retirees with social security income?
Answer: The available data do not allow a precise answer to this question, but it is clear that most private sector retirees have much less to work with than public sector retirees. Half or more go out without any corporate retirement benefits at all and must rely on social security, which gives them a much smaller fraction of their income than a public pension. Given the evidence above as to the rarity of private defined benefit plans and the low payouts from corporate retirement plans overall, it seems clear that even among those who do have some corporate benefits, many will have total income in retirement well below the full 80% replacement offered by public pensions.
Replacement rate is the ratio of starting pension to last working salary level. Here are summaries of several studies that attempt to analyze income replacement rates for private sector retirees:
- As the defined contribution pension declined in the private sector so did the detail offered about them in the BLS compensation surveys. The 1993 survey of medium and large private employers was the last to compute income replacement rates integrating social security benefits with pension benefits. Table 147 indicates that the average total retirement, for defined benefit plan participants retiring at 65 with 30 years of service, combining social security and pension benefits ranged from 84.2 percent of final annual earnings for persons with final annual earnings of $15,000 ($roughly 23,000 in 2011 dollars) down to 49.7 percent for persons with at the $65,000 level (roughly $100,000 in 2011 dollars). For comparison of this 1993 private sector data to 1993 public sector data, see this BLS study.
- One EBRI Analysis compares income replacement rates using data for 2003 from the Survey of Income and Participation (which tends to include more income than other surveys as noted above). The average rate for all workers who have worked for pay and receive a pension is 27% (median amount $8,340), varying from an average of 23% ($6,420) in the private sector to 36% ($12,000) in the public sector. The average combining both sectors is 31% ($9,804) for employees with 20 or more years of service. The replacement rate varies inconsistently with income — $20,000 was the lowest compensation level reflected in the study. If a private sector worker had a pension with a 23% replacement rate, he might have a total income replacement rate including social security near 60%.
- This study in the Monthy Labor Review uses a sophisticated model to estimate replacement rates for defined benefit plans. It does not distinguish public and private sector employees, but it highlights both the effect of longevity on pension earnings and the complexity of the estimating task in the absence of direct survey data.
- For a study highlighting the importance of how one defines income replacement, see this from the Social Security Administration.
Question: Do private sector retirees get cost-of-living increases?
Answer: They do for social security, and that is a very important protection for them. They typically don’t for pensions — in 2003, only 28.5% of persons over 65% had ever received an increase in their pension payment from their private employer. See page 6, Figure 5 in this EBRI note.
Question: Do national data about limited benefits in the private sector correctly reflect Massachusetts conditions?
Answer: Can’t say for sure, but most probably.
Massachusetts is the largest state in New England and in the most recent survey, New England private sector workers had essentially the same low rate of access to defined benefit plans (21%) that workers had nationally (20%). Also the median income for senior headed households is about the same in Massachusetts as nationally.
The American Community Survey, another Census Data Product, does not break out income by persons as the CPS does, but it does provide state specific data.
- Median income for households where householder is 65 or over in Massachusetts: $34,764
- Median income for households where householder is 65 or over in United States: $33,712
Question: Massachusetts public employees pay for the benefits that they receive, don’t they?
Answer: Not really. Public employee pension contributions withheld from pay are pre-tax and aren’t even considered income for tax purposes. Public employee W-2 statements omit the pension contribution amounts in the federal gross income box. As an economic reality, the entire amount of contribution is paid by the government, i.e., the taxpayers. Moreover, the amount contributed is sufficient to pay for benefits received only under favorable economic assumptions. Finally, even under those favorable assumptions, the amounts are insufficient to cover the costs of early retirements for public safety employees who are entitled to retire early.
Most (96%) of private sector defined benefit plans do not require an employee contribution and for the few that do, the median percentage contribution is 5.0%. (Table 3) in the 2010 National Compensation Survey by the Bureau of Labor Statistics. 63% of workers participating in defined contribution plans are required to contribute (Table 8).