Social Security IssuesThe following is the text to the July 1 speech given by Academy Senior~ Pension Fellow Ron Gebhardtsbauer in Providence, RI
Good morning and thank you for joining us. As Vice President Gore said, my name is Ron Gebhardtsbauer and Im the Senior Pension Fellow at the American Academy of Actuaries. The Academy is the non-partisan professional organization for actuaries in the United States and thus we dont take sides on political issues. Instead we discuss the advantages and disadvantages of proposed legislation. We at the Academy would like to thank the President and Vice President, AARP, the Concord Coalition, and Americans Discuss Social Security for inviting us to speak and for sponsoring this conference. Social Security has become a very important issue for us all, because it looks like we may fix Social Security early next year. It is also an important everyday issue for actuaries, because we help individuals, employers, and Social Security with their retirement issues.
Social Security, One of the Most Successful Programs Ever
I have been asked to discuss an overview of Social Security and its financial problems. But first, I
would like to quickly second the statements made by other speakers. Social Security has been one of
the most successful programs of this country. It is probably the primary reason for the dramatic
decreases in poverty rates among the elderly. Poverty rates among Americans over age 65
decreased from 35% in 1959 to about 11% today. This is about the same as poverty rates among
people of working ages. However, they are still pretty high for very elderly people, especially very
elderly women (22%).
Social Security is a very complex program with many benefits, so Ill just hit the highlights. My next chart shows the size of retirement benefits you can expect to get from Social Security based on your average annual indexed earnings at retirement. For example, if your average earnings are $20,000 per year, you will have almost half (actually about 47%) of your earnings replaced by Social Security (or almost $10,000 per year). If your average earnings are $60,000 per year, you will have only 1/4 (actually about 27%) of your earnings replaced (or about $16,000). Thus, you can see two of Social Security's primary goals from this chart. Namely, the tilt in the graph points out Social Security's concern for socially adequate benefits. Social Security provides a safety net for those that have nothing else. (However, people really need to save more in order to maintain their standard of living.) In addition, the following graph shows that the more one pays in, the more one gets from Social Security. This demonstrates their other goal of individual equity. Without this second goal, people might try to avoid paying more in taxes if they knew they were getting nothing for them.
These benefits are payable at the Social Security Normal Retirement Age, which is currently 65. However, starting in just 2 years (the year 2000), this age gradually starts to increase for people born in 1938 and later. I have a chart in my handout which can help you determine your retirement age, and you can take it with you for future reference. Currently, the Normal Retirement Age goes up to age 67 and that applies to people born in 1960 and later. Thus, the Generation Xers in the room will need to wait an extra 2 years (until age 67) to get the benefits on this chart. I should probably also note that Generation Xers will probably live more than 2 years longer than the typical elderly person of today, so they get could get more in total lifetime benefits than current retirees.
Social Security also has early retirement benefits. You can receive a benefit as early as age 62, but your benefit will be lowered to reflect the fact that you will get it for more years. In addition, you can delay your retirement date and thereby get a larger benefit. Pretty soon, the rules will automatically increase your benefit by 8% for every year that you delay your retirement. Thats called the delayed retirement credit.
The heading on this chart also points out another benefit of Social Security - the disability benefit. If disabled, you can get these same benefit amounts - the same as a retiree, even if you become disabled at a young age. It is an insurance benefit. Its value may be much more than you would have paid in.
In addition, your retirement benefits from Social Security are guaranteed, in that they dont depend on how well you invested, they increase every year by inflation and are payable for as long as you live. Currently, you cant buy inflation-indexed annuities from insurance companies and only a few private-sector pension plans in the country have it. This is a very special benefit you have from Social Security. Because of it, you dont need to worry about inflations impact on your benefit or outliving your benefit, no matter how long you live, and you dont have to worry about how to invest your money.
Another insurance benefit is the survivor benefit. I have a slide for it and more in the handouts. Generally, your surviving spouse gets a benefit at least as large as your retirement benefit, even if she (or he) never paid into Social Security. If she (or he) is caring for a child or is disabled, she (or he) can get it at a much earlier age. It could be worth $300,000 for someone with a wife and kids, and would be worth much more than you paid in, if you died when young. The chart also shows that your spouse can also get a benefit in addition to yours when you are both alive. Again, this is true even if the spouse never worked. These benefits are also payable to divorced spouses if the marriage lasted at least 10 years (and the spouse hasnt remarried - generally). This is valuable, especially for the traditional family where only one spouse works. Social Security has several other benefits, but I need to move on to my next topic: Does Social Security have a financial problem?
Social Security Has a Financial Problem
I also have to agree with earlier speakers that Social Security does have a financial problem. The actuaries at the Social Security Administration project that if no action is taken, Social Security will run out of money around the year 2032 (using the intermediate set of assumptions). However, that doesnt mean that Social Security wont be able to pay benefits at that time. According to the intermediate projections, when 2032 arrives, payroll taxes will still be enough to pay about 75% of the benefits. (Another set of assumptions is more optimistic and another one more pessimistic, but most people have agreed that the intermediate assumptions are the ones to base our decisions on.)
Social Security's financial problems are due to the very large baby boomer generation and our longer life spans. When Social Security was first created, life expectancies were less than age 65. Today they are over age 75. In fact, someone who is already age 65 can expect to live into his or her 80's. Due to these longer life spans and the retirement of the baby boomer generation, there will be fewer workers supporting more retirees in the future (unless we make some changes). For instance, today there are almost 3 workers per beneficiary. By 2032, the 3 decreases to just 2 workers per beneficiary. But 2032 is many years away. Why are we so concerned now?
Its because, in 2008, the very large boomer generation can start getting Social Security retirement
benefits (1945 + age 62 = 2008) and that can start causing major problems with the US Budget. In
order to avoid deficits, we may need to have our changes in effect by 2008. (The problems are due
to the interaction of Social Security with the US budget. Currently, Social Security receives about
$50 billion more in payroll tax than it pays out in benefits and administrative expenses, which helps
the U.S. budget appear about $50 billion better than it actually is. (If one also counts the interest
that the Treasury pays on Social Security's government securities, this number is about $100 billion.)
Starting around the year 2008, when the baby boomers start to retire, the $50 billion in extra payroll
taxes from Social Security will start going down, and will reach zero in the year 2013 or so. This
could cause deficits, which means we would have to either increase taxes or decrease government
programs at that point.) But we shouldnt wait until then to decide on the changes. We need to fix
Social Security sooner rather than later. For example, if we fix Social Security next year:
Solving the Problem
So lets talk about some solutions. We can either decrease Social Security benefits or increase taxes (or investment income). On this slide, you will see various options for reform, and how much of the problem they solve. For example, the second option is reducing the annual Cost of Living Adjustment (or COLA) that retirees get. If you reduce the COLA by percent each year, it would solve about 1/3 of Social Security's financial problems. Actually, youll note that none of the options is a silver bullet that solves all of Social Security's financial problem and all of them have at least one disadvantage. Thus, a complete solution requires 2 or more of the options and everyone could be affected, except possibly current retirees.
In addition, if we only make a one-time fix, the system will go out of balance again in 20
years. Well be back here in 2020 discussing it again, as long as we continue to live longer.
One way to avoid the financial problem in the future is to continue the fixes into the future
as we continue to live longer. Examples of fixes that can sustain Social Security for future
generations would be
The first 5 options decrease (or delay) benefits. The first option addresses head on the fact that we are living longer - it would Raise the Retirement Age for full benefits. Currently, Social Security's Normal Retirement Age, or age for full benefits, is age 65. But the Normal Retirement Age starts increasing very soon. Starting in the year 2000, just 1 years from now, the age for full benefits starts increasing for people born in 1938 and later. It quickly levels off at age 66 for 12 years. So for people like me, our normal retirement age is age 66, and thats true for the first half of the baby boomers. The retirement age for full benefits then starts going up again, and finally reaches age 67 for people born in 1960 and later. Thus, those of you here who are Generation Xers will have to wait until age 67 to get full benefits from Social Security - that is, youll have to wait 2 years longer than current retirees did to get full benefits. Of course, since you are expected on average to live more than 2 years longer, you will get at least as many years of benefits as current retirees did (on average).
One option is to increase the retirement age to 70 by the year 2030. Thereafter, this option would continue to increase the retirement age for full benefits (but at a slower rate), in order to keep the system from going out of balance in the future. Who here was born after 1943? Almost everyone here. Genereration Xers will have to wait at least 3 years longer to get a benefit compared with the current rules, although they will still get benefits for more years than people who retired in the early years of Social Security. This affects a lot of people, which is why this option solves over of Social Security's current financial problems. It also would affect all baby boomers like me who would then have to wait until age 67 or later for full benefits. Supporters of this option note that it makes sense since we are living longer, and we are healthier at older ages now. As I mentioned already, it can help solve about of Social Security's financial problem. (In fact, if we raised the retirement age to 73, it would solve all of the problem - but I bet that Congress wont do that.) Opponents of raising the normal retirement age note that it could be difficult for people who have physically demanding jobs and others who cant find work, or for those who are partially disabled (but not disabled enough to get disability benefits). It could also increase the average age of the workforce and raise employer costs for benefits, such as health care. Employers could encourage us to retire by improving our pensions, but that will cost a lot too. Some people question whether employers will hire us at older ages. They wonder if our health is improving as fast as our life span. Supporters cite recent studies, however, that indicate that we are healthier now at age 70 then people were at age 65 when Social Security was enacted. In addition, before Social Security most people worked to age 70 and beyond. Opponents also note that low-income minorities with shorter life spans will be affected more by this provision. Supporters note, however, that they are helped by the progressive benefit formula of Social Security, so they will still receive a better moneys worth on average, than other groups.
By the way, you can still retire at age 62 under the current rules, and if you do, your benefit will be smaller to reflect the fact that you will get your benefit for more years. For example, if and when the retirement age for full benefits becomes 70, then the benefit at age 62 would be about (55% actually) of your benefit at age 70. Thus, an increase in the retirement age for full benefits is a decrease in benefits (except for disability retirees - they would not be affected by an increase in the retirement age) and it does reduce the moneys worth of our contributions, which is true for most solutions to fix Social Security.
The second option is to Reduce the Cost of Living Adjustments (or COLAs) that retirees get each year. Currently, benefits go up by the annual Consumer Price Index (or CPI) so that retirees can buy the same quantity of goods and services each year. However, some people think that the CPI overstates inflation rates. A Congressional Commission (informally called the Boskin Commission), reported that the CPI was too high by 1.1%. Thus, one option might be to reduce the COLA to CPI minus percent. If the Commission was correct, peoples purchasing power would not go down and this could solve about 1/3 of Social Security's financial problems. Pretty powerful just for a % reduction. However, the Bureau of Labor Statistics has recently improved their calculation of the CPI. They expect it to lower the CPI by about 3/4 of a percent. Thus, opponents of this option are concerned that reducing the CPI further by 1/2%, could mean that retirees would fall behind in purchasing power by % each year. This is cumulative, so that after 30 years of this slippage, a retirees purchasing power could have fallen behind by about 15% ( % times 30 years). Thus, it particularly hits the very elderly, where poverty rates are much higher (especially for women). In addition, opponents want the calculation of the CPI to be a technical calculation, not a political decision.
The third option is to reduce benefits by 5%. It would be phased in over 5 years, so current retirees (and those currently eligible to retire) would not be affected. People at all income levels would have their benefits reduced by 5%. This option solves about 23% of Social Security's financial problems. (If benefits were reduced by 11%, if would solve about of the problem.) Opponents note that this is especially difficult on people with low incomes, since they often rely on Social Security for all (or almost all) of their retirement income. They would encourage us to look elsewhere for solutions, such as making the benefit formula more progressive (by lowering the benefits of higher wage earners only, or by instituting a means test, which is the next option). Supporters think everyone should be a part of the solution, even people with low incomes. In response to the concern for low income people, they think that the progressive tilt in the benefit formula is right where it should be and that making it more progressive would make the moneys worth of Social Security even worse for high earners.
The fourth option would be to gradually reduce benefits for those whose retirement income exceeds $45,000 per year. It is sometimes called an affluence test or means test. Once your familys total retirement income reached $110,000 in any year, you would get only 15% of your Social Security benefit (i.e., just a return of your contributions). When I discussed this option with my parents, they thought it could be a good idea. They figured it would only affect the millionaires, not them. When I told them they would be affected (since their total family retirement income was over $45,000), they werent so positive about it anymore. Their total family retirement income isnt much over $45,000, so they would only lose about 1/3 of their Social Security benefit. Thus, how you think about this option, may depend on where you stand. This option solves 3/4 of Social Security's financial problem. (Thats because it affects so many people so much. If it only affected millionaires, it wouldnt help anywhere near as much.) Supporters note this option preserves benefits to those most in need and reduces them for those who dont need them as much. Opponents note that the option hurts people who saved more, a behavior we want to encourage, not discourage. It could also discourage pensions. This option might also encourage abuse. People might hide their income, put their assets in trusts or give it to their kids, so that their Social Security benefit is not cut. In response, the government would write regulations to stop the abuse. Opponents other concern is that an Affluence Test could change the very nature of the Social Security program away from being a universal program where benefits are based on how much you contribute to one based on need. Opponents would rather use the progressive tax system to handle this (which I will discuss later) or make the benefit formula more progressive.
Another option would be to increase the number of years for calculating your benefit from 35 to 40. Currently, benefits are based on your highest 35 years of earnings. Additional years of work beyond 35 years do not improve your benefit much. One option would raise the 35 to 40, which would solve 21% of the problem. If you worked full-time for at least 40 years, this option would not change your benefit much at all. However, if you didnt work full-time for 40 years, your benefit could go down by as much as 12%. Opponents note that this would have the unintended consequence of hurting women who take time out to care for their families. Supporters note it would encourage people to work longer in order to get a better benefit. This would be good for the country because it would create more productivity, and it would help bring in more contributions for Social Security. (This option makes the charge for early retirement more accurate. The current method doesnt reflect the fact that early retirees contribute less to Social Security.) Furthermore, supporters dont want to hurt women who stay at home for child birth and child care reasons. They could remedy this problem by providing women with drop out years for periods when they are carrying or caring for a child.
Solving the Problem - Increasing Taxes
The next 4 options solve Social Security's problems through raising taxes. For example, option 5 suggests we raise the payroll tax rate. Right now, you pay 6.2% of your wages into Social Security and your employer does too. Self employed individuals pay both parts, for a total of 12.4% of earnings. This option would increase the total tax rate by 1% of wages (half to employees and half to employers), so that employees and employers would each pay 6.7%, for a total of 13.4% of wages. Supporters note that this solves almost half of Social Security's current financial problems and that people prefer a tax increase over a benefit decrease. Raising the total payroll tax rate by 2.2% to 14.6% would solve the current financial problems of Social Security (for 20 or 25 years). Opponents ask where the money will come from? Employers will have to raise prices if they can or lower their costs, such as labor costs. Low income people may take it out of their 401(k) contributions and lose their employers matching contribution. Others might have to borrow more or consume less. Opponents also note that we may have to increase payroll taxes for Medicare too, so that total payroll taxes could get much higher in total. In addition, as we continue to live longer, we will have to increase taxes every 20 or 25 years, which would tax future generations more than we were willing to tax ourselves today. It will be too late then for our children to cut our benefits or increase our retirement ages, so they could be forced into paying higher taxes than what we ever paid. Since this option is particularly difficult on lower income people, many would prefer that only higher income people pay more taxes, which is the next option.
The sixth option would raise the amount of wages that are subject to the payroll tax. This year, people pay Social Security taxes on the first $68,400 of their wages. Thats known as the maximum taxable wage base and it goes up every year by the same percent that average wages go up. This option would increase the amount of wages subject to payroll taxes over the next 5 years to $105,000 (which is about $90,000 in todays dollars). This particular option solves 1/4 of Social Security's financial problem. Eliminating the cap and paying taxes on all pay would solve about of the problem. Supporters note that low-income people pay Social Security taxes on all their income while high income people dont, and they can afford a tax increase better than low income people. Opponents note however, that Social Security benefits for high income people dont go up much for each additional dollar that they put in, because of the progressive benefit formula. (Eventually, this change would produce some unnecessarily large benefits for people that may not need them, unless the rules were changed to not count wages over a certain level in the benefit calculation, even if they did contribute on the higher wage amounts. However, if that happened, then some of the individual equity goals of Social Security would be hurt. Currently the more you put in, the more you get out of Social Security.) If people at higher incomes didnt get much or anything more for their additional contributions, then they might find ways around paying the additional taxes, and they would be less likely to support the system. In addition, opponents note that this increases the taxes of businesses, which means they have to either raise their prices if the can, or cut costs, such as labor costs.
The seventh option is to tax Social Security benefits like pension benefits from a private pension plan. Currently, a retired couple with a $20,000 pension and nothing else, would be taxed around $500 or so. But if the income was all from Social Security, there would be no tax on it. This is because you are not taxed on your Social Security benefits if your total income is below $32,000 (or $25,000 if you are single). Above those thresholds, you are taxed on only half of your Social Security benefit. However, if your income is above $44,000 ($34,000 if you are single), then 85% of your Social Security benefit is taxable; not the whole benefit since your contributions were taxed already. (They chose 85% because approximately 15% of your benefit comes from your own contributions which have already been taxed; the rest of your Social Security benefit is attributed to investment earnings and your employers contributions, which have not been taxed yet.) As you can see, this is quite complicated. This option would eliminate the thresholds, which would simplify the calculation a lot. Opponents are concerned that this might hurt low and middle income people. However, Supporters note that low income people will not be touched by this proposal. In fact, 30% of retirees would still pay no income tax due to the exemptions and deductions in the Federal Income Tax system. Only middle income people are affected, and not by a huge amount. Thats why this option doesnt help solve much of Social Security's financial problems (only 14% of the problem). Supporters also see this option as a way for all generations to be a part of the solution, even current retirees, and they note that it simplifies tax laws. They question why two retirees with the same income are taxed differently, just because one person gets their benefit from Social Security and the other doesnt.
The eighth option requires all newly hired State and Local Government workers to
be in Social Security. Some state and local workers participate only in their own pension
systems and dont participate in Social Security. This option would require their new
employees to be in Social Security. Supporters say that Social Security should be universal,
and that most people support this option (except some of those that would be affected).
Since many state and local workers get Social Security anyway through work at other jobs,
they should have to pay their fair share. Opponents note that these workers do fine under
their own systems, so why change the rules. In addition, it would divert employee and
employer contributions from their government plans. This option would bring more money
into the system in the short run, but would solve only about 10 percent of Social Security's
Ive discussed some of the possible solutions. However, there are problems that come along with these solutions. Decreasing Social Security benefits (or increasing the retirement age for full benefits) may put more reliance on the private pension system. It will shift costs to employees and employers. People need a certain amount of income to live and retirement is very much a financial decision. So with smaller Social Security benefits (or later retirement), many individuals will have to work longer (if they can). An older workforce will increase employer costs such as employee health, disability, life insurance, annual leave, and sick leave. If employers dont want an older workforce and the associated additional costs, they can lay off their older employees (always a difficult thing to do) or encourage them to retire by improving the company pension plan, but that will cost a lot too. Due to a huge increase in the number of retirements early in the next century, employers may want to rethink their retirement strategies and encourage employees to stay on (at least part-time). Phased retirement may become popular, but IRS regulations would need to be revised to allow in-service distributions to be payable before a pension plans Normal Retirement Age. In addition, it is quite difficult for employers to increase their Retirement Ages in tandem with Social Security, unless pension law allows higher normal retirement ages than age 65 and relaxes the rules against decreasing benefits. Otherwise, employers will have to calculate 2 separate pension amounts for service before and after each change in the retirement age. This will be very complex. However, it appears that Congress can increase Social Security retirement ages the easy way, but they wont let employers do it. Congress may want to allow employer some of the same flexibility. Finally, decreased Social Security benefits could necessitate changing the non-discrimination rules to reduce the disparity in benefits between low- and highly-compensated employees...
If Social Security COLAs are decreased, it will put more pressure on employer pension plans (at least the Defined Benefit variety) to give greater ad hoc increases to older retirees. It might encourage more lifetime annuity-type benefits. Employers with Defined Contribution plans might still be able to wash their hands of this problem, especially if all ties have been lost with the former employee by paying lump sums and not providing post-retirement benefits of any kind. Employees should prefer Defined Benefit plans more, especially if inflation could be high in their retirement, but they may not be thinking that far ahead.
If BLS lowers the CPI, it would also reduce other government retirement benefits (such as the CSRS and FERS benefits for federal employees) and entitlements, and it would increase income taxes (by not increasing the tax brackets as much). In addition, federal limits on pensions would be smaller in the future. It would also affect the economy, by lowering future expectations of inflation, lowering future wage increases, and interest rates.
As mentioned earlier, a means test would discourage savings and pension plans. It would also mess up offset plans and the rules that integrate pension benefits with Social Security. The employer pension would affect the Social Security benefit which would affect the pension, and on and on. As you can see this would create a circular problem, which could wipe out Social Security offset plans and necessitate a change in the 401(l ) disparity rules. Individuals who were clearly above the means testing threshold would need more income from their employer pension plan or they would need to save more. A means test would also encourage gaming the system. People would accelerate or delay the timing of their employer pension in order to get their full Social Security benefits. If the means test was based on income, people with large pensions would want to receive their benefits in a lump sum, so that they would only lose their Social Security benefit in one year. People with small pensions would not want a lump sum, because their pension would not reduce their Social Security benefit, but a lump sum would hurt them in the year of receipt. If the means test was based on wealth, people with large pensions might want to delay their pension for as long as possible or get it early in a lump sum and transfer it to a trust or child. IRS would have to create some very complex distribution rules which would make the current IRS distribution rules in 401(a)(9) seem simple.
In the Advisory Council proposal, the dependent spouses benefit is reduced from 50% to 33% of the primary workers benefit. This would hurt traditional families. In addition, the proposal would increase the survivor benefit from 2/3rds of the couples benefit to 3/4ths. This would help non-traditional families, but not traditional families (since it doesnt change their survivors benefit amount). This change may not affect pension plans. On the other hand, it could encourage more employees to follow the example of Social Security and elect Joint and 3/4 Survivor benefits. Alternatively, larger survivor benefits from Social Security could encourage employees to feel the survivor is taken care of, so more employees might elect life only benefits.
If we increase Social Security taxes, the moneys got to come from somewhere. Low paid employees may take it from their 401(k) contributions and lose the match. Highly compensated employees also would be restricted because of the non-discrimination rules. If the employee has no pension plan, the increased contribution would have to come from their savings or their consumption. If employers have to pay more into Social Security, they may reduce pension benefits or drop them altogether.
If the wage base is increased or eliminated, it will affect covered compensation and integrated plans. If other forms of compensation besides wages are taxed (such as pensions and health benefits or pension trusts), employers might reduce or drop them (and their cafeteria plans), due to the loss of some of the tax advantages. So you can see there are a lot of different repercussions involved in these proposed options. We have to discuss these unintended consequences before we implement changes such as these.
Using Private Sector Investments
So far Ive talked about either decreasing your benefits or increasing taxes. Another way to help solve Social Security's financial problems is to invest in the private sector. Either Social Security could do the investing or individuals could. These last two options are the most controversial, possibly because they have never been tried before in the United States. Currently, Social Security's Trust Funds can only be invested in government securities. Investing them in the private sector could yield Social Security a higher investment return. It would reduce our arguments about whether the government really saves the money when it buys Treasuries. And it could increase national savings if additional savings were required on top of the current payroll tax (not if it is carved out of the current tax). Sounds like a free lunch! However, its not, and like all the other solutions discussed above, this change doesnt solve all of Social Security's current financial problems. We will still have to raise some taxes or cut some benefits or a little of both. And these statements are true whether Social Security invests in the stock market or individuals do it. Thats because almost 90% of current Social Security taxes are used to pay benefits. The rest of our Social Security contributions create the $63 billion of what some people refer to as the US annual budget surplus. Some people have suggested using this so-called US Budget surplus to cut taxes. However, all of that money is actually from Social Security, so if we invest it in the private sector, then there wont be any money for the tax cut.
So, lets discuss these 2 ways to approach privatization in more detail. Under the first one, Social Security would hire investment managers to invest a potion of its money in the private sector. This option could solve about 40% of Social Security's financial problems. Opponents argue that with their assets reaching maybe 5% of the total market, Social Security investment decisions and stock voting could become politicized. They also worry that a large Trust Fund might tempt Congress to improve benefits too easily. Supporters note that the government already invests in the stock market (albeit on a smaller scale) without these problems and using stock market indexes could avoid the concern that Social Security would manipulate the market. Proxy voting could be delegated to the money managers. In addition, they note that Social Security could get a higher rate of investment return than if individuals did the investing, the administrative and investment expenses would be less, and there would be less risk on individuals. With respect to the concern that Congress might use the money, supporters note that Congress would be less likely to use the money than now.
Alternatively, to avoid the governance concern, individuals could do the investing. Supporters want Congress to require that individuals invest their payroll taxes directly in the private sector ourselves. And then we would reap the better returns ourselves. Workers could have their own individual accounts and control all their own investment decisions. It could be on top of Social Security, as a part of Social Security, or instead of Social Security. However, opponents note that this could have much larger administrative and investment costs than if just Social Security did the investing. In addition, there would be very large transition costs to change over to a totally privatized system. Out of our current Social Security taxes, we would still have to pay the current benefits to our parents, which leaves only 1% of payroll for our own individual accounts. The change over to the new system could be paid for by increased taxes or contributions on top of what we already pay to Social Security. However, most recent legislative proposals havent increased taxes. Instead, they take their mandatory contributions out of what we are already paying into Social Security. If that happens, then less money goes to Social Security and we will have to find more places to cut benefits. For example, if we were to divert 2% of our Social Security taxes to our Individual Accounts, then we would have to come up with a lot more benefit cuts (about twice as much as before) from our prior lists of options. However, those benefit cuts could be offset by the increased benefits from the Individually Invested Accounts, if the stock market does well. However, the Individual Accounts would probably not be enough to help the next generation of retirees fully offset the benefit cuts, because contributions are needed immediately and the advantages of investing in the private sector dont build up right away. Opponents also worry that this option would throw too much risk onto individuals, such as the investment risk, inflation risk, longevity risk (i.e., outliving your money), and leakage risk (i.e., the risk that we would take our money out before retirement). Proponents however, note that they are not suggesting full privatization, only partial privatization. Opponents are still concerned, because the rate of return arguments can be used until Social Security is fully privatized.
Investment risk: For example, if my generation retires when the stock market is down (which could be a likely possibility for the boomer generation), we may not have as much money as we expected and then have to suffer decreased benefits or work longer. The benefit decreases could be much larger than those the notch babies experienced in the 1980's.17 In addition, some people might invest too conservatively and others might invest too aggressively and fail18. Social Security can spread that problem out so that people are affected less. There is also what we call the longevity risk. You may take out too much money too fast, or live a lot longer than you expected and not have any more money. There is also Inflation Risk.. Inflation could go through the roof after you retire. There are some partial solutions for these risks, but they are not complete. For investment risk, some remedies would be investment education, indexes, and restricted investment options, but your expenses would probably still be higher than if one entity did the investing. For longevity and inflation risk, you could be required to buy an inflation-indexed annuity that would pay you for the rest of your life. Of course, that would make Social Security more expensive for women (unless unisex annuities were required). There will be other concerns for women too. Unless dependent spouses have some ownership of the Individual Account, they may not get survivor benefits. This is very important, since the poverty rates for elderly women are among the highest.
Another risk is that people might take their money out before they reach retirement (for eminently good reasons such as health, unemployment, or a childs education). This is sometimes called the leakage risk. The remedy for that is to restrict withdrawals (unless you have enough to buy a lifetime annuity equal to the poverty level). However, I wonder whether Congress can keep these restrictions intact. They havent been able to do it for IRAs and 401(k)s. This risk is similar to the one mentioned above that Congress might use the money for other purposes if the fund got too big. The parallel here is that if Individuals have the money, they may use it for purposes other than retirement. There is also the risk that people would spend it too fast in retirement. Annuities would also remedy this risk, but would Congress mandate annuities for people who dont expect to live long?
There are also the death and disability risks. If you are young, and become disabled, or if you die when young and have survivors, your account will not be enough to provide much of an income. One solution here is to require that people buy disability insurance and life insurance. Alternatively, some proposals only privatize the old age benefit, but dont touch the survivor and disability programs of Social Security. If that happens, then people would actually get the former Social Security disability benefit plus a benefit from the privatized account, which could be too much. They would probably need to reduce the Social Security disability benefit at older ages, because otherwise it could exceed their wage income. However, if the reduction depends on the amount of benefit provided by the privatized account, then those who invested well would be penalized. Thus, another solution would be for individuals to decide how much disability insurance they needed, subject to a mandatory minimum.
Administrative Expenses and Feasibility: Social Security is run very efficiently. Their expenses are less than 1% of benefits. If Individual Accounts are mandated, there will be many implementation issues to work out. It might be difficult for the government to enforce individual contributions to Individual Accounts, so employers may be called upon to collect and forward the money. It might also be difficult for employers to direct funds to any fund, so investment options may initially be restricted. Initially, mutual funds may not be efficient in handling small accounts, so the government may provide a clearinghouse for the deposits. The government could be just a default investment fund until individuals get enough funds to invest in the private sector. Employers may have to identify contributions immediately when sent to the government, which will be more work than now Currently, employers only provide this information on W2s in January of the following year, and 5 million (of the 6 million employers) are on paper. A huge staff would be needed to handle allocation and annuitization questions, but this could be automated. Currently the federal Thrift Savings Plan has less than 2 million participants, Fidelity (the largest Defined Contribution administrator) manages 16 million accounts, and the whole DC industry manages 44 million. They would have to manage 160 million accounts right away. If employers are used, they would also have to help with many workers who they can currently exclude from their pension plans today (e.g., the 17 million part-time workers, 9 million young workers under age 21, and the 27 million new workers who may be temporary), and these people would have small accounts.
So there are many different risks under this individual account idea. There are also remedies, such as requiring inflation-indexed annuities. We need to discuss these remedies to see if they are adequate and whether they are worth using to reduce the risks on the individual. Opponents would suggest that we just stay within the structure of the current system, which avoids all these problems and guarantees everyone their inflation-indexed Social Security pension for life, irregardless of the investment rates in a particular year.
Finally another big concern is that if Congress increases the contribution to Social Security or mandates an additional contribution to an Individual Account, employees may take it from their contributions to their 401(k) plan, and thus lose their match. If lower-paid employees contribute less to their 401(k) plans, then the higher-paid employees will have to reduce their contributions too, because of federal pension law. If the mandate is for additional employer contributions, then employers may reduce contributions to their pension plans.
Papers from the World Bank laud the fact that retirement income in the U.S. comes from more than one source (i.e., the 3-legged retirement stool - four if you count possible earnings)19. Diversification of the sources of retirement income is very important. For example, if the stock market crashes when the baby boom retires, Social Security and traditional DB employer pension plans may be more valuable to employees than Individual Accounts. When the stock market does very well, retirees will benefit greatly from DC plans, personal savings, and individual accounts. When low-income individuals have small savings and pensions, Social Security's subsidized benefit is more helpful. High-income individuals though cant maintain their standard of living on Social Security alone. They need the other legs to do that. If the tax base in the U.S. decreases, retirees can depend more on their advance-funded employer pension plans and personal savings. Thus, there is a very good reason to keep all 3 legs of the retirement stool strong. Congress could save the Social Security leg by harming the other legs and end up with a one-legged stool. We must be careful not to hurt any leg and lose this valuable diversification. Some proposals could hurt the employer leg. In addition, we need to be concerned if all 3 legs put too much risk on the individual (i.e., if Social Security moves toward Individual Accounts, employer plans continue to drop traditional DB plans in favor of 401(k) plans, and the individual leg already has 100% of the risk on the individual).
Thus, a third way to privatize could involve employers as a part of the solution, and it
could solve the governance problem (if the government had the money), the leakage problem
(if either government or individuals had the money), and greatly reduce the risk problems (if
individuals had the money). An employer with a good retirement plan could have the option
to satisfy the individual account mandate for their employees. In fact, it might be easier for
6 million employers with pension plans to meet this challenge, than for 200 million
individuals to each set up their own account. Employers that already have a pension plan
might not have to do anything. Employers without pension plans would not have to start
one (but their employees would have to set up an individual account). For example, if
an employer has an excellent pension or 401(k) plan, then instead of gutting it,
maybe it could satisfy the individual account requirement for its employees.
Defined Benefit pension plans could satisfy this requirement20. Not much has been developed
in this area, and I would be glad to discuss this further.
So as you can see, there are many different ways to solve Social Security's current financial problems. In spite of its current difficulties, Social Security has been a great success. The solutions are there. To implement them will take political courage and the determination of the American people. But it can be done, for our elderly, not only for our parents, but also for us and our children. Often when I give these speeches at town hall meetings around the country, retirees ask if their benefits will be cut? The members of Congress invariably and emphatically state that the answer is no. (Maybe their benefits wont go up quite as fast or they will be taxed a little more, but the benefit wont decrease.) Generation Xers also state their fears that the whole problem will be dumped on them. However, if we solve it this coming year, Generation X will not be stuck with the whole problem. Maybe that is the way democracies like America solves its problems - everyone is a part of the solution. Forums such as this one help us to think more clearly about these tough issues, and Im glad that Americans Discuss Social Security are holding more of them all across the country. Id also like to thank the President and Vice President, and our other distinguished speakers for their leadership on this issue and for joining us today. The actuarial profession looks forward to working with you to find solutions that ensure retirement security for all Americans. I will be happy to answer your questions later. Thank you.