Ch In other countries the replacement rate is often higher. In the Netherlands it is over 90%. This means that after taxes Dutch workers receive

Transcription

1 Ch About Social Security o Social Security is formally called the Federal Old-Age, Survivors, Disability Insurance Trust Fund (OASDI). o It was created as part of the New Deal and was designed in large part to take care of the elderly who had lost their savings in the stock market collapse and subsequent bank runs. o It covers retirees, surviving dependents and also provides disability insurance. o It pays over $500 billion per year in benefits and is the largest government program in the world. o Financing Originally SS was financed by a 2% tax on the first $3,000 of a person s earnings, half of which was paid by the employer. Thus the maximum tax was $60 per year. The 1936 booklet Social Security in Your Old Age outlined how tax rates would increase over time (show slide 1): Finally, beginning in 1949 you and your employer will each pay three cents on each dollar you earn, up to $3,000 a year. That is the most you will ever pay. Currently the Federal Insurance Contributions Act (FICA), the tax which finances SS, is 12.4% on the first $97,500. Again half is paid by the employer. The cutoff is adjusted annually for inflation and other factors. o Eligibility Originally only about half of the U.S. workforce was eligible for SS. Those exempt included farm workers, self-employed workers and anyone who worked for a company that employed less than ten people. Currently there are a few exceptions, but almost everyone participates in SS. In order to receive benefits a person must have paid SS taxes for at least ten years and must be at least 62. o Benefits SS is essentially an annuity. This means that the benefit is paid regularly until the recipient dies. The amount of the payment is based on the Average Indexed Monthly Earning (AIME), the inflation adjusted average of the 35 highest earning years of the person s life. If you don t have 35 years of earnings the missing years are counted as zeros. Then the benefit received, the Primary Insurance Amount (PIA), is calculated as a portion of the AIME (show slide 2). One way to measure the benefit received is the average replacement rate. This is the percentage of pre-retirement income that SS replaces. In the U.S. the replacement rate is about 40% o This varies from about 20% for high-income workers to about 60% for lower income workers.

2 Ch In other countries the replacement rate is often higher. In the Netherlands it is over 90%. This means that after taxes Dutch workers receive a higher income for retiring than for working. Age requirements A worker becomes eligible for full benefits at age 65, the Full Benefit Age (FBA). A worker can collect benefits beginning at age 62, the Early Entitlement Age (EEA); however the amount of the annuity is reduced by 6.67% for each year prior to 65 that he collects benefits. A worker can delay receiving benefits and receive an increase of 6% (grows to 8% in 2008) for each year past age 65. If you work while receiving benefits, your benefits are cut by 50 cents for each dollar you earn over $12,480. This money is returned later with interest. Spouses can also receive benefits based on the workers AIME (even if divorced), as can dependent children of deceased workers. Social Security and redistribution. o One of the initial goals of SS was to help the elderly who had been devastated by the Great Depression. As a result of the depression, about half of elderly American were living below the poverty level. o In order to accomplish this, SS was created as a pay-as-you-go system. This means that the payments of workers contemporaneously funded the benefits of retirees. This contrasts with private pension funds that hold payments in investment vehicles to pay benefits later. Beginning in 1983, however, a trust fund was created to hold SS surpluses. Now FICA payments fund current retirees and also go into a trust fund to pay for future obligations. However, there are problems with this trust fund that we will get into next time. o In order to see the implications of the pay-as-you-go system, let s look at an example from the book. The setup: Suppose we have a world where people live for two periods. In the first period they work; in the second they are retired and depend on SS for income. We will assume that productivity (and thus wages) rises 5% annually, as does the population of young people. In the first period we start with 100 young people who each earn $20,000. Our SS program is financed by a 10% tax on young people that is paid directly to the retirees. The result (show slide 3) In period 1 there are no retirees and no SS. In period 2 there are the 100 original workers who are now retirees. There are 105 young people. Discuss.

3 Ch Discuss periods 3-5. The take away: o The first group of retirees gets a large benefit with no costs. o The benefit received by the middle group depends on the rate of wage growth and the rate of population growth. o The last group, if there is one, pays the cost but receives no benefit. As a result we essentially have a legacy debt carried over from the original generation. To illustrate this point the book talks about Ida May Fuller, the first recipient of Social Security. o She paid $24.75 in Social Security taxes over three years. o She began collecting SS in 1940 and died in 1975 at the age of 100 (show slide). o Over her lifetime she collected $22,889. o By paying her this money we incurred a debt, a legacy debt, that is still with us today. o Redistribution in practice: In order to calculate the actual level of redistribution, Gruber uses a measure called Social Security Wealth (SSW). SSW is the difference between the PDV of a person s expected benefits and the PDV of their expected payments. This measure values benefits received at 65 more than those at 90 because of the discount rate and because you are more likely to receive benefits at 65 than you are at 90 Show slide 4. Why does this happen? The earliest beneficiaries did not pay into SS for all of their lives. The initial tax rate was 2%, now it is 12.4%. The result of this increase is that people who paid in earlier did not have to pay as high of a tax rate, but still reaped the benefits of a higher tax rate. Wage and population growth have slowed over the past 30 years. o Since 1973 wage growth has been about.9% and population growth 1% annually. o Population growth in the U.S. is expected to continue for the foreseeable future, largely as a result of international immigration and the large families that immigrants often have. o Many other countries, however, are in worse shape than we are. For example France has a similar wage growth rate but has a fertility rate significantly below the replacement rate (1.84/2.1) o Europe as a whole is expected to see its population shrink by nine percent by 2050, with Eastern Europe leading the way at a decline of 22%.

4 Ch Another way that SS redistributes is from rich to poor. Note how this has changed over time. It also redistributes from singles to married couples and from dual earning married couples to single income married couples. Women also tend to have higher SS Wealth because they tend to live longer The justification for SS. o Recall the reason we want insurance is to smooth consumption. The question is, does the government have a role to play in providing this consumption smoothing. o Recall that we decided that the value of social insurance declines with the predictability of the event to be insured. Retirement is very predictable. Even if you don t want to worry about saving and investing, there are privately available annuities that will guarantee a particular income for the rest of your life. Then why does the government get involved? The economist s answer is failure in the annuity market. Information asymmetry means that the purchaser of an annuity knows more about his life expectancy than does the seller. As a result, the people who are most likely to buy an annuity are the ones who are most likely to live a long time. Just like in insurance markets, this leads sellers to raise prices, which leads the healthiest people still involved to drop out, which leads to an increase in prices, ect. As a result, when you compare the expected returns to an annuity to the expected returns of investing your money elsewhere, annuities tend to be a pretty bad deal. However, they can be attractive for people who are particularly risk averse. Therefore there is the potential for social security to improve welfare by correcting the failure of the annuity market. o Paternalism is the classic politician s justification of SS: people are concerned that without the government s intervention, people will not save enough for their own retirement. At the moment, people don t save enough for retirement. According to John Kerry, without SS, 52% of elderly American s would be living in poverty. In 1991, the median American approaching retirement age had $16,000 in private pension wealth and $3,000 in other personal retirement assets. o The important question, however, is does social security work? Regardless of the justification, does it smooth consumption or does it simply crowd out other consumption smoothing methods? This is a hard question to address empirically.

5 Ch We can look at the time-series data (show slides 4 & 5). However, poverty may have been dropping anyway. There are not enough changes in SS to see a real effect in the time series data. The problem with using cross-sectional data is that we cannot get an appropriate control group. o Suppose we want to study whether higher expected SS benefits lead to a reduction in savings. o Our treatment group could be one of the groups who will receive higher benefits, e.g. married couples or higher earners. o In order get unbiased results we must find a comparable control group. However, the only available control group would be single people or low-earners. o In this case we might think that the treatment and control groups might have different savings behaviors independent of SS benefits. o This will bias our results. Ideally we could use a natural experiment. o In this case we would like to compare a state which changed its SS benefits to a similar one that did not. o This is impossible, however, since SS is a national program and does not vary across states. In order to address this question, researchers have turned to other countries which have programs similar to SS. o In Italy, for example, recent reforms reduced significantly the expected benefits to young public sector workers while reducing private sector workers expected benefits by much less. o To the extent that private and public sector workers are similar, using DID techniques to study this natural experiment should produce unbiased estimates of the effect of SS on savings. o They find that there is a 30-40% crowd out of private savings. o Thus, in Italy there is some crowding out but it is not complete. o It is hard to say how well international results reflect behavior in the U.S., but we really have little other reliable information. Another way to address the question of whether SS is smoothing consumption is to use what is effectively a regression discontinuity design. We would expect, if individuals are optimally preparing for retirement, that there would be no sudden change in consumption at retirement (draw picture). However, evidence suggests that consumption drops more than 30% when people retire.

6 Ch The implication is that even with the help of SS individuals are not fully smoothing their consumption (as we said is optimal). o Thus, the evidence suggests that SS does provide some consumption smoothing, but that SS also crowds out private savings. Crowding out is one example of SS s unintended effects. Another unintended effect is the way SS changes people s retirement decisions. o A worker can begin collecting benefits at age 62. If he chooses to wait there are four ways that his consumption is affected. He pays an extra year of payroll taxes. He does not receive SS benefits for that year. His SS annuity payment is not reduced by as much because he waited one year closer to age 65. It is likely the case that this year of earnings is higher than his 35 th lowest. Thus, since SS is based on your 35 highest earning years, he gets to replace a low earning year with a high earning one. o The first two ways lead to a decrease in consumption, the second two lead to an increase in future consumption. If the first two dominate then the worker has an incentive to retire at age 62. If the second two dominate then he has an incentive to work an additional year. By comparing the positive and negative effects of working an extra year you can compute an implicit tax on working an additional year. If the tax in positive, then workers would prefer to retire. If it is negative, they should keep working. If it is zero, then they should be indifferent. Gruber and Wise (1999) compute this tax and find that for the U.S., it is zero from age 62 to 65. Thus in the U.S., people should be indifferent between retiring at 62 and retiring at 65 o However, people don t seem to be indifferent. There is some time-series evidence, but it is only suggestive and not conclusive (show slide 7). The 60 s and 70 s were a time of SS growth and declining labor force participation. The more convincing evidence is to look at what is called a hazard rate (show slide 8). The hazard rate is the probability that a particular event occurs in a particular period. o We use hazard rates to study everything from firm bankruptcy to social security. o Here the hazard rate describes the probability that the average person will retire at a given age. o Note that it is less than 10% up until age 62 where it jumps to about 25%. It then falls back down before spiking again at age 65. o This suggests that SS is influencing retirement decisions, but it is not conclusive.

7 Ch o It could be that people believe 65 is the normal retirement age or that employers tend to set 62 as their preferred retirement age. Prior to 1963 men were not allowed to begin collecting benefits at age 62 they had to take them at 65. o Thus we can look at the hazard rate before and after this policy shift to see if there was a change (show slide 9). o This suggests much more convincingly that SS is affecting people s retirement decisions. International evidence from France and Germany suggests that similar behavior occurs there. o In France, where you can retire at age 60 and collect full benefits, the hazard rate at age 60 is over 60%. o In Germany they lowered the entitlement age by 5 years in By 1980 the average retirement age had fallen by almost 5 years (show slide). Taken together this evidence suggests that SS may lead to workers retiring earlier than the otherwise would have. o This is inefficient because we lose the productivity associated with older workers. o SS may encourage people to quit working (show slide) There is however, at least one alternative explanation. o If workers plan privately for retirement then they can choose the age at which they want to retire (e.g. 70, 60, 52). o However, if SS crowds out private savings and workers depend on it for retirement they can no longer retire prior to 62. o As a result, the spikes at 62 and 65 could be driven by people who would prefer to retire earlier but cannot because of the rules of social security. o From a social welfare perspective this is not efficient because they are not able to quit working when the MC exceeds their MB. However, it is probably much more efficient then retiring too early. Social Security Reform o From its inception through the mid 1970s SS generally balanced its budget or ran a slight surplus (show slide 11). This was a result, in part, to constantly increasing taxes. o By the late 1970 s and early 1980 s it was evident that there was a problem. By 1983 the SS trust fund was projected to run out of money by the end of the year. o As a result, the Greenspan commission was created to find a way to fix SS. The commission recommended that SS move away from a pay-as-you go system and accumulate savings in the SS trust fund. They also recommended speeding up previously scheduled tax increases cutting benefits (though benefits were never cut), and expanding the maximum taxable income.

8 Ch As a result, the SS program began to run a surplus and accumulate money in the trust fund. The trust fund money is invested in government bonds. It seems like the trust fund should provide a cushion when revenues drop below expenditures. However, by investing money in government bonds, the SS has essentially lent its money to the government. One way to think of this is that the government has been writing IOU s to itself. When SS goes to spend the money in its trust fund it will have to cash in the government bonds. This means that the government will have to cut spending elsewhere. The result is that SS does not have a pile of money sitting around somewhere. They have promises that the government will repay the money from somewhere else. The implication is that for SS to use its trust fund (or even break even rather than collect a surplus), the rest of the budget will have to cut spending, increase taxes, or borrow money from somewhere else. o The 1983 reforms seem to have put SS on more stable footing, but only for now. Due to longer life expectancies and falling fertility rates the percentage of our population that is retired is projected to grow dramatically (show slide 12) o Another way to look at this is that in 1950 there we almost 8 people of working age to each elderly person while in 2050 that number is projected to be around 2.9 people of working age to support each of the elderly. o The result is the $13.5 trillion of unfunded obligations. Obviously reform is necessary. Options include: Raising taxes. According to the SS office a tax increase of 2.0 percentage points will solve the problem for 75 years and an increase of 3.7 percentage points will solve it forever. This seems like an attractive option; however, given the history of tax increases so far, I m not sure I completely trust this estimate (show slide 13). Expanding the tax base. We could do this by raising the maximum taxable income. o Making it easier for young people to immigrate could expand the tax base. o We could increase the maximum income that is subject to the tax. o However, this would have the effect of increasing the redistribution aspects of the system (show slide 14). Raising the retirement age. In 1950, the average person who turned 65 could expect to live, on average, 13.9 more years. In 2004 that number was 18.4.

9 Ch Despite this, the retirement age has been moved down (to 62 for early retirement) and the full benefits age has only recently moved up. Raising the FBA would also cut benefits to those who take early retirement since their benefits are a function of how early they retire relative to the FBA. Cutting benefits. There are various ways to do this, but all of them are political suicide. These are all ways to reform the current system. There are many who claim the existing system will not work and should be replaced. o Privatization Various privatization plans have been proposed; most of these work by allowing individuals to invest some portion of their payroll taxes in private accounts. This essentially would make social security like a private pension fund. The advantages are: Since much of the money would be invested in private accounts it could not be spent on other things by the government. Private accounts generally provide a higher rate of return than government bonds. It would allow people to choose investment vehicles that match their own risk-tolerance. This would also increase the capital available and lead to long-run productivity growth. The biggest drawback is that someone will lose out. Remember in our example the last group did not receive any benefits after paying taxes. Switching from an unfunded to a funded system means someone will lose. We still have the original legacy debt. Privatization opponents have suggested to seniors that privatization means that they will lose their benefits and be out on the streets. Thus privatization has become very difficult politically. Other potential drawbacks to privatization include: Higher administrative costs. o The cost to run the current system is very low. o Once we allow people to choose their own investments, we incur a lot of additional administrative costs. If people cannot be trusted to save for themselves, can we trust them to invest their own money wisely? In conclusion, SS has be come a very important part of our government and culture in the United States. o However, the system is clearly unsustainable as it is now. o There are various reforms that have been proposed, but each has its own problems.