How much is it going to cost to keep Social Security solvent over the next 75 years? If you listen to the opponents of Social Security reform, not to mention most press accounts, you'll hear the figure of $3.7 trillion.
A New York Times editorial from January 3 stated: "Over a 75-year time frame, Social Security's shortfall is estimated by the Congressional Budget Office at $2 trillion and by the Social Security trustees at $3.7 trillion, a manageable sliver of the economy in each case." In a recent paper for the Center on Budget and Policy Priorities, Jason Furman, William G. Gale, and Peter R. Orszag write, "Over the next 75 years, the deficit in Social Security is 0.7 percent of GDP (or $3.7 trillion) according to the Social Security actuaries."
Yet that $3.7 trillion figure is misleading for two reasons. First, it is a "rolling total" that increases with every passing year, all else being equal. Part of the increase comes from slight changes in various demographic variables. But a good part of the increase comes from the fact that each new report from the Social Security Trustees analyzes a slightly different 75-year period by dropping the current year and adding another year 75 years hence. In so doing, it eliminates a "good" year when Social Security is in surplus and adds a "bad" year when it is in deficit. So, page 10 of the 2003 Report states that Social Security's "unfunded obligation over the 75-year projection period has increased from $3.3 trillion to $3.5 trillion." Page 2 of the 2004 Report notes that the unfunded obligation "over the 75-year period is $3.7 trillion in present value, $0.2 trillion more than the obligation estimated a year ago." If the debate should continue into next year, the reform opponents will probably be saying that Social Security has a shortfall of "only" $3.9 trillion.
The second reason $3.7 trillion is misleading is that it leaves the public with the impression that the federal government will only have to come up with $3.7 trillion over the next 75 years to keep the system solvent. Yet as the quote from the 2004 Report notes, that $3.7 trillion is the "present value." The 2004 Report on page 200 defines present value as
the equivalent value, at the present time, of a future stream of payments (either income or cost). The present value of a future stream of payments may be thought of as the lump-sum amount that, if invested today, together with interest earnings would be just enough to meet each of the payments as they fell due.
Thus, $3.7 trillion is not what is needed over the next 75 years. It is what the federal government would have to invest right now to meet the Social Security's 75-year shortfall.
Since the federal government isn't going to come up with an extra $3.7 trillion to invest tomorrow, a more useful figure in understanding the shortfall is the "cumulative cost." That is the cost to the taxpayers, over a 75-year period, of paying off all the treasury bonds in the Social Security trust fund plus meeting all the obligations of Social Security once the trust fund runs dry. Using data from the 2004 Trustees' report, that number is (sit down) $24.9 trillion in inflation-adjusted dollars! That is an annual average of $332 billion extra that the taxpayers will have to pony up for Social Security. Even that doesn't tell the full story because included are the years 2004-2018 when Social Security is still running a surplus. After 2018, the annual Social Security deficit grows quickly: By 2022 taxpayers will have to pay almost $100 billion above what they will shell out in payroll taxes to meet all of Social Security's obligations. That will reach over $200 billion in 2027, over $300 billion in 2033, and $400 billion in 2046. If you are under 40 years of age, chances are quite good you will live to see all of those dates.
Yet the confusion does not end there. Opponents also compare apples to oranges by contrasting that $3.7 trillion present-value figure with the cumulative cost of reforming Social Security. For example, in that same New York Times editorial, the editorialists write that to pay for President Bush's reform plan "the government would borrow an estimated $2 trillion over the next 10 years or so and even more thereafter."
An accurate comparison would examine both the present value of the Social Security shortfall with that of the present-value cost of a reform plan, such as the Reform Model 2 constructed by the President's Commission on Strengthening Social Security. According to page 128 of the Commission's report, "the total transition investments under Reform Model 2 would be approximately $900 billion in present-value terms." Last time, I checked $900 billion is less than $3.7 trillion. It is also possible to compare cumulative costs. Using data from the Trustees, the cumulative costs of Reform Plan 2 are a mere $225 billion. Any fifth-grader -- even those in public schools -- should be able to figure out that $225 billion is less than $24.9 trillion.
Unfortunately, the reform opponents' inaccuracies are beginning to take hold. Witness this paragraph from a recent article from the Associated Press:
The retirement system faces a projected $3.7 trillion, 75-year shortfall. Bush wants to overhaul the program to let younger workers divert some of their Social Security payroll taxes to personal accounts. But that alone won't fix the problem and could require upfront costs of $1 trillion to $2 trillion over 10 years.
With press like that, reform opponents are going to have little incentive to clear up the confusion.
Author's note: Anyone wanting to know how I calculated the cumulative costs, email me at firstname.lastname@example.org.