In order to get full benefit from this post you would need to open one or more of the following links in separate windows or tabs. So for your convenience I will put them up both as a hot link and as the straight URL. All Tables and Text taken from the 2011 Social Security Trustees Report.
VI.A2 Operations of the OASI Trust Fund, Calendar Years 1937-2010
http://www.ssa.gov/...
VI.A3 Operations of the DI Trust Fund, Calendar Years 1957-2010
http://www.ssa.gov/...
VI.A4 Operations of the Combined OASI and DI Trust Funds, Calendar Years 1957-2010
http://www.ssa.gov/...
While it is usual to treat the Social Security Trust Funds together as seen in Table VI.A4, they are legally distinct and have two different start points, with the Disability Insurance program and Trust Fund added in 1956. Lots of numbers and lots of lessons to be drawn. Starting with some under the fold.
A couple of preliminary notes. The Title is not a typo, the actual Old Age/Survivors Trust Fund was established via the Social Security Amendments of 1939, previous operations being handled out of special accounts at Treasury. Also all dollar figures are in then current dollars, that is not adjusted for inflation. And are expressed in terms of billions to a single decimal point, which is to say to the $100 million mark, years in which income or cost was less than $100 million are marked with footnote 'c' whereas years where there were no such income or cost are marked with a '-'.
The Tables are crowded but not too confusing once you get used to them. We have columns devoted to total income and total cost each followed by a breakdown by category. After that we have three columns devoted to assets, the first being 'Net increase during the year' which is to say the surplus or deficit used in overall budget deficit calculations, the second being total cash balance in the fund(s), and the third expressing that balance as a percentage of the NEXT years cost, this being called the 'Trust Fund Ratio'. Social Security is considered to be in 'actuarial balance' for any given year if the TF Ratio is at 100 or more which is to say at least one year of prospective cost on hand at the beginning of that next year.
Okay lets start with Table VI.A2. We can see right away the Social Security started collecting money three years in advance of what would be the first actual payout of monthly benefits, that is the earliest recipients were not total freeloaders, there was some initial cushion. But if we continue on for the ten years after 1940 and compare total income from column one to increase in assets in the third column from the right we see a curious thing-more than half of all income was retained leading to a Trust Fund ratio in 1944 of 2025 or 20X that of 1945 cost. And as late as 1950 we still see a TF Ratio of 1156 or 10X that of 1950 cost. And in 1956, the 20th year represented in the Table we still see a TF ratio of 371. Which is to say right at the same level as the 'unprecedented' levels we see today. Well they are only unprecedented if you ignore actual precedents in the historical record.
So Social Security was not always Pay-Go from Day One and in fact the whole generation of workers from 1937 to 1956 were themselves 'double-taxed' to pay for current beneficiaries and to build up balances for their own retirement, suddenly the whole idea that early generations were waging war on unborn Gen-Xers starts losing some steam, on balance the Greatest Generation pulled their own freight in peacetime just as they did during the war.
1956 was also the year that Congress added the Disability Insurance component to Social Security and so gave us the combined Social Security system we know today. And though OAS and DI continued to be tracked separately as seen in Tables A2 and A3, the more standard way of representing Social Security solvency was on a combined OASDI basis as seen in Table VI.A4. But one last glance at A2.
In the years before 1950 interest on the Trust Fund had negligible effects on solvency as increases in assets and so TF Ratios were almost entirely due to excess FICA collections while interest contributed in the range of 10-20% to the surplus. But interest became more important over the first half of the 50s and in 1956 made up more than half of the surplus that year. Suddenly surpluses, when they occured, could no longer be counted on to come from contributions, instead they might simply come in the form of interest. Which is crucial because interest wasn't and isn't financed.
Which is where things get interesting. Interest payments to the Trust Fund are simply credited in the form of newly issued Special Treasuries which are not financed by borrowing or any other extraction from the real economy, they are in the word dreaded by all Paulites simply fiat money, straight out obligations on the Federal Government backed only (if that is the word) by Full Faith and Credit. By that same token they are not available for borrowing, there is literally nothing to borrow. What this means is that the Trust Funds can and often have been cash flow negative even as they were running surpluses. And BTW scored as such for the overall surplus/deficit number reported each year.
Meaning that starting with 1957 we have to look at the Table through two or three sets of spectacles. No longer is Total Income - Total Cost a close approximation of Cash Flow and so Surplus, while it still measures Surplus, Cash Flow has to be scored as Income excluding Interest - Cost. So if we switch to Table A4 and the new combined OASDI score we see a surplus and so an increase in Trust Fund balance of $500 million for 1957. But to measure actual cash flow we have to subtract out the $600 million in interest earned but not financed out of the real economy. And what do you know the net is a negative $100 million cash flow. Which in passing puts paid to the common claim that 2010 was the first year EVAH that cash costs exceeded cash collections. Well yeah, if you exclude 1957 and a bunch of years after that.
In looking at the ten years from 1957 to 1966 things look a little different depending on your metric. If we look at TF Ratio in the last column we see a steady shrinkage from above 300 to 96. Which is to say just under the 100 level that meets the standard for actuarial balance, but all things considered not a terrible thing and indeed the TF Ratio returns to levels just over 100 in the years 67-70 which by some measures means a perfectly balanced system. If we look at actual Trust Fund balance we see a $20.6 billion balance in 1966 that is exactly the same level as it was in 1954. But where that represented a full five years of reserve in the earlier year (TF Ratio 500), it was just under a year of 1967 costs (TF Ratio of 97). And if we look at the eight years immediately after 1966 we see a constantly increasing nominal balance even as TF ratios first flatten and then decline. And it is that latter metric that is the official gauge of solvency.
From that latter standpoint 1971 is the actual point of crisis as the TF Ratio dipped below 100 and never recovered and in fact never had a year over year gain. But before we turn our attention to what can fairly be termed the crisis years lets go back and look at the 1957 to 1970 period through different sets of glasses. First as to cash flow. As noted above any year in which interest exceeds increase in assets the system as a whole is cash flow negative. Now this is obviously the case in any year the Trust Fund is in actual deficit, i.e. has negative numbers for 'increase in assets' but is also true when interest comprises all or more than all the surplus. So on close inspection the Social Security Trust Funds were cash flow negative every year but one from 1957 to 1965 even though they remained in official actuarial balance. (In a later post I will show why that is actually the natural state of affairs for a perfectly balanced system).
Something to note here. Cash flow negative does not equal crisis, no one was boo hoo hooing in the 60s about Social Security solvency, that after all is what the strong balances built up from 1937 to 1956 were designed to address, the very function of the Trust Fund is as a reserve fund, a bridge fund, it was not and is not an investment fund, instead it is a mechanism for smoothing out variation in an otherwise pay-go system. On the flip side cash flow positive doesn't mean no crisis, not if the TF Ratio and so your margin is shrinking. And of course the worst of both worlds is where both numbers are moving in the wrong direction. Which is where we were starting in 1971 even as actual deficits didn't hit until 1975.
Summing up we can divide the pre-1983 Trust Fund Operations into four time periods. First we have 1937 to 1950, a period where all the numbers are good and Trust Fund balances and ratios are being beefed up even as benefits lagged. Then we have the period from 1951 to 1961, more or less the golden years as the Amendments of 1950 unfroze benefits and for the first time allowed checks to be adjusted for inflation which while boosting costs still left a Trust Fund with a comfortable Ratio. On the other hand negative annual balances starting in 1957 were a warning sign that Social Security wasn't exactly free, instead keeping it in balance meant honoring a certain amount of the interest accruing with actual cash transfers. Still 1961 to 1970 was a time when the system first attained a fully mature status of steady Trust Fund reserves at or slightly above target. Then we have 1971 to 1982, the true years of crisis.
So the history of the Operations of the Trust Fund prior to the Greenspan Commission of 1982-83 does not well match the simple narrative of balanced Pay-Go system from Day One that some would have it. Instead there were periods not unlike the late 80s when Trust Fund balances were being beefed via actual cash surpluses up in anticipation of future demand, then another when on a cash flow basis the system was more or less neutral, then another when it was technically stable yet cash flow negative, and finally a period of authentic crisis. But in order to see all this you have to stick your nose in the numbers, which I guess if you are still with me means you.