All Things in Moderation

A few years ago I wrote a post asking Is Social Security Promising Too Much?, in which I attempted to demonstrate that the average retiree in 2011 could expect to receive a lot more in Social Security benefits than they ever paid in as FICA taxes. This was a retrospective look, so I used annual historical income levels, tax rates and treasury yields going back 53 years in order to calculate how much would have been paid in, how much compounded interest might have been earned, and how much benefit would be paid out based on those contributions. But because current SS tax rates have changed so much since the inception of the program, it recently occurred to me that it might make more sense to look at the question on a prospective rather than a retrospective basis, in order to see if changes to the program have made it any more viable. That is, rather than looking at a current retiree, I decided to look at a newcomer to the labor market, and calculate for various income levels how much the government would collect from them (at current rates) in FICA taxes over a lifetime of working, and compare it to what the SS benefits calculator says they can expect to receive in benefits when they retire, to see if the deal is a net gain or loss to the government.

The calculator actually makes this very easy, since it reports benefits in constant, 2016 dollars, and it assumes a constant income, also in 2016 dollars, from today until retirement day. So we can easily eliminate the complicating effects of inflation and assume a lifetime of constant dollars and no inflation. The last complicating factor is the issue of discounting future cash flows. $100 today is, generally speaking, worth more than $100 at some point in the future, so one would usually use a current interest rate curve to discount all future cash flows back to today in order to get a net present value (NPV) of all the cash flows. While not strictly correct, for the sake of simplicity I used a single discount factor for all cash flows, but I did look at each scenario under different interest rate assumptions.

So, I looked at a single, 22 year old in 2016, just entering the workforce, and assumed a constant income level for his entire working life, until he was 65. Then I ran the numbers on all income levels, in $10k increments, between $10K and $120k (since both taxes and benefits are capped at $118k of income), assuming current tax rates and benefits levels remain unchanged, and assuming 17 years of total benefit payouts (since the expected lifespan of a current 22 year old is roughly 82 years old). Then I discounted all future tax and benefit payments under various interest rate assumptions, to see if the entirety of the transaction from start to finish is a net positive or negative from the government’s perspective.

Below is a table of various income levels, along with both the annual tax receipts and annual benefit payouts associated with those income levels.

Income         Annual Tax @12.4%            Annual payout at age 65
10,000                     1,240                                           8,988
20,000                    2,480                                          12,348
30,000                    3,720                                          15,552
40,000                    4,960                                          18,756
50,000                    6,200                                          21,948
60,000                    7,440                                          25,152
70,000                    8,680                                          26,976
80,000                    9,920                                          28,476
90,000                   11,160                                          29,976
100,000                 12,400                                          31,476
110,000                  13,640                                          32,976
120,000                  14,880                                         34,248

And here is a table of the net present value (NPV) of all tax receipts and benefits paid for each income level at various discount rate assumptions.

Income      1%               2%           3%
10,000    (46,871)    (18,535)     (2,867)
20,000    (37,277)      (3,084)     14,820
30,000    (26,117)     13,309       33,076
40,000    (14,958)     29,702       51,333
50,000      (3,679)      46,167      69,633
60,000       7,480       62,560       87,889
70,000     32,485       87,286     111,185
80,000     60,740      113,968    135,665
90,000     88,995      140,651    160,144
100,000  117,250     167,333    184,623
110,000  145,505     194,016    209,103
120,000  176,048     222,075    234,415

So at first glance this doesn’t look too bad. With interest rates even at just 1%, only those people in the lower half of the income scale would be expected to produce a net lifetime deficit for the government, while those in the upper half would produce a lifetime surplus. And with median income in the US at $52k we can assume a relatively normal distribution both above and below that income level, which means that, again at 1% interest rate levels, the surplus created by the upper income levels should be enough to cover (and even surpass) the deficit created by those at the lower income levels. Yes, this shows that SS is still essentially a wealth transfer program, not only from the young to the old, but also from the well paid to the less well paid. But at least it appears to be financially sound, and a higher interest rate environment only helps matters. With rates at 3% all but the very lowest income levels represent a positive NPV to the government, meaning that most people will be net contributors to, not takers from, the SS pool of benefits.

There is, however, one problem with this analysis. The only reason we discount future cash flows assuming an interest rate is that, generally speaking, a dollar today is worth more than a dollar a year from now because you can take a dollar today and invest it, producing more than a dollar next year. So, for example, at 3% interest rates, I can turn $1 today into $1.03 next year, meaning that $1 next year, discounted at 3%, is worth only (approximately) 97 cents today. An outflow of 97 cents today added to an inflow of $1 next year would produce a NPV of zero.

That being the case the problem with my above analysis is that the government doesn’t invest its revenues. The government simply spends the money that it gets.  From the government’s perspective, a dollar today is worth the same as a dollar one year from now, because the government can’t take $1 today and turn into anything more than $1 next year.  One might argue that the Social Security Trust Fund does invest its excess funds in government securities, but that is an accounting fiction. The “investment” is not made into any wealth generating venture. The only place to do that is in the private market. There is no growth, no new wealth creation, and as a result this “investment” produces zero real growth. So the proper discount factor to use when analyzing SS cash flows from the government’s perspective is, in fact, zero.

That puts the analysis in an entirely new light. At zero interest rates, which again is the real return produced by SS revenues, SS is a huge loser to the government at almost every income level.

Income          0% DF
10,000          (96,572)
20,000          (99,358)
30,000          (99,545)
40,000          (99,732)
50,000          (99,719)
60,000          (99,905)
70,000          (77,096)
80,000          (48,887)
90,000          (20,678)
100,000            7,531
110,000           35,740
120,000           67,748

Only the very top income earners ($100k+) produce a positive NPV for the government, and everyone at every income level up to $60k (ie more than half of the population) produces a negative NPV of nearly $100k per person. Social Security is a horrible deal for the government, and specifically for future taxpayers as they are the ones who will have to cover the ever widening hole that the SS model produces. If I did deals like this for my company, I would be fired in fairly quick order.

That being said, if SS is a bad deal for the government, it must be a good deal for individuals who are on the other side, right? After all, if the government is losing money on the deal, individuals must be making money.  However, unlike the government, individuals could invest their contributions elsewhere if they weren’t forced to pay them into the SS trust fund, so it makes sense from their perspective to discount future cash flows at current interest rates rather than zero. It is true that, a very low interest rate environment, ie one close to zero, it actually is a pretty good deal for medium to low income earners. The NPV of the expected benefits vs expected contributions for incomes of $60k or less is nearly $100k for them. However, as rates rise, the deal quickly sours. With rates of 1%, nearly half of all income earners are under water on their SS deal, and by the time rates hit 2% only the lowest of income earners, those at $20k or less, have a positive NPV. By the time rates top 3%, pretty much everyone is getting less than they could otherwise have gotten.

So it turns out that SS is not only a bad economic deal for the government (ie future taxpayers), it’s even a bad economic deal for most individual recipients in most interest rate environments. So the question lingers: If SS is not viable program from the government’s perspective, and is a bad economic deal from the individual’s perspective, why in the world do we perpetuate this program?

Filed under: social security | 14 Comments »