In Defense of Social Security Privatization by M. Jed

This one is by reader M. Jed. You might remember that last month, M. Jed had a post on how demographics affect tax collections. Here’s M. Jed:

Let me state up front that I am a proponent of Social Security privatization. My reasons vary, and in researching this, some were validated and others called into question but they included – (1) higher expected rates of return, (2) increased equity ownership and portfolio diversification for those at the lower end of wealth spectrum, and (3) a personal bias towards increased individual responsibility versus social responsibility. Solvency is not and has never been one of my reasons. Frequent commenter Bruce Webb, persistently asks privatization proponents to backup their ideology with numbers and what follows is my attempt.

For those, including myself, that recognize Social Security as a combination of retirement savings and disability insurance, fortunately the SS Administration breaks out the components of the contribution rates. The Old Age and Survivors Insurance (OASI) (original design of the program) component of the contribution rate is 5.3% for each employee and employer. Using the Consumer Population Survey’s Income Statistics, I recreated a life-cycle household earnings chronology, for the median household, which based on my estimates starts at $30,000 at age 22 (median for 15-24 year olds is $28,770) and peaks at $64,000 at age 45, before starting to decline again. The real IRR of SS for the post baby-boom generation is estimated at about 1.8%. With the contribution rate, estimated household incomes, and the IRR, it’s possible to estimate the projected real account value at retirement (I used age 65). For this hypothetical median family, they contribute $247,000 over the course of their career(s) and have a real account value at retirement of $370,000. As I understand the cited source, the 1.8% real IRR only incorporates the OASI component of SS so it is apples-to-apples for the 10.6% employer/employee combined contribution.

Now, most people (even Bruce) acknowledge if one were to have instead invested prior contributions and received the historical equity rates of return, the Social Security account values would be substantially higher. What Bruce doggedly questions is why people assume that historical equity returns are likely to persist in an environment where the required assumption (i.e., that SS faces problems) is that productivity growth will be lower than history. I make no such assumption, however, in researching this post I found that the primary reason OASI has a lower IRR than government bonds is because of the net wealth transfer that benefited initial recipients in a pay-as-you-go pension scheme, since they made no contributions. This wealth transfer needs to be taken into account when projecting rates of return and is estimated as a reduction of 25% of contributions. Note that grossing up the 1.8% IRR for this 25% reduction, is 2.4%, or about the historical real return on government bonds, and slightly better than current long-dated TIPS yields.

So let’s move to equities. There exist >multiple sources that note that historical rates of return on equities are not the correct benchmark, but rather they seem to center on an expected forward real U.S. equity return of 4.5%. I’ve been unable to find projections for foreign equities, but I recognize arguments for either greater or lesser expected returns relative to U.S. markets and also the political firestorm that recommending investing Social Security contributions in non-U.S. based companies would create and thereby default to the 4.5% for the entire equities component. Using a real rate of equity returns of 4.5%, an assumption of an administration fee on equities of 0.25%, and a 25% reduction in the rate to pay for historical wealth transfer produces an effective expected total return on equities of 3.2%. If one were to allocate 50/50, the effective expected SS rate would be 2.5% compared to today’s 1.8%. Over the course of a median career, this results in an expected portfolio at retirement that is 22% higher than the status quo and if one were to allocate 75% to equities it would be 33% higher.

The post-retirement OASI annuity component must also be taken into account, but I’d note that there currently exist a plethora of private products with income-for-life for both the purchaser and spouse that have both market exposure and guaranteed returns. Analyzing these would necessitate a much longer post.

In conclusion, (1) private accounts would produce higher expected at-retirement account values, though nowhere near as high as most proponents suggest and (2) the “free lunch” comes from increased diversification for those who have traditionally not invested in equities, which I personally assume is because most of their savings has gone to Social Security. Finally, by far the best and most balanced summary of SS issues I found was here which addresses the top 10 myths of SS reform perpetuated by both the left and the right, many of which have been discussed at AngryBear.