Understanding Social Security – Part 2

(Click here for Part 1, which sets up this post.)

Take another look at this graph:

It sure would be nice if we were on Path I, wouldn’t it? How do we know what path we’re on, or how could we make an intelligent guess? Where do these lines come from anyhow?

There is a lot of uncertainty about what the future may hold. So three different sets of assumptions are used, with different estimates of productivity growth, birth rate, immigration, etc. All three estimates are fairly conservative, which is good. We’d rather be surprised with something good than something bad, and we shouldn’t blithely assume that the good times today will continue. It’s worth examining some of these assumptions, since they affect so many political and economic decisions about our country’s future. If it turns out all three projections are overly-optimistic, then we’re all in big trouble. If it turns out that we’re on path I, then there is truly no Social Security issue — we might even be able to reduce payroll taxes a bit.

To see the main assumptions that underly these three curves, scroll to the table at the end of this page. It’s well-worth clicking there to get a feel of the ranges we’re talking about.

According to David Langer, model I has historically proven to be the most accurate. (I can’t find the original research, but this interview speaks to it.) He has also found that the numbers that underly those projections are made by political appointees, not independent actuaries and demographers.

All the numerical assumptions are important, but the most important is probably the productivity growth rate of the US economy. The numbers are 1.4% (pessimistic projection), 1.7% (“medium” projection, that’s the basis of the 2042 date), and 2.0% (optimistic projection).

(The productivity number, among other things, helps to address that the ratio of workers to beneficiaries is shrinking. We’re at 3.3:1 right now, and that number is declining. This sounds very worrisome, but the magic of productivity is that each worker produces more value as time goes on. If you truly feel this is an issue, you have to wonder how it is that we went from a ratio of over 100:1 to today’s number without the world ending. In fact, if you look at the historical trends, we’ve been below 4:1 for over 30 years now.)

Admittedly, I am way out of my depth to even try and forecast the future. But I will anyhow. I am encouraged by this quote:

“The federal government’s own Bureau of Labor Statistics estimates that productivity rates in the nonfarm sector improved at a 2.3% average pace from 1947 through 2003. Adjusting for the gap of 0.2 percentage points between the productivity growth of the nonfarm business sector and the economy as a whole still leaves productivity across the economy growing by a healthy 2.1% over the postwar period. That historical record convinces economist Dean Baker, from the Washington-based Center for Economic and Policy Research, that a productivity growth rate of 2.0% a year is a “very reasonable” assumption.”

(By the way, Dean Bakers blog is always interesting. Check it out.)

Below are the annual productivity growth numbers for the USA, since 1948, courtesy of the Bureau of Labor Statistics.

Year Annual Productivity Growth Rate
1948 2.8
1949 3.3
1950 6.7
1951 2.7
1952 1.8
1953 2.3
1954 1.9
1955 4.2
1956 -0.8
1957 2.6
1958 2.2
1959 3.8
1960 1.2
1961 3.1
1962 4.5
1963 3.5
1964 3
1965 3.1
1966 3.6
1967 1.7
1968 3.4
1969 0.1
1970 1.5
1971 4
1972 3.3
1973 3.1
1974 -1.5
1975 2.7
1976 3.3
1977 1.6
1978 1.3
1979 -0.3
1980 -0.2
1981 1.4
1982 -1.1
1983 4.5
1984 2
1985 1.6
1986 3.1
1987 0.5
1988 1.7
1989 0.7
1990 1.9
1991 1.6
1992 4.1
1993 0.4
1994 1.1
1995 0.5
1996 2.7
1997 1.6
1998 2.8
1999 2.9
2000 2.8
2001 2.5
2002 4.1
2003 3.7
2004 3
2005 2.3
2006 2.1

This averages 2.27%. The SSA’s optimistic prediction is 2.0%. The historical average is even more optimistic than the optimistic number used by the SSA! I’m not an actuary, there are undoubtedly other factors at work here, they might turn out to be much more right than I. But all else being equal, you would think the historical average of the last 58 years might be a starting point. Or at least not considered “out of bounds” when it comes to projections.

I don’t know about you, but that makes me feel like there’s a good chance we’re on Path I.