by Bruce Webb
Barkley Rosser and I among others have long claimed that Social Security’s economic models are too pessimistic in light of performance both over the whole post-war period and particularly over the last dozen years. And the numbers looking back are pretty clear, Social Security income/cost ratios have come in better than Intermediate Cost projections and were right up to around August of 2008 (when things went into decline).
These claims have drawn some pushback from the ‘Crisis’ folk including Andrew Biggs. In his most recent post he calls on Doug Elmendorf of CBO to give the explanation. CBO Explains Why Future GDP Growth Will be Slower than in the Past. But first Andrew explains why GDP in and of itself is not the key metric.
This difference comes down to the fact that the Trustees don’t think about “GDP growth” as a single thing, but break it down into its components and project how those components will change over time. In fact, despite Langer’s claim that “The Gross Domestic Product (GDP) is the key economic assumption in estimating costs,” it actually plays no direct part in estimating Social Security’s finances: Social Security doesn’t collect taxes based on GDP, nor does it pay benefits based on GDP.
To some degree this is a straw man argument, critics of the current Intermediate Cost economic model use Real GDP because it is a widely reported and pretty well understood metric that within limits reflects the factors which really determine Social Security solvency, which is to say more people working at higher wages with lower inflation. In any event whatever the source of increased GDP the larger it is the easier it is to supply a basket of goods to a more or less fixed cohort of future retirees, (being that almost everyone who will be collecting Social Security over the 75 year actuarial window being already alive). So granting the point that GDP is not a perfect proxy for Social Security solvency why is it going to be smaller going forward? Elmendorf explains (and Biggs follows up)
Projected growth from 2015 to 2019 is also below historical average growth rates, a difference that is more than accounted for by slower growth in the labor force because of the retirement of the baby boom generation. Over the postwar period, the labor force grew at an average annual rate of 1.6 percent; by contrast, we project it to grow only 0.4 percent per year in the period from 2015 through 2019. As a result, potential GDP grew 3.4 percent per year on average in the postwar period, but CBO expects that it will grow by only 2.4 percent annually (allowing for a tad more productivity growth) in the 2015-2019 period.
In other words, the economy will grow more slowly in the future because the labor force will grow more slowly in the future. No conspiracy needed.
Excuse me for leaving my tin foil hat on. My response to Biggs is under the fold.
A slower growing labor force in the future implies more demand for the labor hours that remain which if the rules of supply and demand work in the way that most economists insist they do should translate into real wage gains (which are directly captured by FICA taxes). Instead we are told that real wage increases will decline permanently to an annual 1.0 to 1.1% rate.
Table V.B1.—Principal Economic Assumptions
Similarly if the economy is sputtering going forward due to a slowing in labor force growth (and so a slip in covered worker ratio) you would think you could counteract that via an increase in legal immigration. Instead we are told that by 2010 legal immigration will stabilize at a rate 21% below that of 2006 and remain at 750,000/year forever.
Table V.A1.—Principal Demographic Assumptions, Calendar Years 1940-2085
Which means that as the population increases from 308 million in 2007 to 481 million in 2085 the proportion of foreign born Americans is projected to steadily drop. This flies in the face of American historical experience, when we needed farmers to open up the frontier we got them from the Scots-Irish and the Germans (that’s when my ancestors checked into America). When we needed laborers to build the railroads we got them from Ireland and China. If we have a growing yet ageing population and a dire need for more night nurses and geriatric care doctors we have a ready supply going forward in Mexico and the Philippines.
Perhaps someone can provide a reasoned explanation why in the face of an economy handicapped by a stagnant labor supply our policy response will be to clamp down on both legal and illegal immigration in the way implied by the data tables. But until that contradiction gets satisfactorily explained I will be keeping that hat firmly down around my ears.
Because it all still looks like a model chasing a desired conclusion. Particularly when Low Cost itself sees immigration settling out an an absolute rate lower than the 2006 peak. This really doesn’t make any real world sense at all.
I want to highlight that last point. Per the 2008 Report legal immigration for 2006 was estimated to be 950,000 with illegal immigration adding another 380,000 for a total of 1,330,000 adding to a total population equalling 306 million. Under Intermediate Cost assumptions those numbers are projected to settle out at 750,000 and 380,000 by 2020 (1,130,000 out of 345 million people) and to an ultimate 750,000 and 275,000 or 1,025,000 new immigrants out of a 2085 population estimated at 481 million. Now these numbers do imply that we ultimately get reasonable control of the borders, which is not a bad thing in and of itself. But it also implies a specific policy decision to limit legal immigration 17% lower in absolute numbers than the 2006 peak. While this maybe would come as good news to the no-growth zealots currently controlling the Sierra Club, it would be an odd response to an systematic economic slowdown caused by low labor supply growth.
And this remains true when we turn to Low Cost. It also suggests we will get better control over the borders with illegal immigration dropping through the 75 year window, although slower than IC. But it also implicitly argues that we will by 2010 set a hard, permanent cap on legal immigration at 960,000. Which date is exactly when Boomer retirement picks up and so the need for cheap labor to care for us in our old age. The notion that we will be beating our heads against the wall wondering where we will get landscapers and nursing home aides come 2020 is pretty odd. Because unless Mexico and the Philippines drop off the face of the planet this shouldn’t really be a problem.
The conspiracy I see is not in Intermediate Cost projections, that could be explained away as the natural caution of actuaries. But in data category after data category it is Low Cost that seems artificially capped both on the economic and demographic fronts. The argument boils down to saying that the near future (say the next twenty years) cannot even on your most optimistic assumptions be as good as the recent past. We probably will not hit every component of Low Cost, that is a function of a model that assumes every factor favorable to solvency tracks together (Low Cost) or alternatively that every factor moves negatively (High Cost). (We indeed would be served better with an expanded matrix of alternatives, something suggested by the SSAB). That is no reason to simply accept that demography has us in a helpless trap that dooms us to no better than 1.1% Real Wage, 1.7% Productivity, and 2.1% Real GDP.
Come on we can do better than that! We are not passive victims here, we can if necessary craft policy in a way that targets economic growth. That is if we want Low Cost outcomes we can implement Low Cost policy, and if that requires relighting the torch on the Statute of Liberty sobeit.