Monday, March 07, 2005

How to Create A Crisis Out of Thin Air: Cooking the Social Security Numbers

Now let’s see. If I wanted to create a scenario where the long term collapse of Social Security would happen, how would I have to construct my economic models to predict such a collapse?

Well since good economic growth will keep the system solvent for the next century or so, I’d have to erase that average annual 3.3% real GDP growth the USA worked hard to earn over the last 40 years. What would do it? So, we’ll just slice GDP growth by 12% now and see how that looks on the spreadsheet. When that didn’t do the job, we’ll just keep on slashing it until over time annual GDP growth is just about half as much as it is today.

But how do we justify that? Well we’ll just slice 10% off the nation’s fertility rate, and then slash the productivity growth rate we enjoy today by over 50% over time. Now despite the fact that the Center for Immigration Studies estimates that 1.3 million immigrants settle in the country each year, let’s just magically reduce that by 30% over time as well. Let’s see what else...?

Oh yes! We’ll just hike up assumed inflation by 17%, from (2.3% today to 2.8% over time). Will people notice that inflation will hardly increase if our economic growth rate is heading the other way? Who cares? Nobody reads these reports beyond the executive summary anyway.

For good measure, we can assume that future unemployment rates will be 10% above the current 10 year average rate of 5.05%. We’ll be safe with THAT assumption, since the Bush administration is doing such a lovely job making that come true.

Well, let’s press F9 and recalculate this spreadsheet and see what we’ve got. Woo hoo! We did it! We now have projected that Social Security will have to pay out more benefits than it has in it's bank account or takes in with taxes 37 years from now! What’s that you say? That’s happened several times before (14 times before, according to the SSA) and the sky didn’t fall? OK, so we present it to the public as 2042 is the year when Social Security will "be flat broke"!

Now what do we do with the other projection we started from? You know, the one with the economic assumptions pretty much in line with what reality has been showing for the last few decades or so? The one that shows that Social Security never runs out of money over the entire time we study, even if we do absolutely nothing to change the system? (See Option I in the Trustee’s Graph below.)

Figure IV.B3.--Long-Range OASI and DI Trust Fund Ratios
[Assets as a percentage of annual expenditures]
from The 2004 OASDI Trustees Report



I know! We’ll call the projection with the realistic assumptions the optimistic "low cost" projection. Then we’ll call our doctored pessimistic version the "intermediate cost" version. We’ll go back to the drawing board and slash immigration rates another 18%, hike up unemployment rates another 18%, drop our fertility rate another 13%, hike up inflation yet another 35%, and lower GDP another big chunk by imagining two recessions and a longer term reduction in GDP growth by another 25%.

Never mind that sustained slow growth and two recessions with simultaneous sustained increased inflation is virtually impossible. What’s important is the first arbitrarily pessimistic scenario will look moderate by comparison. That "intermediate" projection will be the one pundits will use to point to the troubles of Social Security. Even progressives will buy into the idea that we’ll have to do SOMETHING to ‘fix’ Social Security. And that something will happen now, when the right wingers still have control over every branch of government and thus ensure that whatever change happens will be to the right wingers’ liking.

This scenario is not a conjecture. This is exactly what’s happened. The numbers used above come directly from the 2004 Social Security Trustees Report everyone is misquoting and their little noticed chapters describing how they constructed their scenarios. My argument is not necessarily with the policy wonks producing the Trustees’ Report. Sound financial managers can and should examine worst case scenarios in projecting the long term health of funds they are stewards for. My argument is with the politicization of the reports, morphing worst case scenarios into inevitable crises. No one does that unless (1) they are too incompetent to know what they are reading, or (2) they wish to deliberately lie to the public. Either is an impeachable offense in my opinion. Don’t let them destroy Social Security.