John Irons's Blog


Economic News, Data and Analysis

A Tale of Two Budget Forecasts

The Congressional Budget Office (CBO) and the Whitehouse’s Office of Management and Budget (OMB) have each recently come out with their projections about the budget. While the OMB projects a $1 billion non-Social Security surplus this year (and next!), the CBO has forecasted that the surplus will be less than the Social Security surplus.

Even though the debate has hinged critically on whether or not we will be using the Social Security surplus, this is essentially a completely arbitrary yardstick to use.

The projections

Of central importance to budget forecasting are the assumptions that are made about economic growth. The graph below shows the projected surplus (or deficits) from each budget as well as the economic projections. The graphs show the non-Social Security surplus as well as the total surplus.

Where has the surplus gone? Partly to the slowdown in the rate of economics growth, and partly due to the tax cut. For details see the changes in the baseline projections.

The OMB and the CBO projections are remarkably close – the big fight though will happen because of the sign of the non-Social Security surplus in the CBO projections – because of the deficit, the Social Security surplus will have to be used.

What’s so special about using the Social Security surplus as a benchmark? Not much.

The Social Security surplus as tipping point

So what is this benchmark anyway? Well, the way Social Security works is that most of the money that is paid into the system (through payroll taxes) is immediately paid out in the form of benefits to current retirees. Since there is slightly more money put in this year than will be paid out, this creates a surplus.

There is however something very arbitrary about the size of the surplus. It is simply a result of past decisions to start building a trust fund that could then be used to pay benefits in the future. The size of the trust fund, however, is not large enough to alone pay for benefits in the future – this is why Social Security is projected to go “bankrupt” in about 30 years or so.

If we did the accounting right – and took into account the real value of the benefits that are promised to retirees in the future – the real Social Security surplus that we should be running is actually much bigger. That is, if we were to set aside enough money today to finance future retirees, we would have to run much bigger surpluses than we are currently doing.

The debate, then, over whether or not we are eating into the surplus is essentially foray into the land of the arbitrary. If we are really concerned with the long-term stability of Social Security, we should be aiming at a much higher number than the current Social Security surplus.

It’s the politics stupid

The whole reason we are hearing so much about the Social Security surplus is simply the result of political promises over the past year. Just about everyone in sight made the promise to put the Social Security surplus into that “lockbox.”

To the extent that we will go beyond that threshold, those in charge have broken a promise. The budgetary gimmicks of the OMB (changing accounting practices, counting on companies paying their taxes early, revising economic growth higher for next year, as well as 2003-2005, etc., all of which yield a $1 billion surplus this year and, improbably, next year are well) show just how important the perception of breaking a promise about Social Security can be.

For more on the numbers see the CBO or the

Filed under: Economics

Federal Spending: Some graphs

A colleague of mine recently asked me to make a guess as to how much federal government spending went to national defense over the last 50 years. My guess was that it is currently about 15% and was once probably as high as 50%.
As the graph below shows, I was right on the low end, but significantly off on the top end. In 1952, defense spending reached a peak of 84% (!) of government outlays.
I thought it would be interesting to take a look at historical patterns of government spending, deficits, as well as the composition of current outlays. I think the data speaks for itself, so here are the graphs. (If you would like more details, the raw data can be found in the Budget of the U.S. Government .)
Defense as a Percent of Federal Outlays (1946-2000)
Components of Federal Outlays (1990-2000 average)
Components of Federal Outlays (1962-2006 est.)
Federal Receipts, Outlays, and Surplus (1930-2000)
Real Federal Debt held by the Public (1939-2000
Federal Receipts, Outlays, as a Percent of GDP (1960-2000)
Real Federal Debt held by the Public, as a Percent of GDP (1947-2000)

(Note: F. Westhoff generously provided the data for the first and the last 4 graphs on the list above.)

Filed under: Economics