Blame the Recession
Every percentage point of movement counts to the budget.

Mr. Bartlett is senior fellow at the National Center for Policy Analysis
January 14, 2002, 8:00 a.m.

The first thing to know is that all budgets necessarily must incorporate a lot of assumptions about future economic activity -- gross domestic product growth, inflation, interest rates, and the like. In the near term, economists can try to forecast these variables, but even highly paid private-sector economists are often wrong just a few months ahead. Since the federal budget goes out 10 years, obviously a great deal of guesswork is involved.

The second thing to understand is that the U.S. economy and the federal budget are so large that very tiny changes in assumptions can create multi-billion dollar errors. GDP is about $10 trillion, for example, which means that every tenth of one percent of growth is worth $10 billion. Since the federal government takes about 20% of GDP, 0.1% higher growth adds $2 billion to the Treasury.

Thirdly, this impact rises over time. Last year's budget estimated GDP at $17.5 trillion in 2011. This means that an error of 0.1% 10 years from now could cause federal revenues to be higher or lower than estimated by $3.5 billion that year.

Finally, it is important to remember that when forecast errors are made over a long period, they compound very rapidly. The result is that estimates of cumulative surpluses or deficits can rise and fall by huge amounts very quickly from small changes in economic assumptions. Every budget has a table that allows some of these magnitudes to be estimated. The one from last year's budget shows the following:

— If real GDP growth was 1% less than expected in 2001, federal revenues would be reduced by $9.6 billion and government spending would be $2.1 billion higher for things such as unemployment compensation. Thus the surplus would be lower than estimated by $11.7 billion.

— The effects of lower-than-expected growth in just a single year have permanent effects on the budget. One percent lower growth in 2001 would cause revenues to be $20.9 billion less and spending to be $7.3 billion higher in 2002. And this will be the case even if growth this year is exactly what was predicted.

— By the year 2011, revenues will be $35.9 billion lower and spending will be $29.7 billion higher because growth 10 years earlier — in 2001 — was just 1% lower than estimated. Over the entire period, the cumulative budget surplus will be reduced by $475 billion.

— These figures are symmetrical, meaning that if growth is 2% less than expected, they are doubled. Conversely, if growth is higher than expected, future surpluses will be higher to the same degree.

— With growth being about 2.5% less last year than originally estimated, it means that revenues are going to be $52 billion lower and spending is going to be $18 billion higher this year. And that will be true even if we get the 3.3% growth originally expected. Over a 10-year period, the recession will lower combined surpluses by at least $1.2 trillion.

— If growth is continuously overestimated over a longer period, the budgetary impact becomes staggering. If real GDP growth comes in 1% less than expected for 10 straight years, the cumulative reduction in surpluses rises to $2.6 trillion.

The most important variable in estimating long-term growth is productivity — output per man-hour. Productivity growth plus population growth will approximately equal real GDP growth. From the mid-1980s until the mid-1990s, productivity only grew about 1% per year, and economists thought this was about as good as we could expect in the future, as well. But since 1995, productivity has spurted to about 2.5% per year, and many economists now think that we may be able to do this well in coming years, too.

Thus, the long-term real GDP growth rate has essentially risen by 1.5% per year. Projecting and compounding this increase out into the budget estimates explains almost all of the budgetary improvement that we have observed lately.

People should keep these orders of magnitude in mind when they hear people like Daschle say that lower surpluses are due entirely to last year's tax cut. Even if the recession is already over, its continuing effects on the budget will lower future surpluses by more than the tax cut. Should the recession linger, its effects will be even larger.