Predictions Are Hard: CBO and the Art of Forecasting the Future of America

Derek Thompson
February 5, 2013

The Congressional Budget Office has a glamorous job (at least among Washington wonks), which is also an exceedingly difficult job (for anybody).

Every year, it's required to project the next decade in federal budgets. This year, that means it must construct a tableau of the American economy in 2023 -- the year when Justin Bieber turns 29, George Clooney qualifies for early Social Security benefits, and a child born in the W. Bush administration can legally drink. To determine tax revenue, the CBO must project such mysterious figures as the share of national income that will go to labor and share of employers who will drop their health plans. To determine spending levels, it must project everything from the future of consumer price inflation to the rate of unemployment and retirement.

These reports (their charts, in particular) are often presented and quoted with sort of Delphic reverence. But to me, the most interesting part of their projections are the moments where the CBO says whoops! -- where the office revises past predictions, not because Congress passed a new law, but because the country changed in a way they didn't expect.

So follow me as I scroll, scroll, scroll down to page 56 in the belly of Appendix A, and you'll find this paragraph (underlined sentences are my emphasis) ...

Medicaid and Medicare. " In recent years, health care spending has grown much more slowly both nationally and for federal programs than historical rates would have indicated. (For example, in 2012, federal spending for Medicare and Medicaid was about 5 percent below the amount that CBO had projected in March 2010.) In response to that slowdown, over the past several years, CBO has made a series of downward technical adjustments to projections of spending for Medicaid and Medicare. From the March 2010 baseline to the current baseline, such technical revisions have lowered estimates of federal spending for the two programs in 2020 by about $200 billion--by $126 billion for Medicare and by $78 billion for Medicaid, or by roughly 15 percent for each program."

In other words: Whoops!

To CBO's great surprise, the growth of health care spending has slowed. Maybe it's the recession. Maybe it's something more. Either way, it's big. Cutting $200 billion from Medicare and Medicaid is "about double the revenue the government would generate by raising the Medicare eligibility age from 65 to 67," as Sarah Kliff points out. Medical inflation, heal thyself.

The least responsible of the deficit hawks are occasionally fond of pointing out that America's "real debt" isn't our $14 trillion in actual debt, but our $87 trillion in "unfunded liabilities" to Social Security, Medicare, and federal pensions. It's true that if you add up all projected spending and projected revenue for these programs over the next 75 years, the gap is ungodly. But, as I've said, this shortfall is exquisitely sensitive to just about every demographic trend you can imagine, plus many trends in the future of economics and health care that we cannot imagine. The fact that the government's top number-crunchers are obliged to report figures from the future doesn't make them oracles. It just makes them dutiful forecasters, using present-day trends to make useful, but ultimately impossible, predictions about the future of the country.

The further out you see budget projections, the more you should question their veracity. Instead, repeat to yourself seven simple words. Everything can change, and it usually does.

More From The Atlantic