The folly of long-term projections

Ezra Klein makes a great point about long-term budget projections in a column today titled "Don't Worry About Deficit That Will Heal Itself":

[T]here’s something ridiculous about extrapolating current trends all the way
out to 2080. By that point, we’ll probably either be robots, the servants of
robots or a bit of both. Either way, the health-care system will probably
undergo dramatic change.

Look what happens when you turn back the clock 70 years from today. That
puts you in 1942, the year John Bumstead and Orvan Hess first saved a patient’s
life using penicillin. There were no pacemakers, oral contraceptives or
chemotherapy. Water wasn’t fluoridated, and health insurance was a niche
product. Imagine trying to predict the trajectory of today’s health-care system
from that vantage point. How incredibly, hilariously wrong would we have
been?

In part for that reason, we don’t balance the budget for 70 years at a
time. Indeed, we usually don’t even balance it for 10 years at a time. Instead,
we muddle through, striking deals that are smaller than wonks like, but
sufficient to keep us out of the woods. That’s what we did in the 1990s, which
featured deficit-reduction bills in 1991, 1993, 1995 and 1997. We’ll probably
follow a similar path in the decade to come.