February 2, 1997

Many Unhappy Returns at I.R.S.


[F] or once, we can do something about government waste
and I.R.S. ineptitude, quickly and cheaply.

The Internal Revenue Service disclosed its latest
incompetency on Thursday when it conceded that the $4
billion it had spent on developing modern computer
systems was a waste. Part of that money, $284 million,
was given to Lockheed Martin, the defense contractor, to
turn paper tax returns into electronic images. The
agency now says it has to scrap this effort, at what an
I.R.S. official described as an "astronomical cost."

Instead, the agency now desperately proposes to contract
out the processing to private employees, who will get to
see the tax returns of their fellow Americans in order
to enter the data manually.

What a gargantuan foul-up! At the core of the debacle
were philosophical errors about the right way to use
computers. First, the I.R.S. must have believed the hype
from some people in the computer-science establishment
who hope to make computers think like people. The truth
is that while computers are very good at doing certain
menial tasks, they are still not able to perform many
tasks that humans find simple, like reading numbers on a
paper tax form.

Second, the I.R.S. ignored the inexpensive,
decentralized Internet in favor of an outrageously
expensive, centralized custom solution. Large numbers of
Americans are already sending around digital information
on the Internet every day. All kinds of commercial
firms, from banks to bookstores, are routinely
conducting transactions over the Internet, which is
faster, more secure and costs less than using the post

By contrast, all that the I.R.S. has offered taxpayers
is a little-publicized service requiring electronic
returns to be filed through approved intermediaries, who
charge as much as $25. As a result of this bizarre
policy, the I.R.S. is still receiving 90 percent of its
tax returns on paper forms.

If the I.R.S. could shed the twin myths that
centralization and artificial intelligence are
achievable or even desirable, it could benefit from the
Internet, just like everyone else.

Jaron Lanier is a visiting scholar at Columbia
University's computer science department.

Copyright 1997 The New York Times Company

Posted here by permission of the New York Times.

Go back to Jaron's home page.