If you’ve been following
The Picket Line for a while,
you’ll know that from time to time I check in to see how the
is doing with its database modernization effort.
naturally, was an early-adopter of database technology. But unfortunately for
them, as a result their data is now very much enmeshed in a web of
IRS-specific, archaic programs and databases that are written in the Computer Science equivalent of “dead languages.”
So they decided to do a complete upgrade and bring things up to more modern
1997, they budgeted $3 billion for
a 15-year project (after an earlier, $3.3 billion effort had failed). The
budget quickly jumped to $8 billion, but even so,
by the time the first billion was spent,
five years in, the project was already 40% over-budget for what little it had
managed to accomplish and some milestones were more than two years overdue.
As late as
announced that it hoped to launch the Customer Account Data Engine
( CADE) portion of the project that year, but
the launch date kept being put off: to 2003, then
the agency dropped the contractor they’d
been working with and decided to start over. Only portions of
CADE have ever gone live (it currently
processes the simplest 30% or so of personal income tax returns). And the
agency’s inability to stick to its software modernization schedule
is costing the government serious
They tried throwing more money at the problem, they tried
radically scaling back the project
requirements… nothing worked. By 2007, they
were still overbudget and missing deadlines, and the following year auditors
noticed that the project, to the extent it worked at all,
had serious security vulnerabilities.
Now, according to
Government Computer News, they’re throwing in the towel.
They’ve halted work on the project (“pending a strategy review”) “because of concerns over increasing complexities in system development.”