The Wall Street Journal
January 18, 1995, p. A14.
Get Smart -- Eliminate the CIA
By Angelo Codevilla
Over the past several years, U.S. intelligence agencies in
general and the CIA in particular have proved themselves
incompetent in peacetime and of little use in conflict.
Stripped of their mystique and lacking the capacity to
reform themselves, these organizations are virtually in
receivership.
The maladies ailing the intelligence community are
numerous. Independent quality control was never more than
a pretense, and competition among intelligence agencies was
nonexistent. Producers of intelligence -- rather than the
soldiers and diplomats who have to use it -- have also
become its judges. All this has spawned a complex of
habits, procedures, mentalities and people too entrenched
to be repaired and too noxious for any part to form the
nucleus of a new, healthy system. Hence, we should take
Sen. Daniel Patrick Moynihan's advice, and rethink our
intelligence from the ground up.
A good place to start is with the fact that about half of
the $28 billion U.S. intelligence budget pays for units
directly controlled by military commanders, which routinely
provide precise information for the armed forces'
operations close to the front lines. The Treasury and State
Departments also have their own intelligence units, which
fit their needs quite well. So why do we need a national
system headed by the CIA?
The original justification for the creation of the CIA in
1947 was that intelligence would be best if its gathering
and evaluation were divorced as much as possible from the
operating departments of government -- State, Defense, etc.
-- and placed under the president. This judgment has turned
out to be wrong. Because presidents have relied on the CIA
to run the system, the result has been a system dominated
by the priorities of the producers -- not the users -- of
intelligence.
A basic failing is that the CIA has primary responsibility
for intelligence and none at all for real world events. The
CIA prefers to place its career employees in U.S.
embassies, where they pretend to be employees of other
parts of the government. Such "case officers" must
acknowledge that they are gathering information for the
U.S. Another disadvantage is that they don't speak foreign
languages well.
And unlike successful reporters, they virtually never know
the substantive fields about which they are seeking
information. Thus it is unsurprising that they are usually
outdone in economic reporting by economic reporters, in
military reporting by military reporters, and so forth.
The Aldrich Ames case shows how much more highly the CIA
values the smooth functioning of its system than what the
system produces. Mr. Ames handed the KGB the capacity to
shape the intelligence flowing to top U.S. officials during
the endgame of the Cold War. Thus disinformation made
presidents and secretaries of state more vulnerable to
Gorbomania than the average citizen informed by newspaper
accounts.
How could the CIA fail to notice the fishiness of reports
generated by a network controlled by the other side? The
same way that, in the 1970s and 1980s, the agency had
failed to notice that it was passing along reports from a
network of agents in Cuba totally controlled by Castro's
DGI, and from a network in East Germany all but a few of
whose agents were working for the Stasi. In other words,
while the Ames case was unusually destructive, it was a
typical example of bureaucratic sclerosis.
In the Gulf War, intelligence worsened the farther one got
away from the front lines. The national system headed by
the CIA misperceived the nature of Saddam Hussein's regime,
failed to grasp the obvious signs of attack, and has yet to
learn Saddam's military and political reasoning. Our
imaging satellites failed to find mobile Scud launchers,
and our communications intelligence antennas failed to shed
light on the diplomatic intercourse between Saddam and the
Soviets. National analysts misjudged Iraq's nuclear
program, and were fooled by elementary camouflage. Gen. H.
Norman Schwarzkopf's public belittling of CIA-run
intelligence was matched by unprintable epithets from field
commanders.
What happened in the Gulf would have happened in any
conflict because the intelligence community's cameras and
antennas were conceived, and its people trained, on the
CIA's assumption that cooperative competition with the
Soviet Union would last forever and that the basic designs
of weapons would not change. Thus cameras, for example,
were optimized to take pictures of fixed installations
rather than to keep track of attacking military forces or
mobile missile launchers.
Long before the Soviet collapse, however, it had become
clear that the CIA made bad bets. The age of mobile
missiles arrived long ago, and modern weapons are defined
by the software they contain rather than by observable
features. So what's the point of, for example, analyzing a
radar signal that a computer can change in an instant?
Divorce from operational responsibility also tends to make
the reports that flow to top officials less valuable than
the information used to compile them. (In any given
subject, the CIA delivers a consensus of the system's
several agencies. It takes far more time for a paper to go
through the interagency process than for someone to write
the paper. Considering the elementary errors and ignorance
that often come out, it is clear that the conferees do not
spend much time fact-checking. Intelligence analysts become
spin doctors, concerned not with facts but with pushing
policy makers in the direction of their parent agencies'
prejudices. Hence the ultimate irony: A system whose
ostensible reason for being was to eliminate from
intelligence the parochial interests of tank drivers,
diplomats, bomber pilots, etc. ended up aggregating the
prejudices of the analysts -- prejudices unrelieved by the
sobering prospect of having to carry out the policies they
are pushing for.
The CIA has maintained a monopoly on judging the quality of
the system's operations and products. It does not heed
presidents, much less their appointees. A decade ago, the
agency ignored President Reagan's executive order to
reorganize counterintelligence. Two decades ago, President
Ford, shocked by how far intelligence estimates were
diverging from reality, asked a group of distinguished
outsiders (the B team) to see whether the intelligence
community's data on Soviet nuclear forces could support
conclusions different from those of the insider analysts.
The B Team, despite resistance from the agency, came up
with results far superior to the insider A Team's.
A better intelligence system should be built on a model
radically different from the 1947 original. Each of the
major departments of the U.S. government (State, Defense,
etc.) should be responsible for gathering and evaluating
the information it needs to operate in the new world
disorder. Intelligence, in short, should be franchised out
to its consumers. There is reason to believe that the
departments would do better without the ClA's tutelage than
with it. In the past, the armed forces have asked to deploy
officers who speak foreign languages, who could blend in
with the local population, and who would be experts in the
military fields on which they were reporting. U.S. military
leaders have also clamored for satellites whose products
they could use. Each time, the CIA made sure such requests
were denied. If those requests had been granted, the
country would be better informed.
In all this there is a need for some central coordination.
The several agencies have to mesh their quest for agents
abroad, lest they stumble over each other. The information
that any part of the government collects must be available
to properly cleared people in all other parts, so that any
and all analysis can be based on all the facts.
Fortunately, maintaining a central registry nowadays
requires computers, rather than the bureaucratic monster
that arose a half century ago.
Finally the president of the United States' own
intelligence needs should be provided by his own staff.
Among its duties should be to make sure that all the
agencies get each others' estimates. The availability to
the president and other top decision makers of contrasting
estimates from through out the government would stimulate
better performance all around. So, while there is a role
for a central intelligence agency in a system based on
consumer sovereignty, there is none for the CIA.
Mr. Codevilla, a fellow at the Hoover Institution, is the
author of "Informing Statecraft" (Free Press, 1992).
End
