WSJ on CIA Dump

[John Young <jya@pipeline.com>]

   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