Four years ago (after decades of debate), the United States began to release the total amount spent on strategic intelligence each year. In 2007, it was $43.5 billion. Now the same is being done for military intelligence. Add military intelligence and the 2007 total goes to $63.5 billion. That total has since increased to $80.1 billion for the past year. For over half a century, releasing these numbers were believed useful to potential enemies. But post-Cold War discussions with former Soviet intelligence officials indicates this was not the case.
We now know that intel budgets in the late 1990s were (adjusted for inflation) about $33 billion. So the war on terror has nearly doubled spending on "civilian" intelligence agencies (CIA, NSA and so on). Military intelligence, primarily DIA (Defense Intelligence Agency) and the individual services, increase that cost by about 60 percent (from $49.8 to 80.1 billion dollars a year). That's about half of what is spent on intelligence collecting and analysis worldwide. Most of it is spent on expensive hardware, particularly space satellites and computer networks. Few people really know what all this intelligence effort is capable of, because so much of it is highly classified. There is believed to be a lot of waste involved. This is usually the result of duplication and resistance to sharing information.
There are fifteen American civilian and military intelligence agencies. Efforts to get everyone to cooperate and produce a better and more effective product have had little success. Intelligence people are good at hiding their secrets, and for many senior intelligence officers, their careers depend on doing a convincing job of pretending to cooperate. Too often, they don't.
The basic problem is that there are several different types of intel specialist, and each believes their contributions are more important than everyone else's. You have the geeks (who use satellites and electronic gear to take photos and collect electronic messages) and the operators (the people who go into the field, run spies and do the James Bond stuff). And then there are the analysts, who get raw intel from the geeks and operators, and try to make sense of it. The geeks see themselves as the successors of the operators, and over the last four decades, the proportion of actionable intelligence (stuff you can use) that comes from the geeks has gone to over 80 percent. The operators don't deny that, but also point out that in the field, you do things geeks can't, like capturing Islamic terrorists, and intercepting couriers carrying messages that the geeks can't pick up using their gadgets. The analysts have the most balanced view, and often wish they had more operators to follow up on murky leads the geeks have uncovered. The problem with the geeks is that even the best satellite photos are not as revealing as an operator on the ground, right in the middle of the situation. The same with electronic eavesdropping, which often provides only fragments, while an operator interrogating a terrorist can provide much more information. The geeks and operators used to fight about money, but the geeks won that battle decades ago. Since September 11, 2001, the operators have a blank check. The only problem is that you can hire a lot of the needed geeks right away, while it takes years to train a useful operator. Police detectives and private investigators have many of the skills used by operators, but recruiting from this community has never been very successful. Basically, the CIA, and other agencies, have to recruit and train their own, or hired retired military operatives.
This spotlights another problem. While most of the geeks are in one place; the NSA (National Security Agency), the operators and analysts are in many other agencies. Each of the armed forces, plus the Department of Defense itself, an intelligence agency (one for each service, plus the DIA/Defense Intelligence Agency). Add in the State Department, Homeland Security, FBI and a few others, and you have a hell of a coordination job. No one wants to share contacts or information, lest the other agency somehow pollute the source. That's easy to do. If the army has a bunch of agents in Iraq, recruited with great effort, the last thing they want to do is let the CIA know who these guys are. The Iraqis or Afghans spying for the army know that if the wrong people find out what they are doing, they are dead. If they suddenly find out that another bunch of Americans, from the CIA, are on to them, they may just quit the spy business while they are still alive. This is a legitimate fear, and the reason why local police are reluctant to share such information with the FBI or Homeland Security. Informants are the most important tool operators have, and these valuable sources of information can disappear.
But it's more than mistrust between agencies, often it's downright dislike. The FBI and CIA have had a hate/hate relationship for half a century. The various military intel outfits have always been competitive. The CIA sees the Department of Defense intelligence operations as wasted effort, while the military intel types see the CIA as a waste of money. Changing all this will take more than time, it will take a few minor miracles.