THE MILITARY INTELLIGENCE America needs is being sabotaged by an enemy within: the U.S. Army, Navy, Air Force and Marines.

Narrow parochialism, with each service putting its own interests ahead of those of the nation, is crippling the effort to coordinate military intelligence begun 21 years ago with the creation of the multi-service Defense Intelligence Agency.

This parochialism might be called the "Wampanoag factor," after the Navy's first warship powered primarily by steam. When she was commissioned in 1868, the U.S.S. Wampanoag was a revolutionary ship, nearly twice as fast as any other warship afloat. Instead of applauding this technological breakthrough, however, the admirals of the day tried to scuttle it.

A board of naval officers declared in 1869 that the Wampanoag was "a sad and signal failure and utterly unfit to be retained in the service." Why? They feared that a steam- powered ship would make life too easy for men at sea. Sailors who spent their time shoveling coal rather than climbing the rigging of a sailing vessel might lose "that combination of boldness, strength, and skill which characterized the American sailor of an earlier day," the officers warned. The Wampanoag was laid up, converted to other uses and finally sold. It was two decades before our Navy built a ship that matched Wampanoag's performance.

The Defense Intelligence Agency, like the Wampanoag, was a good idea that went bad because it threatened the culture and prerogatives of the military.

Now it's time to make DIA work as the strong, centralized intelligence organization it was intended to be. It's time to abolish the service intelligence units. And to make these reforms work this time, it's essential to give the Joint Chiefs of Staff real power to break the parochialism of the individual services and run the military efficiently.

The DIA was created in 1964 by Secretary of Defense Robert S. McNamara. Previously, all military intelligence work had been done in the separate intelligence arms of the four military services under the direction of the four service chiefs of staff. McNamara believed that the secretary of defense and the Joint Chiefs of Staff deserved to have intelligence that was not suspect of any service bias and that was immediately and directly responsive to their needs.

McNamara pared down the service intelligence organizations and transferred many of their responsibilties and people to the new organization. But in just four years the service intelligence organizations had grown back to at least their former size. Moreover, they were attracting the best people. Military officers understood that they would advance their promotion prospects more by staying at home, within their own service intelligence organizations, than by serving in DIA.

The result has been an awkward mismatch of tasks and personnel. The DIA has had responsibility for many important areas of intelligence, but it hasn't had the capability to produce what was needed. The intelligence units of the various services, with better people, inevitably began to infringe on the DIA's areas of responsibility. Often the DIA simply defaulted to the services.

Consider the issue of Soviet naval capabilities, an intelligence topic of vital concern. How good would the Soviet navy be if pitted against our Navy's ships, submarines and aircraft in combat, especially against our ballistic missile submarines on which our deterrent posture largely rests? Much of the intelligence data in this arena is collected by the Navy, but the Navy refuses to share much of it with the rest of the intelligence community.

The DIA is in too weak a position to dispute this Navy viewpoint. It doesn't make much difference because the DIA seldom disputes assessments made by the individual service intelligence organizations in their areas of specialty anyway. But it's bad for the country when only the Navy, with its own vested interests, is our sole interpreter of the Soviet navy's capabilities.

Intelligence estimates can be crucial in lobbying Congress for military spending. For instance, if the Soviet Navy is regarded as very good, the Congress may be willing to fund a bigger and better Navy of our own. If it's judged to be rather mediocre, the Congress may find a large U.S. Navy is not necessary.

A similar situation exists with respect to intelligence input to our estimates of Soviet strategic nuclear forces. The DIA, let alone the service intelligence organizations, simply cannot rise above the parochialism of the military services.

In 1978, while I was CIA director, the agency did a study of the balance of strategic nuclear forces. Instead of just comparing numbers of weapons, it compared the potential destructiveness of the two arsenals. It showed how many military targets like hardened silos for ICBMs, or how many cities the Soviets could destroy if they attacked us first; the study then compared how much destructive potential each arsenal would have left, the Soviets after firing a lot of theirs and the United States after losing a lot of ours to the Soviet attack.

The point of this exercise was that if we had enough firepower left to do grievous damage to the Soviet Union after absorbing a first strike, it would hardly be worth their while to attack. In fact, the results showed that if the Soviets initiated a surprise attack on all the U.S. missiles, submarines and bombers they could locate, our surviving destructive potential would be just as great as what they would have left -- and sufficient to level all urban areas of the Soviet Union. Unfortunately, this analysis did not produce the answers the Pentagon wanted in order to justify its MX missile program. Thus, a fight began.

For the next three years the DIA argued over and over that the CIA's analytic technique was invalid. They never said why. Rather, they fell back on the argument that the CIA was operating in an area that should be reserved to the military, hardly a meaningful way to judge whether or not the analysis was beneficial to the country. One year, though, the DIA did a thorough study of its own. Unfortunately for the DIA, the results were nearly identical to the CIA's. When asked to include those results along with those of the CIA in that year's report on strategic forces, the DIA refused.

What all this amounts to is that the DIA simply cannot stand up to the pressures of the individual services. The services don't want the DIA to produce analyses that might jeopardize their programs before Congress. The DIA goes along because it doesn't want to upset the applecart. The independent military intelligence voice that McNamara sought is thus pretty well muted today.

The DIA's reputation as a pushover for the individual services compounds its problems. Consumers of DIA intelligence products (and there are many both inside the military and outside) are likely to suspect bias in the DIA's work. That hurts all DIA products, whether they are in fact biased or not. And, it often leads the CIA and the State Department's intelligence organizations to over-compensate in their work for what they anticipate the DIA's bias will be.

For example, the DIA has consistently estimated that the Soviet Backfire bomber was designed for intercontinental missions, which would increase the threat to the U.S. and the need for larger retaliatory or defensive forces of our own. The CIA has held that it was not so designed and does not have intercontinental capability. The truth is probably somewhere in between. Namely, that regardless of the designer's intent, the Backfire has intercontinental range, but only under very particular conditions.

Because the DIA is self-conscious about living within the shadow of the more capable CIA, it often takes contrary positions just to assert its independence. In 1977 it dissented from a CIA study that said the Soviets were going to have difficulty within a few years in pumping sufficient oil out of the ground to meet their needs. DIA's only explanation was that the Soviets had built pipelines large enough to take a greater amount of oil than the CIA estimated they could produce. In time the CIA was proven correct.

The weaknesses of the DIA hurt not only the military. Our intelligence system is built around competitive analysis. That is, we like to have several independent reviews made of important intelligence topics. This concept breaks down when the DIA cannot hold up its end of the competition by producing meaningful analyses. More often than not, when the DIA does produce a differing view, it cannot -- or will not -- support it.

Ideally, competitive analysis should produce reports that state the several differing views and compare them. When the DIA holds a dissenting view, though, it is nearly impossible to obtain a succinct statement as to why it holds that view. It simply submits a statement that "The DIA disagrees with the conclusion that . . . " and no amount of cajoling will bring forth more. That, of course, is because the dissent is driven by the pressure of political considerations, not intelligence.

The answer to the problem of military-intelligence parochialism is to disband the service intelligence organizations and place all military intelligence under the DIA. As a practical matter this would prevent the best intelligence people from avoiding the DIA. More important, it would create a multi-service environment for defense intelligence work where parochialism would not be the accepted norm -- and where there would be watchdogs to deter it.

In the present service intelligence agencies, it is easy to become so absorbed in your service's perspective as to lose objectivity. Most everyone around you is thinking similarly. If you call all of the questionable shots on one side, you're not likely to be challenged. And, parochialism is more likely than not to meet with the approval of your superiors. That isn't because they deliberately seek to skew the intelligence product, but because they may not be aware of how many questionable and one-sided calls have already been made at the working level.

It's time to put the DIA back on the track McNamara intended by making it the center of the Pentagon intelligence process. Service officers assigned to a revitalized DIA would still have their service biases, but those biases would cancel each other out in a strong multi-service organization.

Revitalizing the DIA is worthwhile on its own merits because a strong military input to our national intelligence effort is highly desirable and we do not have that today. Moreover, consolidation of military intelligence in the DIA might be a good model for solving some of the more wide ranging problems of the JCS.

The JCS is to strategy and planning what the DIA is to intelligence: a centralized, multi-service organization with large responsibilities and inadequate independence to carry them out. The JCS staff must contend with four separate military strategy and planning organizations of the individual services. Just as the service intelligence organizations dominate the DIA, the service planning groups virtually dictate what service officers assigned to the JCS planning system can say. It's an enormous duplication of effort driven by the most flagrant parochialsim.

The secretary of defense could simply disband the service strategy and planning organizations and force all strategy and planning into the staff of the JCS. Or, Congress could consider forcing this issue by mandating the demise of service planning organizations. The objective would be to place our national military planning in an atmosphere where objectivity would be expected and where there would be checks on blatant parochialism. The service officers there might still be parochial, but at least they would have the opportunity to take a broader viewpoint.

The parlous state of our military intelligence today needs attention, but it also should remind us that the problem of military intelligence is a microcosm of the problem that haunts the Joint Chiefs of Staff system. Without a reshaping of the JCS, we will have neither good strategy nor good intelligence.