Relationship between the political system and causes of war (questions)

[John T. Fishel]

Hi Entropy--

Good post. I will only take issue with you on an aspect of the rational actor approach/model. This is, as Graham Allison pointed out in his classic Essence of Decision, the realist model of international relations that has long dominated the field (Thucydides through Morgenthau and the Military decision Making Model). It tends to explain and predict better than any other single model as long as (1) the analysis is rigorous and (2) the analyst has a solid understanding of the explicit goals of the target. For example, if one had read Hitler's Mein Kampf and assumed that Hitler believed what he said in the book then all his subsequent actions were a rational strategy to achieve some essentially irrational goals. Use of the model would, under those conditions, have been an effective predictive tool - as it was for Churchill (who used it implicitly). Like Adam Smith's economic theory (which has been expanded but never overturned) the rational actor model assumes that individuals rationally seek to satisfy their interests. This works in the aggregate in economics. In IR/intel/policy it depends on a little more before the whole thing comes together. Again, Allison showed that much of what the rational actor model does not explain/predict can be explained by 2 other models: (1) organization theory and (2) an individual political model. But what he does not do is to connect the 3 models and show how they mesh in the way that Adam Smith did. For Smith the individual rationally sought to maximize his economic interest. He and his successors added the notion of the firm doing the same thing. The result was the entire economy rationally maximizing its economic interest moved "as if by an invisible hand." Realist IR theory historically starts with the state and never moves down. Allison made the move downward. The final step (and why the rational actor model explains/predicts better than any other single approach) is to bring it back up - individuals acting rationally in organizations tend to move the state into rational policies (with the caveat noted above in the Hitler example). So, the danger you note is not inherent in the model but rather in the way the analyst uses it - or misuses it, as the case may be. One other caveat - in some (few) cases what is rational for theindividual or organization is not rational for the state (or the economic system as a whole - hence the need for regulation).

The only other comment I have is related to what technology can do. Your thoughts on that express most of the dilemmas and positive results very well. (I left the intel analyst business before the advent of the PC so I only experienced deluges of paper.) But what computer technolgy can do for the analyst is to call up all reporting on a given subject. Much of what we got from Humint was F6 (old term for "source has never reported before - reliability unknown - and validity of info unknown"). this stuff was often discarded unless the analyst saw it being reported again from a different source - in which case he could/should upgrade the source and the info. But given paper files, it essentially depended on the analyst to remember. Now a click of the mouse can bring ups all reports on the subject or from the source for comparison. A good capability of technogy, I think, as long as we know enough to use it.

Cheers

JohnT

[selil]

Good post. I will only take issue with you on an aspect of the rational actor approach/model. This is, as Graham Allison pointed out in his classic Essence of Decision, the realist model of international relations that has long dominated the field (Thucydides through Morgenthau and the Military decision Making Model).

Correct me if I'm wrong, but in essence what you're saying is the "rational choice model" is predicated on the "point of view" of the subject?

Is this part of the Durkheim philosophy?

My thought is that in some cases we talk about Tilly, and others who look at resource mobilization and extending that into concepts such as driving group think, and that resulting in collective action by intelligence analysts. This would be totally counter to independence of thought and analysis. Perhaps a simpler way to state it would be "rush to judgment".

[Entropy]

J. Wolfsberger,

But I will stick to the assertion that however good the other sources are, they can never completely replace HUMINT.

I would take that further and suggest that, in general, the same can be said of any of the intelligence disciplines. Of course, much is situational and in a given situation one "int" may be inherently better at providing the needed information than another.

John Fishel,

For the most part I agree with you about the rational actor model in theory, but in practice it all too often results in failure. Even when we do have "a solid understanding of the explicit goals of the target" and use rigorous analysis failure is still frequent, but more often the model is simplistically used and is little more than a cover for mirror-imaging.

As a case study, we can look no further than Saddam and Iraq. Dr. Gerrold Post's psychological profile of him back in 1990 proves reslilient today with the benefit of hindsight, yet we still critically misjudged what he saw as his greatest threats and his WMD decisionmaking - two factors that directly lead to war.

So even if the rational actor model is judiciously applied huge errors in analysis still occur. The reason for this, in my view, is that analysts are inevitably forced to put themselves in an adversaries shoes to predict what they would do in given circumstances. The record of such prediction is not a good one unfortunately.

More often, however, the model is not applied rigorously in day-to-day intelligence production because most analysts (particularly current analysts) simply don't have the level of expertise and intimate knowledge of the adversary to make even educated guesses. In these situations mirror-imaging is most often the rule rather than the exception and the model provides a false sense of security and false analytical rigor.

This reminds me of Occam's razor, which is a dangerous tool for the intelligence analyst because of deception.

Finally, I agree with you completely regarding HUMINT reporting. Fortunately, the system no longer uses inappropriately precise designations like "F6" - instead using a plain-language source description, but it's still difficult for the all-source analyst to determine the strength of the information in HUMINT reports. I hope those producing HUMINT reports understand how critical their insights and source descriptions are to the all-source analyst.

[John T. Fishel]

No, I'm not saying that the model always looks at the state being analyzed from that state's point of view. In fact, it is rare that the analyst has that kind of information. Most tyrants do not write their intentions into books as did Hitler. The problem for the realists like Chamberalin and Daladier was that they had not read Hitler or assumed that he did not mean what he said and, therefore, assumed his rationality was like that of any other European state. Most realist analysis operates that way asumming states act in their "objective" interest.

No, I wasn't referring to Durkheim. I haven't read enough of him to be comfortable citing him.

Finally, groupthink is always a danger but rigorous analysis and constantly questioning one's assumptions can minimize its negative effects. Not only do we have Irving Janis' groupthink concept but we also have the more conventional view that several minds are better than one alone. Both are, IMHO, true.

Cheers

JohnT

[John T. Fishel]

I think that you and I are in basic agreement. What I am suggesting is not that the rational actor model - rigorously applied, of course, - is always a good predictor/explainer but rather that it is simply the best one available. I tend to think about its efficacy arbitrarily as 80% accurate. I then posit that org theory adds another 10% predictive capability, and the political model another 5% (Remember these are my arbitrary assigned values - may be more but probably are less.) That leaves 5% of the cases unexplained or worse, wrongly predicted. Compound that with Sam's introduction of "groupthink" and your challenge of mirror imaging and you increase several times the nimber of cases mis-predicted/explained.

That said, what is the alternative to using the rational actor model? To go back to AP's questions and dilemma, he's stuck with it. So, he needs to use it as rigorously as he can and supplement it with things like Allison's Models II and III, testing it for evidence of Janis' groupthink. Sorry AP, you don't have to do all that; it's not a doctoral dissertation - just be aware of the possibilities and pitfalls.

Cheers

JohnT

PS I did like the A - F, 1 - 6 scale. Used well it was a good shorthand but it never was scientific and F 6 was most often misinterpreted as meaning bad info instead of unknown quality of both source and info.

[AmericanPride]

Dr. Fishel,

I think the assumptions of realism will be sufficient for the scope of the paper and the class. The paper is in essence a research paper to explain, by theory, the causes of war. But I wanted to take it a step further and focus on a particular component of decision making within the framework of the theoretical model. The problems of realism are clearly visible, but I think if I shift the emphasis away from the pace of the decision-making process toward the consequences of its design (in relation to strategic intelligence) for states, I will be better off. In this way, I can skirt the issue of technology, addressing only how it relates to intelligence capabilities, and focus on how states use strategic intelligence.

[Ken White]

...
PS I did like the A - F, 1 - 6 scale. Used well it was a good shorthand but it never was scientific and F 6 was most often misinterpreted as meaning bad info instead of unknown quality of both source and info.

you could see how many F6s were laying about in comparison to all others that came in -- kept one sorta honest...

[Entropy]

John,

Thanks again for your comments.

I do believe the rational actor model does have utility but it's always important to recognize it limitations, and, in fact, the limitations of attempting to apply any model to particular circumstances. That said, I think much depends on the character of the nation and it's particular decision-making process. Ideally one would want models tailored to the peculiarities of each nation, but that brings up another set of problems.

Americanpride,

A potential problem you may run into is that decision-making processes vary widely. For simplicity's sake, you might want to consider reducing your scope further to look at a particular category or archetype of national decision-making process.

[AmericanPride]

John,
Americanpride,

A potential problem you may run into is that decision-making processes vary widely. For simplicity's sake, you might want to consider reducing your scope further to look at a particular category or archetype of national decision-making process.

Well, I'm aiming to illustrate in my conclusion that states that can collect, analyze, and apply strategic intelligence, however they manage it, more rapidly will be more successful. I'm going to tailor my thesis somewhat to reflect the slight change in emphasis.

[Ken White]

Which is most important...

[Jedburgh]

Well, I'm aiming to illustrate in my conclusion that states that can collect, analyze, and apply strategic intelligence, however they manage it, more rapidly will be more successful. I'm going to tailor my thesis somewhat to reflect the slight change in emphasis.

Again, I refer you to Knowing One's Enemies, the book I recommended in my first post in this thread.

The studies clearly demonstrate that the personalities and individual biases of the national level leaders - the consumers of strategic intelligence - often have a greater influence upon strategic decision making than does the quality of intelligence or the rapidity of its acquisition.

In theory, I agree with your premise - that rapid acquisition, analysis and dissemination of accurate intelligence to national decision makers increases the likelihood of success. However, I have to state that the historical record puts the emphasis on the word "likelihood". Throughout history, national leaders have often dismissed, cherry-picked or simply refused to consider solid intelligence in favor of other information that played to their personal biases, or simply contrary recommendations from close confidantes whom they trusted more than their intelligence apparatus.

......Another one that previously slipped my mind is For the President's Eyes Only, which is focused on the President as a consumer of intelligence, and looks at the office in that perspective from George Washington to Bush I. However, the first book is really the one that speaks directly to your topic.