Why Was the Total War Institute’s Report That “Japan’s Defeat Was Inevitable” Not Accepted?

1. The “Institute” that could give an answer and the organization that could not receive it

On a whim I went out of my way to see the film “The Night Before War.” The film is based on Naoki Inose’s “The Defeat of Summer 1941.” Eight months before the Pearl Harbor attack, young people gathered at the Total War Institute, a direct agency of the Prime Minister, formed a mock cabinet that crossed ministries, the military and the private sector, and simulated the outcome of a Japan‑U.S. war. After building up classified data and debating, they arrived at the answer “Japan would inevitably lose,” and that report was effectively rejected at a briefing to the then Konoe cabinet. The film depicts this whole episode.

While watching, two thoughts occurred to me. One was whether we would repeat the same mistake. The other was, if a situation arose that forced us to create a venue like the Total War Institute, could we actually create it, and would it be taken seriously?

However, I realized the latter point almost immediately. The Total War Institute was actually created. In 2001, long after the war, under the vision “Consider a policy Plan B for Kasumigaseki,” the independent administrative agency RIETI (Research Institute of Economy, Trade and Industry) was established. Sixty years earlier, the government of the time had set up the Total War Institute as an organization that could produce an inconvenient “Plan B” for itself. So why was that answer not accepted at the time?

2. Was the government that rejected the report wrong?

Here I want to be careful not to dismiss the logic of the government that rejected the report as mere foolishness. In the film, several army officers argue that “desk‑top exercises cannot capture the contingency of reality” and “war does not proceed as calculated,” which is partially correct. Moreover, the troublesome point is that no matter how much you improve the accuracy of a simulation, you cannot completely refute that argument. In other words, the claim “reality exceeds simulation” can always be invoked.

There are two aspects to why the Total War Institute could be set up. It was a “mock” cabinet, so its answer had no binding force. That is why they could write “Japan would inevitably lose” without self‑censorship. Yet that very harmlessness also became the reason for dismissal as “just a desk‑top exercise.” The freedom to give an answer and the freedom not to have to listen to an answer both arise from the “mock” design.

Needless to say, the very design of the briefing itself was flawed. When the logical conclusion of the Institute would deny the very atmosphere that supported the government’s own “southward advance” policy, it was an inevitable reaction for the government not to accept it.

If that is the case, what we should consider is not the resolve or conscience of the Institute or the government, but rather the very structure of the system: who, when, and in what form will receive the result, and what responsibility will be taken when it is not received?

3. Things become decided even though no one decides them

One overlap between the film and reality is that no person who clearly makes the decision appears at any point in the film. No one decides “to fight.” The weight of facts already set in motion and the pressure of the atmosphere cause things that no one has explicitly decided to become settled. Even if you look for a decision‑maker, there is none. In Masao Maruyama’s “Thought and Action in Modern Politics,” he described the mental pattern of militarist rulers as a system of irresponsibility consisting of submission to the status quo and retreat into authority—exactly what appears in the drama. Responsibility floating in a void is not the result of someone evading it; it is the consequence of a structure in which events proceed even without anyone taking responsibility, independently of any particular individual.

What must also be examined is the behavior of those in power. They look up, or they are swayed by the prevailing atmosphere. Those who are swayed do not eliminate the deliberative space of the Total War Institute; they keep it as a “decoration.” It serves as a façade and, more importantly, allows them to avoid pulling the floating responsibility toward themselves.

If we raise the level of abstraction and look at the outcome of the Total War Institute’s activity, we see that people were carefully selected, given data, made to discuss, and asked to report. Yet the conclusion that emerged was never sincerely taken up by the government.

One more point: the government of the time did not ignore the report because there was no data suggesting inevitable defeat; rather, because the data had been gathered, they could choose not to look at it. The three phenomena—responsibility floating in a void, pandering to the atmosphere, and the choice not to look—are perhaps why the Total War Institute was meaningless.

Turning to the present, as someone involved in practical citizen participation in local governments, I often see a similar pattern to the outcome of the activities described above. A deliberation that functions only as a decoration is more troublesome than having no deliberation at all. If it is absent, the lack is obvious; but a decoration appears to be fulfilling the requirement. I thought I was watching a story from eighty-five years ago, but midway I could no longer see it that way. Frankly, should I call it sadness or melancholy? In any case, I do not want to be satisfied with merely flaunting that feeling; I want to continue observing more closely.

4. The precision of the input itself does not generate the justifiability and legitimacy of the process

This structure links to a point that struck me strongly when I recently read "Technological Republic." What the U.S. defense-tech company Palantir, led by the author of the same book, Alex Karp, is developing is an ontology—a mechanism that describes the world as a collection of entities and relationships—and its customers include government agencies and militaries of various countries.

Alex Karp earned a degree in social theory in Frankfurt, studying a tradition that runs from Habermas to Adorno. For example, Habermas, in “The Structural Transformation of the Public Sphere,” asks who gives reasons to whom—essentially the inter‑subjective response that defines a process. By contrast, Palantir’s product is a device for describing how to represent the world, focusing intensely on improving the quality of the input that feeds the process. At the same time, the company positions the justifiability and legitimacy of the process that uses its product as something outside its own responsibility. In other words, Palantir seems to keep a deliberate distance from the question “who takes responsibility for the decision?” This distancing is not unique to Karp; it is a daily reality in my own industry.

In our industry, the way we keep distance takes this form. We gather good opinions, organize the points of discussion, and deliver them in an easy‑to‑read way. That is our work, and how those opinions are actually handled belongs to the judgment of the local government, which is the client. Given the contractor‑client relationship, this line‑drawing looks reasonable. However, if we take on the input that goes into the process—namely, the improvement of the quality of citizen opinions—while adopting a stance of not being involved in the legitimacy or authority of the process itself, then, aside from scale and client, the attitude and structure are the same as the one we saw earlier at Palantir.

To be clear, this does not mean that improving the input has no meaning. In the area of citizen participation that our company works on, residents’ voices often remain invisible, discussion points are not organized before the debate ends, feedback is not returned, and a loop of further questioning never materializes. Because of such reasons, the quality of the output frequently suffers, and what we are working on is precisely the improvement of those inputs.

But that improvement only shows its effect when the bottleneck of the whole process lies in the quality and quantity of the input itself. So, can we say that the report by the Total‑War Research Institute was limited in both quality and quantity? They had abundant classified national data, elite experts from various fields gathered, and time to discuss exhaustively. Yet even if they had further refined the input and raised its quality and quantity, the outcome of the briefing would probably not have changed.

The more the venue for participation is set up, the more the act of opening it becomes a convenient proof of legitimacy for the organizer. And when a participation venue is well prepared, it also raises the chance that it functions merely as a “decoration.” A crude decoration is recognizable as such, but a sophisticated one is hard to distinguish. What separates a decorative venue from a genuine one, the author argues, ultimately comes down to how difficult it is for decision-makers to avoid responding to and giving feedback on the input that was presented. Let’s call this the “binding nature” of the process. When the decision‑maker’s discretion to respond is left entirely to their own will, even the most intricately crafted venue can become a decoration. The justifiability and legitimacy of the process are not supported by the precision of the input itself, but by this binding nature.

5. Procedures for Leaving a Response

So, what procedural steps can be taken to enable responses and feedback? I would like to consider this in line with the three phenomena described in Chapter 3: responsibility floating in a void, going along with the atmosphere, and the choice to not see.

Regarding responsibility floating in a void, how about countering it by keeping the process open? Here, “keeping the process open” means that the course of the discussion and how the points raised were handled become visible to others, and that visibility itself becomes the binding force of the product derived from the process. If the issues are brought into an open forum, regardless of whether they are reflected, it becomes easy to later demand an explanation of how those points were dealt with.

Before detailing the practice of going along with the atmosphere, I want to make clear that this film does not portray the government officials who received the report as foolish people. The cabinet did not fail to understand the Total‑War Research Institute’s report. In the movie, Hideki Tōjō, after commending the institute’s research effort, replies that “we didn’t think we could win the Russo‑Japanese War, but we did; wars can bring unexpected outcomes.” The film also shows them riding the current while harboring doubts and guilt. It depicts not a decision made by those who knew nothing, but by people who understood to some extent yet still could not stop it.

Why, then, were they unable to take the Total War Institute’s report seriously and halt the atmosphere pushing toward war? Why did they pander to that atmosphere?

Because the very people who try to “stop the atmosphere” are the ones who get labeled (e.g., as cowards). Continuing a direction that has already started carries no label. No one made the decision, so no one is later held accountable for the continuation. However, if the option to “stop once” is placed on the table, the responsibility of the person who raised that option begins to be questioned. In other words, under a structure where responsibility floats in a void, only those who try to stop it end up pulling responsibility toward themselves from that void. As written in Chapter 3, the government presumably did not want to take on that burden. Those who tried to stop it could be held responsible in the ultimate form of being reassigned to the front lines. In such a situation, going along with the atmosphere becomes less a sign of weakness than a rational choice.

The third phenomenon—the choice to not see—can also be explained by the same framework. If one looks at inconvenient data, one must personally decide whether to intervene in the ongoing action based on that data. Intervening invites responsibility; not intervening invites scrutiny for having watched without acting. Remaining unaware protects the individual. Hence, they choose “not to see.”

When we organize the analysis this way, we see that the three phenomena share a common bias. Continuing the flow is never questioned, while attempting to stop or to look—actions that diverge from the flow—draw a label onto the individual and demand responsibility. This bias is the foundation that turns deliberation into a decoration. If that is the case, the role that procedures should play becomes clear: to balance that bias from both sides. If the handling of each raised point remains visible to others, then both not looking and rejecting become actions that can later be asked to be explained. In other words, continuing can also become subject to questioning. Conversely, stopping becomes somewhat lighter. If a record accumulates showing how each point was handled, the decision to stop can be justified as a choice aligned with the accumulated record rather than as a heroic judgment of an individual.

Just to be clear, this is not a call to bind power. What the author asks of decision‑makers is not “agreement” with the participants in the deliberative forum, but “response” to them. In other words, participants’ conclusions do not have to be constantly obeyed by mayors, administrators, or legislators. However, if those in decision‑making positions do not follow the conclusions, the reasons for not doing so should remain as words visible to the participants. And having those reasons recorded also protects the decision‑makers themselves. When later questioned, they can show, as an organization, what issues were present at the time and why they chose the conclusion, even after the individual who made the decision has left. It is not for pursuit, but rather a provision for tracing.

However, this mechanism has an order. Even if you try to leave a record or give a reason after entering the decision phase, the proposal itself gets caught up in a power struggle with the atmosphere. If you say “we should stop” after the move toward war has begun, those words become a demand for someone to decide. Yet, if before the move begins it had been stipulated what result would require a pause, then stopping would not be an individual decision but a step according to the pre‑arranged procedure. The weak tone of the institute members was not because of the content but because the timing of the words was the reason. So why would a pre‑made “promise” have made a difference? I will consider that in the next chapter.

6. Why must a “promise” be made in advance

Here I want to use Hannah Arendt’s notion of “promise” from The Human Condition as an auxiliary line. Arendt raised promise as a capacity that humans use to counter their inability to foresee the consequences of their actions. In an unknowable future, a promise creates partial certainty. It is not that we obey because the forecast is accurate; rather, because we do not know whether it will be accurate, we bind each other in advance, following that order.

Applying this auxiliary line shows why a “promise” must be made before events occur. After the result “Japan would inevitably lose” appears, the very judgment of how to handle that result becomes a stance on whether to go to war. Saying we take it seriously puts us on the cautious side of war; saying we treat it as a reference puts us on the side of promotion. In other words, trying to discuss the “handling” after the result is known inevitably turns the discussion into a dispute over the result itself. Conversely, if the discussion happens before the result, a rule that says “whatever the result, we will treat it this way” does not threaten anyone’s position yet.

Of course, I am not trying to retroactively blame the Total‑War Institute for not having made a promise – that would be a tactless thing to say. History is cruel, but when the institute was established, the push toward southern expansion had already begun. In that situation, trying to set a rule for "what result would make us stop" would have been taken as a stance questioning the flow. I wrote earlier "before the result appears," but more precisely it had to be "before the flow toward a decision is created." The institute had simulation capability and the wisdom of its participants, yet its very history meant the timing to make a "promise" had already passed. In other words, by the time the institute was born, that moment had already elapsed.

Reaching this point connects with Chapter 4 mentioned earlier. Chapter 4 argued that the legitimacy of a process is supported not by input precision but by binding force. That binding force originates from the “promise” that should be made in advance. If a promise exists, even a rough simulation can bind decision‑makers; without a promise, even the most precise simulation cannot bind them. The pre‑made “promise” creates binding force, and that binding force underpins the legitimacy of the process.

In addition, this promise should be recorded as an explicit form, not merely as a mindset of the participants. Who made it with whom, where it is recorded, whose eyes will see it if broken. A mindset alone can be swayed by atmosphere at critical moments, but a formal record cannot simply be denied. Designing that form, and actually materializing it, is a responsibility for those who create deliberative spaces, and I reflect on it deeply.

7. Conclusion

When the need is not apparent, such a “promise” may seem excessive. It can be seen as unnecessary effort that delays progress or development. Of course, even if a “promise” is made, it may lose substance and become mere ceremony. Moreover, the observation that it is ultimately people who invoke a promise to halt movement is probably correct.

Nevertheless, what this paper has examined is how a few conscientious individuals can confront a dominant “atmosphere” without endangering themselves, by establishing an appropriate mechanism.

If we interpret the lesson of the Total‑War Institute as a story where the absolute number of courageous individuals was insufficient, we will repeat the same mistake. The film depicted that even when the correct answer emerges, the structure itself prevents it from being accepted. Structures must be addressed by structures. The author continues to ruminate even after leaving the screen.

Post‑script: After finishing the draft, I thought that this argument would likely hold even if we replace “Total‑War Institute” with “Akimaru Agency.”