First, what is a hero class? Well, there are specific job functions that allow a person to be viewed as heroic by virtue of their daily activities within that job function. Typically, these are privileged classes in the sense that workers within them are viewed as being better, in some way, than the typical member of the society. In western societies these classes are the Firefighters, Police, Paramedics, Doctors, and Military Personnel.
Why these classes are viewed as being particularly heroic is largely a narrative question. We have narratives throughout our social experience that enforce the heroism of these types of worker. These narratives contain the logic that these individuals are engaged in professions where they contribute to the good of the society in some way. However, this logic applies to a far greater class of worker than just these classes, so there must be additional, less manifest criteria. In any event, these categories of worker comprise a class that will from here be called the hero class.
A note on the analysis that appears below: The terminology used may appear to be imprecise. It is not explicitly connected to any particular ethical theory. It is my view that the terms used can be internalized to a given ethical system such that they make sense within that system. This will work for a large number of ethical systems. Whereas my language often suggests a consequentialist framework, it may be recast as deontological analysis by agreed upon replacement mechanisms that replace object values with rule arrangements.
How much greater is the value of the life of a hero than that of other classes of person? Certainly, the value of a hero is treated as being much greater than that of a criminal. For instance, if a law enforcement official suspects any degree of danger to his person, he can usually justify the summary execution of a criminal. This is the basis of the ethical justification for killing of criminals extrajudicially.
Foreign nationals are valued depending on their occupation and nationality. Generally speaking, Foreign military personnel and police officers are not given heroic distinction at all. Additionally, peasants, blue collar workers, the unemployed, day laborers, and non-professional white collar workers (basic civilians) are viewed as less valuable than professionals, businessmen, capitalists, and academics (prestige civilians). Peoples from countries outside of the United States, Europe, or the British Commonwealth (western) are also generally viewed as less valuable.
Thus, in order to preserve the life of a single hero, as many as one hundred civilians can be killed or allowed to die, provided that they are basic civilians from non-western countries. In this scheme, a prestige civilian is probably worth about 10 basic civilians, and a western civilian is probably worth about 10 non-western civilians. An empirical study could elucidate this ratio.
How did this arrangement come about? There are three aspects to it: first, the power of unions bargaining for occupational safety standards for these classes of peoples leads to restrictions on activities in times of crisis. Secondly, political leaders often perceive that allowing deaths of many heroes will be politically unprofitable, so they will make decisions that prioritize the lives of heroes. Third, the casualties among less valuable classes of people are heavily discounted in various ways.
On the first point, unions are successful in their attempts to secure benefits for the worker classes they represent largely as a result of the political dynamic in place. Hero classes generally receive a favorable political dynamic, so the interests of their unions become more successfully advanced than that of other workers represented under collective bargaining. Safety is typically a highly valued concern for these classes. Thus, the result is the establishment of procedures that reduce hazard levels. While not a deliberate trade off between the lives of non-heroes and heroes, it is this in effect because efficacy, and therefore net life-savings, becomes a secondary criterion to safety at least some of the time.
On the second point, Political leaders recognize the importance of preserving the lives of heroes, in part because some of the blame for hero deaths ultimately falls upon them. When heroes die, it carries extra poignancy to a society that idolizes them. Additionally, politicians seek to appease their unions, lobbyists, and advocates. Specifically, in the industrial supply for these industries, it is more profitable to utilize advanced technology, often on the basis of preserving hero life, and politicians will follow the money in this regard. Additionally, popular sentiment is more likely to follow stories of individuals or small numbers of people than large, statistical quantities. Popular sentiment also views the lives of typical people as being nearly worthless - this is the basis of our attitude toward strangers.
Regarding the third point, while the death of many people might carry a larger absolute number, people do not register these deaths in the same way because of various discounting mechanisms. First, media is often not present in the decision making of heroes. Thus, information on what actually happens is frequently unavailable. Secondly, people have a tendency to deny that heroes' decision making or operational pattern of action can actually be anything but ideal. Thus, the view that heroes (not necessarily at the individual level but through their institutions) can make a calculation that puts their lives above those of who they are saving is a source of cognitive dissonance for the average person, who will therefore avoid conscious consideration of the problem.
Turning now to the question of what hero institutions should strive to achieve, consider that as an alternative to this actual, unequal value arrangement, there was an ideal in which all human life is treated as being of equal value.
Under such an arrangement, one must consider relative effects on human action. Thus, the policy of police to less often shoot criminals might encourage them to become more confrontational on average, or less likely to flee from police. However, much sociological work suggests that this is of limited effect. Similarly, military interventions would probably not suffer from significantly increased resistance if military actions were made less brutal. In this regard, it is important to count accurately, since some of the changed effect comes from differences in immediate conditions, which should not be included, as the question is whether the net quantity of such decisions increases, not how they individually pay out.
Within a given operation, there are logistical constraints. The death of a hero prevents their presence at a later time in the operation. In this regard, two types of operation can be distinguished.
In an emergency operation, attrition is important because replacement cannot occur within a sufficiently brief scope, so that each death hinders efforts through to the remainder of the operation. Thus, in this case, the initial assessment of operation length and frequency of life-saving trade-off should help determine the degree of risk that the hero should initially undertake. As the operation approaches its conclusion, the hero should gradually be exposed to greater risk, until such point that the expected value is maximized with no risk avoidance premium at the termination of the operation.
In an operation that is planned in advance, arrangements can be made to replace any hero that is killed. Thus, the expected value should be maximized throughout, with no risk avoidance premium at all.
Expected value is maximized in terms of net human life. Thus, for the hero that has no risk avoidance premium, operational possibilities should be ranked based on expected lives saved - expected lives lost. A hero must be indifferent between his own death and that of another person. The risk premium is simply an additional value that is placed on the life of the hero, thus the calculation when there is a risk premium takes the form of expected lives saved - expected lives lost - risk premium * risk of hero death.
In conclusion, let an operation that has a net gain of human lives be considered good. Let an operation that has a net loss of human lives be considered evil. When the perspective is set such that the status quo is the current operational framework, operations are then defined as changes to policies and corresponding changes to lives saved or lost.
It is clear that policies that allow police to shoot people with near impunity, or allow militaries to bomb civilians in pursuit of political goals are evil. The resistance, even by unions (such as police unions) to policies aimed at increasing accountability of heroes, are policies that enforce the unequal valuation of lives between heroes and non-heroes. The honoring of military personnel who have committed atrocities in foreign states similarly enforces this inequality. Even the use of the term "hero" to describe activities that in any other light would be horrific is testament to the relativity and self-serving nature of our popular thinking regarding good and evil.