concurring:
I agree with the result reached in the Opinion of the Court, though convinced that we have been given an unnecessary task with respect to the issue of prosecutorial argument for general deterrence considerations in deliberations on a criminal sentence.
As to the principal Opinion, I would emphasize the implicit holding that the sentence which the appellant offered to accept is relevant to the issue of prejudice. That proffered sentence is one useful benchmark when determining, as we did, that the error could not have deprived appellant of any opportunity to “beat his deal” at trial. The identification of such indicia of prejudice is imperative because the existence of error “. . . does not automatically require a rehearing on the sentence. United States v. Peters, 8 U.S.C.M.A. 520, 25 C.M.R. 24 (1957). What must still be considered is the effect of the errors.” United States v. Mosely, 1 M.J. 350 (March 19, 1976). In short, the identified error here had no effect on the real result of the total trial-level proceedings. That it might have affected the inchoate sentence of the trial court is *838but one consideration on the road to the final determination of “effect.”
We have decided this case in accordance with the law announced by our higher tribunal. While so doing, we may also offer individual observations on the state of the law. United States v. Heflin, 23 U.S.C. M.A. 505, 50 C.M.R. 644, 1 M.J. 131 at note 6 (1975). Mine follow.
The Mosely decision found error in an argument by the prosecutor for a sentence sufficiently severe to deter others from similar action. Two bases were assigned for that finding of error. First, “The aspect of deterrence oriented to the general public appears in the maximum punishment prescribed for the offense.” Second, military practice has established that the notion of general deterrence is not to be utilized “. . . in a way that allows the accused to be punished more severely than he justly deserves.” 1 M.J. at 352.
The court in Mosely indicated that the military practice was consciously different from civilian practice. Such a difference could be proper; the sentencing of persons found guilty is clearly a legal function, but there are influences on such decisions that are not exclusively “legal.” These influences must be dealt with, whether the sentencing authority is a judge, jury or other body.1 Principally, the propriety argument turns on the ethical interests served by punishment and reasoned choices among such competing interests is the business of both judges and ethicists.2
Judge Clause’s opinion shows that Mosely is a watershed, and I have two points concerning the choice made by the Court of Military Appeals therein. First, the general deterrence supplied by the criminal law does not come from a statement of maximum punishments. Secondly, general deterrence is an appropriate consideration in individual cases, i. e., a “sentencing factor.”
I
As to the function of a table of maximum punishments, I believe that such legislative acts are self-imposed restraints on the potential for excess extant in the coercive power of the state. The statement “We can do no more to you than ‘X’ ” has little potential to affect human conduct in the person addressed. This remains true even when the statement is “We can do as much as ‘X’,” but there is no experience to make it real. However, a statement that “Upon the doing of ‘X’, ‘Y’ will result” is a price tag permitting individual assessment of costs. Thus, to the extent that time in jail is recognized as a cost of doing ‘X’, deterrence will come only from a table of minimum punishments or “streetwise” awareness of what courts actually do.3
A table of maximum punishments meets one philosopher’s requirement for acceptance of a coercive mechanism to be imposed by the state, i. e., “. . . it is obviously essential to define precisely the tendency of its operation.” Rawls, A Theory of Justice (1971) p. 241. It may be surprising to be quoting John Rawls in an argument for greater control of individual conduct, but he recognized that “. . . even in a just society it is reasonable to admit certain constraining arrangements to insure compliance. . . . ” Id., p. 577. Rawls was not speaking of specific mechanisms, but he was speaking of specific results. Even in a near-ideal society some human tendencies can only be influenced by the prospect of certain and unfavorable outcomes upon de*839viant behavior. Id., § 39. However, the power of the state to impose those outcomes must be rigorously defined in advance. That is the function of pre-set maxima.
To return to the present world of courts, law and criminals, courts can best view tables of maximum punishments as buffers to the Eighth Amendment. Even the earliest tables were but “indicators” of an appropriate punishment for an offense which is attended by aggravating circumstances. The punishments listed were never a promise. Article VII, Executive Order of March 10, 1920 in M.C.M., 1920, p. 285.4
This element of uncertainty permits the proposition that such tables are a deterrent to be tested by Mr. Justice Holmes’s “bad man” theory of law. Just as obsolete criminal statutes do not today prevent blasphemy and expectoration on public thoroughfares, the tables are too remote from everyday affairs and uncertain in their application to influence the “bad man.” Thus, whatever deterrence is provided by the criminal law must come from an aspect of its operation other than restraints imposed on sentencing authorities.
II
Deterrence theory has a place in military sentencing procedures today, just as it does in civilian practice. That place is both proper and necessary; proper because it does not offend against the maxim that a sentence ought to fit the individual criminal and necessary because the notion of deterrence is fundamental to our basic concepts of the criminal justice system in the United States.
The statement that application of deterrence theory results in a sentence higher than that which “otherwise would have been imposed” proceeds from a view of the criminal and his act which assumes that they can, somehow, be treated as separate from the society in which they existed when the act was committed and in which they continue to exist. In this sense, crime is a social act, an act denominated criminal because it has adverse consequences for others than the actor. When such an act is found to have been committed, the burdens of the wider consequences also fall upon the actor according to the demands of fairness.
Given that calling one criminal severely to task will deter others from doing the same, each criminal then incurs the risk of becoming an occasion for society’s lesson-teaching. Thus, there is no “otherwise would have been imposed” that might be considered. Punishment removed from the societal context is totally inconsistent with the view of crime as a social act.
This analysis does not deny the relevance of individual circumstances, nor does it impugn the desirability of the maxim as a sentencing objective. There is an ordering to this business which now becomes clear. The criminal act in society is assessed as such, then the individually relevant considerations are to be adduced and applied toward the achievement of the goal announced by the maxim. There are patent weaknesses in the system, but only this ordering gives proper account to all that needs to be considered.
Thus, it is possible, logically and practically, to accommodate deterrence theory to the desire for “individualized” sentencing. Obviously, the analysis leading to that conclusion depends on the assumption that some kinds of criminal sentences do deter some people from some kinds of crime. This is not the place for a review of the vast literature on that proposition, so I will content myself with two observations in support of the assumption. Deterrence theory is the best of the ethical bases for punishment, the rest are marginal or totally unsatisfactory. Secondly, deterrence theory is a basic part of civilian criminal jurisprudence and there is no reason for its abandonment in our military system.
Saying that deterrence theory is the best of the ethical bases for punishment is saying only that it is the least assailable of them all. Data about recidivism weakens the idea of individual deterrence; retribu*840tion is just not an acceptable idea today, and it cannot exist in a system which also pretends to be rehabilitative; prevention is also made unpersuasive by the figures on recidivism and is objectionable as speculative punishment; finally rehabilitation is not a basis for punishment, it is a distinct social response to aberrant behavior.
General deterrence also involves undemonstrable predications about human behavior, but the theory is as hard to disprove as it is to prove — for the same reasons. It has been said that the total criminal process would be “bankrupt and indefensible” if general deterrence didn’t work. Gaylin, Partial Justice (1974), p. 18. That is patently extreme, involving as it does a near-total rejection of all the other bases which few could do. However, the strength of the theory is in its generality; its foundation is in common sense and there is some evidence to support it. Bailey and Smith, Punishment: Its Severity and Certainty, 63 J. Cr L & Criminology 530, 531 (1974).
This view of deterrence theory as a grundnorm of the criminal law is supported by the ABA Standards Relating to Sentencing Alternatives cited above. The Standards contain a stated preference for the avoidance of “total confinement,” i. e., uninterrupted jail terms, but would allow them on any of three grounds. The one relevant here is that failure to confine “. . . would unduly depreciate the seriousness of the offense. . . . ” Standards, supra, p. 81. The Commentary to that Standard gives two examples of “white collar” crime and a “crime of passion” in which no confinement would appear indicated, if the prisoner were viewed in vacuo. Nevertheless, an opinion is offered that confinement is indicated because of the adverse social consequences of a failure to do so. Id. p. 107.
Thus, there are both negative and positive reasons to reconsider the implications of Mosely. Even though each case may be tested for prejudice, the testing process is lengthy and the determinants of a sound decision are not always available. Egregious error introduced by prosecutorial misconduct can always be reached on general grounds. Other prosecutorial conduct which conforms to community expectations and which is consistent with traditional and current legal theory ought to be tolerated.
. See Frankel, Criminal Sentences: Law Without Order (1972), wherein Judge Frankel says: “Questions as to the purposes of punishment have, naturally, engaged philosophers, and we would do well to consult more the disparate things philosophers have taught.” p. 57.
. The ethical bases for the state’s punitive intrusion upon personal liberty are (1) retributive, (2) general deterrence, (3) specific deterrence, (4) preventive, (5) rehabilitative. ABA Standards Relating to Appellate Review of Sentences, Approved Draft, 1968, p. 126. The choice among these in civil society is for the judge. Coppens, Moral Philosophy (1924), p. 195.
. General deterrence is but a part of the argument for minimum sentences. ABA Standards Relating to Sentencing Alternatives and Procedures, Approved Draft, 1968, p. 146.
. The earliest Order ís in the 1917 Manual.