A Notice Theory of Just Punishment

Here is a draft of a long simmering paper examining the circumstances in which the offender's knowing that he would be subject to a severe penalty before electing to commit a crime makes just the imposition of that penalty.

A Notice Theory of Just Punishment

Lawrence Crocker Draft 6/19/14

There float around courthouses various versions of the following bit of folk criminal justice theory, “If you can’t do the time, don’t do the crime.” Implicit in this is the idea that it is fair to punish an offender with any penalty in the code, a code he easily could have looked up.The theory places no upper limit on prison terms, short of what the legislature sets in its even handed concern for victims and offenders.

For many years I thought that this theory, though having some superficial appeal, was based on wholly fallacious reasoning. Then I began to think that there might be something right about it. What fallows is what I think is right.

A. Bev’s Case 

Bev, an investment banker late to an important client meeting, parked her car on the street and ran towards the lobby of the office tower. A parking enforcement officer yelled after her “You, just parked in a red zone.” Bev replied, “O.K..” Parking Officer: “That’s going to be $300.” Bev: “I hear you,” as she disappeared into the building.

Assume that Bev did indeed park in a red zone – a zone marked out for special parking fines because of its importance to traffic flow. High fines had been found necessary to deter illegal parking in this part of the city. Assume as well that Bev paid the $300 ticket left on her windshield.  Bev’s making the $300 payment is a paradigm case of just punishment. Certainly it is hard to claim that there was anything unfair in the way Bev was treated – despite the high dollar amount.

In saying that this is a paradigm case of justice in punishment, I specifically avoid saying that it is the only such paradigm. There is also justice, of a different sort, in the informal punishment of socially snubbing someone who snubbed you in a similar circumstance. In general, talionic (“eye for eye”) punishment in terms of equality with respect to harms caused, rights violated, or risks imposed, may also be paradigmatically just. Sometimes, however, talionic punishment, e.g. for a small theft or minor assault, may be insufficient to give much deterrence effect. If we relied upon talionic punishment alone, even on sophisticated and extended versions, we might have too little crime control.¹ This criticism extends to my “Upper Limit of Just Punishment,” Emory Law Journal, 1992, which I now think tells only part of the story of justice in punishment, and even that part has some errors.

I also use “paradigm” to avoid setting out necessary and sufficient conditions for just punishment. When we have a paradigm instance, as in Bev's case, we can be confident that the punishment is just. As we deviate from the paradigm along various dimensions, punishment may well become less just, without their being any but an arbitrary line between the not very just and the wholly unjust – although some punishments are paradigmatically unjust, as well.

Back to Bev. It is worthwhile to inventory the features this case has, and those features it may well lack. First, Bev may not have caused any harm. Although red zone parking is proscribed because it generally obstructs traffic, vehicle flows may have been unusually light on the day of Bev’s meeting. No backup at all may have resulted from her illegal parking. Additionally, Bev might be an amateur student of traffic matters who was very well aware that her illegal parking would, this day, cause no harm. We may be tempted to say that it was too bad that Bev paid so steep a fine for doing something she knew to be harmless, but we would not think that her being required to pay was unfair. These reflections show that what is going on here cannot be accounted for by a retributive theory of punishment that focuses either on resulting harm or mental culpability respecting resulting harm.

Bev also need not have expected that she would in fact be fined. She might have thought the officer was bluffing and would not write the ticket. She might have thought that if she got a ticket she could get it fixed by her friend at City Hall. She certainly need not have anticipated actually paying the fine, perhaps believing that scofflaw enforcement was too lax to worry about. So she need not have anticipated either that she would be charged with an offense or punished.

It has been suggested that the offender’s consent is what justifies punishment.² Bev’s conduct does bring to mind aspects of the consent family of concepts, but what she did was at most a distant cousin of true consent. There was no agreement that Bev entered into with the officer. Bev did not say, “I agree that I am liable for a $300 fine,” or “I consent to your writing a ticket that will require me to pay $300.” She did not waive $300 worth of her right to her personal property, and there is nothing from which any such waiver could be inferred.

What could Bev’s “I hear you” be fairly taken to convey? Here this once common expression, now well past its prime, can be taken at face value. Bev is acknowledging that she hears, and understands, that if she does not move her car she risks being held legally liable for a $300 fine. “I hear you,” because it did not dispute the statement of the uniformed officer, might also be said to express, or at least imply, her understanding that the officer was speaking authoritatively as a representative of the state in stating the fine amount.

What is morally crucial here is not an agreement or any sort of quasi-agreement or waiver, it is something that Bev now has good reason to believe, and it is what Bev does. Being fully aware of presumptively reliable information about her potential legal liability if she does not move her car, Bev elects to face the consequences. She goes into the building rather than moving her car and embarrassing herself by being tardy to the meeting. That she would rather risk the liability for a $300 fine than the displeasure of her clients is what makes it so clearly fair to carry through on the fine.

B. The Key Features.

It seems to me that the following are the conditions involved in making punishment fair in the way that Bev’s fine is fair.

1. B had fair notice that she would be liable to the consequences if B did a particular act.
2. B elected so to act.
3. The person making the threat was appropriately situated, as a moral matter, to make the threat to B and to carry through on it.

C. Standing to Punish.

In considering some of the implications of this paradigm for punishment theory and practice, I want to take up the third feature first. Consider a variation on Bev’s case. Ben, an investment banker late to an important client meeting, parked his car on the street and ran towards the lobby of the office tower. A 20 year old wearing gang colors yelled after him “That’s one of our parking spots.” Ben replied, “O.K..” Gang member: “That’s going to be $300.” Ben: “I hear you,” as he disappeared into the building. An hour later, when Ben emerged from the building, he was set upon by three gang members who grabbed his billfold, riffled it, and fled, leaving the billfold behind on the sidewalk. Ben found that $300 in cash was missing from the billfold.

I introduce this variation for two purposes. First, it helps dispel any lingering intuitions that Bev’s conduct amounted to some form of agreement, consent, or waiver. Bev did nothing more that could be relevant to these concepts than did Ben. There is little temptation to say that Ben tacitly consented to parting with his $300 or that he waived $300 worth of his property rights. What Ben did was to express his understanding of the nature of the threat directed at him. Ben could have moved his car, but instead elected to face possible consequences from the gang rather than the displeasure of his clients.

We do not, of course, think that the $300 was justly taken from Ben. At most we have a grudging respect for the gang for being so scrupulous as to carry through only on their explicit threat rather than to keep the entire contents of Ben’s wallet or to beat him to a pulp as well. What the gang did was in that minimal respect not unfair. Ben got just what was threatened if he parked in that spot. But overall what happened to Ben was unfair because what was threatened was a robbery.

My second, and chief, purpose for giving the Ben variation is to examine what sets the two cases apart. Bev’s penalty came at the hands of those who had moral standing to exact such a penalty; Ben’s did not. Bev and Ben do not differ with respect to their ability to elect to avoid threatened consequences by moving their cars. They need not differ in their assessment of the probabilities of their loss of the $300. They each could have considered it very likely, very unlikely, or somewhere in between. Though I did not describe the officer’s shouts at Bev as threats, that would be as appropriate a description as it was in Ben’s case.

 

The salient difference obviously was that the officer threatened Bev with the consequences of a legal liability and the gang member threatened Ben with robbery. The parking ordinances were enacted by a legitimate government. If they were not, if for example a block organization decided to establish red zones and to have its agents hand out tickets, without any authority to do so, then it would be a different case. (If Bev paid such a ticket under the misapprehension she was required to do so by law, she would have been defrauded, not punished.)

Legitimate state authority can justly establish legal liabilities including liabilities carrying punishment. Control of crime is a core state responsibility, and with that responsibility comes state prerogatives. Some punishment theorists essentially stop here. Having the responsibility to control crime, the state, perhaps via some social contract reasoning, has a right to punish at least up to the severity drawn from utilitarian calculation. My view is that state legitimacy is a necessary condition on such punishments as incarceration, but that severity is subject to justice constraints, and so features (1) and (2) come into play if we are looking towards Bev case type justice.

It is important that Bev's case paradigm depends upon state legitimacy in a way that the talionic paradigm of just punishment does not. If you, in your fun loving way, push me off the dock into the lake on Monday morning, it is not unjust for me to return the favor on Tuesday. My giving you fair warning that I will shoot your dog with my pellet gun if he crosses my yard again, however, does not make my subsequent shooting paradigmatically just.

The question exactly how much legitimacy a state must have to enforce what sort of laws is a difficult one. I take it that the Nazi state lacked the legitimacy to enforce its racial laws, and all punishments under those laws were unjust. Yet if the Reich's traffic regulations were enforced in a generally impartial fashion, it is arguable that its traffic fines were not unjust.³

An otherwise legitimate state may also act beyond its authority in meting out punishments. Any penalty for conduct that no one could have guessed would be considered criminal, as in ex post facto prosecutions, is a clear example. These sorts of punishments, however, are already excluded by the first feature of the paradigm.

Proper publication will not serve to bring within the paradigm punishment pursuant to a law criminalizing conduct that ought to be beyond state regulation. For example, a state with otherwise unquestionable legitimacy would arguably act ultra vires were it to criminalize, perhaps under the heading “countenancing treason” the thinking of thoughts critical of the government or its officials. One does not have to hold too exotic a contractarian theory to believe that some prohibitions are properly beyond the reach of even the most legitimate of states.

It is also arguable that certain sorts of punishment are always illegitimate even in support of otherwise proper criminal prohibitions. Torture and death are the most obvious candidates. The arguments that these are ultra vires turn on specific features of torture and capital punishment, their cultures and consequences. I will not examine those arguments here.

The chief way in which we may suspect that liberal states punish beyond their authority is in giving prison sentences disproportionately long for the crime. Here, however, the best arguments seem to run from the injustice of the punishment to their ultra vires character rather than the other way round. In exploring the moral limits of criminal punishment, I will assume as a background matter that the state is otherwise legitimate.

 

So, the question whether a state could properly enact a law requiring its citizens to vote is one that goes directly to state legitimacy. If such a law is permissible, however, whether its violation could properly be punished by a year in prison will, I think, turn on matters more specific to punishment theory, to be discussed below.

D. The Quality of the Notice

If a punishment is to be justified as in Bev’s case, is it a problem for condition (1) if the offender does not in fact know the range of the penalties he faces for the offense he is about to commit? Although many career street criminals are remarkably well informed on the statutory penalties for their favorite crimes, white collar criminals, although often college graduates, are typically oblivious as to such matters.

For a malum in se offense, an act of obvious wrongness, the offender pleading lack of actual notice of the penalty, should usually not receive much sympathy. He could have looked the penalty up on the internet. It is true that the calculation of a Federal Guidelines level can be tricky, and is now, in principle, truly only a guideline for the sentencing judge. Yet, it is fairly straightforward to find maximum sentences, and, with a little patience and acuity, one can find probable sentences. Of course, the typical real world offender will not have had the express notice Bev received. If one is considering committing an offense, however, it is certainly rational, if unusual, to look up the penalty. There is, in short, a fairly good “constructive” notice of the penalty for anyone considering committing what is obviously a crime.

The malum in se offender who does not bother to look it up can complain of the severity of his sentence only if it departs too far from what anyone could have anticipated. The $3000 parking fine and the 75 year term for shop lifting would be examples, assuming that there was so much other compelling news that these draconic penalties became law with little media attention. These extreme penalties may well be undeserved, in the way that two eyes for an eye would be underserved. That is not the defect I am concerned with here, however. There would (or could) be failure to give good notice because of the unexpectedness of the departure from the regime of existing penalties, more or less publicly known. This upper limit will vary depending upon the severity of the regime of punishment as a whole. In a generally draconian system, the offender would have to anticipate greater severity in the penalty he omitted to check in advance.

For malum prohibitum offenses, which are not moral wrongs, the offender has a compelling argument if he did not know that there was any penalty at all. Once, however, he knows that his projected conduct constitutes an offense, he must, again, inquire or forego any claims of unfairness for any penalty that does not depart too far from what could be anticipated.

I conclude, then, that the fact that real world notice of penalty usually falls well short of the notice Bev received does not disqualify the paradigm as a device for analyzing, and potentially justifying, real world criminal sentences. Requirement (1) notice does not make all actual sentences unjust under the paradigm. It has that effect only in the (relatively rare) cases in which the penalty for an offense is more severe than the defendant could reasonably have anticipated given an average citizen's understanding of the background level of the existing system of criminal penalties.

E. Election

Real world Bevs will rarely receive unintuitively high penalties in those cases in which requirements (1) and (2) are fully satisfied. Informed of, and believing in the existence of, say a $2000 fine, the displeasure of her clients would be outweighed, and Bev would move her car.

As a conceptual matter, however, (1) and (2) may initially seem consistent with indefinitely severe penalties, at least if there has been fair notice of the severity. Alexander argued this point against Nino’s “Consent” theory of punishment, a theory which, despite its name, understood punishment in terms of conditions fairly similar to (1) and (2). Although Alexander had the better of that argument, I think that Nino in the end was right. An upper limit is implicit in what makes punishment fair in cases like Bev’s, and in the punishment theory that results from its generalization.

What gives this paradigm its force in justifying punishment simultaneously implies an upper limit on the severity of punishment. Yes, feature (2) of the paradigm is consistent with some long prison sentences. The logic of (2), however, keeps the paradigm from blowing the lid off penalties altogether.

Let us take a closer look at why we think it fair that Bev should pay the $300. It is because she let herself in for it in a fully voluntary and informed way. It was her own choice. This is not to say that she chose the penalty or even accepted the penalty, but she did choose to take on a risk of that penalty with eyes wide open.

Some caution is needed here, however. We do not always insist that people take the consequences of risks they have chosen even with their eyes wide open. If you agree to a liquidated damages clause in a contract, the amount of which is far greater than any reasonable estimate of expectable damages, courts will refuse to enforce the clause upon your default. Other contract clauses too are unenforceable if they are so extreme as to be unconscionable. Someone in Bev's position has, as a moral matter, a similar defense against very severe penalties. Surely what Bev did in taking on the risk of liability committed her no more tightly than does a signature on a written contract.

It is worth examining, then, a factually close analogy in Bev’s case to an unconscionable contract term. Suppose that the parking officer, somehow, persuades Bev that there is actually a $300,000 fine for parking in her chosen space. The business consequences of Bev’s being late are at most a few thousand. Therefore if Bev now fails to move her car, she is acting irrationally, very irrationally. How would such irrationality affect the moral force of a penalty limited by feature (2)?

That moral force clearly would not be affected if Bev’s decision is only irrational in that she made a slight miscalculation. If the expected utility of moving her car would have been marginally better than that of her arriving punctually to her meeting, she fails to be a rational economic maximizer. This level of irrationality, however, does not call into question the voluntariness of her election. The parking judge would be right in rejecting Bev's defense of miscalculation.

The irrationality must be more serious before it has moral weight. Bev’s competence to make decisions must be in some question, at least as a temporary matter. If we do get to that level of irrationality, however, it eviscerates the moral force of penalties supposedly supported by (2).

Suppose that someone in Bev’s position is otherwise acting bizarrely, for example, by engaging in a one-sided shouting match with a street light. He is told by a parking officer of the amount of a traffic fine, and replies, “Thank you, Admiral.” Here we simply would not know what to make of his declining to move his car. He may not understand the warning or appreciate its significance for his future. He may not really be able to move his car. In any event, we lose our confidence that his not moving his car ought properly be described as an “election.” We would still lack that confidence even if it turned out that the defendant was sane in terms of the criminal law definition of insanity.

A level of rationality is essential, then, for (2) to give rise to any presumption that the penalty is justified. As the severity of the penalty reaches heights that cast serious doubt on that rationality, Bev’s paradigm loses force. In the absence of any data going to the states of mind of offenders, it is no easy matter, however, to determine what level of sentence is so high as to make it doubtful that there has been an “election.” Consider, for concreteness, a middle level assault. What I have in mind by “middle level” would be, for example a bare fist beating that inflicts some severe pain and causes the victim to miss three or four days of work. By the end of three weeks, however, the victim is entirely back to normal with no physical deficits or scars.

Suppose that B is clearly about to attack A and that an assault as described would ensue. A bystander yells at A, “You could get life in prison if you beat him up, and there are police just down the block.” B, who indicates that he understands and believes the bystander, then throws the first of several punches.

Whether it is fair to send B to prison until his death depends upon B’s state of mind when he decides to go ahead with his assault. If, for example, A is a lifelong enemy, who has done truly awful things to B and B’s family, then B might well elect, with full cost-benefit rationality, to face even the certainty of life imprisonment to beat A. Indeed, B might be grateful for the opportunity. If A is a lesser object of hatred, but B has not much fear of prison, in fact is indifferent as to whether he is in or out, B might also elect to go forward with the assault with eyes wide open. Offenders with highly idiosyncratic preferences like these do exist, but even among offenders they are uncommon.

Far more common is the typical bar brawler, too drunk and too angry to factor into his decision making the disadvantages of a lifetime caged. His assault does not look very much at all like Bev’s election. We would not want a penalty for middle level assault so high that we can expect many of the offenses must have been committed by those who did not take the penalty into account.

The whole analysis becomes more complicated when we include the important real world fact that most offenders expect not to be caught. It is not that they prefer the announced prison sentence to foregoing their crime. Their cost-benefit preferences may be fairly normal, but, inasmuch as they will never be apprehended, and certainly not convicted, as they believe, those preferences make no difference.

We do not have much sympathy for the brawler who believes he can certainly duck out through the alley and outrun the police. We count his decision, though stupid, as sufficiently voluntary to count as an election. At least we do up to a point. It is fair enough if done in the face of a one or two year sentence, perhaps even a five year sentence. But it does seem hard to support the fairness of a sentence of life in prison on the basis of the risk having been elected on so flawed a probability analysis.

What if B knew that he was surrounded by police, what would we make then of his confidence that he could dodge through and out sprint them? No doubt we want to convict him. We might even say that he elected the risk of life in prison. His analysis of the risk was so divorced from reality, however, that it no longer resembles Bev’s. As offender incompetence at probability analysis increases, the paradigm again applies less and less well.

It therefore makes no difference whether our angry, drunk brawler throws his punch after hearing about life in prison because he cannot bring himself to focus on costs and benefits at all or because he cannot begin to do the probabilities. In neither instance can we comfortably say it is just that he receive the penalty of which he was informed. So, although occasionally a close psychological investigation would reveal that it was fair to administer a very high threatened penalty to middle level assailant, far more frequently the criminal conduct, in such cases, could not be traced to an election that looks voluntary on Bev’s model.

If, then, we were legislators trying to use this paradigm in setting a maximum penalty for our middle level assault, we should conclude that life imprisonment would risk injustice in too many cases.

Going towards the other extreme, consider a nine month sentence. Doubtless most fully rational Bs would not commit the assault if the announced penalty were nine months in jail, but committing the offense with knowledge of this penalty does not imply a level of irrationality so extreme as to make it unfair to administer the threatened sentence. We think, “Yes, people can commit such assaults reasonably voluntarily in the face of a nine month sentence.”

How high can the penalty go before we lose this assurance? In the absence of any data, this is no easy question. My own concerns arise in the neighborhood of a two year announced sentence. At that point, I begin to have serious doubt that a middle level assault could be sufficiently like an “election” for it to be fair to administer the threatened sentence – except in such rare cases as the lifelong enemy. I do not, however, put great stock in my intuitions on this point. I anticipate others, asking the same questions would come up with somewhat different answers.

That this paradigm will not produce a neat and uncontroversial sentencing table should not be too dispiriting. Nothing else does either – certainly nothing else that can purport to be fair to the offender, nothing that would meet the Kantian prescription against using the offender as a means only. 

Desert in the extended family of an eye for an eye potentially might give us more precision, but for a wide range of less serious offenses and middle level sentences, there seems to be no plausible concept of such desert that gives us enough crime control. Its sentences will not be severe enough to control crime through deterrence, incapacitation, or rehabilitation. Bev’s paradigm shows a way we can get penalties with more teeth for smaller thefts and the less violent of violent offenses. It does so, however, while still retaining some upper limit on the severity of punishment. Not every severe sentence becomes just by the simple expedient of being passed and published.

1 See Alan H. Goldman, “The Paradox of Punishment,” 9 Philosophy & Public Affairs 42 (1979).

2C.S. Nino, “A Consensual Theory of Punishment” 12 Philosophy & Public Affairs 289 (1983)

3 Yet should a member of the Weise Rose have felt moral compunctions about ripping up a parking ticket if their VW had been ticketed while they were out putting up anti-Nazi posters? Generally, however, even the least legitimate of de facto states probably have authority to punish rape and non-political murder, theft, and other offenses that are considered criminal everywhere.