I am moving all my files to a different and, for my purpose, better, server.
In 10 seconds, you will be automatically directed to the new site.  Please note the URL.
If you immediately want to go to my new home page, click here
If you immediately want to go to this page on the new site, click here

Sorry for the inconvenience.

Collateral Consequences of Punishment:
Civil Penalties Accompanying Formal Punishment

Hugh LaFollette

[pdf version]

Journal of Applied Philosophy, 2005, 241-61.
Journal of Applied Philosophy

Feel free to read or link to these essays. But please do not place copies of them on another server. 

When most people think of legal punishment, they envision a judge or jury convicting a person for a crime, and then sentencing that person in accordance with clearly prescribed penalties, as specified in the criminal law. The person serves the sentence, is released (perhaps a bit early for A good behavior@ ), and then welcomed back into society as a full-functioning member, adorned with all the rights and responsibilities of ordinary citizens.

If only it were so. For upon conviction of a felony in the United States, criminals face substantial collateral consequencesC civil penalties, which, unlike fines, prison time, or probation, are not specified in the criminal law and are rarely mentioned during sentencing. These penalties are assessed not just while a felon is incarcerated or paroled or on probation, but after completion of her sentenceC and often permanently. There are important questions about the propriety of all collateral consequences, including those assessed against a felon serving her sentence. However, I am concerned here only with post-sentence collateral consequences (hereafter, just A collateral consequences@ ).

The best-known example of such consequences is the disenfranchisement of felons. At the time of this writing, seven states permanently bar all felons from voting, and another five permanently disenfranchise some felons. Because of its political significance (it likely determined the outcome of the 2000 presidential race) and theoretical importance, this is surely a significant instance of collateral consequences. But it is far from the only one. The official listing of collateral consequences requires more than 140 pages. Obviously, I cannot discuss them all. But here are some important legal disabilities convicted felons face in the United States. These (or their effects, in the case of the first two) continue post-sentence.

The scope and significance of these collateral consequences show that the real world of punishment is far different from the one most people imagine. In this world a felon= s debt to society is rarely paid in full. For these felons the Mark of Cain is permanent. Can this be justified?

Of course we could ensure that no collateral consequences are post-sentence by making all penalties part of the normal sentencing process. Then they would no longer be collateral consequences but punishment proper. None of their elements would be post-sentenceC since the sentence would never endC although they might be post-incarceration or post-probation. However, that would not avoid the fundamental issues I discuss. For then the question would be: should significant elements of formal sentences extend beyond incarceration and parole?

Three Views of Collateral Punishments

Some people are unaware that there are any collateral consequences of punishment; few know their scope or significance. I suspect that is what some advocates want. They fearC rightly, I argue laterC that the public might reject them if they were fully aware of them. Whatever their motives, some policy makers obviously endorse the use of collateral consequences else we wouldn= t have them. But most support and employ them unthinkingly, oblivious to the need to explain what they are and how they are justified; still less do they ask whether such penalties could be part of a coherent comprehensive penal theory. Consequently, it is not surprising that there is no accepted explanation of what collateral consequences are, and even less agreement about how they (or some subset them) could be justified. This lack of agreement makes them elusive targets: when someone criticizes one account, advocates may then defend them in another way.

To focus debate, I will classify competing accounts of collateral consequences of punishment and critically assess possible justifications for them. Obviously I cannot consider every collateral consequence; still less can I evaluate all possible justifications. Therefore, I will inevitably pass over certain complexities. I do so, though, because I think that focusing exclusively on one or two collateral consequences will blind us to significant problems pervading the practice. That is what I am most concerned to expose. Finally, although my focus is on collateral consequences in the United States, I believe the arguments are applicable to uses of these penalties elsewhere.

There are three primary ways collateral consequences are conceived:

  1. They are punishments in all relevant respects, although, at least within the United States, they differ from ordinary punishments in several significant (presumably irrelevant) respects.
  2. They are not punishments but direct consequences of criminal behavior.
  3. They are neither punishments nor direct consequences of criminal behavior but state action to protect citizens from risks posed by former felons.

We will have some difficulty drawing precise boundaries between these accounts since they bleed into and blend with one another. Nonetheless, they are sufficiently different that we can theoretically distinguish them. However, we should not assume that all collateral consequences must be justified in the same way. Some might be justified in one way; others, in another. Even so, we still need a explanation of these penalties= role within a comprehensive penal system. The elements of a criminal justice system (police, prosecutors, judges, prisons) are interrelated. Changes one place can significantly effect other elements. That is why we should insure that any collateral consequences do not undermine important penal elements of the system or make the criminal justice system as a whole inefficient. Finally, we should also see how the penal system fits within the larger economic, political, and legal system; we must insure that the whole system is (generally) just.

To evaluate the use of collateral consequences, I focus on three cases, identify their most plausible justifications, and evaluate them. I will generalize from these cases to identify plausible select uses for such penalties while isolating some pervasive worries about them. I do not argue that all collateral consequences are ill-conceived or morally illegitimate. Even were I inclined to defend such a viewC and I am notC that would require a monograph, not an essay. What I will argue is that no single justification is both plausible and capable of defending more than a smidgen of current practice. No combination of justifications fares much better. We need to rethink the practice, and impose collateral consequences only in isolated cases. Those collateral consequences should be clearly specified, widely known, and integrated into a coherent penal system.

Three Cases

Here are the three cases: (1) denying social benefits to former drug felons, (2) permanently disenfranchising all felons, and (3) baring former felons from practicing a particular profession if their crimes were committed while working in that profession. I choose these cases for two reasons. First, they represent the range of collateral consequences felons face. Second, each represents a different way of conceiving and justifying collateral consequences of punishment. By examining all three, we can gain a better sense of the proffered justifications for and problems with collateral consequences.

Denial of Social Benefits to Former Drug Felons

Current United States law denies welfare and food stamps to all drug felons upon completion of their sentence, unless the state in which the felons live has specifically exempted itself from this legal provision. While the original version of the law specified the same punishment for all drug felons. Congress later amended the law so that the length of time that felons lose such benefits varies somewhat with the nature of their offense and the number of previous drug convictions. Other laws bar them from living in public housing, except under a few tightly prescribed circumstances. Still another law denies student loans to anyone convicted of a drug felony unless she completes an approved drug rehabilitation program.

Although certain peripheral features of these penalties differ from ordinary punishments, they are typically described as A punishments.@ Rightly so, I think. If these can be defended at all, they are best construed as forms of punishment. Some advocates might defend these penalties as either a consequence of criminal behavior or a form of risk preventionC two accounts of collateral consequences I discuss later. However, that later discussion should show why those approaches could not justify the denial of these social benefits to drug felons.

If they are justified as punishments, then they must be defended using one (or some mixture) of the standard justifications (and their variants). That is, these penalties would have to be (1) punitive measures employed to deter potential felons, (2) penal treatment that felons deserve for having violated the law; (3) forms of communication to the criminal, designed to make prisoners repent: to condemn their own misdeeds and to seek reunion with the community; (4) means of rehabilitating criminals, or (5) a form of restitution. Let me first canvas the last four rationales since I think they are the least plausible justifications. My discussion of retribution will be somewhat extended since it will isolate important problems with collateral consequences generally. I end the section by examining the most plausible alternative: deterrence theory.


Some might claim that felons deserve to be deprived of these social benefits. Retributive theorists often explain this idea of desert using the metaphor of the Social Contract. Once someone has broken a contract, e.g., by violating the law, the other party (the state) is freed from that contract: it no longer needs to grant to the criminal (all?) the rights it grants to ordinary citizens. Put differently, by their actions felons have forfeited (some of?) their rights. Typically that means they should be physically removed from society (incarcerated). Some retributivists might claim that drug felons also deserve to lose access to the polity= s welfare, housing, and education benefits.

There are three problems with a retributive defense of these penalties. One, although we talk about felons forfeiting their rights, critics have argued that punishment cannot be explained as a forfeiture of rights. More relevant to the current inquiry, even if felons forfeit some rights, they don= t forfeit them all. Even in the U.S., prisoners retain the rights not to suffer cruel and unusual punishment, limited rights to free religious expression and free speech, and the right to due process. Since prisoners maintain some rights, retributivist advocates of collateral consequences must explain why drug felons deserve to forfeit these rights to social benefits. It seems unlikely that defenders could provide such an argument, especially since we have good reason to think most drug felonies are, as a class, less serious than other feloniesC -felonies not treated so harshly.

Why do I say they are less serious? While there is nothing to be said for murder; there is something to be said for allowing people to act as they wish (e.g., use drugs) even if it is harmful to them. Those arguments show that making drugs illegal is, even if ultimately justified, undesirable inasmuch as they are (at least partly) paternalistic. In contrast, laws against murder, robbery, rape, etc. are not paternalistic at all; these acts are criminal because they harm others, full stop. This explains why, drug crimes are generally less serious crimes, and, as such, drug felons deserve less severe punishment than violent criminals. Yet our current policies punish drug offenders more harshly.

Of course there are drug crimes and there are drug crimes. Someone= s puffing weed in her home is, on any account, less serious than selling crack to elementary school kids. But, what makes the latter serious is that it harms highly vulnerable people. That, and not the fact that drugs are involved, is what makes them merit severe punishment. Drugs are merely the tool whereby children are harmed.

Two, the previous argument suggests why felons in prison do not deserve the loss of rights available to other felons. It is even more obvious that they do not deserve the loss of these basic social benefits after release or even permanently. Defenders might contend that this would not be a worry if such punishments became a formal part of the sentencing processC for then the sentence would never end even if incarceration did. Put differently, these civil penalties would no longer be post punishment. I fail to see, though, how this explains why drug felons should lose access to these benefits long after they are released from prison.

Three, these collateral consequences clash with a key element of retributive theory (and, indeed all viable theories of punishment), namely, that the punishment should be proportional to the crime. As Duff puts it, @ the imposition of disproportionately harsh punishments cannot be justified.@

According to a robust account of proportionality punishment should A fit@ the crime. The problem is spelling out what this means. So I shall leave this richer notion asideC except to rebut one objection I consider later. I will focus instead on the uncontroversial A relative@ sense of proportionality which includes three elements:

C Rank-ordering. More serious crimes should be treated more seriously.

C Spacing. Much more serious crimes should be punished much more seriously.

Using these elements we can specify three ways in which the current uses of collateral consequences are disproportionate.

  1. They violate the principle of rank ordering: these penalties are arguably stronger than any drug felon deserves. Any drug felony is less serious than first degree murder, yet the murderer will not, upon release, be denied access to these social benefits. It is true that some drug felonies are more serious than others. However, as I argued earlier, thoseC like selling cocaine to childrenC which are especially bad are bad because they harm members of a particularly vulnerable group. Yet that is not the reason these felons lose access to social benefits.
  2. They violate the principles of spacing and parity: the recreational user receives only a slightly smaller penalty than a dealer.
  3. They violate the principles of rank ordering and spacing: often punishment for a second conviction is relatively less stringent than for a first conviction, since one can permanently lose these rights only once. So if the latter convictions are to be treated more seriouslyC and we generally assume they should beC then repeat felons should face stronger penalties, not lesser ones.

There are two related objections to this argument. First, an objector might claim that I am focusing on a single feature of the offender= s punishment rather than the array of penalties they faceC their A penalty package.@ I thereby overlook a larger perspective from which the penalties for drug felons are not disproportionate. Although drug felons may lose social benefits, rapists and murders spend more time in prison. Hence, violent criminals= A penalty package@ is greater than that of drug felons. Despite first appearances, these penalties are not disproportionate.

However, someone raising this objections faces a dilemma: either these penalties are disproportionate, or, prior to their passage, penalties for all non-drug felons were disproportionate. For when these collateral consequences were added for drug felons, there was no systematic increase in penalties for other felons. So unless this objector wants to argue that prior to the advent of these penalties, non-drug felons were treated unduly harshly relative to drug felonsC an implausible claim on its faceC then my contention stands: these collateral consequences are disproportionate.

Second, an objector might avoid the charge that these penalties violate the principle of rank ordering by increasing penalties for other felons. To rebut this objection I must now call upon the more robust notion of proportionality. Such a loss of social benefits, on top of often lengthy incarceration, is greater than what these felons deserve. Making penalties for other felons still harsher will not make this problem vanish. After all, punishments in the U.S. are already stiffer than those in other developed countries. Thus, the burden of proof rests on those who want to boost penalties. They would have to explain why the U.S. alone has the appropriate (robust) sense of proportionality, while the rest of the developed world systematically divvies up unduly lenient punishments.

If these arguments are correct, then since these penalties are clearly not proportional, those defending collateral consequences retributively must abandon either these collateral consequences or the principle of proportionality. They cannot plausibly abandon the latter since it is an essential feature of their theory. Retributivists cannot defend these collateral consequences.


The communicative theorist claims that A criminal punishment should . . . communicate to offenders the censure they deserve for their crimes, and thus to bring them to repent their crimes, to reform themselves, and to reconcile themselves with those they have wronged.@ These aims are accomplished, at least in part, by letting the criminal know that her behavior has separated her from the rest of society. Punishments that reasonably communicates that separation can be justified; any punishment more severe would be unwarranted. Since we have no reason to think drug felons will grasp the separation if and only if we deny them access to these benefits long after incarceration, then these penalties are excessively severe.

Still some post sentence collateral consequences might be a suitable way of communicating separation to one who committed an especially heinous crime (perhaps, treason for monetary gain). However, this would be the exception, not the rule. Such penalties would be appropriate only for the most serious crimes; they would be aimed at specific felons not a large class of felons. We have no reason to think these penalties are appropriate for manyC let alone allC drug felons. After all, they are more easily reintegrated into society than most felons; they are far less likely to return to crime. Yet for some inexplicable reason we burden them with penalties other felons do not faceC penalties which will make their rejoining society more difficult.

Put differently, such penalties clash with core aims of communicative theories. The very considerations that make temporary denial of some rights plausible symbols of a felon= s separation from society make regaining those rights plausible symbols of her reintegration into that society. Conversely, barring her from access to these social benefits for years after releaseC and sometimes permanentlyC tells her that despite her having completed the sentence, she is, and will continue to be, separated from society. There is nothing she can do to repair this rift.

This is unfortunate even if one is not a communicative theorist. Reintegrating former felons should be a significant aim of a penal system, especially in the United States which has the largest percentage of its population (2.2%) under criminal supervision of any country in the world. At the end of 1999, there were 59 million criminal records in this country (some of these will be for the same felon in two or more states). The number of people with criminal records well exceeds 10% of the country= s population. The U.S. can ill afford to maintain such large numbers of its people with records and under penal supervision, to have so many people who cannot contribute. Yet the U.S., which most needs to reintegrate felons, burdens drug felons with these penalties, and thereby diminishes their chances for reintegration. Even if denying such benefits could be abstractly justified, it would be demonstrably imprudent.

Finally, communicative theorists endorse the principle of proportionality. As we saw in the previous section, these collateral consequences are disproportionate in at least three different senses. Together these factors explain why communicative theories cannot plausibly countenance permanently barring all drug felons from access to welfare, food stamps, housing, and education.


For similar reasons, rehabilitative theories could not justify denying these social benefits to drug felons. According to this theory, the aim of punishment is to rehabilitate criminalsC to prepare them for life as productive citizens. Those aims cannot be advancedC and are instead, substantially hinderedC if former drug users are denied access to certain significant benefits available to other citizens. The loss of these benefits is not a serious problem for felons who are well-educated and financially well-off. But, of course, most felons are poor, undereducated, and ill-trained. Nearly half of the people incarcerated in state prisons didn= t finish high school and less than 11% of them pursued any education beyond high school. In contrast, nearly three-fourths of all Americans graduated from high school, and more than one-fourth completed college.

That explains why former felons are typically ill-trained for many jobs. These civil penalties make it likely that they will not gain needed training. Other collateral consequences further restrict their access to jobs (e.g., jobs with the state and federal government). And twenty-seven states revoke the driving licenses of some or all former drug felons; that makes it still more difficult for them to find jobs. Finally, new government regulations prohibit the Legal Services Corporation from using public or private funds to help former felons legally defend their interests, seek to regain their access to lost services, combat discrimination in employment, etc. These forces conspire to ensure that former felons= chance of finding gainful employment is slim. It is unsurprising that many think their only viable option is to return to crime. These penalties promote recidivism, not rehabilitation.


This is the least known theory of A punishment,@ in large part because it rejects the very idea that we should respond to crime with punishment. Punishment, advocates claim, does not work, is horribly costly, and completely ignores the victimC except as an afterthought. They claim we should jettison punishment as we know it and seek, instead, to provide restitution to victims. It is clear this theory could not countenance these collateral consequences. These penalties are purely punitive.

In sum, three theories of punishment (communication, rehabilitation, and restitution) are wholly incompatible with these (and arguably most) collateral consequences. And, although retributivism might justify some, it could not justify those under discussion. There are no good reasons why drug felonsC and only drug felonsC deserve to lose these social benefits. Thus, if these collateral consequences are to be justified as forms of punishment, they are best seen as forms of deterrence. THAT is exactly the argument most often given for them.


These collateral consequences were adopted in the mid 1990s as one instantiation of the A get tough@ approach to crime that emerged in the 1970s. The most common argument for these penalties was that they would deter people from committing drug crimes. Senator Phil Gramm, in proposing laws to permanently bar drug felons from receiving most federal benefits, claimed A if we are serious about our drug laws, we ought not to give people welfare benefits who are violating our nation= s drug laws.@ The rationale is clear: if we have what we take to be a serious crime problem and current penal measures are not controlling it, then we must look for alternative measures. In today= s political climate that typically means more severe punishments.

Why, though, were drug offenses singled out for these penalties? Clearly Senator Gramm thinks all drug felonies are supremely serious since he did not advocate similar (let alone harsher) penalties for rapists, murderers, armed robbersC and certainly not for embezzlers, price fixers, or other white collar criminals. Apparently Congress must have agreed with him since they passed his amendment after two minutes of discussion.

As we saw earlier, however, we have no reason to think drug felonies are especially serious crimes. Indeed, we have good reason to think them less serious than many felonies not subject to special collateral consequences. Nonetheless, even if they were especially serious, we should dispense these penalties only if (1) we have good reason to think that they will deter better than A mere@ incarceration, and (2) we are confident there are no compelling reasons to think they are unjust. I do not see how defenders can make good on either claim, let alone both.

Do They Deter?

Do we have reason to think these policies are successful deterrents? Immediately we can see one reason for thinking they are not. These provisions are not part of the criminal law, but are included in the Social Security and Education codes. Such consequences are rarely a formal part of sentencing. U.S. courts have even held that attorneys for those accused needn= t inform their clients about them. Hence, many (potential) drug users may be unaware of these penalties. People cannot be deterred by threats of punishment of which they are unaware.

Still, after sufficient time has passed, more people will inevitably learn about them. People will hear about these penalties from family, friends, or acquaintances who were denied social benefits after release from prison. However, since these policies have been around for less than a decade, it is likely that few people know about them. Therefore, it is unlikely that they now deter.

We also have empirical evidence that these penalties do not deter potential drug crimes. If they did deter, we would expect that drug use and the number of people arrested for drug felonies would have declined since their passage. It is unclear that drug use has declined; it is certain that drug arrests have not. The U.S. Department of Health and Human Services= annual report on drug use suggests that the rate of drug use has increased since the passage of these laws. Admittedly this claim must be qualified since in 2002 the department began measuring use differently; therefore, there is no straightforward way to measure trends. Nonetheless, they did find that new drug use is on the rise among young people, and noted that A in the past, increases and decreases in incidence usually have been followed by corresponding changes in the prevalence of use . . . @ Since that report, a study by the University of Michigan rejects this inference: it claims drug use has declined slightly. Slightly, but not much.

What is clear is that drug arrests have not declined. In fact, the percentage of people in the criminal justice system (both incarcerated and on parole) convinced of a drug felony has increased, with the largest increase being those convicted of simple possession. Together these considerations suggest that these penalties are not successful deterrents.

Perhaps, though, we are moving too fast. Perhaps drug felonies would have increased even more without these added measures. However, that is not a plausible supposition. Violent crime decreased by 50% since 1993, and property crime decreased by more than 60% since 1975. It is highly unlikely that drug crime alone would have increased while other crimes decreased dramatically, indeed, declined far more than any decline in drug use suggested by the Michigan study mentioned earlier. If nothing else, these considerations place a heavy evidentiary burden on those who advocate such policies on the grounds that they deter drug use. It is difficult to know how that burden could be met.

Finally, deterrence theory should also consider consequences other than a short term decline in crime. Some punishments might deter potential criminals in the short run, but do so in a way that lessens the chance that released felons will become law abiding and productive citizens. If so, then such punishments increase the incidence of crime long-term, thereby heightening human suffering. Given earlier arguments, we see why these penalties are more likely to increase crime.

Are They Just?

An acceptable deterrence theory must be constrained by considerations of justice. Not just anything which might deter is morally permissible. Lopping of the legs of jaywalkers might be an efficient deterrent, but it is neither appropriate nor proportionate. Penal means should be both. We should not have inappropriate penalties, even if they are not excessive. We should not forbid criminals from eating with a spoon, require them to walk with a cane, or require left-handed prisoners to write with their right hands. Nor should penalties be excessive: we should not deny prisoners freedom of religious expression or visitors. Barring extraordinary circumstances, such means would be unacceptable. Permanent loss of important social benefits is likewise extreme, and, is neither appropriate nor proportional. Sans compelling arguments that it is an effective and essential deterrent, then even advocates of deterrence theory should reject these penalties. And we have no such arguments, certainly not any compelling ones.


If permanent disenfranchisement can be justified, it will not be justified as a form of punishment. First, just as retributivism, communication, rehabilitation, and restitution cannot justify denying social benefits to drug felons, they cannot justify permanently disenfranchising all or even most felons. Permanent disenfranchisement is not deserved, is disproportionate, works against rehabilitative ends, is an inappropriate form of communication, and does not offer restitution to victims. Perhaps, though, disenfranchisement might be justified as a deterrent. Indeed, some defenders of permanent disenfranchisement claim the threat of disenfranchisement would deter. However, the threat of disenfranchisement is a less plausible deterrent than is the threat of the loss of social benefits. Yet, we saw earlier, despite first appearances, these threats are not effective deterrents.

We can say more. To the extent that punishments can deter directly, they do so by changing potential criminals= calculations of costs and benefits. However, if in making that calculation, someone is not deterred by the threat of five years in prison, the additional threat of losing the vote is unlikely to deter her, unless she especially treasures voting. However, we have no reason to think criminals (or aspiring criminals) are more likely to prize voting than are non-felons. Rather, we have reason to think they care about voting less. They are more likely to be poor and uneducated, and less likely to believe that they benefit from the state. Since non-felons in the U.S. are less likely to vote than are citizens of most democratic countries, we can safely infer that U.S. felons won= t especially prize voting. Threatening to take away the vote will not be an effective deterrent.

Perhaps, though, denying former felons the franchise might deter them indirectly. If the state shows, by instituting strict collateral consequences, that it finds certainly criminal behavior unacceptable, its citizens could be deterred subconsciously. The problem, however, is that it is unclear how threatening disenfranchisement might plausibly have this effect. We have four reasons for thinking it would not. One, for the reasons already given, criminals are less likely to care about the state at all, and therefore, less likely to be moved by what the state wants. Second, since criminals assume they won= t get caught, an added threat is rarely enough to stop them from committing the crime. Third, this subconscious mechanism works more plausibly if it is used to convey to potential criminals that some small number of crimes are especially heinous. It is less likely to be effective if applied to most or all felonies equally. Fourth, collateral consequences cannot deter indirectly or directly unless most people are aware of them. Since many citizens do not know their nature, scope, or duration, they cannot deter.

These flaws lead me to think that the most plausible strategy for defending permanent disenfranchisement is to see it not as a punishment at all, but as a consequence of criminal behavior. Arguments for this way of conceiving collateral consequencesC like arguments for retributivismC often employ the metaphor of the Social Contract. The criminal, by violating the law, breaks the social contract and therefore is no longer morally part of the community. Consider the following example. A school has an honor code strictly forbidding cheating. The honor code is not just a set of rules that give the school its cachet; it is a statement about what it is to be a member of that academic community. When a student cheats, she is expelled from the university, not as a punishment, but as a consequence of her behavior. The expulsion simply acknowledges that the student, by her action, has separated herself from the community of which she was once a member. Of course such expulsion, especially if it is seen as a likely consequence of cheating, may have the result of deterring potential cheaters. But that does not mean it is a punishment.

As a way of understanding why cheaters may be expelled from school, this is sensible. Perhaps, in the end, the policy of automatic expulsion is counterproductive, unfair, or in some other way inappropriate. Nonetheless, it is plausibly construed as a consequence of the student= s behavior. Someone might argue analogously that incarcerated and former felons face a similar range of collateral consequences: by their criminal behavior they have separated themselves from the community. The loss of the vote is one such consequence.

This approach could be explained by exploring the relationship between rights and responsibilities. John Deigh argues:

To retain the right [which implies responsibilities] and so to continue enjoying the freedom it secures, one must remain competent and fit to perform those responsibilities. Where continual failure to perform them is evidence of incompetence or unfitness, as it is in many cases, such failure makes one liable to forfeit the right, a forfeiture that results in the loss of the freedom or at least the guaranty of its continued enjoyment that the right represents.

Perhaps the most obvious example is the impeachment of a public official. By abusing the powers of the office, the official shows herself unfit for that office. This approach might also be further applied to the right to vote.

Consider too how, typically a citizen in a democracy can forfeit his right to vote. The citizen, by virtue of this right, has a kind of legislative authority in that he exercises the right either by participating directly in the enactment of laws that govern the members of his society or by participating indirectly in their enactment through election of those members who constitute the society= s legislature. Consequently, a serious breach of the lawC specifically a criminal offense because it shows that the citizen is unwilling to abide by the laws in the enactment of which he can participateC disqualifies him from being a legislator or elector. It shows him to be unfit for assuming the responsibilities for either office, and for this reason, he forfeits its essential right.

The voter= s right is taken away not as punishment, but because her felonious behavior demonstrates that she is unfit to carry out the responsibilities of her office (as a voter) and therefore (temporarily) loses the right associated with those responsibilities. Although Deigh primarily suggests that temporary disenfranchisement might be justified (although even here he expresses misgivings), someone could extend his arguments to try to justify permanent disenfranchisement.

Although we might imagine an isolated case in which permanent disenfranchisement is arguably justified (say, for an official who rigged elections for money), permanently disenfranchising all felons is not. The right to vote is relevantly different from the right to be President, a judge, or even a university student. These latter A rights@ are more properly understood as privileges, or rights that depend upon demonstrated competence. In contrast, the right to vote is fundamental. It is A the essence of a democratic society@ and A any restrictions on that right strike at the heart of representative government.@ That right cannot be denied without showing a compelling state interest. Defenders of disenfranchisement have not shown such an interest.

Moreover, such a penalty is disproportionate in all three senses mentioned earlier: (1) it is excessive for most crimes, (2) it treats all felonsC no matter how different their crimeC the same, and (3) repeat offenders are punished more lightly than first time offenders (since they can be permanently disenfranchised only once). Admittedly, the requirements of proportionality are not as strong when penalties are conceived as consequences of criminal behavior rather than as punishments proper. Nonetheless, something like that principle is still at workC as I explain in more detail in the next section.

Finally even if someone could, following Deigh, use this argument to justify temporarily disenfranchising some felons, it could not plausibly justify permanently disenfranchising all felons, and certainly could not justify many, let alone all, the collateral consequences now in place in the U.S.

Prohibiting Some Felons from Practicing Certain Professions

A dentist convicted of having sex with his patients under anesthesia might be forever forbidden from practicing dentistry, not as a form of punishment, but to prevent him from harming future patients. Someone convicted of vehicular homicide might be forever forbidden from driving, not to punish her, but to protect drivers, passengers, and pedestrians from her reckless driving. These actions are not punishments but A civil risk-prevention measures.@ While punishment is determined by the person= s past behavior, A disqualification as civil measures . . . are addressed to the offender= s future behavior . . . @

These civil disqualifications are justified as A part of wider powers to regulate the profession in the public interests.@ The state has the responsibility to protect the public from harm. One way it meets this responsibility is by limiting membership in professions (medicine, law, engineering, etc.) to those it deems competent. The state can legitimately deny licenses to anyone it plausibly deems unsuitable to the task.

The idea that some felons might pose special risks to others after their release is highly plausible. If the nature of the harm a former felon poses is significant and probable, then we have someC and perhaps compellingC reason to selectively restrict their activities. It seems plausible, for instance, to forbid a surgeon who killed his patients to harvest their organs for sale from ever practicing surgery again. Furthermore, if it is true, as many people believe, that pedophiles are never cured, then we won= t want released pedophiles teaching elementary school.

Such preventive measures are plausible (even if not ultimately defensible) if we have firm evidence that these people pose significant risks to the public. We now have such laws, e.g., a person cannot work for the federal government if she used federal monies to lobby Congress. More generally, courts have the authority to restrict a felon A from engaging in a specified occupation, business, or profession bearing a reasonably direct relationship to the conduct constituting the offense.@

Nonetheless, even advocates of these policies insist that we must be careful. In moments of cool reflection we recognize our tendency to make unfounded predictions about the risks former felons pose. Norval Morris admits that our capacity to predict dangerousness is weak, no better than one in three. His conclusions are duplicated in study after study.

The Flood Committee reported in 1981 . . . that no study had revealed a prediction method which could successfully identify one actual offender for every two predicted to reoffend, and that most studies could do no better than one actual reoffending person for every three persons predicted to be dangerous. The findings of a large-scale study conducted by the Home Office were even less impressive. . . . Numerous other studies confirm this problem of overprediction.

Or, to take a specific example mentioned earlier, the common view that pedophiles are never cured is far from ironclad. According to the Department of Justice, people convicted of sex crimes as a class are, short of those convicted of homicide, least likely to reoffend. Perhaps this is one instance of our proclivity to simply assume former felons are dangerous. Too often we do not even pretend to make informed judgments of risk. We restrict felons simply to make ourselves feel safe. For instance, twenty-nine A states have no standards governing the relevance of conviction@ for determining whether a former felon poses a sufficient risk to deny him an occupational license.

Despite these worries, I think the use of collateral consequences as risk prevention measures are eminently plausible in select cases. However, this admission does not advance the widespread use of collateral consequences. Indeed, it hinders it. Such penalties are plausible only if selectiveC when they are imposed against those posing specific and demonstrable risks. This approach would not justify forbidding pedophiles or lascivious dentists from being lawyers or accountants. More generally, since most former felons do not pose such risks, these arguments cannot justify automatic and sweeping collateral consequences against all felons. Perhaps for these reasons, defenders of collateral consequences rarely defend them in this way.

General Problems with Collateral Consequences

I have not canvassed all consequences, let alone all possible justifications for them. However, the three cases I examined represent the range of collateral consequences and embody the three standard justifications given for them. Therefore, I think we can generalize tentatively from them. We can identify some plausible arguments for select collateral consequences, and we can specify an array of problems facing most collateral consequences.

The use of collateral consequences as risk prevention measures are plausible when the harm we are preventing is serious and the risk the former felons pose is significant. Other collateral consequences may be justified as consequences of criminal behavior. But, these uses would be plausible only if we can see how and why the criminal= s behavior disqualifies her from having certain rights, as it does, say, in cases of impeachment. This account would, at most, justify a hint of the collateral consequences former felons now face in the United States. Conversely, the use of collateral consequences as a form of punishment would rarely, if ever, be justified.

The former arguments explain why the extensive use of collateral consequences is unjustified. They fall prey to five objections. One, as currently practiced in the U.S., they violate the requirement of publicity. Two, even if the practice were modified so that they fulfilled this requirement, their sweeping use would be neither justified nor likely to be supported by the public. Three, although on the last two proffered justifications, collateral consequences need not be strictly proportional, they must be proportional to some degree. They are not. Four, the arguments given for these practices are confused and conflicting. Five, most of these practices have significant racist effects in the United States; that alone gives us reason to jettison them.

Violates the condition of publicity

Collateral consequences as practiced in the U.S. violate the principle of publicity central to a justified legal system. At first glance it might appear the principle of publicity is irrelevant to the current debateC that the principle simply forbids punishing someone for an act that she could not plausibly know was against the law. However, I think the import of the principle is easily extended to the current debate. To explain why, recall Fuller= s classic statement of the principle:

Certainly there can be no rational ground for asserting that a man can have a moral obligation to obey a legal rule that does not exist, or is kept secret from him, or that came into existence only after he acted, or was unintelligible. . . . Government says to the citizen in effect, A These are the rules we expect you to follow. If you follow them, you will have our assurance that they are the rules that will be applied to your conduct.

Laws should be public so that citizens can guide their behavior by their knowledge of it. The rationale for this requirement likewise demands that penalties be public: ignorance of the penalties for violating a law is akin to nonculpable ignorance of the law. It is not just that people need to know about possible penalties before deciding if violating the law is worth risking the punishmentC which is true enough. More relevantly, the severity of punishment is the main way the state conveys to its citizens the relative seriousness of an offense. If we set a $5 fine for parking at an expired meter, we A tell@ prospective A lawbreakers@ that this offense is not very serious. If a citizen then parks at a meter without paying, only to discover that the state not only fines her $5, but, as a collateral consequence, confiscates her auto, she would rightly be upset. Not only would this collateral consequence be disproportionate to the offense, the state would have violated the requirement of publicity.

Since most people are unaware of the existenceC and certainly the range and natureC of collateral consequences, then the state has violated the principle of publicity. People in the United State are not ignorant of these penalties because they fail to pay attention to the law. Citizens have to work to discover them. They are a hodgepodge of regulations, buried in footnotes and added as amendments to non-criminal legislation, are administered by diverse state and federal agencies, and are rarely codified in the criminal law. Even many legislators are unfamiliar with them since amendments to major bills often receive skimpy debate. Phil Gramm= s amendment denying drug felons access to social benefits was debated for only two minutes.

Of course legislatures could insure that collateral consequences did not violate the requirement of publicity. However, I doubt that many advocates of the first two uses of collateral consequences want this since I doubt that these measures would be supported by the majority of citizens, legislators, or courts if most people knew about them; I will shortly explain why.

Finally, unless the condition of publicity is satisfied, we cannot obtain evidence of these penalties= putative benefits. Those who defend these practices, either on grounds of deterrence or to protect the public from risks, must provide evidence of their value. Otherwise, the use of these draconian measures is unjustified.

Disproportionate Response

A central element of every defensible theory of punishment is that any punishment must be proportional in some sense. Although what this means differs from theory to theory, minimally it means that punishments should be constrained by considerations of parity, spacing, and rank ordering. Hence, most, and perhaps all, collateral consequences defended as forms of punishment will run afoul of some (or multiple) elements of this principle. Some, like the denial of social benefits to drug felons, are so obviously disproportionate that no standard theory of punishment could countenance them. For similar reasons, permanent disenfranchisement of all felons is also unacceptably disproportionate, at least if this is thought to be a form of punishment.

Suppose, though, such penalties are not punishments but consequences of criminal behavior or risk prevention measures. If so, then they need not be proportional in the strict ways that punishments must be. Nonetheless, such measures are constrained by a A non-desert notion of > proportionality= , involving the idea of suitability to means to ends: drastic interventions in people= s lives ought not to be resorted to in order to achieve relative unimportant goals.@ Put differently, this weaker sense of proportionality forbids the imposition of collateral consequences which substantially violate considerations of parity, spacing, and rank ordering. Thus, although less stringent than the constraints on ordinary punishment, they will still prohibit placing much greater civil penalties on an entire class of felons whose crimes are clearly less serious than those of other felons.

Thus, even on this weaker sense of proportionality, the denial of social benefits to all drug felons is clearly disproportionate. Few drug felonies are as serious as many other felonies which do not receive such penalties. Furthermore, although we might be able to justify temporarily disenfranchising felons as a consequence of criminal behavior, permanent disenfranchisement would be unacceptably disproportionate, even in this attenuated sense. Heaping this penalty on all felons treats wildly disparate crimes with the same penalty.

The Public Finds Extensive Collateral Consequences Unacceptable

If the public were aware of the scope and seriousness of most collateral consequences, they would not support them. Although we should be cautious in making such predictions, this one is eminently plausible. In 1998, the Sentencing Project issued a report attacking the practice of disenfranchising felons. At that time only a few citizens were aware that most states disenfranchised incarcerated felons and that 13 states permanently disenfranchised felons. Five years later, eight states had weakened their laws disenfranchising current and former felons. The upshot was dramatic. In just five of these states, A 471,000 persons gained access to the ballot box.@ Now that the public is aware of the practice of disenfranchisement, A 80% of them support restoration of voting rights for ex-felons who have completed their sentence . . .@ Since the publication of this second report, two other states also ceased permanently disenfranchising felons. All these changes occurred quickly and at a time when the public tended to be unsympathetic to criminals.

From this I infer that although the U.S. public might support some collateral consequencesC e.g., those requiring the names of released sex offenders to be made publicC most would not support the range of consequences now the norm in the United States. I cannot imagine that most people would support denying released offenders access to student loans and many jobs. They would conclude that those policies increase the chance that former felons will return to crime, and are therefore, self-defeating. This is not a surprising conclusion since no other developed country has collateral consequences even remotely resembling those in place in the United State. When civilized people know about and reflect on these penalties, they find few of them justifiable or tolerable.

Confused and Conflicting Rationales

Of the three major justifications, the least defensible one treats collateral consequences as a form of punishment. Yet that is the most commonly proffered justification in public debate. It seemed natural to Senator Gramm to defend his proposal as a form of deterrence, and, given the brief debate, his arguments apparently did not strike other Senators as wrong, let along wrong-headed. However, as my earlier arguments showed, the claim that these collateral consequences are justified punishments cannot be sustained. If these penalties are justified, it is by some means other than that typically given by supporters.

Conversely, the most plausible justifications for some collateral consequences, are infrequently used. Some collateral consequences might be plausibly justified as consequences of criminal behavior. Others may be plausibly justified as risk prevention measures. But these latter justifications are plausible only for particular criminals and select crimes. Together these accounts could justify only a minuscule number of collateral consequences. They ill serve as justifications for collateral consequences now imposed on former felons in the United States.

Racist Effects ofC If not Motive forC Collateral Consequences

Collateral consequences affect more African Americans than whites, and they affect them more profoundly. African Americans are imprisoned at ten times the rate of whites. Thirty-two percent of black males will, at some point in their lives, be incarcerated. Since not all felons are incarcerated, even more blacks are affected by these civil penalties. The denial of social benefits for drug felons (the highest portion of incarcerated blacks) seriously undercuts their chance for full membership in American society. The loss of the vote also disproportionately squelches their political voice. In 1998 13 percent of black makes could not vote, and in seven states, one in four Afro-American men is permanently disenfranchised. They had no say inC and thus in some sense no stake inC civil society.

Many disenfranchisement laws were openly racist: they were passed between 1890 and 1910 specifically A to increase the effect. . . . on black citizens.@ Several states passed legislation with the express intent of limiting the number of black voters, and in Mississippi that purpose was explicitly approved by the state= s supreme court. This practice is most common in the states of the old Confederacy. From this historical evidence we cannot automatically infer that there is no non-racist justification for the practice. Nonetheless, since there are no compelling defense for automatic permanent disenfranchisement of all felons, and the history of the practice is morally tainted, we should conclude that it is, at core, racist. Especially when defenders of these practices only response to the worries about the racial impact of collateral consequences is that these penalties wouldn= t have such an impact if Afro-Americans weren= t so prone to crime. What Clegg and others ignore is that the same factors that make the effects of collateral punishments fall disproportionately on the shoulders of African Americans are the same factors that make them more likely to resort to crimeC and especially prone to be convicted of crime.

However, we needn= t rest objections to these policies on their being motivated by racism. It is enough that their effects fall disproportionately on African Americans. The 1982 Amendments to the Voting Rights Act specify that a law can be discriminatory even if we have no evidence that it A was designed and maintained for a discriminatory purpose.@ All that matters is that its effect, as determined by the A totality of circumstances@ are racist. Policies that sustain the unjust treatment of blacksC by ignoring the legacy and ongoing effects of discriminationC are likewise racist. African-Americans have been systematically deprived of income, equal education, and wealth for centuries. The effects of this discrimination did not die with the end of slavery or Jim Crow. This history does not make African-Americans resort to crime, but it does substantially limit their options. Since they have fewer options, it is not surprising that more poor, uneducated, and black people more often resort to crime.

Moreover, it is not just that blacks commit crimes in higher numbers, they are more likely than whites to be caught, convicted, and incarcerated even when committing the A same@ crimes. Although African Americans (which make up 12% of the U.S. population) are no more likely than whites to use drugs, they represent 35 percent of those persons arrested for drug crimes, and 53 percent of drug convictions. Given our history of racism, the effects of the War on Drugs and the policies of excluding felons from social benefits were easy to foresee. They are part of racism= s continuing legacy.

These factors converge to give us compelling reason to find the racial effects of the current widespread use of collateral consequences politically and socially intolerable. Any policy that so substantially sets back their interests should be rejected, unless there is an overwhelming argument in support of that practice. Except for select cases, there is no compelling argument.


The arguments for collateral consequences range from the wildly implausible through the mildly implausible, to the seemingly plausible. However, those that are plausible would, even if wholly defensible, justify only a limited number of collateral consequences, and then for only select felons. The current practice cannot be justifiably sustained. If we conclude, upon careful examination, that we wish to retain some collateral consequences for select felons, that is sensible. But if we do, we must make these penalties a formal part of the criminal law, and incorporate them into the sentencing procedure as they do in Germany. For as long as these collateral consequences stay hidden, it is all too easy for them to continue without good reason.


Thanks to Antony Duff, Eva LaFollette, Nora Demleitner, Pat Kain, Chris Pincock, and other members of the Philosophy Department at Purdue University for thoughtful comments of on earlier incarnations of this paper.