POSNER, Circuit Judge.

Robert Presley, the defendant in a trial lasting eight days, was convicted by a jury of heroin and related gun offenses and also of being a felon in possession of a gun. See 18 U.S.C. §§ 922(g)(1), 924(c)(1)(A)(i); 21 U.S.C. § 846. We affirmed the convictions and sentences of two of his codefendants in United States v. Cooper, 767 F.3d 721 (7th Cir.2014), an opinion that contains a full description of the crimes of the defendants, including Presley. The judge sentenced him to 440 months (36.67 years) in prison and imposed 17 conditions of supervised release. The appeal challenges the conviction and the prison sentence.

The principal challenge to the conviction concerns a fourth codefendant, Norman Breedlove, who testified against Presley pursuant to a plea agreement. Shortly after the trial, Presley, who was housed in a cell near Breedlove's cell, moved for a new trial on the ground that he had overheard Breedlove saying things that suggested that his testimony had been "false and perjurious." Shortly before Presley filed his motion, the district judge at the request of Breedlove's lawyer had ordered Breedlove examined to determine his mental competency. On the basis of the results of the examination the judge ruled that Breedlove was suffering "from a mental disease or defect rendering him ... unable to understand the nature and consequences of the proceedings against him or assist properly in his defense." The judge ordered him committed for "a reasonable period of time ... to determine whether there is a substantial probability that in the foreseeable future" he would "attain the capacity to permit the proceedings [against him, pursuant to his plea agreement] to go forward."

A guard at the jail in which Breedlove was held submitted an affidavit stating that Breedlove had admitted to the guard that he had testified falsely against Presley — explaining that he had made a plea deal with the government (requiring him to testify against Presley) because one of his heroin customers had died of an overdose and he was afraid of being charged with murder if he didn't do whatever the government wanted him to do. Presley argued that he was entitled to a new trial because Breedlove had either given perjured testimony or had not been competent to testify.

The judge denied Presley's motion for a new trial. He said that it was unclear whether Breedlove had been mentally incompetent at the time of the trial and whether if so this would have made him incapable of testifying truthfully, and that in any event any error in allowing Breedlove to testify had to have been harmless because of the overwhelming other evidence of Presley's guilt, evidence recounted in our previous opinion. The judge's rulings on the Breedlove matter were sound.

Presley also questions the amount of heroin that the judge determined the conspiracy to which Presley belonged to have been responsible for while he was a member: at least 1 kilogram a month. For purposes of determining Presley's sentence the judge assumed that he had been responsible in toto for only 1 kilogram — a safe assumption, amply supported by the evidence.

The only questionable feature of the judgment is the length of the sentence — almost 37 years, though it is within the applicable guidelines range because of Presley's very lengthy criminal history. Presley was 34 years old when sentenced and the Bureau of Prisons has calculated his release date as 28.5 years after sentencing. The Bureau calculates release dates by subtracting from the sentence (1) credit for the defendant's pretrial custody, a known period, and (2) the maximum possible credit that the defendant might earn for good behavior in prison, which is merely a possibility. If Presley earns the maximum possible good-time credit he'll be almost 64 years old when released. If he earns no good time he'll be almost 69. And after release he'll undergo five years of supervised release, which like parole is a form of custody because it imposes significant restrictions on the supervisee. Yet all that Presley's briefs say about the length of his sentence (besides the challenge to the drug quantity) is that "this amount of incarceration far exceeded a sentence under 18 U.S.C. § 3553(a) that was sufficient, but no greater than necessary to achieve the sentencing goals set forth therein." Without elaboration — and there is none — this perfunctory argument fails to establish a basis for resentencing.

Nevertheless we have an independent responsibility to make sure that the sentencing judge justified the imposition of a sentence of such length. The judge pointed out that Presley is a career offender, that he began his criminal career when he was 16, that he was a large-scale heroin dealer, and that he had committed disciplinary violations in previous incarcerations. What the judge failed to consider was the appropriateness of incarcerating Presley for so long that he would be elderly when released. Criminals, especially ones engaged in dangerous activities such as heroin dealing, tend to have what economists call a "high discount rate" — that is, they weight future consequences less heavily than a normal, sensible, law-abiding person would. John Bronsteen et al., "Happiness and Punishment," 76 U. Chi. L.Rev. 1037, 1060 n. 115 (2009); Yair Listokin, "Crime and (with a Lag) Punishment: The Implications of Discounting for Equitable Sentencing," 44 Am.Crim. L.Rev. 115, 124 (2007); Stephanos Bibas, "Plea Bargaining Outside the Shadow of Trial," 117 Harv. L.Rev. 2463, 2504-06 (2004). Just as $1000 to be received 30 years from today is worth less to a person than $1000 received today (at an annual discount rate of 10 percent its present value is only $57), so the prospect of being in prison at age 60 is less worrisome to a 30 year old than the prospect of being in prison today — and the higher his discount rate, the less worrisome the prospect. The length of a sentence therefore has less of a deterrent effect on such a person than the likelihood that he'll be caught, convicted, and imprisoned. A. Mitchell Polinsky & Steven Shavell, "On the Disutility and Discounting of Imprisonment and the Theory of Deterrence," 28 J. Legal Studies 1, 4-6 (1999); see also Paul H. Robinson & John M. Darley, "The Role of Deterrence in the Formulation of Criminal Law Rules: At Its Worst When Doing Its Best," 91 Geo. L.J. 949, 954-55 (2003). An increase in the length of a sentence may therefore add little additional deterrence, since every sentence increment is an increment in future, not present, punishment. Linda S. Beres & Thomas D. Griffith, "Habitual Offender Statutes and Criminal Deterrence," 34 Conn. L.Rev. 55, 62-65 (2001).

The sentencing judge in this case did not refer to any of this literature, and in any event gave no reason to think that imposing a 37-year sentence on Presley would have a greater deterrent effect on current or prospective heroin dealers than a 20-year or perhaps even a 10-year sentence, or that incapacitating him into his sixties is necessary to prevent his resuming his criminal activities at that advanced age. Sentencing judges need to consider the phenomenon of aging out of risky occupations. Violent crime, which can include trafficking in heroin, is generally a young man's game. Elderly people tend to be cautious, often indeed timid, and averse to physical danger. Violent crime is far less common among persons over 40, let alone over 60, than among younger persons. According to Bureau of Justice Statistics, "Prisoners Entering Federal Prison," www.bjs.gov/fjsrc/var.cfm?ttype=one_ variable & agency=BOP & db_type=Prisoners & saf=IN, only 1.18 percent of persons entering federal prisons in 2012 for drug crimes were in the 61 to 70 age group. That is another reason to doubt that very long sentences reduce violent crime significantly.

There needs finally to be considered the cost of imprisonment to the government, which is not trivial. The U.S. prison population is enormous by world standards — about 1 percent of the nation's entire population — and prisons are costly to operate because of their building materials (steel especially is very expensive) and large staffs. If the deterrent or incapacitative effect on criminal propensities fades sharply with time, the expenses incurred in the incarceration of elderly persons may be a social waste. The Bureau of Prisons reports that 17.5 percent of the federal prison population is over the age of 50, 2.7 percent between 61 and 65, and 2.2 percent older than 65. Federal Bureau of Prisons, Inmate Statistics: Inmate Age, www.bop.gov/about/statistics/statistics_ inmate_age.jsp.

"Aging prisoners," defined as prisoners 50 years old or older, cost the federal prison system about 8 percent more than younger prisoners, and these excess costs, mainly medical, rise with age. See Office of the Inspector General, U.S. Dept. of Justice, "The Impact of an Aging Inmate Population on the Federal Bureau of Prisons," May 2015, https://oig.justice.gov/ reports/2015/e1505.pdf. The Inspector General's study also finds that only 8 percent of inmates aged 60 to 64 who were released between 2006 and 2010 were rearrested for new crimes within three years, compared to 19 percent who were 50 to 54.

We are not suggesting that sentencing judges (or counsel, or the probation service) should conduct a cost-benefit analysis to determine how long a prison sentence to give. But the considerations that we've listed should be part of the knowledge base that judges, lawyers, and probation officers consult in deciding on the length of sentences to recommend or impose. There is no indication that these considerations received any attention in this case. We do not criticize the district judge and the lawyers and probation officers for the oversight; recognition of the downside of long sentences is recent and is just beginning to dawn on the correctional authorities and criminal lawyers. Neither the Justice Department nor the defendant's lawyer (or the probation service) evinced awareness in this case of the problem of the elderly prison inmate. Some judges, however, have drawn attention to the problem in their opinions. See, e.g., United States v. Johnson, 685 F.3d 660, 661-62 (7th Cir.2012); United States v. Bullion, 466 F.3d 574, 577 (7th Cir.2006); United States v. Howard, 773 F.3d 519, 532-33 (4th Cir.2014); United States v. Payton, 754 F.3d 375, 378-79 (6th Cir.2014); United States v. Craig, 703 F.3d 1001, 1002-04 (7th Cir.2012) (concurring opinion). It is time we all awakened to it.

There is much that federal sentencing judges are required to consider in deciding on a sentence to impose — maybe too much: the guidelines, the statutory sentencing factors, the statutory and regulatory provisions relating to conditions of supervised release, presentence reports, briefs and arguments of counsel, statements by defendants and others at sentencing hearings. But in thinking about the optimal sentence in relation to the problem of the elderly prisoner, probably the judge's primary focus should be on the traditional triad of sentencing considerations: incapacitation, which prevents the defendant from committing crimes (at least crimes against persons other than prison personnel and other prisoners) until he is released, general deterrence (the effect of the sentence in deterring other persons from committing crimes), and specific deterrence (its effect in deterring the defendant from committing crimes after he's released). A sentence long enough to keep the defendant in prison until he enters the age range at which the type of criminal activity in which he has engaged is rare should achieve the aims of incapacitation and specific deterrence, while lengthening the sentence is unlikely to increase general deterrence significantly if the persons engaged in the criminal activity for which the defendant is being sentenced have a high discount rate; for beyond a point reached by a not very long sentence, such persons tend not to react to increases in sentence length by abandoning their criminal careers.

There will be cases in which the heinousness of the defendant's crime or other factors argue compellingly for a sentence longer that needed to achieve the goals served by incapacitation and general and specific deterrence. But in run of the mine cases consideration of those goals is likely to argue for a relatively lenient sentence.

Nevertheless we are not disposed to reverse Presley's sentence, given that it is within the guidelines range and that the defendant has made only the most perfunctory argument (really an assertion rather than an argument) for reducing it. But we invite the district judge to consider resentencing the defendant in light of the concerns that we've expressed in this opinion about the elderly-prisoner problem. We are mindful that the judge may already have concluded that Presley's record requires incapacitation and specific deterrence well into his 60s. Apart from his many convictions, Presley told his probation officer that he'd been shot ten different times since he was 16 years old — and those were just the shootings when he was on the losing side of a fight. He was still using a gun for criminal purposes at age 32 and between the ages of 16 and 32 had held no job other than that of dealer in illegal drugs. The judge may have concluded not without reason that Presley is one of the outliers who will still be dangerous when released in his (early or late) 60s. But it is at least possible that were the judge to reexamine in light of this opinion the sentence that he imposed, he would shorten it. And he'll surely want to reexamine the conditions of supervised release that he imposed, in light of recent decisions of this court questioning a number of them. See United States v. Kappes, 782 F.3d 828 (7th Cir.2015), and cases cited there.

Although as mentioned we do not think the judge committed reversible error, we are ordering a limited remand to enable him to consider whether to resentence the defendant in light of the concerns expressed in this opinion.