Unfortunately, it is difficult to identify which prisoners are wrongly incarcerated, and it would take an enormous investment of professional expertise and money to produce that information. However, we could make valuable progress on this issue by offering appropriate incentives for attorneys to identify some of these wasteful incarcerations, thus saving public money and serving the ends of liberty.
The Scale and Cost of U.S. Incarceration
First, let us quantify the enormous public cost of our prison system. The rate of incarceration is profound: the United States has less than 5% of the world’s population, yet it houses 25% of the world’s prisoners. On average, each inmate costs taxpayers roughly $36,286 per year. When we multiply that figure by the 2.3 million state and federal inmates incarcerated (one in every 110 adults), we get a cost of $83.5 billion, which means the average taxpayer pays $260 each year for incarceration. If you include the nearly four million additional people under criminal justice supervision (such as parole), the cost grows further.
While a significant portion of that money is needed to keep crime rates low, every dollar that goes to criminal justice is one less dollar spent on education, infrastructure, or medical research, or returned to the people in the form of lower taxes. At a time of deficits, we should ask, should all of these 2.3 million inmates be locked up? The answer is no.
Wrongful and Unnecessary Imprisonments
Until now, much of the debate on reducing incarceration has focused on changing the law to reduce the severity of prison sentences, especially for non-violent drug crimes. These efforts are valuable. However, even under the current law, there are thousands of people incarcerated in both state and federal prisons who should be released; they are actually innocent or have overserved their sentences. Without resolving the larger questions about criminal justice policy, perhaps everyone can agree that these incarcerations are a waste of taxpayer money. Even a government with no commitment to liberty would want to minimize such wasteful spending.
Incarceration waste exists because our pretrial and trial procedures do not perfectly sort those who should be imprisoned from those who should not. Although it is difficult to estimate the rate of wrongful conviction, several scholars have offered rigorous and conservative estimates in the 3–5% range. With 2.3 million people incarcerated, those estimates suggest there are as many as 100,000 people locked up who do not belong there, costing some $3.5 billion annually.
The Innocence Project, a nonprofit legal organization committed to exonerating wrongly convicted people, has succeeded in having 337 cases overturned using DNA evidence alone. Those prisoners served, on average, 14 years for crimes they did not commit. A University of Michigan project has compiled a registry of 1,747 legal exonerations, many of which involved the death penalty because our legal system does focus some critical scrutiny on those convictions.
Thanks to that work, we know that people are imprisoned because of false eyewitness testimony, unreliable forensic science, ineffective assistance of counsel, or the misconduct of police or prosecutors. Many more are imprisoned for reasons that are technically sound but that defy common sense.
Shoddy forensic science / To appreciate the shoddy science that results in many wrongful conviction, consider fire science, which is often used in arson investigations. Fire science has advanced greatly in the past 20 years, and those advancements have called into question earlier “science” used by supposed arson experts.
One victim of that pseudo-science is Han Tak Lee of New York. He was sentenced to life imprisonment for the 1989 arson murder of his 20-year old mentally ill daughter. The fire science evidence at the heart of the prosecution’s case, undisputed at the time, has since been conclusively discredited. Lee served 25 years of the sentence, until a judge granted his petition for relief in 2014. That is 25 years lost to Lee, his family, and his community, as well as 25 years of incarceration costs borne by society.
Lee’s case is not unique. At least 55 arson convictions are now being reviewed in light of improved fire science. Meanwhile, arson investigators continue to use questionable methods—such as accelerant-sniffing dogs—that have not been scientifically validated.
For decades, behavioral scientists have known that police line-ups are unreliable if done in traditional ways, with only a few suspects and police administrators who know which ones are the real suspects.
Like fire science, the understanding of hair analysis has advanced and changed significantly in recent years. In April 2015, the Federal Bureau of Investigation acknowledged that examiners in its microscopic hair comparison unit had overstated forensic matches in ways that favored prosecutors in more than 95% of the 268 trials that analysts had reviewed.
Bite mark evidence has also come under critical scrutiny. Texas recently banned the use of such evidence in trials because it is now considered far too unreliable. Thousands of individuals nonetheless sit in prison, convicted on bite mark evidence.
There have even been scandals about fingerprints and bullet striations—techniques that depend on subjective judgments of similarity but are essential to many criminal prosecutions. Only now are the courts beginning to ask whether these forensic “sciences” have actually been tested and shown reliable, and whether the analysts used appropriate techniques to preserve their objectivity against the clear preferences of police officers with whom they often work in close collaboration.
Erroneous and biased eyewitness testimony / Perhaps the most powerful, yet most unreliable, evidence that leads to wrongful convictions is eyewitness testimony. For example, in 1998 the police created a sketch based on eyewitness testimony of “the Bronx rapist,” a suspected serial rapist who terrorized the New York borough over the previous year. The sketch resembled a young man named Tyron Hicks—so much so that his own parents turned him in. He was convicted and served 10 years in prison before DNA testing proved him to be innocent.
For decades, behavioral scientists have known that police lineups—the mainstay of television crime dramas—are unreliable if done in the traditional ways, with only a few suspects and police administrators who know which ones are the suspects. In 70% of wrongful conviction cases in the Michigan registry alone, an eyewitness falsely fingered the defendant. Last year, the National Academies of Science called for reform of this practice.
One particularly disturbing feature of our criminal justice system is that prosecutors are allowed to offer incentives—including leniency and cash—to witnesses in exchange for testimony favorable to prosecutors’ cases. In any other context, the offering of such incentives would be illegal as witness bribery, but the courts have held that prosecutors are not “persons” for the purposes of the federal bribery statute. We should not be astounded if people given incentives to lie will do so.
Failures of process / Another reason for incarceration waste is that the adversarial process failed in the original trial proceedings. The vast majority of criminal defendants depend on state-funded counsel, but public defender offices are woefully underfunded and overworked, regularly receiving half the funding of prosecutors in the same jurisdictions. With that sort of economic thumb on the scale, it’s not surprising that cases on the margin are biased toward convictions.
For the few individuals who demand a trial, the procedures are also cramped. In civil litigation, the courts long ago rejected “trial by ambush”—the use of surprise witnesses and evidence—and instead allow extensive processes of discovery so that each side must submit all the evidence to vigorous scrutiny. But in criminal trials in most jurisdictions, the government does not have to disclose evidence that could impeach a witness until after the witness has testified. Thus, defendants have no real chance to challenge the evidence used against them.
Of course, most defendants—including some innocent ones—plead guilty as part of agreements with prosecutors. Defendants choose to do this, in part, because prosecutors have almost limitless discretion to stack charges, making the risk of a conviction at trial profound. By pleading guilty, the defendant usually receives a reduced sentence and thus faces a lower risk-adjusted cost. Many of the documented wrongful convictions were the result of confessions and plea deals.
When a plea agreement is reached, the trial judge is supposed to question the defendant to make sure he is really guilty. In reality, this practice is little more than kabuki theater. Both the prosecutor and the defendant play along, to avoid the waste of time and onerous penalties imposed on defendants who go to trial.
Changes in law / When a court strikes down or narrows the scope of a state or federal statute, as sometimes happens, prisoners are suddenly rendered retroactively innocent. Yet, many continue to serve time for these offenses that are no longer crimes.
For instance, federal law makes it a distinct crime to use a firearm as part of a drug crime. For many years, prosecutors interpreted this law to include defendants who merely had a gun at the crime scene. Then in 1995, in Bailey v. United States, the Supreme Court held that the term “use” required prosecutors to prove active deployment of the gun—e.g., pointing or firing it. A federal commission later estimated that between 1,500 and 2,200 federal defendants per year had been convicted under the broader reading of the statute. Many of these people continued to serve time for the charge until they challenged the conviction in court.
In recent years, the Supreme Court has taken a more aggressive approach to the First Amendment, the Second Amendment, and the Commerce Clause, resulting in decisions that should alter convictions. The Court has also begun to pare back some of the absurdly broad federal statutes under the rule of lenity, which holds that ambiguities in a statute should be resolved in favor of the defendant so long as it does not violate legislative intent. Thus we are likely to see more statutory and constitutional exonerations.
Along these same lines, Congress or the Federal Sentencing Commission sometimes changes the sentencing laws or guidelines in ways that retroactively shorten sentences. For example, lawmakers recently reduced the differences between prison sentences for crack versus powdered cocaine, a disparity that affected racial minorities. Many who advocate for changing the national policy of mass incarceration call for similar legislation that could lead to current prisoners going free or getting shorter sentences, in addition to changing sentencing policy.
Post-Conviction Litigation
Across these categories of cases—wrongful convictions, failures in process, and changes in law—there are thousands of prisoners who should be released and millions of tax dollars that should be saved. To prevent these sorts of problems in the first place, we need reform at every stage of the criminal justice process, from arrests and pleas to appeals and eventual probation. On each of these dimensions reforms are happening, and some jurisdictions have gone further than others.
For individuals who are currently incarcerated, however, the only options are clemency by the executive or post-conviction litigation in the courts, also known as habeas corpus (named for a common-law procedure that has since been displaced by statutes). In habeas, a convict has one last opportunity to challenge his conviction or sentence.
Post-conviction litigation is potentially the only meaningful form of review in the vast majority of cases. When a plea or trial proceeds on an obsolete legal theory and the law is changed thereafter, habeas is appropriate. Several federal circuit courts have also held that claims of ineffective assistance of counsel cannot be raised on direct appeal at all; they can only be raised during post-conviction proceedings. More generally, because sentencing risk makes it rational for many accused to plead guilty regardless of their guilt, the shoddy forensic science, the government-purchased eyewitnesses, and the overly broad readings of the criminal statutes have never been challenged in court, much less reviewed on appeal.
Prisoners can hardly be blamed for exercising their rights themselves when they cannot retain expert counsel. Like asking prisoners to perform their own brain surgeries, it is hardly surprising that most petitions fail.
Need for expert counsel / Post-conviction litigation is a shambles, however. Given centuries of doctrine and occasional intervention by Congress, habeas is one of the most convoluted areas of the law, with a gamut of procedural tricks and traps that can mire worthy claims. This extreme technicality exists because courts have a bona fide interest in the finality of decisions and review is costly. Accordingly, current habeas law strictly limits relief to constitutional errors and other problems that lead to a complete miscarriage of justice.
Since states and the federal government provide virtually no financial support of post-conviction litigation for non-capital defendants (i.e., defendants whose do not face the death penalty), prisoners are left to navigate the extremely complex legal domain alone. In the federal court system for the year 2013, 92% of prisoner petitions were filed pro se, and most of them proved to be legally frivolous and, too often, outright incomprehensible. Yet these prisoners can hardly be blamed for exercising their rights themselves when they cannot retain expert counsel. Like asking prisoners to perform their own brain surgeries, it is hardly surprising that most petitions fail.
In capital cases, 92.9% of the petitioners have attorneys, but in non-capital cases only 7.7% do. A comparison of outcomes shows that capital defendants are 35 times more likely to get relief than non-capital defendants. While it is possible that this difference in outcomes is because capital cases are more prone to error, or that courts are more receptive to capital defendants, we cannot ignore the lack of skilled attorneys for non-capital defendants as a factor. For the vast majority of prisoners, there is no meaningful review of their convictions.
Incentives for counsel / Under current law, most prisoners probably deserve to be there, and there is no simple algorithm for identifying which ones don’t. The challenge is to separate the wheat from the chaff, and that requires professional skills and the investment of both time and money. Currently, to do this sorting, we largely depend on charity, luck, and pluck, which is no way to run a multi-billion dollar government enterprise.
A better approach would be for the government to increase funding for public defenders so they can do more post-conviction litigation. Some public defenders already have in-house innocence projects. Still, funding for public defenders’ offices is notoriously scarce, the salaries offered for these cases often fail to attract the best attorneys needed to undertake such complex work, and the overworked offices naturally triage in favor of new cases.
Of course, we could spend more on public defenders. But as a centrally planned solution, it’s hard to assess the optimal level of investment. Prior reform efforts suggest that additional spending on public defenders may also be politically infeasible because it is often viewed as providing a service for criminals.
Instead, governments should consider using a contingent-fee system for post-conviction counsel. Attorneys would only receive this fee if they successfully show that a prisoner’s continued incarceration is wrongful. The fee could be based on a simple proportion of the estimated amount the government would save by stopping the incarceration—perhaps 50% of those costs. Or, the system could be set up like the statutory fee paid to civil rights attorneys, taking into account a reasonable hourly rate multiplied by a factor to recognize the low chances of prevailing. In the False Claims Act, passed during the Civil War to root out fraud by government contractors, and the more recent whistleblower statute that the Internal Revenue Service uses to expose tax evaders, we have precedents for paying financial rewards that align the interests of knowledgeable individuals and the government.
The advantage of a contingent fee is that it gives attorneys an incentive to search for worthy cases and bring them to prosecutors and the courts, which is exactly what a cost-conscious government needs. Unlike desperate and unskilled prisoners representing themselves, attorneys would have no incentive to clog the courts with frivolous claims for post-conviction relief. Any such claim would require the investment of time and money without promise of return. Instead, we should expect a small industry of specialist attorneys to develop, at first focusing on the low-hanging fruit, but then becoming more specialized to identify entire categories of cases where review is most promising.
Unfortunately, this proposal seems to conflict with an ethical rule against criminal law attorneys receiving payment from clients only if the clients are found innocent. This rule is a longstanding provision of the model rules of professional conduct adopted by most states. It is motivated by the belief that individuals who succeed in proving their innocence should not have to subsidize the defense of guilty individuals who lose in court. Yet, the proposed arrangement for post-conviction relief is better conceived as a government bounty for freeing the innocent. The model rule is thus irrelevant.
Prosecutors often view efforts to identify wasteful convictions as meddlesome—at best, attempts to overturn years of hard work to put bad guys away, and at worst, attacks on the prosecutors’ character and competence.
Economists and accountants can provide more precise estimates of the optimal fee level, accounting for both the marginal costs of incarcerating a prisoner as well as the transaction costs for prosecutors to respond to petitions and judges to decide them—costs taxpayers must also bear. The transaction costs are one-time costs, not continuing on for years or decades as the costs of incarceration do. And the costs are largely the same regardless of whether a frivolous, incoherent petition is filed by a prisoner himself or an adept petition is filed by a skilled attorney.
Role of the prosecutor / These transaction costs could be minimized and incarceration waste could be resolved more quickly if prosecutors would cooperate in the process. Since the vast majority of people go to prison on the basis of quick-and-cheap plea agreements, one might hope that the vast majority of post-conviction challenges could also be resolved with the consent of the prosecutors. After all, they are ostensibly guardians of justice (and fiduciaries of public money), not merely zealous advocates for incarceration.
Unfortunately, prosecutors often view efforts to identify wasteful convictions as meddlesome—at best, attempts to overturn years of their own hard work to put bad guys away, and at worst, attacks on the prosecutors’ own character and competence. In fact, prosecutors have strenuously objected to convicts even testing the physical evidence for DNA that could exonerate them. Not only are prosecutors’ self-conceptions at stake, but they also operate in a vacuum, not internalizing the costs of incarceration. Local prosecutors have almost complete autonomy and the executive branch is fragmented, with another agency bearing the expenses of imprisonment.
A small number of prosecutors’ offices—24 nationwide as of 2016—have set up special departments consisting of one or a few attorneys to review convictions. The very creation of such “conviction integrity units” is an acknowledgment of the problem, and at least creates a vehicle for cases to be reviewed. Compared to having the same prosecutors that put someone away then review the case for error, these designated units may be more objective (though they are still prosecutors working within prosecutors’ offices). Still, the fervor and success of these programs vary widely. Like the inspectors general who work in other federal agencies, these units would benefit from greater independence from prosecutors who are trying to put people away.
Regrettably, many prosecutors loathe the idea of considering that a person may have been wrongfully convicted. But just as the original plea agreement was based on the strength of the case (with stronger cases leading to longer terms of imprisonment), when new information weakens the case, prosecutors should agree to a shorter term of imprisonment. Under current procedures, such a negotiation often requires a joint motion to set aside the original conviction, and an agreement for the defendant not to contest new, reduced charges. When a reduced charge leads to a sentence of time-served, it may lead to immediate release. Courts should be amenable to these processes, just as they are for original plea deals.
Objectivity of the courts / Currently, when a habeas case goes to court, the deck is stacked against it. First, in the federal system and many states, the habeas statutes send a prisoner back to the same judge who presided over the conviction in the first place, requiring the prisoner to persuade the judge that he or she committed a grave error. We know from the social sciences that judges will see these cases a second time with a bias toward upholding their original decisions. Confirmation bias is the documented tendency for individuals to cling to prior beliefs, regardless of new evidence.
To make matters worse, in the federal system the defendant isn’t allowed to appeal unless a court grants him permission to do so. This provision is peculiar to habeas law and provides a second chance for confirmation bias to kill a valid case. In the rare instance that an appeal is granted, the defendant must go back to the same court of appeals—and often the very same panel—that denied his direct appeal (if any). It’s no wonder why the majority of these petitions fail.
If legislators were serious about reducing unnecessary incarceration, they would ensure that fresh eyes review a post-conviction case. Some may argue that a new judge would lack familiarity with the facts of a case, but that’s a weak argument. While facts can be provided to a new judge, nothing can remove the confirmation bias from the mind of the prior judge.
Procedural hurdles / A variety of legal rules also tilt the scales against liberty and thus perpetuate incarceration waste. Although the habeas statute specifically allows for challenges to convictions or sentences that are “in violation of the Constitution or the laws of the United States,” amazingly some federal circuit courts have held that there is no basis for relief in cases where a person is serving time for a crime that, post-conviction, it has been shown he did not commit. That is, courts reject actual innocence as a basis for relief.
Finality is valuable from a judicial perspective, but it can make for stupid government policy. What interest does the state have in spending money to incarcerate innocent people?
Another obstacle is the statutes of limitations. Under federal law, prisoners only have a year to conceive a basis for relief and file a petition, even though prosecutors typically had at least five years to bring the original case. The idea of a statute of limitations is to quickly dismiss stale claims from being heard, even if the claim itself is valid. Yet, the passage of time does not make innocent prisoners any less innocent, nor create value for the state paying to incarcerate them.
These sorts of doctrines make sense in the civil context, where the parties are simply adversaries. Oddly, in the criminal context the prior judgment of conviction is causing one party (the government) to pay for the other party’s housing, food, medical care, and security. Thus, there is a shared interest in overturning a wrong judgment.
These sorts of doctrines should be reconsidered to facilitate the objective review and disposition of credible post-conviction cases identified by properly incentivized counsel. The modern conception of criminal justice needs to incorporate not just the judicial interest in finality, but also the administrative and economic interest in efficiency and accuracy.
Conclusion
We have suggested that post-conviction review could be a meaningful way to reduce incarceration waste if the state and federal governments decide to provide smart incentives to attorneys to screen and develop such cases. We have also suggested reforms in prosecutors’ offices and streamlined procedures in the courts, to make the ultimate decisions more accurate and less biased toward incarceration.
These reforms will help on the margins. Still, much more profound reforms to criminal law, procedure, and sentencing policy will be required to solve the problem of wrongful mass incarceration generally.
A wise government would at the very least seek to identify those prisoners who do not belong in the prison system and provide contingent funding to attorneys who successfully identify those prisoners. Each guilty plea or trial outcome costs the state hundreds of thousands of dollars, and also exacts a heavy toll on the defendant, his family, and his community. The idea of incarceration waste is a modest reminder that the government and some of the imprisoned have aligned interests in identifying those who should be released.
