Get Informed

Subscribe to our newsletters for regular updates, analysis and context straight to your email.

Close Newsletter Signup

The Perverse Incentives of Punishment

Etowah County Sheriff Todd Entrekin

The Perverse Incentives of Punishment

Todd Entrekin, the sheriff of the small Alabama county of Etowah, recently found himself in the national spotlight when an Alabama newspaper discovered that over the course of three years he pocketed at least $750,000budgeted for feeding the people detained in his county jail. While the inmates in his jail ate meat from a package labeled “not fit for human consumption,” the sheriff bought himself a $740,000 beach house.

And it was all seemingly legal, thanks to a 1911 Alabama law that many sheriffs interpret to mean that whatever funds they don’t spend on their jails they can keep for themselves.

The story is horrific on its own terms, which is why the actions of a small-town sheriff — Etowah County has a population of about 100,000 people — quickly made national news. It is yet one more example of the almost countless ways in which our criminal justice system dehumanizes those it touches.

What happened in Etowah, however, highlights a deeper flaw in our criminal justice system. Much of the harm and destruction the system causes is exacerbated, if not often directly caused, by the complex web of financial obligations that criss-cross the convoluted morass of agencies that we too-simplistically call our “criminal justice system.”

Insufficient attention is given to how obscure contractual terms and budgetary decisions — things that fall fully within the responsibility of the public sector — shape how criminal justice actors behave. To be clear, values and attitudes matter a lot: Many officials inarguably view those who come within their control as undeserving of compassion, if not less than human, and treat them accordingly. But everyone from private prison managers to elected sheriffs to county commissioners also pay very close attention to the fiscal incentives they face.

In some cases, like in Etowah County, the incentive is clearly stated and transparently problematic. And there are examples from other states as well. Take Missouri. One prominent revelation in the wake of the Ferguson protests was the extent to which municipal officials encouraged various police departments across St. Louis County to impose fines in order to fundlocal governments.

Another example from St. Louis County, however, demonstrates how less-obvious contractual provisions encourage punitiveness. Following the Ferguson protests, prosecutors aggressively went after people who had participated in them, or just journalists who covered the protests, even as they dropped charges or kept losing cases. It would be easy to talk about aggressive prosecution or problematic desires to silence the press, but the explanation is far more contractual. Many municipalities in St. Louis County contracted prosecution services to local law firms (something that happens in about 15 percent of U.S. counties), and the contract didn’t include a cap on payments. More prosecutions, more money. Because local officials wrote bad contracts.

Similarly, with only a few exceptions, the attacks on private prisons generally miss the point, overlooking the significance of bad contract incentives to focus instead on the seeming evil of the “profit motive” — as if the $30 billion in wages and benefits going to public sector correctional officers isn’t a profit motive itself. But that’s a separate issue.

The conventional argument against private prisons runs something like this: States pay private prisons a per-prisoner per diem, and the prisons respond by cutting services and staffing and training and food in order to get their per-prisoner costs below that per diem. The people running the prisons then take their per-prisoner savings and invest them outside the prison, and lobby hard against reforms that would reduce prison populations, since that would cut profits. And the lack of staffing and training provide a second, more insidious benefit — they likely increase recidivism rates, which increases the number of people returning to prison and thus raises profits too.

This is an awful story. And it happened in Louisiana almost entirely without the help of private prisons. The state government paid local (public sector) sheriffs a per-diem to house state (public) prisoners in county (public) jails, and the sheriffs cut jail costs, diverted the savings to fund their (public sector) departments outside the jails, and lobbied against reforms.

Public sector officials behave exactly like private sector ones when given the same incentives. It’s the incentives, not the “profit,” that matter. So if we change the financial incentives people face, we may be able to get better outcomes. Pennsylvania recently introduced new contracts that reward private halfway houses that beat recidivism targets, and cancel the contracts of those who fall short; so far, they seem to be producing good results. Similarly, Australia just opened a private women’s prison with a strong recidivism provision in its contract, and the U.K. has done the same.

Not all of the fiscal incentive problems arise from explicit contract provisions, however. Sometimes they result from the baffling structure of our criminal justice system. In fact, what we call our “criminal justice system” is not a system, but rather a morass of city, county, state, and federal systems, all of which interact with each other in ways that are frequently perplexing at best, and incomprehensibly counterproductive at worst.

I don’t know to what extent the poor design of these overlapping institutions stems from intentional malfeasance, malicious indifference, or just genuine incompetence or inattention. My guess is that in many cases they arose haphazardly and unintentionally over time, and that no one has the interest or incentive now to try to fix them. But whatever their origin, and whatever explains their durability, they create terrible incentives that likely play major, but generally underappreciated, roles in driving the everyday failures and pathologies of punishment we see today.

Just look again at Alabama’s 1911 jail-food law, the product of a time when sheriffs lived in the jails, had their wives cook the prisoners’ food, and received no salary outside of the fees they could collect. It’s not like the state is unaware of the law’s problems; the legislature simply hasn’t been able to amend it. And while you might think the opposition comes from greedy county sheriffs looking to buy summer homes, the real resistance has come from the state’s association of county commissioners. The flip side of Alabama’s law is that while sheriffs get to keep any unspent funds, they are also personally liable for any shortfalls — and the county commissioners don’t want to assume that risk, even as some sheriffs actually push to fix the law.

In other words, the law persists because fiscal responsibility is fractured even within a county, and that creates strong incentives to make sure “someone else” has to take responsibility. Obviously greed and malicious indifference, if not actual malice, matter too, but the impacts of far-more-mundane fiscal obligations are quite significant.

A far more common example of how fractured responsibility encourages harshness and severity is what I call the “prosecutorial moral hazard problem.” Prosecutors, as reformers have started stressing in recent years, are mostly county-elected officials, who are mostly paid from county funds. Jails, too, are paid for by the county, as is often probation. Prisons, however, are paid for by the state.

This creates a powerful incentive to be harsh, since it is actually cheaper for a prosecutor or county-elected judge to charge or impose a felony sentence rather than a misdemeanor. A felony sends the defendant to state prison, and thus off the county books, while a misdemeanor would keep him in county-funded jail or probation.

California is the one state that has confronted this moral hazard problem in any serious way — and, quite tellingly, it alone is responsible for over half of the national decline in prison populations since 2010. One component of its complex realignment reform is that counties must now bear the cost of locking up certain less-serious offenders, even when they are convicted of felonies. Data indicate that this sort of cost-internalization has worked in the past, and it seems to be working now in California (with no real impact on public safety). Unfortunately, few states seem willing to follow in California’s footsteps.

This sort of moral hazard problem likely explains another policy failure, or at least an inefficiency, that we frequently see. The data is clear that policing is far more effective than incarceration at reducing crime, and yet we have over-invested in prisons, compared to policing (and many, many other interventions, including those outside the realm of law enforcement altogether). And at least one reason is surely that local officials have a strong incentive to push for tougher sentencing laws: They get to appear tough on crime for their local constituents while pushing the costs onto a different, and better-funded, level of government. The city would have to pay for more police, but the state picks up the tab for the longer sentences.

Now, there are some efforts to address these misaligned incentives. The Justice Reinvestment Initiative (JRI), for example, is trying to target some of these, by bringing state and local government officials together to work to shift funding away from what works poorly to what works better. But JRI is just a small part of the overall reform effort; in general, inter-governmental fiscal incentives, while quite important, seem to get far too little attention. Most reforms aim to change the basic criminal and sentencing laws, not the deeper sets of financial and political incentives that shape how police and sheriffs and prosecutors and others use the unavoidable discretion they will always wield.

But to end on a quirkily optimistic note, just as our failure to account for financial incentives causes us to miss a lot of problems, it also sometimes causes us to overstate them. Take civil asset forfeiture, which allows the police to seize property they think was involved in a crime (such as a car used to transport drugs) even if no one is convicted of the crime. Asset forfeiture is hated by reformers on the left and right alike, and it is often accused of encouraging police to target drug crimes just for profit.

There is much wrong with civil asset forfeiture, but its impact is likely overstated—again because of jostling, conflicting financial goals and obligations. In this case, the police do not determine their own budgets — those are set by local city and county legislatures and executives, all of which have a lot of funding obligations besides law enforcement. Two economists produced results that suggest these other agencies tend to cut police budgets to offset forfeiture earnings: For every dollar the police seize, subsequent budgets are offset by about 40 or 50 cents, sometimes almost dollar for dollar.

In many ways, I fear that these sorts of powerful financial incentives are distinctly treacherous because they are so technical and mundane. Often what is shocking and emotionally gripping is less important than the tedious stuff chugging away in the background, and so that tedious stuff gets a pass.

I saw this during the 2016 presidential primaries, when Hillary Clinton was attacked for her support for the 1994 Crime Bill whose provisions were provocative but whose impact was slight. Ignored in the debate over the Clinton legacy on crime was the far more significant Clinton-era Prison Litigation Reform Act (PLRA), which restricts prisoners’ ability to challenge terrible prison conditions in federal court by denying them the ability to sue until all “administrative remedies as are available are exhausted.” Bureaucratic issues like the “exhaustion of administrative remedies” is not a gripping topic, but these technical provisions of the PLRA enabled California to under-fund its prison system for years, to the point that the Ninth Circuit held that approximately 60 prisoners died from preventable deaths per year. The PLRA killed people, but the law received almost no attention whatsoever during the 2016 campaign.

The public-sector financial incentives are victims of the same inattention. They aren’t exciting to talk about, but their impact is real and powerful, and reformers need to direct far more energy toward changing them.

Two States Just Made It Easier to Take Babies Away From Mothers Who Use Drugs During Pregnancy

Alex Nuñez / Flickr

Two States Just Made It Easier to Take Babies Away From Mothers Who Use Drugs During Pregnancy

M. thought she was doing the right thing. She had become dependent on opioids, but when she learned she was pregnant, she immediately tried to enroll in a medication assisted treatment (MAT) program. MAT is the standard of care for treating people with opioid use disorder — especially pregnant women, as quitting opioids too suddenly during pregnancy can result in complications.

Despite several phone calls to treatment providers in her county and surrounding counties, M. could not find a provider who was willing to accept her. Many substance use treatment services don’t accept pregnant women, or are otherwise inaccessible to them, despite federal and state regulations that require prioritizing them for treatment. In order to to avoid the risk of withdrawal and possibly harming her fetus, M. did what many pregnant women in her situation do — she continued to use illicit opioids.

As her due date approached, M. found a hospital that claimed to specialize in the care of babies who were exposed to opioids in utero and traveled a far distance to give birth there. She told her treating physician about her opioid use during pregnancy, not realizing this could precipitate call to Child Protective Services (CPS). CPS immediately removed her newborn from her care, largely based on evidence of her opioid use. In the subsequent months, M. saw her son once a week at best. When she visited him, she was distressed to find he often had seemingly untreated rashes. She struggled every day with the unimaginable pain and grief of separation from her newborn. As of last year, they were still apart, and M. has since lost touch with her lawyer.

More mothers may soon know M.’s pain. At least two states, Arizona and Kentucky, have just made it easier to terminate the rights of mothers who use controlled substances while pregnant. Arizona’s legislation, which became law in April, permits termination of a mother’s parental rights, either immediately when her newborn is born or within one year of her newborn’s birth, depending on how chronic the illicit drug use appears to the court. Kentucky’s legislation, which also became law last month, permits termination of a mother’s parental rights if her newborn exhibits signs of withdrawal, known as neonatal abstinence syndrome, as the result of illicit opioid use, unless the mother is in substantial compliance with both a drug treatment program and a regimen of postnatal care within 90 days of giving birth.

Terminating a mother’s rights to her newborn is an especially brutal drug war tactic that research and experience show will inflict far more harm than good on the children and families it allegedly aims to protect. Such policies are rooted in stigma and gross indifference to what the best available science tells us about how to compassionately and effectively serve pregnant women struggling with drug use disorders and their families.

“The legislators behind these laws are essentially creating a capital offense for women who give birth despite having used an illegal drug,” said Erin Miles Cloud, a senior attorney at the Bronx Defenders Family Defense Practice, who has represented numerous parents facing termination of parental rights proceedings. “Termination of parental rights is a mechanism by which families are turned into strangers, all contact and personal identity is erased, and families are destroyed forever. In that way, this legislation’s attempt to punish mothers and fathers acts as a civil death penalty for families, for which children will pay the ultimate price.”

Supporters of these laws justify the surveillance, policing, and punishment of drug-using mothers by referencing two assumptions that were touted and then discredited during and after the “crack baby” scare. First, they say the developing fetus and newborn will be harmed by prenatal exposure to illicit substances. Second, they claim that drug use during pregnancy is a reliable indicator of parental unfitness.

While a robust body of literature supports a causal connection between prenatal exposure to alcohol or tobacco (or lead or poverty for that matter) and negative postnatal health outcomes, the scientific literature has not conclusively demonstrated any long-term negative effect of prenatal exposure to opioids. Neonatal abstinence syndrome (NAS) is itself a treatable and transient condition, and a growing body of literature confirms that one of the most effective treatments for NAS is keeping the newborn and mother together in a soothing environment while encouraging skin-to-skin contact and breastfeeding. So the very condition that Kentucky believes should trigger fast-tracked termination of parental rights is in fact often most effectively treated by close, consistent contact between the mother and newborn.

The research is also clear that the results of a drug test alone are not an appropriate proxy for determining parental fitness. Millions of parents who use drugs or have substance use disorders parent their children well. Studies have found that babies exposed to cocaine in utero are not at greater risk of maltreatment as young children than similarly situated babies. Other studies have found that babies exposed to cocaine in utero perform better on several developmental measures when left with their mothers than do those removed to foster care.

The research that does claim to link substance use to maltreatment of children is not authoritative. It often relies on caseworker-confirmed reports of child maltreatment, despite the well-documented propensity of caseworkers and family courts to base findings of child neglect on evidence of drug use alone. And the scientific literature that suggests that substance use produces social cognitive deficits in parenting is in its infancy and rather underwhelming. For example, one study found that parents who use opiates find babies less cute — while conceivably notable, this finding does not meet the legal standard for terminating a parent’s rights.

Medical and public health authorities warn that women who fear losing their babies upon seeking medical care will be deterred from seeking the care they need. Community after community has seen this in the aftermath of local crackdowns on drug-using pregnant women — fewer women seek prenatal care and substance use treatment, even after the local authorities decide to change course. Experts recognize that pregnancy and childbirth present an especially opportune moment to connect a woman with services, including substance use disorder treatment — yet threats of punishment only isolate pregnant women.

Supporters of these laws blame the mothers for the dissolution of their families, arguing that the mothers need only enroll in a drug treatment program to retain custody of their children. This claim disregards the fact that drug addiction is a health issue with biological, behavioral, and genetic dimensions, and similar to other health conditions, it does not respond wellto artificially imposed treatment timelines or mandated treatment. These same legislators would be hard-pressed to take babies away from mothers who were struggling with medical treatment for hypertension or diabetes. This claim also disregards the serious gaps in access to treatment for substance use disorders, especially for poor people, and the absence of a social safety net to even support struggling families in the first place.

“Instead of punishing women for the chronic condition of addiction, something that is unethical, ineffective and inhumane, we should invest instead in the expansion of women- and family-friendly treatment resources,” said Dr. Mishka Terplan, a professor in obstetrics and gynecology and psychiatry and associate director in addiction medicine at Virginia Commonwealth University.

While medication assisted treatment (MAT) is covered under Arizona’s newly expanded Medicaid programs, Arizonans still struggle with access to MAT treatment. As is true across the nation, rural areas have limited, if any, access to MAT. Many substance use treatment centers do not offer MAT due to the stigma associated with it, and those centers that do provide MAT have not necessarily fulfilled their responsibility to prioritize pregnant women. Meanwhile, private insurance companies have found ways to circumvent Obamacare requirements on MAT coverage.

This is all exacerbated by a series of measures Arizona has undertaken to constrict its social safety net, leaving families struggling with inadequate access to cash, food, housing, child care, and transportation. The vast majority of Temporary Assistance for Needy Families (TANF) funding, or federal welfare dollars meant to help the poor, is not spent on assistance to poor families, but rather on placing and keeping children in the foster care system. Arizona also has the strictest TANF timeline in the nation, kicking families off welfare after one year. Not coincidentally, in the aftermath of the 2008 financial crisis, as Arizona’s social safety net shrank, its foster care numbers soared. Arizona now has one of the highest foster care placement rates in the nation.

Kentucky similarly struggles with poverty, poor access to substance use treatment services, and a child protective system that spends more resources on placing children in foster care than keeping them with their families. Kentucky ranks amongst the last in the nation in rates of childhood poverty. The federal Department of Health and Human Services recently released a report reviewing Kentucky’s child welfare system. It found that parents face long wait lists when accessing substance use treatment services and often don’t have the means to pay. The report also found that Kentucky’s child protection agency was not making enough efforts to prevent removal or re-entry to foster care. Indeed, Kentucky has one of the highest rates of child removal in the country.

Such inhumane responses to drug use can only exist because they are almost exclusively reserved for poor people and people of color. The overwhelming majority of parents prosecuted by the child welfare system are poor, and parents of color are overrepresented. Several studiesdocument that hospital staff disproportionately drug test and report to child protective services low-income women and women of color. Indeed, few, if any, middle- or upper-class women who use drugs during pregnancy will ever experience a child abuse and neglect proceeding, let alone a termination of parental rights — though drug use is common among people of all socioeconomic levels. Many advocates I interviewed in the course of my own research on this subject described the surveillance of pregnancies and non-consensual drug testing performed on pregnant women and their newborns as comparable to stop-and-frisks for young men of color.

Despite progress in the criminal legal system, the drug war remains almost unchallenged in the child protection system, and it is wreaking havoc on families. Before losing contact with her lawyer, M. told me her story via email. “In the family court systems, there is no … respect for medicine or science,” she wrote. “This is all done under the cloak of what is in the ‘best interest’ of the child — but that is ironic, because they are hurting my son.”

More in Explainers

The Hidden Police Violence Epidemic Behind a ‘Swatting’ Death

A SWAT team prepares for an exercise.
Oregon DOT

The Hidden Police Violence Epidemic Behind a ‘Swatting’ Death

On December 28, 2017, 28-year-old Andrew Finch of Wichita, Kansas, opened his front door to a horde of shouting police officers. Ten seconds later, he was fatally shot in the head — yet the officer who pulled the trigger isn’t the one being charged with his death.

The events of that tragic late December day were set in motion by three Call of Duty gamers fighting over a $1.50 wager. One gamer, Tyler Barriss, prank-called Wichita 911, saying he was suicidal, had killed his father, and was holding others hostage at the address that one of the other gamers had provided as his own. But the target of the prank didn’t actually live there; Andrew Finch and his family did.

At approximately 5 p.m. that day, 10 Wichita police officers and three deputies surrounded the house. When Finch exited the home to see what the commotion was about, he threw his hands into the air in surrender. But then, after becoming startled, he put one hand down for a second. That’s when a police sniper fired from 50 yards away, killing him.

Immediately after shooting Finch, the officers ordered his family to come out of their home with their hands up, and even forced his young niece to walk over her dying uncle’s body, according to a civil rights lawsuit filed in January in federal court on behalf of Finch’s mother. The family members were then handcuffed and forced to wait outside in below-freezing temperatures for over an hour, interrogated, and released without explanation.

The prank caller, Barriss, was tracked down in Los Angeles less than 24 hours later and charged with involuntary manslaughter, giving false alarm, and interference with a law enforcement officer. He faces 11 years in prison. In the wake of the incident, reporters from national media outlets like the New York TimesCBS News, and NPR wrote stories about “swatting,” a gamer trend where players like Barriss call in fake emergencies to 911 which sends SWAT teams to a rival’s house as a prank.

The headlines, however, obscured the fact that there was no justice for Finch. On April 12, four months after the shooting, Wichita District Attorney Marc Bennett announced that he would not bring charges against the officer who killed Finch. The city has refused to name the officer who fired the fatal shot.

According to the Finch family complaint, his killing at the hands of the Wichita police was unusual only because it stemmed from a prank. Police shootings are disproportionately common for a city its size, the complaint says. Wichita has a ratio of one shooting for every 120 officers, about 11 times higher than the national average. Wichita officers were involved in at least 29 shootings between 2010 and 2015, 15 of which were fatal. Nonetheless, the DA has determined that every one of these shootings was justifiable. And in 95 percent of police shootings there, Wichita officials have also shielded the officers’ names from the public.

“This pattern and practice of concealing misconduct and concealing the identities of officers involved in misconduct encourages officers to believe that their unconstitutional behavior will be protected,” the Finch family said in its complaint, “and that they will suffer no discipline, thereby emboldening them to act with impunity.”

The Finch family argues in its lawsuit that even if the police faced a true hostage situation at Finch’s house, shooting whoever came to the door violated departmental policy. But despite clear Wichita Police Department guidelines regarding how volatile situations involving mentally ill suspects should be handled — including instructions that “in a stressful situation, a police member’s first reaction should be to determine whether the objective can be accomplished without the use of a weapon” — the Wichita DA has only once determined that an officer-involved shooting violated department policy over the past two decades.

Perhaps this is because the city’s police shooting investigations are hampered by strict requirements in its police union contract. According to data compiled by the Black Lives Matter-affiliated Campaign Zero, Wichita allows officers to meet privately with union representatives who can coach them on what to tell investigators before being questioned about a shooting. The city also allows officers to record their own interrogations, tightly restricts what interrogators can say or do in questioning, and erases officer misconduct files. Interviews are conducted by the officer’s co-workers, rather than by independent or state investigators.

“How is it remotely justice when the wrongdoers are investigating themselves?” Finch’s mother, Lisa, asked City Council in March.

The Finch family has continued to fight for justice, even after the DA announced his decision to not charge the officer who fired the fatal shots, with some measure of progress. A newly formed citizens advisory board will review the case, and the department says it is conducting an internal investigation.

“The issue of my innocent son being shot through the screen door is not going away any time soon,” Lisa Finch warned the council.

More in Podcasts