GUEST JOINT REVIEW: Homicide, by David Simon

Homicide: A Year on the Killing Streets, David Simon (1991; Picador, 2006).

The following is an email exchange between Jane Psmith and Charles Lehman of the Manhattan Institute and

The Causal Fallacy

, edited slightly for clarity.

Jane: In late 1987, Baltimore Police Commissioner Edward J. Tilghman — who was, at the time, dying of a brain tumor — agreed to permit Baltimore Sun reporter David Simon to spend a year with the nineteen men of Lieutenant Gary D’Addario’s shift in the homicide squad. Beginning on New Years Day of 1988 (and over the objections of the captain of the homicide unit, the deputy commissioner for operations, and the detectives themselves if they had been asked), Simon was granted the rank of “police intern” and allowed to read case files, follow along on calls, sit in on interrogations, and once, probably inappropriately, to help pin a fleeing suspect when a detective got stuck in his seatbelt on his way out of the unmarked car. The book Simon produced at the end of his year would become the basis for the TV show Homicide: Life on the Street, and eventually (along with Simon’s 1997 book The Corner: A Year in the Life of an Inner-City Neighborhood) for The Wire and the erstwhile journalist’s second career as a TV producer and showrunner. Even setting aside its eventual impact, though, Homicide is a remarkable book: an unsparing, relentless, and ultimately deeply affectionate portrait of the homicide detectives and the institution they serve. It’s also just really fun, which is why I felt no qualms about making you read it.

Simon is a hell of a writer, transitioning beautifully between vivid descriptive passages, lucid explanations of the investigative process, and lyrical meditations on the nature of the job. There’s an entire beautiful passage on this, pages long, that comes just after our cold open at a crime scene in “whatever is left around Gold and Etting that resembles a neighborhood.” Simon begins with, “This is the job: You sit behind a government-issue desk on the sixth of ten floors in a gleaming, steel-frame death trap with poor ventilation, dysfunctional air conditioning, and enough free-floating asbestos to pad the devil’s own jumpsuit” and rolls straight through with the power of a runaway locomotive ‘til:

Because in a police department of about three thousand sworn souls, you are one of thirty-six investigators entrusted with the pursuit of that most extraordinary of crimes: the theft of a human life. You speak for the dead. You avenge those lost to the world. Your paycheck may come from fiscal services but, goddammit, after six beers you can pretty much convince yourself you work for the Lord himself. If you are not as good as you should be, you’ll be gone within a year or two, transferred to fugitive, or auto theft or check and fraud at the other hand of the hall. If you are good enough, you will never do anything else as a cop that matters this much. Homicide is the major leagues, the center ring, the show. It always has been. When Cain threw a cap at Abel, you don’t think the Big Guy told a couple of fresh uniforms to go down and work up the prosecution report. Hell no, he sent for a fucking detective. And it will always be that way, because the homicide unit of any urban police force has for generations been the natural habitat of that rarefied species, the thinking cop.

It’s a little bit a book about the cases, especially three big ones that recur throughout the narrative: John Randolph Scott, shot while fleeing the police but for whom no one in the department will claim responsibility; Gene Cassidy, a patrolman who miraculously survived being shot in the head but whose case becomes a personal crusade for his homicide sergeant friend, and especially eleven-year-old Latonya Wallace, found raped and eviscerated in a Reservoir Hill alley one February morning. In Simon’s postscript to the 2006 edition he describes that investigation as “the spine of the book,” and it is, but the real meat of the book is the homicide unit itself: the guys, the way they act, the jokes and the arguments, the camaraderie and the conflicts and the pressures — internal and psychological, external and institutional — that shape the way they solve Baltimore’s murders.

It’s also, identifiably, an old book, not just because the technology is dated but because society and policing have changed. There’s a point in March where a freakout over the homicide clearance rate, standing at only 36%, reaches the level of the commander of the whole Criminal Investigations Division: heads may roll, jobs may be lost! And while 1988 eventually closes with 72% of homicides cleared, by 2022 it was all the way back down at 36% for the end of the year. Meanwhile, 1988’s 237 homicides (31 per 100,000) became 2022’s 333. But pretty much everything I know about policing comes from old books; how much has changed?

Charles: As the crime wonk here, I feel obligated to put the book in a little historical context for readers who don’t spend all day thinking about violent murders.

Simon is writing at the peak of what Jill Leovy memorably labels the “big years” of American murder. (Homicide is an obvious inspiration for Levoy's book, Ghettoside, although I’m not yet sure which to declare the better of the two.) Starting in the mid-1960s, the U.S. homicide rate rose more or less inexorably for three decades, almost doubling to its peak in 1991. This was part of a broader, again multi-decade, crime wave, which is a major driver of everything from white flight to the Reagan Revolution. By the early 1990s, crime is arguably the most important issue to the electorate, great citizen movements to retake the cities start to spring up, and Bill Clinton retakes the white House for the Democrats more or less by becoming a Republican on the issue of crime.

Against this backdrop, it makes sense that Simon would want to write a book about the criminal justice system. What I think is particularly interesting, though, is his choice of subject. Throughout the book, Simon makes clear that there's a hierarchy of policing. The beat cops on the streets are at the bottom, and above them are the detectives. The detectives in turn are ordered based on the seriousness of their investigations, with something like vice at the bottom, and homicide at the top. And this is a view of things extremely typical of the period. Throughout the crime increase, the general consensus was that the beat cop had little to do with controlling crime, and that policing should be mostly reactive—responding to a crime scene, finding clues, putting together cases. As James Q. Wilson and George Kelling put it in “Broken Windows,” “as the crime wave that began in the early 1960s continued without abatement throughout the decade and into the 1970s ... [s]tudies of police behavior ceased, by and large, to be accounts of the order-maintenance function and became, instead, efforts to propose and test ways whereby the police could solve more crimes, make more arrests, and gather better evidence.”

But, time for the second half of the story. Starting in the early 1990s, the homicide rate (and crime rate more generally) plummeted. By 2014, its local nadir, homicide rates are slightly below where they were in the 1960s. What happened? That’s a topic of hot dispute. But the story that I find persuasive (based in large part on Frank Zimring's The Great American Crime Decline) is three part. About half of it is the Baby Boomers aging into, and then out of, their peak crime-committing years. Another small but appreciable fraction is mass incarceration, although those effects have diminishing marginal returns. And then a big part of it is, more or less, the ascendancy of the beat cop. Under Clinton, we put maybe 60,000 new officers on the street (thanks, Joe Biden!). Beat cops in major cities (the archetypical example is New York) also get way more proactive about preventing crime, the fruit of Kelling and Wilson's “broken windows” work. Those investigators, meanwhile? The homicide clearance rate declines more or less continuously from 1960 to, well, today. In other words, America enjoys historically low homicide rates today for reasons totally orthogonal to the theory of the problem implicitly presented by Simon.

So that's the national story. Let's talk about Baltimore.¹ In 1988, when Simon wrote the book, Baltimore PD reported 234 murders to the FBI, equivalent to about 30 murders per 100,000 city residents. Last year, Baltimore recorded 335 murders, equivalent to 59 murders per 100,000. Clearance rates, as you mentioned, are basically non-existent. I chart both figures in the plot above, and what you can see is that both are long-term trends. Clearance rates have been steadily dropping for years, and homicide rates actually rose as they declined elsewhere. The big jump at the end of the homicide rate series is 2015, when the death of Freddy Gray causes mass protests and a spike in homicide that hasn’t abated. But the problem was there before that.

What’s most remarkable here is that, for Baltimore, the great homicide decline basically didn't happen. Instead, crime keeps getting worse and worse, and clearance keeps getting worse and worse. People don’t really want to work for BPD anymore—I met a cop recently who drives an hour and a half each way from his home in Baltimore to work for DC police, because he’d rather work here than there.

Given that context, I suppose to me the book is interesting for two reasons. One is as a story about how we thought (wrongly, in my view) about crime control in the 1980s. But more than that, two is about the mismatch between the cold calculations I just did and our deeply felt, instinctive response to the stories of crime and punishment that Simon lays out. If investigation is just meaningfully less important than walking the beat, then why is it so intuitive to us that the homicide investigation unit is where the most important stuff happens?

Jane: I’m going to come at that question in sort of a roundabout way, because I think Simon — or at least the homicide detective whose Weltanschauung he’s more or less adopted for the duration of the book — has an answer, and it’s not what you might expect. (Generic “you” here, I don’t think you were surprised.) Because none of the process of investigating and prosecuting a homicide is really about the victim.

There’s this phrase that pops up a couple of times in the book: the real victim. To a cop every dead body is, technically, “the victim” — the victim was shot in the head, the victim was stabbed, the victim was found in the trunk of an abandoned car — but most of them aren’t actually, you know, victims. The standard-issue Baltimore homicide victim is someone like Gregory Taylor, a thirty-one-year-old heroin addict selling baking soda to other junkies at $10 a cap until someone takes exception to the con and puts a dozen .22 slugs in him, or Roy Johnson, a Jamaican from New York who shows up with a bodyguard to shake down a man who owes him cocaine money but ends up on the wrong end of a burst from his own bodyguard’s Ingram when the other man uses him as a human shield. In other words, they’re stupid drug murders, mostly of people who kind of had it coming and who could very easily have been the killer if the situation had been only slightly different. (Among other things, this book is a great reminder of how unbelievably stupid most homicides are.) The real victims, on the other hand, are much rarer, but they’re innocent children like Latonya Wallace:

And like Latonya Wallace, there are those rare victims for whom death is not the inevitable consequence of a long-running domestic feud or a stunted pharmaceutical career. Poor souls like Henry Coleman, a forty-year-old cab driver who picks up the wrong fare at Broadway and Chase; and Mary Irons, age nineteen, who leaves a downtown dance club with the wrong escort and is found cut up behind an elementary school; and Edgar Henson, thirty-seven, who is leaving an east side 7-Eleven when a group of teenagers announce a robbery and then, without warning, begin blasting away. The gang takes two dollars in food stamps, leaving behind a quart of milk and a can of Dinty Moore stew. And Charles Frederick Lehman,² fifty-one, a Church Home hospital employee whose last moments on earth are consumed by the carry-out purchase of a two-piece extra crispy dinner from the Kentucky Fried Chicken on Fayette Street. Lehman doesn’t make the twenty feet between the restaurant door and his Plymouth; he is found spread-eagled on the rain-soaked parking lot, his wallet gone, the contents of one pocket spread across the asphalt, the chicken dinner lying in a puddle near his head.

The detectives’ attitude is different, more respectful, when it comes to the real victims — “[i]t’s one thing,” Simon writes, “for a detective to cut up with the uniforms when they’re gathered around some dead yo; it’s another entirely to behave that way when the cases involves a young wife with her blouse pulled up, her throat slit open and her husband waiting in the company lot” — but it’s not like they try harder for them. They’re already trying as hard as they can on all the dumb drug murders! Because it’s not actually about justice for the victim: even the story of poor blind Gene Cassidy, the one case in the book where the victim isn’t a corpse, falls strangely flat when we come to his shooter’s conviction. Identifying and punishing the perpetrator can’t bring back what’s lost.

They do it because a killer has taken something that was not his to take, and even if the life taken was that of a dope fiend named Pee Wee who owed his dealer $35, well, it wasn’t really Pee Wee’s to throw away either. It belonged to Someone Else. And despite a judicious dose of irony, Simon really does imbue the entire process with religious undertones: the book’s epigraphs are Deuteronomy 21:1-9 and a forensic manual’s description of gunshot wounds, and he quotes New York homicide commander Vernon Geberth’s line “Remember, we work for God” before calling it “both maudlin and pompous.” But that’s more or less how he (and by implication a homicide detective) sees it too, when he pauses to think about why it matters.

The book opens with a long passage about what it’s like to be murder police; he closes it with a similar one on what it’s like to be a murderer, wrapping up with:

In four months, you’ll be a trivia question to these men. In four months, when the carbon-sheet court notices show up in their mailboxes, the men who took your freedom will look down at your name in computer-embossed type and wonder who the hell you are: Wilson, David. Jury trial in part six. Christ, they’ll think, which one is Wilson? Oh yeah, the double from Pimlico. Yeah, that brain-dead with the story about the Jake.

In time, your tragedy will be consigned to an admin office file drawer, and later to a strip of microfilm somewhere in the bowels of the headquarters building. In time, you will be nothing more than a 3-by-5 index card in the suspect name file, packed into the T-Z drawer with about ten thousand others. In time, you will mean nothing.

But today, as the wagon man checks your cuffs and checks his paperwork, you are the precious spoils of one day’s war, the Holy Grail of one more ghetto crusade. To the detectives watching you leave, you are living, breathing testimony to a devotion that the world never sees. To them, you are validation for honorable lives spent in service of a lost cause. On this fading December afternoon, you are pride itself.

But this raises some interesting questions about what happens when it’s a cop who kills someone. Because if homicide investigation is important not as crime prevention but as balancing the cosmic scales of justice, well — are the scales knocked out of whack by a police-involved shooting? The police department’s authority to use lethal force depends on the answer being “no.” A cop is allowed to take what Pee Wee’s irritated dealer was not, but to allow him to do that we must believe that he’ll know whether it’s appropriate and necessary, infallibly, with perfect judgment — and of course that’s impossible, because police officers are human beings and will make mistakes. But a myth is necessary.

Simon comes to the department at a moment of institutional transition from one myth to another. It used to be that a cop would stand by his shooting and in turn the department would stand by the cop, even if the shooting was bad; the myth was that police did not make mistakes so anything the police did was necessarily the up-and-up even if it involved obvious racial animus or brutality. Over the course of the long 1970s, though, the world changed and the department changed with it, and the new myth is that only bad cops make bad shootings. “If the shooting was good,” Simon writes, “you were covered, though even the most justified use of force could no longer occur in Baltimore without someone, somewhere, getting in front of a television camera to say that police murdered a man. And if the shooting was borderline, you were probably still covered, provided you knew how to write the report. But if the shooting was bad, you were expendable.” And so the drama of the John Randolph Scott shooting, still unsolved, because while the officer who was chasing him couldn’t have shot him, whoever did was probably another cop who fled the scene because he didn’t think the department could — or would — protect him. And of course the transition has only continued since then…

Homicide has a lot to say about, well, homicides, but at its core I think it’s a book about the relationship between individuals and institutions: the ways they give us meaning and purpose, channeling our disparate selves towards a unified end, but also necessarily conscript us as cogs in a machine with its own inertia.

Charles: I think you are correct that the book understands homicide investigation to be basically retributive in function. But that is, in the context of how we think about policing and criminal justice in contemporary America, actually a fairly controversial thing, so I want to dwell on it for a second.

People — and by people I mean the liberals who have set the criminal justice agenda for the last several decades — do not like retribution. It is usually described as backwards, atavistic, primitive. This is because it is at odds with a utilitarian theory of justice, and indeed, criticisms of retribution stretch back to the earliest utilitarian legal reformers. In this view, the goal of the criminal justice system is the prevention of crime. That means through deterrence, before the crime happens; and incapacitation and rehabilitation after it happens. Which one you prefer is often a matter of mood affiliation. But they are really all, to abuse a metaphor, three sides of the same coin. (Stephanos Bibas dwells on this a little bit in his very good The Machinery of Criminal Justice.)

It's harder to make a case that retribution has an instrumental function such that it satisfies utilitarian concerns. To be fair, retribution is deterrent, i.e. tit requiring tat is about stopping further bad behavior. Thus the theory that the state monopoly on violence is mostly about stopping cycles of retribution, by ensuring that transgressions are met with insurmountable power. But this account of instrumental retribution breaks down in the cases in which retribution is most called for. When someone is brutally murdered, there's no victimizing left to be done, so punishment has no deterrent function. Indeed, murderers are relatively unlikely to murder again, a fact that progressive non-profits use—in the example par excellence of anti-retributivism—to argue that they should get shorter sentences.

I can make that argument, and yet almost nobody will concede “yes, the dead person can never be harmed again; and yes, the murderer might never kill again; therefore he should not be punished.” And this is because everyone is an intuitive retributivist, and it takes a lot of work to meme yourself out of it! And this is — to return to the actual book — why we root for Simon's characters: not because they are bringing down the crime rate (as I have argued, they are not), but because they are on the side of justice, of restoring the cosmic order.

One of the topics Homicide is really interested in is the process of interrogation, of getting criminals to admit to their offenses. As Simon observes, this used to be a fairly brutal affair! But at least in the precinct he's working with, in the time he's working, it's really not. And Simon is, bluntly, enough of a lib that he would have told us if it were. Instead, there's some great discussion of the cardinal rules of interrogation (homicide rule #4: “An innocent man left alone in an interrogation room will remain fully awake, rubbing his eyes, staring at the cubicle walls, and scratching himself in dark, forbidden places. A guilty man left alone in an interrogation room goes to sleep”). And he shows us a lot of the tricks. There’s a hilarious story about administering a “lie detector” test by preloading a Xerox machine with paper that says “false,” “true,” and “false” and then having the suspect touch the machine and press the copy button at the right time.

But the funny thing about criminal interrogations is that—and this is also a point that Simon drives home—the only thing the offender has to do is keep their mouths shut. If they do, they're basically free! (Rule #8: “In any case where there is no apparent suspect, the crime lab will produce no valuable evidence. In those cases where a suspect has already confessed and been identified by at least two eyewitnesses, the lab will give you print hits, fiber evidence, blood typings, and a ballistic match.”) And at root, the reason that offenders do end up talking is not really intimidation. It's usually that the cops give them a way to frame it like their actions, to which they admit, weren't really wrong. In other words, why do offenders talk? Because they know they've done something wrong, and they want to admit to it, while avoiding their sense of guilt. In other words, the chief tool that homicide has to get people to talk is the fact that even homicide offenders are retributivists.³

To return to the point about institutions, the idea that you need to have an institution responsible for addressing retributive matters is far older than utilitarianism. And while we would prefer to think that we have exceeded that need, everyone still basically intuitively believes in it. Indeed, even in the case of an officer involved shooting — we understand that sometimes the guy deserved it (because he was an active threat) and sometimes he didn’t, and we try to adjudicate that. The shifting norms around how OISs are treated are really shifting epistemic standards — from a cop’s word being law, to a cop’s word being questionable, to (today, often) a cop’s word being less than worthless. But everyone still agrees that there’s a question of right and wrong there, and on an atavistic level we root for the guy who is on the side of right!

Jane: That Xerox lie detector bit is so great that it opens the fifth season of The Wire (here’s the clip). In fact, most of the really good stories in Homicide ended up in The Wire sooner or later; the famous cold open of the whole series, for instance, comes pretty much verbatim from the book. And yet, curiously, neither the book nor the Homicide TV show ever reached the level of cultural salience that The Wire did. Of course part of that is that you’d never catch someone like Barack Obama enthusing about a series on network TV, but there’s a bigger reason too: the liberals who set the criminal justice agenda and write the columns about prestige cable shows don’t want to understand why crime actually happens. They want to believe it’s simply due to a lack of material resources, which we can solve by giving people money and/or stuff. And all of us, when it comes to entertainment, would prefer something like a Mafia flick: the bad guys should have clear agendas and long-term plans. They should be men whose violence we decry but whose agency and archaic honor codes nevertheless appeal. I mean, I certainly don’t want to watch a TV show about a bunch of teenagers who shoot a guy over $2 in food stamps, but that’s the reality of most crime in America. It’s all hookers stabbing each other for the last cap of heroin, or young men stealing a Kia to go joyriding, or a guy with a gun and poor impulse control getting mad, and none of that is going to be solved by an expanded safety net. Though I suppose that since 1988 the heroin thing has been solved by fentanyl! Womp womp.

If nothing else, this book should be required reading for anyone who thinks we should defund or abolish the police. Are the police perfect? Obviously and definitionally not, they’re human beings and a police department is a human institution. Is it a good idea to pretend that they’re perfect, the way the Baltimore did in the 1950s and 60s? Probably not. But there are a lot of stupid impulsive people out there who do horrible things to other human beings out of malice or wounded honor or simply because it seemed like a good idea at the time, and there has never been a society without people like that.⁴ I have no patience at all for claims that if we only radically restructured society enough it wouldn’t happen; it’s always going to happen, at least a bit, and when it does people are going to notice and care. They’re going to recognize that the cosmos is out of whack. They’re going to want it fixed. And if there’s not an impersonal institution like a police department there to do the job, well, that’s how you get lynch mobs. Personally, I’ll take the cops.

1. The data for this chart come from the following sources. For 1964 to 2019, homicide and clearance counts, as well as population counts, come from the FBI's Uniform Crime Report, via Jacob Kaplan's excellent crime data tool. Because of the NIBRS transition, this data series ends in 2019. So 2020, 2021, and 2022 homicide counts are from the Baltimore Sun, as are 2020 and 2021 clearance rates. 2022 clearance rates are from Fox Baltimore. Population estimates for those years are from the Census Bureau. Also, I am breaking all the rules by giving you a two-y-axis chart, may Tufte forgive me!

2. Creepy coincidence or veiled threat from the lady who told you to read this book? You be the judge.

3. This, by the way, is the thing about the final interrogation of Latonya Wallace’s likely killer, “the Fish Man,” who near the end of the book is interrogated by a professional interrogator. He almost breaks...but he doesn’t! Because he's a child rapist and a sociopath! No moral compass!

4. Though there are certainly more of them doing it more often in some times and place than in others, because culture matters!