In early December 1984, an undercover police officer named Marcellus Ward met with a pair of heroin dealers above a candy store in southwest Baltimore. Ward had planned to make a buy, then an arrest. But when Drug Enforcement Administration agents stormed into the building, one of the dealers panicked and shot Ward to death. The next day, Kurt Schmoke listened to the recording from Ward’s body wire. A friend of the slain detective, Schmoke was then Baltimore’s 35-year-old chief prosecutor. The incident, he would say later, prompted him to rethink the drug laws he had spent six years enforcing: Setups and stings and jail terms hadn’t curbed the violence associated with the drug trade, let alone reduced drug use.

In 1988, soon after his election as Baltimore’s mayor, Schmoke proposed easing many drug laws and repealing others–in part to undercut the black market, in part to focus resources on reducing demand. It was startling talk from a big-city mayor, and Schmoke’s call for decriminalization got him on Nightline and on the front page of the New York Times. But of course no mayor can decriminalize drugs. Schmoke was soon overshadowed by national drug czar William Bennett, who preferred escalation of the drug war, not reform.

In 1992 Schmoke returned to the drug issue–this time with a strategy that went beyond rhetoric. That year he appointed Peter Beilenson, then 32, a young and relatively inexperienced doctor, as health commissioner. The two officials then set about increasing capacity for drug treatment, pledging to continue until they reached “treatment on demand.” In the past three years, the city’s treatment slots have doubled, making Baltimore a case study for the promise–and problems–of universal drug treatment.

Aside from Baltimore, San Francisco is the only locality to even set a goal of guaranteed, immediate treatment for any drug user who wants it. But San Francisco has the luxury of booming tax rolls, while Baltimore is characteristic of declining cities with the most pressing drug problems, the most need for change and the least resources. With an estimated 60,000 addicts, the city regularly ranks first on lists of drug-related emergency-room visits and per capita violent crime. The city has lost at least 200,000 people since 1970, and its population drops by another 1,000 each month. Boarded up or burned-out homes scar most blocks–in total, 40,000 are abandoned. When I accompanied police Lieut. Michael Kundrat on his evening shift through the Western District, he pointed out each of the area’s fifteen active streetcorner drug markets, which he largely blames for forty-five murders (out of 314 citywide) last year. “If you ask what is one of this city’s biggest problems,” he says, “there’s no question that it’s drugs.”

When Beilenson went over the budget in 1995, he found that the city was missing its best opportunity to ameliorate that problem. Of $15 million spent annually on treatment, only $350,000 was coming from the city. And the city’s 4,000 slots could serve only 12,000 people a year–a pittance compared with the total addict population. Last year Schmoke ordered city departments–including police, housing, social services and health–to come up with budget cuts totaling $10 million over three years. That, plus new state money and grants from the local Abell Foundation ($1 million a year for three years) and George Soros’s Open Society Institute ($2 million), pushed the total drug-treatment budget to $32 million and the number of slots to 7,500. At the end of this year, the slots are expected to number 8,100.

Critics note that it’s taken Schmoke more than a decade to reach these goals. Others say that if he deserves credit for the drug-treatment plan, it is for hiring Beilenson and giving him wide latitude. Beilenson grew up in West Los Angeles the son of a state legislator (later congressman). His interest in drug policy–as with his other focus, universal healthcare–is that of the pragmatic liberal, not an experienced user. (Beilenson says he’s never tried illegal drugs. When he notes, “I know what a speedball is. I know what a cooker is. I know what ice is,” he sounds like a proud student.)

The soft-spoken Beilenson quickly shook up a system that had grown lethargic. Two years ago he spent an afternoon calling the city’s public treatment centers, telling them he was “Todd Jackson,” a three-times-a-day heroin user for four years, and asking “When can I start treatment?” Only one of the twenty-three programs he called would even put him on the waiting list; the others insisted he call later or come in person. He visited three other treatment centers but none would let him through the doors–telling him, via intercom, to come back when they were open for intake. So Beilenson ordered the programs to coordinate their intake, and he created a citywide referral line.

Far more important is the added treatment capacity. The waiting list for outpatient counseling–which is least in demand–has dropped to zero. Waits for methadone and inpatient programs, which account for two-thirds of slots, still range from about a week for some clinics to as much as a month for others. (Despite the citywide referral line, many people still contact programs directly.) When Beilenson makes the case for more slots, he always returns to the bottom line: The cost of treatment ranges from $3,500 a year for a methadone slot to about $35,000 a year for an inpatient slot (with a twenty-eight-day program, this would serve twelve people). Meanwhile, a drug user supporting a $50 habit might easily need to steal $300 worth of property each day. (Baltimore police estimate that fenced goods sell for one-sixth of their retail value.) That, plus the costs ranging from broken car windows to security guards to AIDS treatment to prison cells, adds up. In a finding that has been repeated elsewhere, the California Department of Health under Governor Pete Wilson followed addicts before, during and after treatment and found that every dollar spent on treatment yielded seven dollars in savings.

Which is not to say that treatment yields permanent abstinence–from either drugs or crime. University of Maryland studies have found that people in outpatient treatment programs commit crimes, on average, forty-one days a year, which seems high until it’s compared with the addict population not in treatment–for whom the average is 248 days.

Schmoke says he now prefers the word “medicalization” to “decriminalization” but that his goal is still to treat “drug abuse primarily as a public health problem and not primarily as a criminal justice problem.” And the refrain of the city’s police commissioner, Thomas Frazier, is that “we can’t arrest our way out of this problem.” Still, Baltimore police have hardly laid down their arms. From 1988 to 1998, Maryland’s prison population climbed from about 13,000 to more than 22,000; a significant percentage of the increase was the result of drug arrests in Baltimore. Last year city police made 90,000 arrests, 85 percent of which were for drug possession, sales or some offshoot of the drug trade.

Hundreds of those arrests came near the corner of Monroe and Fayette in West Baltimore–a corner made infamous by the recent book The Corner by David Simon and Edward Burns. A city-run needle-exchange van comes to Monroe and Fayette each week. When Beilenson took me to visit the program, we met a woman who embodied his wish for drug treatment–and his frustration. Wearing a vacant look and moving in jerk-steps, she pushed a handful of needles into a canister and picked up a package of new ones. I followed her onto the street, where Beilenson joined us. Jackie, who wanted to be known only by her first name, is 37 years old and looks twenty years older. She has five children and uses between $50 and $200 worth of heroin and cocaine a day. She’s been in jail five times, for soliciting prostitution and for “boosting” (shoplifting). She spoke with evident anguish about this–about the jobs she had lost, about “what I put my parents through.”

“Have you thought about getting into treatment?” I asked.

“Oh yes,” she said. “If I could right this minute, yes I would. Yes I would. Next time I get the chance to get into treatment…” She trailed off. The meaning of “right this minute” was clear enough. At that moment, Jackie was in withdrawal–the low point of the addiction cycle and a time when treatment referrals can be most successful. Beilenson gave her a phone number to call, but her interest in treatment seemed unlikely to last beyond the next score.

As we drove away, Beilenson made his diagnosis: “This is the typical addict in Baltimore. In and out of jobs. Boosting. Prostituting. Did you see the yellow in her eyes? She has serious liver issues–and if she doesn’t get taken care of she’s going to end up very sick. What I would have liked to do is say, ‘Look, we’ve got treatment for you. Right now. This afternoon. Here’s where to go, and here’s a token for a cab.'” But Baltimore can’t even offer immediate care to people who are begging for treatment. Despite shorter wait lists, psychiatric hospitals report that addicts regularly feign threats of suicide or violence to gain entry. When the city added a twenty-four-hour staff to its treatment hotline, it turned out that 80 percent of after-hours calls came between 7:30 and 8 am. The morning’s slots are dispensed beginning at 8 am, and the callers were trying to be first in line.

People who already want treatment, Beilenson argues, shouldn’t have to compete for slots. Everyone benefits if they get off the streets as quickly as possible. But the real advantage with immediate treatment would apply to people like Jackie–people who struggle with an addiction but, in the five or ten or thirty days it takes for a slot to open, are likely to have a change of heart, switch addresses, lose their phone or even land in jail. And an ideal treatment system would go even further, cajoling people who might be amenable to treatment but who would never seek it out.

Just how much will Beilenson need to make treatment-on-request a reality? He suggests $30 million to $35 million, on top of the current budget of $32 million. But a precise figure is impossible because no one knows how many people would seek treatment if it were easily, immediately available. Even the total number of addicts is unknown–60,000 is just the best guess. Finally, adding capacity to the existing system is only the beginning. At the Glenwood Life Counseling Center, which offers methadone and therapy for addicts, three positions went unfilled for months. The reason, says executive director Frank Satterfield, is that the starting salary for a job that requires a bachelor’s degree and offers a slim chance of raises is only $21,000. Satterfield says he would need to pay $27,000 to be minimally competitive. Applied to treatment programs across the city, such salary increases would require $2.4 million. Even fully staffed, Glenwood can barely stay on top of administering 340 patients, soon to increase to about 500. With rare exceptions, therapy is done in groups. Vocational training, family counseling, good psychiatric care–these services, which Satterfield says he took for granted when he first entered the field in the early seventies, are not possible with his budget.

The irony is that these deeper needs will only become apparent once the first steps are taken. In other words, the further Baltimore moves down the path to “medicalization,” the longer the path seems. And though political obstacles have so far been few–there is no Republican Party to speak of in Baltimore–that may change.

A hint of future conflict came last summer, when Beilenson expressed interest in a plan by Johns Hopkins researchers to conduct a trial in heroin maintenance. The idea was to replicate a now-legendary Switzerland study in which heroin addicts who had failed in repeated treatments were given injections of the drug in a clinical setting. The results were impressive (though not necessarily definitive): Many in the test group found housing and employment, and the percentage committing crimes went from 59 percent to 10 percent. In casual conversation about such a study in Baltimore, Beilenson told a Sun reporter–while they watched their kids play soccer–“It will be politically difficult, but I think it’s going to happen.” The Sun printed these remarks, and sharp criticism came quickly–from Governor Parris Glendening, City Council members and even the Mayor. “This administration has no intention of initiating a heroin maintenance program,” Schmoke told the Sun. But in a recent interview, Schmoke struck a different tone. He had to rebuke Beilenson, he explained, because the commissioner floated a radical idea without building consensus. But, Schmoke continued, “if you ask me what should be a part of a good public health system to combat substance abuse, many different treatment options should be part of that system–plus this last one, medical distribution of certain drugs.”

This puts Schmoke and Beilenson on a collision course with state officials, most notably Lieut. Gov. Kathleen Kennedy Townsend, who said such an idea “undermines [the] whole effort” of convincing young people that “heroin is bad.”

Indeed, while city officials try to build a model program of treatment on demand, Townsend is staking her reputation on a different model: coerced treatment. Last year the legislature approved Townsend’s proposal to administer twice-a-week drug tests to 25,000 people on probation or parole. Under the $2.9 million program, positive tests for cocaine, heroin or marijuana would trigger a series of escalating sanctions–for a midlevel offender, for example, the first failed test would lead to two days in jail, then five, ten, thirty, forty-five and finally a return to court for parole violation. “You don’t have to want treatment for it to work,” says Adam Gelb, Townsend’s policy director.

Researchers and treatment providers have mixed opinions on coerced treatment. “If you have leverage, you ought to use it,” says Dr. Robert Schwartz, who directs the University of Maryland School of Medicine’s division of alcohol and drug abuse, and who consults with the Open Society Institute in Baltimore. Still, Schwartz urges that coerced treatment not squeeze out voluntary programs–which seems a real possibility were Townsend to get her way. Beilenson urges the state to spend more on both forms of treatment, arguing that the big savings will come from keeping people out of the criminal justice system in the first place.

The idea that drug users belong in prison has in the past few decades been carved deeply into US politics and won’t be easy to change. In New York, for example, a recent study found that the state spends $680 million a year to lock up nonviolent drug offenders, and yet the legislature has stalled on even modest reforms of its draconian Rockefeller drug laws. Those laws, adopted in the early seventies, were quickly adopted by other states and by federal officials. Now, the damage is apparent even to many hawks. “We have a failed social policy, and it has to be re-evaluated,” national drug policy director Gen. Barry McCaffrey said this year. “Otherwise, we’re going to bankrupt ourselves.”

Of course McCaffrey won’t propose the next logical step, which would be to eliminate mandatory minimums and other harsh measures and redirect that money into treatment. During his 1992 campaign, Bill Clinton pledged to enact “treatment on demand.” He quickly abandoned the promise, not just because it would be expensive but because it runs contrary to a winning political formula: more arrests, longer sentences, more jails.

The experiment in Baltimore challenges that formula. Officials there are driven by the basic instinct that drug users need help, not punishment. But they also believe the policy makes fiscal sense. And they know they’ll have to support that view with hard numbers. That’s why Beilenson contracted with a team of independent social scientists (from Johns Hopkins and two other universities) to measure the crime, health and income of three groups of addicts in Baltimore: those who get treatment immediately, those who have to wait and those who get no treatment at all. The results of the $2.8 million study are expected in 2001. If it confirms Beilenson’s argument that “treatment saves money and treatment on request will save more,” the study could turn an iconoclastic experiment into a model for other cities and states. Today, the idea of universal, immediate treatment on a national scale is hard to imagine. But then, so was Governor Nelson Rockefeller’s idea that possession of two grams of cocaine deserved fifteen years in prison.