The March 1982 Atlantic article called "Broken Windows" by George L. Kelling and James Q. Wilson is a darned interesting artifact 32 years later.  It begins with an experiment with community policing and foot patrols in Newark, New Jersey in the mid-1970s.  We are, at that point, seeing the start of the use of technology in law enforcement and, of course, the start of globalization and the hollowing out of America's cities that resulted.  The idea of increasing foot patrols in high crime areas struck many as nonsensical.  Foot patrol was a punishment.  Good cops got cars.  Crime would not be deterred by a bunch of flatfoots.

Indeed, foot patrols had no effect on crime rates, report Kelling and Wilson a finding that they allow "may be taken as evidence that the skeptics were right."  But, they said, there were ancillary benefits.  The neighborhoods with foot patrols reported more favorable opinions of the police.  The police on foot patrols reported greater job satisfaction.  Crime rates remained the same but citizens said that they felt safer.  That feeling, argue Kelling and Wilson, shouldn't be underestimated:

"Outside observers should not assume that they know how much of the anxiety now endemic in many big city neighborhoods stems from a fear of "real" crime and how much from a sense that that the street is disorderly, a source of distasteful, worrisome encounters."

Now, Wilson writes a little about his own work on the history of policing which, he says, evolved from a the night watchman's role (maintaining order) to the modern role of crime solver.  Wilson clearly sees maintenance of order as a more vital function. Here is how he characterizes police actions in big cities during periods of mafia gangs and the like:

"Young toughs were roughed up, people were arrested "on suspicion" or for vagrancy, and prostitutes and petty thieves were routed.  "Rights" were something enjoyed by decent folk, and perhaps also by the serious criminal, who avoided violence and could afford a lawyer."

Kelling and Wilson aren't necessarily endorsing that state of affairs but that pesky issue of rights has consequences later on and it has big consequences decades after their recommendations were adopted by urban police forces around the country. Kelling and Wilson want the police to be able to enforce unwritten community standards.  That, they believe, leads to a sense of order, unity among neighbors and crime deterrence.  The problem is, these standards really are unwritten.  They aren't laws.  

Kelling followed a Newark foot patrol officer for a long time to see how he performed his job. The neighborhood was both a transportation and business hub. The officer, named "Kelly" in the article, has divided the locals into groups of "regulars" and "strangers."  The regulars, he divided into "decent folks" and "drunks or derelicts."  Drunks or derelicts were not to be hassled so long as they "knew their place."  For example, he'd allow public alcohol consumption if the bottles were in paper bags and if the drinking happened on side streets not on the main roads or bus stops.  Panhandling from people using mass transit was an arrestable offense.  "If a stranger loitered, Kelly would ask him if he had any means of support and what his business was; if he gave unsatisfactory answers, he was sent on his way." Further,

"Sometimes what Kelly did could be described as "enforcing the law," but just as often it involved taking informal or extralegal steps to help protect what the neighborhood had decided was the appropriate level of public order.  Some of the things he did probably would not withstand legal challenge."

Of course, there would be no legal challenge unless Kelly made the mistake of hassling somebody with money -- a mistake he would be unlikely to make.  Operating within the neighborhood, with the support of locals (especially business owners) Kelly is quite safe to deviate from the law.  Somebody with a complaint against Kelly would be a huge disadvantage.  Witnesses would side with Kelly.  Pursuing a complaint would be an expensive and difficult endeavor.

What Kelling and Wilson admit, near the conclusion of the article, is that Broken Windows policing is going to have a lot of unjust individual outcomes.  A lot of people are going to be hassled, cited, searched. questioned or arrested just for being in the wrong place, not for committing crimes that have actually harmed other people.  What we have here is, in essence, a Communitarian argument, at odds with the socially libertarian outlook of many cosmopolitan, urban dwellers.

"This wish to "decriminalize" disreputable behavior that "harms no one" - and thus remove the ultimate sanction the police can employ to maintain neighborhood order - is, we think, a mistake.  Arresting a single drunk or a single vagrant who has harmed no identifiable person seems unjust, and in a sense it is. But failing to do anything about a score of drunks or a hundred vagrants may destroy an entire community."  

The authors go on to say:

"...we have become accustomed to thinking of the law in essentially individualistic terms.  The law defines my rights, punishes his behavior and is applied by that officer because of this harm.  We assume, in thinking this way, that what is good for the individual will be good for the community and what doesn't matter if it happens to one person won't matter if it happens to many. Ordinarily, these are plausible assumptions. But in cases where behavior that is tolerable to one person is intolerable to many others, the reactions of others - fear, withdrawal, flight - may ultimately make matters worse for everyone, including the individual who first professed his indifference."

The problem here is that the police officers have to interpret community standards in order to enforce them.  The officers would also have to deal with the evolution of those standards. The officers, as the only sanctioned users of force in the community, would also get a pretty big say in what those standards are, whose voices count and when and how those standards could change.

I think you can see, in just my summation of the article, the roots of how Broken Windows in New York turned into "Stop and Frisk," and other unacceptable forms of harassment.  You can also see, just in the scenarios Kelling and Wilson either observed or imagined, that the police would need to be wise enough to only sparingly use their powers against people with the economic means to fight back.  It is only safe to take extra-legal steps, after all, against those least likely to be able to use the system to fight back or to assert their rights.

Though they don't mention drugs in the article, Kelling and Wilson argue forcefully against the decriminalization of victimless crimes like public drunkenness, prostitution and "pornographic displays." They seem to believe that standards are important even if they are irrational, unfair, or not applied equally to all members of a community. It is, in fact, unwritten and somewhat loose standards that empower the police under this model.  

An example from the article is that elderly people in a neighborhood may well experience fear and anxiety when walking past a gang of rowdy teenagers.  That anxiety is what Kelling and Wilson would want the police to address. A typical scenario might involve the foot patrolman approaching the teenagers, talking to them for awhile, getting them to quiet down and even ordering them to disperse.  It might be that under the law the teens have not been loud enough to violate any noise statute, that they have a constitutional right to gather in public and that appearing as if they might commit a crime is not the same as committing a crime.

From a civil libertarian point of view. these kids should not have to explain themselves to a police officer, nor should they have to follow orders to stop engaging in legal activity.  The very encounter with the police should not happen as it could escalate into an actual crime or result in a search or some other issue. Kelling and Wilson would not be worried about any of that.

This might be why so many individual cases of Stop and Frisk abuses fell on deaf ears for so many years. The policymakers behind such police work have internalized the Kelling and Wilson mindset which simply does not concern itself much with the particulars of individual police encounters so long as order is preserved.