This article’s Criticism or Controversy section may compromise the article’s neutral point of view of the subject. Please integrate the section’s contents into the article as a whole, or rewrite the material; see the discussion on the talk page. (November 2010)
According to most criminologists who speak of a broader “backlash”, the broken windows theory is not theoretically sound. They claim that the “broken windows theory” closely relates correlation with causality, a reasoning which is prone to fallacy. David Thacher, assistant professor of public policy and urban planning at the University of Michigan, stated in a 2004 paper that:
[S]ocial science has not been kind to the broken windows theory. A number of scholars reanalyzed the initial studies that appeared to support it…. Others pressed forward with new, more sophisticated studies of the relationship between disorder and crime. The most prominent among them concluded that the relationship between disorder and serious crime is modest, and even that relationship is largely an artifact of more fundamental social forces.
It has also been argued that rates of major crimes also dropped in many other U.S. cities during the 1990s, both those that had adopted “zero-tolerance” policies and those that had not. In the Winter 2006 edition of the University of Chicago Law Review, Bernard Harcourt and Jens Ludwig looked at the later Department of Housing and Urban Development program that re-housed inner-city project tenants in New York into more orderly neighborhoods. The broken windows theory would suggest that these tenants would commit less crime once moved, due to the more stable conditions on the streets. Harcourt and Ludwig found instead that the tenants continued to commit crime at the same rate.
In a 2007 study called “Reefer Madness” in the journal Criminology and Public Policy, Harcourt and Ludwig found further evidence confirming that mean reversion fully explained the changes in crime rates in the different precincts in New York during the 1990s. Further alternative explanations that have been put forward include the waning of the crack epidemic, unrelated growth in the prison population due to Rockefeller drug laws, and that the number of males aged 16–24 was dropping regardless due to the shape of the US population pyramid.
 Drawbacks in practice
A low-level intervention of police in neighborhoods has been considered problematic. Accordingly, Gary Stewart writes that “The central drawback of the approaches advanced by Wilson, Kelling, and Kennedy rests in their shared blindness to the potentially harmful impact of broad police discretion on minority communities.” This was seen by the authors, who worried that people would be arrested “for the ‘crime’ of being undesirable”. According to Stewart, arguments for low-level police intervention, including the broken windows hypothesis, often act “as cover for racist behavior”.