segunda-feira, maio 11, 2009

Broken Windows

A propósito do editorial do director do "i" e desta resposta de João Gonçalves, impora notar que, como em tudo o que sucede nos Estados Unidos, há quilómetros de estantes sobre o assunto. Alguns dos trabalhos em revistas peer-reviewed mais recentes e/ou mais citados:

Broken windows: New evidence from New York City and a five-city social experiment- "evidence (..) provides no support for a simple first-order disorder-crime relationship as hypothesized by Wilson and Keller."

Carrots, Sticks, and Broken Windows - "it is important to emphasize that arrests for felonies have the largest effect on felony crimes and that the effects of broken-windows policing, although significant for some crimes, are not universally significant, nor are they of great magnitude."

The irony of broken windows policing: A micro-place study of the relationship between disorder, focused police crackdowns and fear of crime: "This study provided evidence for the broken windows policing hypothesis that disorder leads to fear of crime. Examining fear of crime among citizens at a micro-place unit of analysis revealed that higher levels of both perceived social disorder and observed physical disorder led to significantly higher levels of fear of crime. For those advocating broken windows policing approaches, however, the data also suggest reason for caution. Those living in an area that received the extra police presence (controlling for other factors in the model) also reported higher levels of fear."

E dois estudos que interpretam os mesmos dados chegando a conclusões diferentes. O primeiro é o mais citado de sempre sobre o tema:

Systematic social observation of public spaces: A new look at disorder in urban neighborhoods:
"observed disorder did not match the theoretical expectations set up by the main thesis of “broken windows” (Wilson and Kelling 1982; Kelling and Coles 1996). Disorder is a moderate correlate of predatory crime, and it varies consistently with antecedent neighborhood characteristics. Once these characteristics were taken into account, however, the connection between disorder and crime vanished in 4 out of 5 tests—including homicide, arguably our best measure of violence."


Spuriousness or mediation? Broken windows according to Sampson and Raudenbush (1999): "In sum, Wilson and Kelling's broken windows theory argues that disorder causes social withdrawal and low informal social control, which in turn causes crime. Sampson and Raudenbush, 1999 interpreted their own results as not supporting the broken windows theory because disorder did not have a direct effect on crime after controlling for collective efficacy. The problem with this interpretation is that the broken windows theory never argued that disorder had a direct unmediated effect on crime. According to the broken windows theory, disorder undermines informal social control, which in turn leads to an increase in crime. This is the core premise of the broken windows theory, a premise that Sampson and Raudenbush's results did not contradict."

E uma última frase neste último artigo que diz tudo:

"While debate in the popular press continues, academics are pushing forward in their attempts to understand the link between disorder and crime. Several articles have been published in recent years showing empirical evidence both in support of and against the broken windows theory (...) Thus far, no final consensus has been reached."

Era fácil se fosse só as broken windows, não era?

P.S. - Só para dizer que, não sendo especialista do assunto, até simpatizo com a tese. Mas os resultados das experiências dos psicólogos sociais, como acontece tantas vezes, não são fáceis de replicar e comprovar quando se sai de ambientes mais artificiais e se tenta generalizar...

P.P.S - Enviada por um leitor, uma referência adicional de Steven Levitt, tornada famosa devido ao argumento sobre os efeitos de Roe vs. Wade na criminalidade: Understanding Why Crime Fell in the 1990s.

Sem comentários: