'Stop and Frisk' is Fair and Balanced

New York Minorities More Likely to Be Frisked

So blares the headline in the NYT.

The lede graf explains a bit more:

Blacks and Latinos were nine times as likely as whites to be stopped by the police in New York City in 2009, but, once stopped, were no more likely to be arrested.[Emphasis Added]

The article has what I would call a moderate bias against the stop-and-frisk procedure. But the article's lede tells you all you need to know - the police are obviously using similar approaches to choosing suspects to stop for all ethnic groups, otherwise the arrest rate by group would be different. For example, if the police were biased against non-whites in deciding who to stop, we should find a far higher arrest rate among the white suspects. I know I don't need to explain this to my highly select readership, but if the police were just willy-nilly stopping blacks and Hispanics but only stopping truly-bad whites, then the stats should show lower arrest rates for minorities vs. whites. Since the stats show equal arrest rates, we can commend the NYPD for doing a really good job of policing.

Does the article even remotely suggest this? Not at all - they do present this telling info up-front, but no interpretation is provided for the reader to understand its implication. How many Times readers understand this - 10%? At one point later in the article, the equivalent arrest rates are again mentioned, but as part of a lead-in to a critique of the practice as probably biased!

So this article is an example of the opposite of "burying the lede" - where the article's most newsworthy item is buried down in the text somewhere. This article's lede - that the statistics of stop-and-frisk show no evidence of bias - is presented at the opening as it should - but is essentially ignored and even disputed in the rest of the article.

In fact, towards the end of the article, a fact is brought up which further undermines the case for bias, but is presented as supporting it:

The Center for Constitutional Rights also studied poststop outcomes. It found that officers frisked more people in 2009 than a year earlier but that the rate of frisks for blacks and Latinos was much higher than it was for whites. It found that the police used force in 24 percent of stops — drawing a weapon, say, or throwing people to the ground. The police used force in 19 percent of the stops involving whites but in 27 percent of stops against Latinos and in 25 percent of those involving blacks.

Since the arrest rates were the same, this suggests that police were stopping more unhinged or disorderly characters among minorities, requiring more force on their part to subdue them. The discrepancy here is not particularly dramatic, but is presented as damning evidence of bias, when the opposite interpretation is more logical.

Overall, I counted 32 paragraphs - many are just one line. Eight explicitly support the police policy, 11 oppose it. The others simply present facts. However, several of the neutral paragraphs are used as lead-ins to opposing views. The article essentially presents the police, on one side, and the Center for Constitutional Rights on the other. There are 5 other commenters cited - of these, only one, Heather McDonald, supports the police.