Friday, February 01, 2008

Venkatesh's Gang Leader For a Day and IRBs

Yet another IRB post to kick up some kerfuffle between Daniel Goldberg, Bruce Boyden, and maybe 1L+:

This is the book by Sudhir Venkatesh, who basically embedded himself in gangs to do his ethnography for his dissertation.

This is the exchange with Venkatesh on Slate.

Tyler Cowen thinks the book is interesting, but "evil."

Over at my new favorite academic blog, Scatterplot, Jeremy Freese has my favorite comments on the book (but read the comments to the post too):

1. If you are going to test a survey in an obviously dangerous public housing project, do not have your first question be arguably the worst survey question I have ever seen. (here)

2. If you hear a person planning a drive-by shooting of another person, you are under a legal–as well as, some of us might add, ethical and moral–obligation to do something about it. (This lesson, according to the author, was not learned until four years had been spent doing fieldwork.) (here)

3. If you are told information by interviewees, you should not divulge that information to criminals you know already are extorting money from some of the interviewees. (here)

Remember that exchange between Bruce Boyden and Daniel Goldberg over whether there was any First Amendment right to conduct surveys and interviews, which pretty much is what makes up an ethnography?

Remember the article mentioned by Bruce by Philip Hamburger on how IRBs requiring prior approval unconstitutionally limit speech?

Well, how about this case, you guys? Seems like an interesting as-applied case to me. This also, as Jeremy notes, some other interesting legal issues, like whether Venkatesh was legally (and morally) obligated to alert the authorities of imminent unlawful activity.

From Scatterplot, here are some choice comments to kick off the discussion:


My Ph.D. institution was already well into the new stronger rendition of IRBs by that point, and there was a project not long after I arrived where a student was told he could not use ethnographic data collected for his dissertation because he tried to obtain retrospective IRB approval instead of approval beforehand. It’s just odd to see someone talking about a case from roughly the same time in which there seems to be no IRB presence and instead the only stricture anyone seems to care about is the law. (And with more vulnerable participants, to boot!) I’m not sure if Chicago sociology just had/has a FAR more blithe attitude toward human subjects protections than what I’m used to.


I think that to the extent the field cares about ethnography as a research approach, we should be advocating less, not more oversight from IRBs. (Which themselves reflect changes more in the university’s concern for culpability from the state rather than the heightened desire to police morals.) Didn’t Becker once argue that any research will always ultimately be guided by his/her own conscience? And in any case, name a time that a person has truly been hurt by a social scientist’s study…


I agree that IRBs are perhaps a red herring or mostly for legal cover but ummmm… it sounds as if some of the low-level hustlers in the Venkatesh study were ‘hurt’ as a direct result of the study (link #3 above). So, I can name one.

I think we underestimate the harm that results when we are aware of something problematic and choose to do nothing. Would the problem/harm have occurred if we weren’t present? Probably yes, but don’t we have an obligation to prevent it when we become aware of it? That’s Becker’s point — IRB approval doesn’t absolve us from behaving with conscience and responsibility in the field.


“Yet would SV’s contribution to urban sociology be derailed if he blew the whistle?”

Uh. Are we seriously debating this? Are “contributions to urban sociology” really worth weighing against the protection of human subjects from obvious potential harm?


The definition of a “human subject” from the federal rules for human subject research, 45 CFR 46.102(f), is “a living individual about whom an investigator(whether faculty or student) conducting research obtains 1) data through intervention or interaction with the individual, or 2) identifiable private information.” Private information includes information about behavior that occurs in a context in which the participant can reasonably expect that no recording is taking place and information the participant has provided for a specific purpose and can reasonably expect will not be made public (for example, a medical record).

Research is defined in the federal rules as “a systematic investigation, including research development, testing and evaluation, designed to develop or contribute to generalizable knowledge.” 45 CFR 46.102(d). We construe this to include formal investigations, from which the results will be publicly disseminated, and pilot and exploratory research. The definition also includes research undertaken by students for the purposes of independent study, theses, or dissertations. Generally the definition does not extend to: 1) classroom activities which are designed to educate students about research methodologies or simulate research activities, or 2) activities conducted to improve the quality of teaching. (See, Student Research Policy)
Both Jay & Jeremy: Campuses differ in just how broadly they interpret the second paragraph over in the social science & humanities end of the campus.

Jeremy, again:

Just so we’re clear, if an enterprising graduate student were to hear a gang leader describe a plot to kill a specific person, and that person was subsequently killed, the issue would not be whether the student is an “unethical ethnographer,” but whether they are an “accessory to murder.” My understanding is that even attorney-client privilege does not apply to the case of a client talking about a plan to commit a future crime, although a more legally inclined reader may wish to confirm or correct that.

(In Gang Leader for a Day, there is nothing saying that the gang leader actually killed anyone then or at any other time, nor do I believe it said the drive-by shooting(s) alluded to were intended to kill, so maybe sometimes such things are just meant to scare or whatever.)


I certainly understand the issue of decisions made in the heat of the moment and appreciate reflectiveness about it afterwards. I mean, the only reason one can criticize any of this is because Venkatesh provides an open description of it, which is to be commended. (Granted, he provides the open description in the context of a book that’s going to make him a lot of money, but still.)

Even with all this, though, the durations of naivete he provides in his account simply astound me. I mean, I get that many legal/ethical/moral ramifications of fieldwork will not be immediately apparent, but a person spends four years (FOUR YEARS!) in the field without apprehending that there are legal/ethical/moral implications to having research participants tell you about crimes (SHOOTINGS!) that they are committing or planning to commit–an issue that the reader appreciates immediately and that olderwoman teaches in her undergraduate methods course.

Likewise, I certainly understand somebody letting something told to them in confidence slip in conversation, but a person spends three hours (THREE HOURS!) recounting the off-the-books income for dozens (DOZENS!) of interviewees to a gang leader and a corrupt official that he already knows regularly and ruthlessly shake down various people in this housing project for money. My jaw drops, I scratch my head, I pace around my apartment all agitated.

I’m not myself an ethnographer, but I know plenty, and within the context of how I’m used to hear them talk about their work, these stories just astonish me. Then, on top of it, this is going to be the most publicly visible statement out there of how sociology ethnography works. It’s very disconcerting to me.


( I am moving today and am unsure when I'll have internet access again. It depends on whether AT&T hooked my new apartment up.)