Sunday, January 17, 2016

Political Scientists Complain about Lost Exemption

Political scientists and law professors are protesting the proposed elimination of the current exemption for research about public officials and candidates for office. The drafters may have eliminated the exemption by mistake, thinking that such research was exempted by other provisions of the proposed rule, which turns out not to be the case.

OHRP can't explain why it cut the exemption

Since 1981, 45 CFR 46.101(b)(3)(i) has exempted “research involving the use of educational tests (cognitive, diagnostic, aptitude, achievement), survey procedures, interview procedures, or observation of public behavior that is not exempt under paragraph (b)(2) of this section, if: (i) the human subjects are elected or appointed public officials or candidates for public office.” I was never able to nail down the exact origin of this exemption, but my best guess is that it was a response to the criticism leveled by political scientist Ithiel de Sola Pool, one of the leaders of the 1978–1981 social science campaign against IRB regulation.

The 2011 ANPRM proposed that “The existing six exemption categories would be retained as part of the new Excused category.” (76 FR 44518). Since the ANPRM did not warn anyone that they might lose the public-official exemption, none of the 2011 ANPRM comments seems to have commented on its value.

The 2015 NPRM’s proposed rule, however, does not include the exemption. As the NPRM states, “The rationale for this change in the proposed NPRM is that it does not seem appropriate to single out this category of subjects for different treatment in this way.” (80 FR 53951) In other words, the drafters eliminated the exemption because they wanted to eliminate the exemption.

Is so circular an explanation permissible under the Administrative Procedure Act? As Jason Schwartz and Richard Revesz note in their 2014 report, Petitions for Rulemaking, "‘bare conclusions,’ such as simply asserting that existing rules are adequate, will be found ‘not responsive’: where a petition raises alternative regulatory options, it “merit[s] some brief explanation of why the agency did not find it desirable to consider those alternatives.” But the NPRM was not a response to a petition, so I need some administrative law experts to weigh in.

OHRP officials offered a longer, but more tangled, rationale at the November 18 Town Hall in Philadelphia. As Peregrine Schwartz-Shea, Dvora Yanow, and Daniel Levin, note in their NPRM comment, Julie Kaneshiro, Deputy Director, OHRP, thought that interviews with public officials could proceed under the proposed exclusion at §____.101.(b)(2)(i), while Ivor Pritchard, Senior Advisor to the Director, OHRP, thought they could go ahead as “activities that involve focusing on the specific individuals about whom one collects information,” which presumably means §____.101.(b)(1)(ii). But, as Schwartz-Shea et al. note, the first applies only to research that could not reasonably damage the subjects’ reputation, which is inappropriate for research that may hold public officials and candidates to account. And Pritchard’s 101(b)(1)(ii) applies only to “oral history, journalism, biography, and historical scholarship activities,” not research in law and political science.

In any event, political scientists are angry and increasingly organized, and 91 comments on the NPRM mentioned the public official exemption. I have not read them all, but a quick sampling suggests that most or all of them want to preserve the exemption.

At stake are interviews, not experiments

In her January 7 Vox article, “The Obama Administration Is Quietly Trying to Make It Harder to Study Public Officials,” Michelle Hackman claims that the current value of the exemption lies not in its easing of interviews, but rather because

the newest wave of research, which involves using randomized field experiments to gauge politicians’ behavior, relies heavily on the exemption.

Take, for example, one recent study that tried to determine whether state legislators discriminate against their black constituents. The researchers sent state legislators emails from fake constituents, requesting help with registering to vote. The emails were identical but for the constituent’s name, which the researchers varied to sound white (Jake Mueller) or black (DeShawn Jackson). They found that, on average, white legislators responded less often to the emails sent by constituents with black-sounding names. And black legislators did the exact opposite.

Were that study conducted on private citizens, the researchers would probably have been required to ask for their prior consent, and politicians are unlikely to agree to participate in a study that measures how racist they are.

I’m not sure I buy this. For one thing, the current exemption covers tests, surveys, interviews, and observation. It’s not clear that it covers deceptive experiments of the kind Hackman describes.

And for all of their faults, IRBs do seem to understand that they can and should waive the informed-consent requirement for this kind of “audit study.” See, for instance, Katherine L. Milkman, Modupe Akinola, and Dolly Chugh, “What Happens Before? A Field Experiment Exploring How Pay and Representation Differentially Shape Bias on the Pathway Into Organizations” (2015) and Benjamin Edelman, Michael Luca, and Dan Svirsky, “Racial Discrimination in the Sharing Economy: Evidence from a Field Experiment” (2016), neither of which involved public officials or candidates for office, and both of which got IRB approval.

So while eliminating the exemption is a bad idea, and the NPRM’s non-explanation might not satisfy the Administrative Procedure Act, I think what’s at stake are the same interviews that Pool fretted about in 1979.

Finally, I question Hackman’s evocation of Obama’s name in her headline. Both Kaneshiro and Pritchard have been at OHRP since the George W. Bush administration. If they are the ones behind the proposed elimination, why name Obama?

No comments: