I agree with Casey Mulligan that participants in government surveys should be paid, and I think it should be part of the code of ethics for commercial pollsters to compensate their respondents also.
As Mulligan points out, if a survey is worth doing, it should be worth compensating the participants for their time and effort.
P.S. Just to clarify, I do not recommend that Census surveys be made voluntary, I just think that respondents (who can be required to participate) should be paid a small amount.
P.P.S. More rant here.
What were the reprocussions of the professor study? Did anything come of the outcry?
Have we any results on how our paying the respondent might influence who responds? Personally, I hate answering surveys, but I do it out of respect for scientific endeavour – a sense of “civic duty”. This might have fled if I was to be offered a monetary compensation. Didn’t the blood donor studies show something similar?
If you convert a mandatory (presumably randomly-sampled) national survey with an optional / voluntary survey, does this not make the bias problem significantly worse? I’m under the impression that self-selection is a much more difficult problem to deal with than adjusting for pseudo-random sampling from a population. I’m tying ‘survey’ here with ‘census’ since in Casey Mulligan’s case he is talking about wide-scale census economic surveys. As far as I’m concerned, when it comes to government agency-requested census forms, no reimbursement is necessary whatsoever — it’s part of being a resident of the country in question.
My IRB, and I assume others, prefers uncompensated surveys on the grounds that they are less coercive. The reasoning seems to be that if you aren’t paying a respondent, he or she will be/feel more free to decline to participate in the survey and/or skip questions than if he or she is paid. Also, the extent to which payment creates coercion also differs by the social and economic position of the respondent: if you’re making minimum wage, $20 for a half hour of your time is far more difficult to pass up than if you’re, say, a tenured economics professor at Chicago.
In theory, you could pay people regardless of whether they actually completed the survey, but that creates the wrong incentives. And a tiered payment system in which poor people earn less, nominally, than rich people for completing the same survey doesn’t seem like a great solution.
* I’d assume, that, after his handling of the 2008 crisis and a number of subsequent blunders, Casey Mulligan would be considered a thoroughly failed economist by now. (Of course, I’d assume that with regard to the entire Chicago school, and I’d be wrong there.)
* It’s a basic fact straight out of Freakonomics that, as soon as you start paying people for something that they normally would voluntarily do for free, their performance often goes down. If you start paying *and* you make it voluntary to respond, that introduces a massive source of bias, alluded to by the article where it says “the sample size would need to be increased to offset biased response rates.”
* The analogy between free & paid survey respondents vs. drafted & volunteer soldiers is extremely strained.
* Even if the government could afford to cough up the dough to increase the sample size and to pay some scientists to quantify the bias, it would be a big leap to say that private researchers should do the same.
This is not the consensus on blood donation. See for example, this post on Marginal Revolution:
http://marginalrevolution.com/marginalrevolution/2011/12/rewarding-altruism-blood-for-money.html
With Andrew’s P.S., this proposal seems analogous to jury service. Everyone has to remain in the sampled population but we compensate those that bear the unfair share of the costs. There seems some merit to that. Still, drafted soldiers were paid so the mercenaries vs. slaves polemics would still remain.
I have long thought that air travel could be made less awful if those (randomly, not fore cause) singled out for additional scrutiny were compensated for their time.