Reminder: For a comment to be considered it must be accompanied by your full name: a first name only or a pseudonym is not normally accepted. Please limit your comment to 1,000 characters (including spaces), and also avoid epithets and personal attacks.
The article by reporter Blythe Bernhard, “Smoking ban would not hurt state’s bars and restaurants,” appeared in the August 8th, 2013, St. Louis Post-Dispatch, the study making something of an impression. It was also reproduced on this blog.
One pro-smoking critic of the study was Michael McFadden, who wrote to the study’s primary author, Dr. Brett Loomis, with a list of questions. He followed up with reminders but never received a reply.
Although I don’t necessarily agree with Michael McFadden, or have reason to doubt the validity of Dr. Loomis’s study, I offered to post Michael’s abbreviated list of questions on this blog to give them a public airing, as below:MoGASP has allowed me to respond to the August blog entry on the CDC study by sharing a condensation of the email I sent, and re-sent, to Dr. Loomis, the study author. Although my questions about his work were simple, straightforward, and clearly deserving of a response, none was made. I feel that such non-response is both poor form and also speaks to the overall judgment of the validity of the original research.
I thank MoGASP for helping to “balance the books” on the blog that he presented in good faith on the basis of the initial press release and publicity, although it’s unlikely this “balance” will achieve anywhere near the amount of internet and media coverage of the original.
Dear Dr. Loomis,
Can you help me with some specific quick questions to help me understand your CDC study better so I can represent it fairly? Thank you!
(1) Your WVA analysis showing restaurant improvement used NAICS coding separating bars and restaurants but the separation was unclear in the other seven states:
“We found no significant association between smoke-free laws and economic outcomes in restaurants and bars in 7 of the 8 states.”
Clearly you are saying that neither restaurants nor bars were hurt, but the word “and” rather than “or” opens the possibility you were speaking only overall.
Did you actually find no significant losses in restaurants OR bars anywhere?
(2) If there were no significant changes in any state’s bar OR restaurant revenue other than WVA, it is informative but what about non-significant changes? If 12 of the remaining 13 sectors all had non-significant gains or all had non-significant losses, that result would carry some meaning.
What was the actual breakdown?
(3) Since bar losses would ordinarily be expected, did you test for significance with a one-tailed 90% test?
If not, why not?
(4) You say, “we combined data from all counties for which data were available, whether smoke-free or not, and compared the average effect of smoke-free laws in counties that contain smoke-free communities with counties that have no smoke-free communities.”
Am I correct in understanding you compared changes in bar employment between 2000-2010 in counties that banned smoking as compared to bar employment in that period for the ten counties that continued to permit smoking?
Is that what you compared and found no differences?
And, related to that:
(5) Did you examine changes in bar employment in counties where smoking had become banned only in restaurants?
If so, did you run separate analyses then for (A) no ban counties, (B) partial ban counties, and (C) full ban counties?
(6) Is the grant proposal for this research publicly available?