What’s the value of all those hospitality industry studies?

Interest groups

The details show that there are more or less three interest groups:

1. Groups interested in showing that there is no economic damage (pharmaceutical industry, health industry, governments)
2. Groups interested in showing a negative effect (tobacco industry, hospitality industry)
3. Independent researchers like universities if they are not paid by one of the two above mentioned interest groups.

The division into three interest groups was mainly based on the funding. The greater part of the studies was done by the anti-tobacco lobby (56%), 34% of all studies had ties with the hospitality or tobacco industry and 10% could be considered independent (2% had an unknown source of funding). Note that independent and university studies are more likely to be funded by the traditional heavy funders (the antismoking groups) than by prosmoking forces.

Table 2: Categorisation	Interest category
Anti-tobacco	Independent, University, Unknown	Free Choice
#	#	#
Funding Category	23	.	.
Government
Pharmaceutical Industry	8	.	.
Tobacco Related	.	.	23
Non Governmental Organizations	28	.	.
Hospitality Industry	.	.	13
Independent funding	.	4	.
Unknown	.	2	.
University	.	5	.

It becomes very interesting if you take a look at the results of the studies while keeping those interest groups in mind. There is almost a 100 % correlation between interest group and results (confirmed by a variance analysis). We see that 100 % of the studies done by anti-tobacco researchers does not find any negative effect of smoking bans. The majority of the anti-anti-tobacco studies shows a negative effect after a smoking ban has been introduced.

Table 3: Positive/negative impacts found per funding category	Conclusion of negative Impact?	Total
No	Yes
%	%	%
Funding Category	100	–	100
Anti-tobacco
Free Choice	6	94	100
Independent, University, Unknown	71	29	100

(- = no data available)

It’s very clear that anti-smoking studies suffer extremely from what we call Publication Bias. Whoever pays for the study decides which results are wanted. The publication bias alone is already enough reason to conclude that anti-smoking studies are unreliable.

Publication policy

The Australian researchers were very proud to mention that almost all studies that showed no negative effect of smoking bans had been published in scientific journals and had been peer reviewed. Peer review means that researchers who have not taken part in the study review the results of a study. That’s why peer review normally stands for an independent review.

Table 4	JournalType
Economic journal	Hospitality biz related	Medical or Health related
#	%	#	%	#	%
Funding Category	.	.	1	4	21	84
Anti-tobacco
Free Choice	1	4	.	.	.	.
Independent, University, Unknown	.	.	1	4	1	4
Total	1	4	2	8	22	88

One would expect that this kind of study would be published in economic journals, which in fact didn’t happen. Almost all the studies were published in anti-smoking friendly journals. Table 4 shows in what kind of journals the results of the studies were published in when researchers mentioned that the studies had been peer reviewed. 84 % of all studies were published in medical and/or health-related magazines. This means that the studies have been peer reviewed by colleagues with the same negative attitude towards smoking. The statement that the quality of these studies is high because almost all of the studies have been peer reviewed is meaningless.

The only study of all 106 analyzed studies that was published in an independent economic scientific magazine was tobacco-industry related and found a negative economic result because of smoking bans.