Monday, 5 August 2013

Empty casinos don't need ambulances

Stanton Glantz—junk science's gift that keeps on giving—has produced another study purporting to show a decline in acute illnesses after a smoking ban. This time, however, he's not looking just at heart attacks but at ambulance calls as a whole. So that's every serious bar room fight, every case of alcohol poisoning, every road accident and, yes, every heart attack.

His claim is that ambulance calls fell by 20 per cent in Colorado after casinos were forced to ban smoking in January 2008 and he has no doubt as to why this was...

"The research shows strong evidence of a significant drop in ambulance calls due to less secondhand smoke exposure,” Glantz said.

In fact, secondhand smoke was never measured, nor is there any record of why the ambulances were called. In keeping with his notorious Helena heart miracle, Glantz looked only at one small, rural community (Gilpin County) rather than the whole of Colorado. If he had looked at the whole state, it would have provided a much larger sample and would have allowed less scope for cherry-picking.

Did ambulance calls really fall by 20 per cent? Although I would never trust any of Glantz's figures until I've seen the raw data with my own eyes, he provides a graph which suggests that a decline did indeed occur (the casino smoking ban is shown by the middle red line; the line to the right of it shows when casino hours were extended in 2009). Note that these are ambulances going to and from the casinos themselves, not the whole area.

But here's another graph. It was published by the Denver Post after the smoking ban had been in place for six months. It shows revenue in the Colorado casino industry which, as it says, was "languishing in its worst downturn ever" in 2008.

The story of what happened to these casinos after the smoking ban was enacted is predictable enough. This is from June 2008...

Colorado's mountain casinos are mired in their worst downturn ever, walloped by a smoking ban, high gas prices and a tightening economy that has cut into discretionary spending.

While other states with commercial casinos, such as Nevada and New Jersey, are also struggling, Colorado's 10.7 percent revenue drop during the first four months of 2008 is the second-largest in the nation, according to an analysis by The Denver Post.

"We might've had a few blips before, but nothing like this," said Marc Murphy, co-general manager of Bronco Billy's Casino in Cripple Creek.

...Murphy said he believes 75 percent of the current downturn is attributable to the statewide smoking ban that extended to casinos Jan. 1.

Lower revenues in the casino business can only be the result of fewer people coming in and/or people staying for shorter periods of time.

Golden resident Jane Painter said, between puffs, that she cut her trips from twice a week to once every couple of weeks because of the smoking ban.

"It's not so much fun anymore," said Painter, 68.

Inside the casino, Denver resident Paul Williams, 48, said his trips have dropped from twice a week to three times a month.

Not only did revenues fall but four casinos went out of business altogether.

Colorado has three fewer casinos operating this year, since four have shut down, while a new one recently opened in Cripple Creek.

The rapid decline of the Colorado casino industry has been well documented. There is no doubt that things went down quickly and sharply. For example...

Thanks to smoking bans initiated in early 2008, Colorado and Illinois have sustained major dips in gambling revenue, according to experts on a panel Wednesday at the Global Gaming Expo.

Gaming-tax collections in the Centennial State were off as much as 25.4 percent in the second quarter year over year in some jurisdictions, and as much as 53 percent in the third quarter year over year once you throw in the effects of a new, graduated tax, said Lois Rice, executive director of the Colorado Gaming Association.


Colorado casinos took the second-largest revenue hit in the industry in 2008 with a 12.3 percent drop, according to a new report by the American Gaming Association.

Only Illinois suffered a bigger drop with a 20.9 percent decline. The American Gaming Association attributed the states’ struggles, in part, to smoking bans that took effect last year.

...For 2008, Colorado casinos posted revenue of $715.9 million, down from $816.1 million. They generated $88.4 million in tax revenue for the state, down 23.4 percent from $115.4 million in 2007.

Things got so bad that the government agreed to help them out.

Colorado regulators are giving tax relief to some casinos hurt by revenue declines ranging in the double digits after smoking was banned.

The Colorado Limited Gaming Control Commission approved the new formula Thursday. Casinos that have revenue between $4 million and $15 million a year will pay less next year.

...Casino revenue has declined every month this year in wake of a smoking ban that took effect in January.

Of course, it is impossible to completely disentangle the effect of the smoking ban from other factors, such as the recession, but exact quantification is not necessary to demonstrate the fatal problem with Glantz's latest study. It is only necessary to show that casinos in Colorado saw a large decline in attendance at exactly the same time that the smoking ban came into effect. That such a drop-off occurred is beyond dispute. If you've got fewer people in the casino, obviously you're going to have fewer people needing ambulances.

Incredibly, Glantz does not control for the fact that attendance dropped off after the smoking ban. Nowhere in the study does he acknowledge that the fall in ambulance calls from casinos coincided with "the worst downturn ever" in the local casino industry. How could he? Not only would it screw up his attempt to show that smoking bans reduce ambulance call outs, but it would also undermine his previous attempt to show that smoking bans don't harm the casino business. Oh, what a tangled web we weave.


Ben said...

How can Glantz get away with all these shoddy studies? doesn't his University have a reputation to defend? Couldn't his University make better use of its money than paying a salary to Glantz?

JohnB said...

If what Professor Glands and other antismoking activists are claiming about secondhand smoke is correct, that it is such an incredible hazard, then there should be a fleet of ambulances parked at the entrance of every “smoking permitted” casino that provides around-the-clock - 24/7 - ferrying of the “smoke-affected” to nearby hospitals. Why didn’t Professor Glands check if there are such ambulance fleets at casinos that permit smoking? If there was such a fleet, he would then have to also be able to distinguish between those requiring an ambulance ride that are “smoke affected” from those suffering from the [coronary] effects of the excitement/anxiety of winning/losing from gambling.

Rursus said...

Correlation and causality appear often (in public health) like identical twins. Who does not know them closer, may become confused.

Scientists should explain such facts, not obfuscate them!

Shame on Glantz!

Carl V Phillips said...

Glantz does make one important contribution to science, as a shibboleth. If someone cites him or seems to believe what he writes, you can safely conclude that they have no understanding of science and ignore the rest of what they say.

Unknown said...

SG is a santimonious twit with delusions of grandure. May he spend eternity in that special level of hell normally reserved for child molesters and people that talk in the theatre.

Ivan D said...

Stanton Glantz has tenure at UCSF. On the basis of his shameful output we should not trust anything whatsoever that emanates from that establishment. Universities might be reasonably expected to adhere to certain minimum standards and I would expect those standards to include dishonorable removal of those who are quite happy to print untruthful nonsense in support of their political obsessions.

handymanphil said...
This comment has been removed by the author.
handymanphil said...

Two flies were crawling up a wall when the man, who was smoking at the time, swatted one..... it died. Glantz then declared to an awestruck scientific world that he had cast iron proof that SHSS (Second Hand Swatting Smoke) kills!
Such is the wonderful standard of this lunatics methodology :(

Rursus said...

Wow... Surprise!

My comment came through the quality control and validation of S. Glantz' blog:

and was answered, too.

Christopher Snowdon said...


Glantz says:

"The facts that (1) there was a change in ambulance calls not coming from casinos but no change from casinos when the law applying everywhere but casinos, combined with (2) a change of similar magnitude in casinos but not elsewhere when the law was extended to casinos strongly supports causality."

On point (1), take a look at the two graphs in the study. The drop in admissions in casinos is clearly visible in the bottom graph, as is the subsequent rise when hours were extended. For the 'not casinos' call outs (which are a smaller number because Gilpin County is a very small place which relies on the casino industry), no comparable decline is visible. Insofar as there was a drop-off, it mainly happened after the smoking ban in casinos, not after the general smoking ban. That doesn't fit Glantz's theory, but it fits the fact that there were fewer people coming and going to Gilpin after the casino smoking ban and therefore fewer people in the area to have traffic accidents, fights, heart attacks etc.

It's important to note that the lines Glantz has put on his graphs are *not* the trendlines of actual call outs, but are two different predicted plots based on his own model. If you want to see what actually happened, look only at the dots (if you can see them - Glantz has a habit of covering them with his quasi-trendlines).

Point (2) is just a restatement of his original conclusion and amounts to saying 'call outs to casinos fell and therefore causality is supported'. It doesn't add anything.

As a point of interest, if you look at the casino/ambulance graph (shown in the post above), you can see that 2003 also saw an unusually low number of call outs to casinos. It's not immediately obvious because - again - Glantz's 'predicted' line draws the eye away from what's actually going on in the data, but look how far out Glantz's line is from the data in 2003! You can see that the actual call outs were much lower than average, never rising much above 60 a month. The explanation for this comes in the other graph in the post above showing that 2003, like 2008, was a time when casino revenues dropped (and, therefore, fewer people were in casinos). The drop in both revenue and call outs was less severe in 2003 than in 2008, but this evidence of two variables being aligned in 2003 supports what I say in this post - it's about the attendance not the secondhand smoke.

FXR said...

There is only one secret to living a good life;

When a politician tells you, he is here to protect you. or that he has a "duty to protect your children"[Mein Komph et al]

Prepare to defend yourself or run.

FXR said...

In order to evaluate the norm, you have to first define the norm. Such as a simple calculation to define the ratio of ambulance calls per-capita within a much larger but similar group. Then compare the two timelines equitably, as the most simplistic and reasonable comparison method. Did he in fact do this, or did he simply search for and cherry pick the data because it fit his desired conclusions?

The answer should be obvious by his reputation and by his reinforcement of that reputation found herein.

Glaze-centric scare-vertizing, exists as always, in the familiar or typical hysterics and emotionally charged blather providing nothing useful, to anyone beyond those who make a lot of cash by his fanatical promotions.

FXR said...

With the realities of "Godwin's Law" already well established and proven, we face another unavoidable fact. The length of time in which that fact exists, precipitously increases the rise of Fascism in our politics, having now garnered the perfect defense. Defending the "Big Lies", by denials that produce themselves.