Well, we're certainly on the same side of the fence in the big issue. Normally, I'm anti-smoking.
But, I hate to base any of my judgements on possible fallacies, and I intend to put the common data into doubt as being unsubstantiated anti-smoking propaganda; I believe the corrected data will be a much less convincing argument for your case. Nevertheless, it should still be against smoking. It is clear to see the health benefits from quitting, so we should expect a higher death rate, but the figures quoted are entirely unreasonable and fall at the very maximum end of the scale of taking bad assumptions.
I dislike medicine because of the strong political spin on it. Many scientists, who I would trust absolutely if they were in other fields, make silly mistakes in not checking their assumptions or trying to understand what they mean. Including the assumptions of the papers they build on. They simply, create a model and plug the numbers in. Since the analysis of the data is governed by the papers they put out, if they are wrong in their premises and assumptions, there is no point at which they are corrected.
So, to correct a few errors in your judgement and analysis;
I'm not saying we should actively encourage people to make a bad choice, I'm not unhappy with the current anti-smoking campaigns.
I am against banning products because they may proselytise people, and statistically do, to smoking.
That's the weaker logical equivalent statement of banning cigarettes because they encourage people to smoke, by their existence.
Should we ban cigarettes because people may be enticed to buy them and smoke due to their existence?
Should we ban flavoured cigarettes because people may be enticed to buy them and smoke due to their existence?
It is the same logic, and the same argument, on a smaller scale, and I cannot accept that, because some people,
knowing the risks, choose to smoke and smoke in private. They are not hurting me, why should I restrict their freedom to damage themselves?
actually, the statistics are that 1/2 of people who start smoking smoke for at least 20 years. and 1/2 of people who smoke for 20+ years die of a tobacco-related disease. so if you start smoking, your odds are 1 in 4 that you will die as a direct result of that. about 20% of american adults smoke - that's millions and millions of people, just in the u.s. tobacco-related diseases were the leading cause of death in the u.s. in 2000 - 435,000. "sexual behaviors" caused only 20,000 - so your chances of a bad outcome from cigarettes are in fact much higher than for promiscuous sex.
Let's examine this in greater detail, because I'm a bit of a mathematician, as you may have heard.
First, let us consider that death is not the only consequence of both sex and smoking. Indeed, I would refer to unplanned pregnancies moreso than health problems as major issues with sex.
Fun fact: recent studies have estimated that nearly 49% of pregnancies in the states are unplanned, and only about half of those are aborted or avoided in some manner or form. Which comes to about 2 million unplanned pregnancies. Fun.
500,000 babies born, unplanned, to teenage moms in the states every year.
3 million teenagers with new sexual diseases.
Okay, okay.
Now, let's examine smoking: What is considered a tabacco related disease? Particularly ones that relate to death.
I'm finding cancer as the main one, supposedly relating to 425,000 deaths per year, of smokers, through cancer considered related to smoking, including throat, larynx, lung, mouth, nasal, etc. etc.
First, I'm going to look up incidences of these types of cancer in non-smokers; I'm finding that "87% of lung cancer cases are caused by smoking", in a report by the surgeon general; This obviously is not correct.
Ah. Right in the report, he notes that anti-smoking groups, such as Mormons and Seventh-day Adventists have less instances of lung cancer. The numbers he give work out to about 86%, so I assume that's where he got the numbers from.
Well, no. That doesn't work. Yeah, everyone else is probably exposed to 2nd-hand smoke, I agree. But, those religious groups has less
recorded cases of lung cancer. Do you see the difference here? They refuse diagnosis on a religious basis; Of course they will have less recorded cases of cancer if they choose, under normal circumstances, not to be tested for it.
Of course, that's not the only study that quotes near 90% . .. Let me check a few others and see what tactics they use to isolate non-smoking populations.
I'm going to look for a less political report and find another number for lung cancer incidence among non-smokers.
Oh, fun, I just found the original paper behind the 90%. Guess what? It counts all deaths of lung cancer and the such as smoking deaths. That's no good. I'm going to see if I can find a paper closer to my means that quotes 90%; That is to say, one that separates lung cancer due to smoking, and lung cancer not due to smoking. I doubt it though, I've found a few quoting 40-50% with that condition imposed, and they seem to be reasonable in their assumptions.
Annoyingly enough, I found a few quoting the original paper as saying that 90% of lung cancer is due to smoking; That is NOT the statistic it derives. It derives that amoung the smoking population, 90% of cancer deaths are related to the respiratory system.
As smoke of any kind can be clearly shown to weaken the lungs, deaths are expected from it, but this does not give us any linkage to incidences of cancer, merely WHICH CANCER
kills smokers. Not which ones they get. (They have a much lower rate of skin cancer death, for example, than the normal population, and a only slightly increased, (<10%), chance of dying from cancer than the normal person. The type of cancer that they die from is different, that's all. That doesn't even tell us they get that cancer more often, just that they die from a type of cancer more often)
Okay, found a paper that tells me the incidence of smoking deaths. 24/122 expressed as a fraction (I'm not sure why?) over a period of 10 years. The assumptions in this paper detailing these deaths are:
Death, in a person who has previously smoked at anytime in their life, caused by any disease or disorder that has been possibly linked to smoking, is counted as a smoking death
Given that assumption, unsurprisingly, the amount of smokers to the normal population is . . . 18.5% (For all calcs, I'll be picking numbers relevant to Canada, btw. We have better statistics here to base judgement on)
24/122 is about 19.6%.
So, 18.5% of the population is responsible for about 19.6% of deaths. For a time period of sampling 10 years. Same 10 years for both papers.
And then, we're really pushing those deaths. They are only connected; those people might've died of the same causes or similar causes regardless of their smoking habits. Indeed, looking at the difference in cancer, I feel quite safe in saying that the above statement is true for a statistically significant portion of the population.
Some other papers quote the smoking deaths, on the same assumption, as high as 24%, others as low as 17%. I'm going to trust statistics canada on the percent of smokers though, stats canada is pretty amazing.
Okay, that's all very interesting, but I haven't found what I wanted to find. I want to find the incidence of respiratory cancers in smokers and non-smokers. Not the death rate, the incidence.
Then, using that, I want to find the amount of smokers who would've contacted lung cancer anyway, and reduce the death statistic to the corrected form. I have clearly shown already that it cannot be the 90% quoted from other sources. That value does not even come from considering which people get lung cancer, and it is definitely not what what studies say it is. Which is horrible.
Then, having the death statistic for cancers in particular, I want to extend that to find the same statistics for other major deaths amoung smokers and correct each of those, preferably even with age profiling. The number of "17.6 years taken off your life" deals with the uncorrected statistics, and therefore deals with the idiotic assumption that people would not have otherwise contacted a disease linked to tobacco. This is the kind of assumption I reference at the very top of my post. It's a pity, because I like the organisations putting out these studies.
Oh, fun fact: list of diseases/disorders linked to tobacco? If we subtract the bacterial and viral from the full list of recognised and well-defined physiological disorders/diseases, etc., and then consider any linkage, no matter how small, we find that "tobacco related" covers over 3/4 of the list. Most of the papers I'm wading through make no note of how strongly linked a disease has to be to be considered "tobacco related", and it is abundantly clear that many of them are using the maximum statistic.
Let's assume a rather stronger case, where there is desired at least a "strong correlation", (it's hard to find exact values of correlation, but I do trust these scientists, so where they say they have a strong correlation, I shall trust them), to link tobacco and the disease, and then we only have a scant few diseases to consider.
(afternote: never did find any study which only considered strong correlations that actually considered strong correlations,. Found a lot that said they did, but no, no, they did not. Looking at the lists of "strongly correlated" conditions they list. Influenza? Really? Related to smoking by strong correlation?)
Oh, look at this gem of idiocy!
Lung cancer is the leading cause of
cancer death in Ontario for both men and
women. In 2002, there will be
approximately 6,100 deaths from lung
cancer in Ontario: 2,700 women and
3,400 men.
(Canadian Cancer Society 2002 Statistics)
• In 2002, there will be more than 18,000
deaths from lung cancer in Ontario related
to smoking
Can anyone tell me what's wrong with that?
Also in the same article, I find that the instances of lung cancer death in women has raised to 4 times the amount from 1972 to 2002. Now that's funny, because, having still the number sheets open from before, I can confirm whether that is true. It is true. The funny part? The amount of women smokers has been decreasing by percent, (same numbers sheets), since 1950-ish. Not the most steady of declines, but the moving average is constantly decreasing. More lung cancer deaths. Less smoking. Make of that what you will. Maybe cigarettes have become tons more dangerous in the last 50 years. (actually, that'd be where my bets lie. That smoking and smoking habits have become more dangerous. Alternatively, if I had a regionated statistic, I bet I'd find that near smog centres, (large cities, pollution producing factories), the lung cancer rate raises enormously, and I hypothesise that these statistics would show a much higher connection to smog than to cigarettes).
(afternote: i forgot to say. the deaths were absolute, while the smokers was percentage. But check the population of canada between 1972 and 2002 . . .
22,219,560 and 31,940,700 respectively. Population climbed, yes, and the population of smokers as an absolute rose, yes, but it did not nearly quadruple!)
The same article notes that lung cancer directly links to lung cancer deaths, with a lag of twenty years. If you'll note the stats in the paragraph above, that is not true. Obviously. Let's stop reading this heavily erroneous article.
Still have not find papers relating to incidence of lung cancer in healthy and smoking populations. I find plenty related to deaths, but as I explained above, that does not correlate to getting more lung cancer; simply that the lungs are more heavily effected by it in smoking populations. Which should be obvious.
(afternote: another 30 minutes . . . )
I'm going to give up searching at this point.
Conclusion:
When the statistics are thoroughly examined, and the studies with erroneous assumptions are discarded, I find much less convincing data in favour of anti-smoking. Still none in favour of smoking.
The conclusion is that smokers are weaker against, and die more often, from diseases and conditions that effect the lungs and respiratory system.
I could not find papers detailing the incidence of said diseases and conditions, only papers detailing the cases of deaths. So the study is incomplete.
When the data was examined, it was discovered that ~20% of people, (smokers) are responsible for ~20% of deaths. (smoking related deaths). The percentage of smokers dying from non smoking related deaths is next to negligible because of the large number of conditions that are related to smoking.
Nevertheless, that 20% does not include, for example, car crashes, so yes, there are significantly more deaths due to smokers, but it is not huge. Considering the data, I would raise the number of smokers dying to no more than 25%, or 1/4 of the deaths. That is also an exaggeration. I can guarantee it is not that high.
Actually, given this, let us take your statistics, that 1/2 of people smoking die over a 20 year period, and that half again die over the next 20 years.
Let's, uh, let's take the chances of you dying in twenty years: oh look, it's up around the high 30%'s
Averaged around the total of your lifetime of course; you'll have less risk when you're young than old, etc. etc. But your smoking study ALSO averages it throughout the life. Funny, that. It'd be more convincing if it considered a younger population picking it up. We expect old people to die.
Let's take the chances of you dying in 40 years. Well, we already have a 20 year stat, lets approx and say 30% + 30% * (100% - 30%) on the low and 40% + 40% * (100% - 40%) on the high. 51%-64%. It's somewhere in the high 30%'s, so bets are it's above 60% over 40 years.
Does that 75% look a bit less compelling put into context? 75% over 60% doesn't seem that much larger. Nor does 50% over ~37.5%ish
Yeah, smoking is damaging, and I never disagreed that it is, but your statistics are propaganda. Not science.
Sex is indeed, a larger contributor to death, disease, and unplanned lifetime complications than smoking. This is due, yes, to the larger number of people having sex. But even compensating for that, (dividing by 5 We can agree that 100% of the population having sex is not that far off of an assumption? Even so, it will overcompensate, which is only bad if it shows that what I am saying is not true) it is still a larger contributor to disease and unplanned lifetime complications. (and if I wasn't overcompensating, the numbers would be higher.)
Now, with smoking in the proper context and framework, do you understand, why, even being against the tobacco industry, I would advocate for peoples freedoms? Spend extras on educating people about tobacco, (properly, please, give them less room to doubt your research. When there are more holes in the 'research' presented than a sieve, it presents a weaker picture, not a stronger one. Despite the fact that the numbers may be stronger when lied about).
I'm going to go to bed, I spent two hours digging up and examining research, rejecting it all, (All of it, I only accepted the statistics from some sources, ie: all of that from statistics canada because I have math crush on them, but statistics =/= research, and the statistics were not as through as I was hoping for), in an attempt to form a proper rebuttal to your argument.
Which consisted of ad hominem, glitter numbers taken out of context to seem amazing, taking my arguments out of context, etc. etc.
Really, I said it was impossible to overdose on nicotine through smoking, (admitting that is is possible and common through other means), and you send me to a link that says that nicotine overdose is possible through chemical means, but not purely through cigarettes. Good job.
Here the references used in coming to the conclusion:
Statistics Canada, current data, back to 10 years.
Canadian Cancer Society
American Anti-cancer Association, data release March 2003. (so you see that the bias from my sources is against cigarette smoking, one would hope.)
Have a good night, and I hope you learned something about putting statistics in context, and trying to understand what they are really implying through their assumptions.