US STATES WITH MORE GUN OWNERS HAVE MORE MURDERS

Reuters ^ | 12/04/02 | Reuters - Charnicia E Huggins

[AnAmericanMother]

Amen!

My husband used to ride the Pennsy into Newark, walk along Broad to the Lackawanna, and ride it out to Bloomfield. He was merrily strolling down the street during the riots . . . nobody bothered him though.

. . . of course this had nothing to do with the fact that he is 6'6", 250 #, and always carried a shillelagh with the head drilled and filled with molten lead . . . just a coinkidink . . .

[Pitchfork]

Given that the Journal in which this article appeared is a respected, peer reviewed academic journal I suspect that is is VERY unlikely that the author made the kinds of errors that some of the self-appointed experts on research design and statistics on this forum have accused them making. The sheer ignorace of these postings is grating. You may not agree with the findings but unless you have some background or have read the article slinging around baseless accusations of shoddy or biased research is juvenile.

Public health researchers from Harvard generally understand the concept of statistical control. The mere notion that that they would have simply compared the raw numbers without accounting for population size, or failed to inlcude relevant control variables, is ABSOLUTELY INANE.

Just reading this summary and the way they talk about 'likelihood' I suspect that some type of MLE-estimation method, such as LOGIT was used. LOGIT is designed to accomodate dichotomous dependent variables (0 = murdered, 1 = murdered) while estimating how inclusion in a particular sub-groups (high or low gun states) increased the likelihood of being in either of the outcome categories. LOGIT allows for the inclusion of control variables I am certain they were incorporated into the analysis, no peer reviewed journal would accept a study without them.

The only real criticism that can be leveled at the research is the point in the last sentence about the direction of the causal arrow. However, causal time order questions can be addressed through trend studies and time series analysis. This would require data on gun ownership and murder rates over time to test whether increased ownership preceeded or followed murder rate increases. Obviously controls for the broader crime rate in each area would have to be included along with a host of demographic variables. These studies are sophisticated and complicated. And sometimes the estimation techniques are controvertial but to simply reject out of hand qulity research because you don't like the result is sheer ignorace.

[meyer]

Public health researchers from Harvard generally understand the concept of statistical control. The mere notion that that they would have simply compared the raw numbers without accounting for population size, or failed to inlcude relevant control variables, is ABSOLUTELY INANE.

I don't doubt that researchers at Harvard know the concept of statistic control. However, I haven't seen too many leftists that haven't been quite selective when they try to sway public opinion. Perhaps they are counting on others not having privy to the raw numbers that they purport to have. Given the politics of today's institutions of higher learning, particularly in the areas of "soft" science, I would be awfully suspicious of any studies that come out of today's colleges. In particular, I am suspicious of studies that claim to have no conclusive verdict while they make numerous implications throughout their press release/abstract based on apparently incomplete and arbitrarily selected and divided data.

Frankly, I think that rather than sling your own baseless accusations at other folks on this forum who have shown other data, you ought to try to show the errors presented by those you criticize. That is the best way to arrive at the truth.

[TigersEye]

Go to post #119, take the link and read the article. Then come back and tell us how reliable and believable research from Harvard is.

[Pitchfork]

Forgive my incivility but quite frankly as a researcher I am becoming increasingly distraught over what I think are twin movements undermining progress in the social sciences. 1) the new left's post-modern and post-postivist turn that rejects the notion of causality, brands empirical research as uselss, and contends that all truth is relative and 2) the new right's anti-intellectualism and increasingly frequent resort to unfalsifiable bias claims against anyone who happens to hold a position at university and produces findings that don't accord with current right-wing dogma.

Karl Popper the ardent anti-communist and philosopher of science introduced the notion of falsifiability partly to undercut the Marixt unfalsifiable 'false consciousness' axiom. Yet we seem the same kind of logic being raised to shoot down otherwise rigourous social scientific research. The author of this study is at Harvard, Harvard is populated predominantly by liberals (although I can point to a number of rather conservative scholars), therefore this research is liberal crap. Even if the liberal reseacher intends to be objective he/she will invariably skew the results. However, it does not follow that this researcher is liberal OR that all liberals fudge data any more than all conservatives would. The paper's results might not conform to your beliefs but that does not justify a claim of bais or incompetence. The bias brush has a broad swath and each stroke colors both the target, the discipline, and the process of scientific investigation.

When all science findings are questioned as arising out of some liberal or conservative ideology the enlightenment ideals of human progress through knowledge and research will have died. (this is the post-modernist dream by the way). Everything will be political and polemical.

[Godel]

"It is possible, for example, that locally elevated homicide rates may have led to increased local gun acquisition," they write.

Exactly. People always confuse correlation with causation.

[TigersEye]

unfalsifiable bias

There was a comedian in the '60s that used to do that kind of shtick. Barney Googenheimer? Naw, I think he was a Red Skelton character. There was a guy whose whole routine was double talk. Funny stuff.

[relee]

Yes, stats can be manipulated, or at least "cherry-picked" to support ones' own conclusions.

From post 45
...Regionally, the association exists for victims aged 5 to 14 years and those 35 years and older....

Why did they exclude ages 15-34? More cherry picking!

[DarthRaven]

That article is full of anti-gun bovine excriment!

I've known many of West Virginians in my life time and alot of them are gun owners. However, I believe that West Virginia is definately one of the top five states in heart disease! WV has the least restrictive gun laws yet the lowest violent crime rates as compared to most other states. On the other hand, MD has draconian gun laws yet is still #1 in Robbery and #3 or 4 in other violent crimes (like murder, rape, etc.) Need to check statistics in the FBI Uniform Crime Reports for the past ten years.

[DarthRaven]

That article is further proof that doctors and other health care professionals need to stop playing politics and focus on preventing life-threatening medical mistakes!

[Travis McGee]

Nice research.

Things that make you go "Hmmmm...."

[sneakypete]

Still, Miller's team notes that it is not clear whether the higher rates of household gun ownership caused or resulted from the increased number of homicides.

And there is the "yeah,BUT..." that allows them to spin facts into lies and still save their professional reputations.

[ServesURight]

BUMP

[sargon]

Miller and his team investigated the association between homicide and rates of household firearm ownership using 1988-1997 data collected from the nine US census regions and the 50 states.

Who gives a crap about whether people say they have a gun in their home? That's not scientific proof. They are comparing a well known rate with a highly speculative rate, and failing to consider myriad other factors. There certainly is no causality here, although that seems to be the propagandist intent of most such studies. The presence of a gun in a household would probably tend to be under-reported in areas where their were stricter controls.

Such reporting would also necessarily be influenced by peoples' perception of the of local laws.

This would have to skew the results of any such study.

So, I'm not impressed with the "science" here. They can find out how many households have guns only by doing a house-to-house search...

[wardaddy]

Amazing how easy it is to doctor "stats" to justify one's original premise. Is this an example of nearly all university research and publishing today?

[Travis McGee]

For sure. Did you read the data posted at 47? There is a MUCH more direct relation between race and murder, but that is ignored.

[meyer]

Forgive my incivility but quite frankly as a researcher I am becoming increasingly distraught over what I think are twin movements undermining progress in the social sciences. 1) the new left's post-modern and post-postivist turn that rejects the notion of causality, brands empirical research as uselss, and contends that all truth is relative and 2) the new right's anti-intellectualism and increasingly frequent resort to unfalsifiable bias claims against anyone who happens to hold a position at university and produces findings that don't accord with current right-wing dogma.

The primary reason I spoke out against what I perceived as incivility is that 1) there are a few folks here that also work in research, and 2) some of the evidence produced here by freepers is quite factual and easily shows the flaws in this story. Whether the story is an accurate depiction of the research or not is debatable, but without access to what has been found, it isn't possible to tell.

I do find it quite interesting that while you concede that the left may be promoting relativism as opposed to objectivity, you seem to imply that any non-leftist's attempt to point out when this may be the case is merely parrotting the "current right-wing dogma". Perhaps there is some left-wing dogma at work in this very thread.

Karl Popper the ardent anti-communist and philosopher of science introduced the notion of falsifiability partly to undercut the Marixt unfalsifiable 'false consciousness' axiom. Yet we seem the same kind of logic being raised to shoot down otherwise rigourous social scientific research. The author of this study is at Harvard, Harvard is populated predominantly by liberals (although I can point to a number of rather conservative scholars), therefore this research is liberal crap. Even if the liberal reseacher intends to be objective he/she will invariably skew the results. However, it does not follow that this researcher is liberal OR that all liberals fudge data any more than all conservatives would. The paper's results might not conform to your beliefs but that does not justify a claim of bais or incompetence. The bias brush has a broad swath and each stroke colors both the target, the discipline, and the process of scientific investigation.

OK, to be fair, just because the study comes from Harvard doesn't mean it is leftist propaganda. But considering that some 90% of acedemia is from the left and Harvard seems to play pretty hard-left at times (note their stance on, say, the military or the middle east), I would say that it is much more likely that their studies are leftist propaganda as opposed to their being objective. Perhaps the good doctor would care to share his figures and methodology so that the rest of the real world might see.

When all science findings are questioned as arising out of some liberal or conservative ideology the enlightenment ideals of human progress through knowledge and research will have died. (this is the post-modernist dream by the way). Everything will be political and polemical.

Sadly, in the social sciences, I think we are beyond that point already. While most hard sciences still tend to remain objective, the preponderance of what I've seen in the social sciences shows a certain leftist bias. Frankly, there's been so much mis-information from that side of the aisle the over the last several years that all such research needs to be scrutinized. And, it ought to be scrutinized publically if it is to have influence on public issues.

[challenger]

Mr. Pitchfork, your anger is misdirected, since you have apparently not read either their current article, nor their February, 2002 article < http://www.hsph.harvard.edu/press/releases/march.pdf >. If anyone is debasing the prestige and credibility of science, it is surely those, like Harvard's School of Public Health, who dangerously abuse science to create political polemics and shamelessly manipulate the voters. Your anger should be directed at Miller, et al, for their carelessly chosen or deliberately deceptive basis for their study--i.e., presumed rate of firearms ownership, using unreliable proxies. Go back to my post 115 for more information on that idiocy, and read SoDak's reply to post 115 to see how common firearms ownership is. Read the HSPH article of Feb. to see their rankings of firearms ownership. Then compare those rankings with the FBI "Uniform Crime Reports" rates of homicide. You should also examine the FBI data on the jurisdictions where the murder rates are highest--every one of them is a major city where legal rates of firearms ownership are low. Could it be that there is something other than "easy availability" at work here? In case you are unaware, the Harvard School of Public Health has been 100% opposed to firearms ownership for many years and their "research" reflects this fact. They have never once conducted a firearms study that included cost-to-benefit ratio for all aspects of firearms ownership, such as self-defense.

Furthermore, as another Freeper has pointed out, the "researchers" have lumped states together so that they won't have to explain the massive discrpancies among individual states, such as the low murder rates in Wyoming and West Virginia which have high firearms ownership. Imagine if the headline had been, "Two of the states with the highest firearms ownership rates have some of the lowest murder rates in the United States". Accurate, but doesn't fit the political agenda of HSPH, so you'll never see that headline on their press releases.

[Pitchfork wrote]
Forgive my incivility but quite frankly as a researcher I am becoming increasingly distraught over what I think are twin movements undermining progress in the social sciences. 1) the new left's post-modern and post-postivist turn that rejects the notion of causality, brands empirical research as uselss, and contends that all truth is relative and 2) the new right's anti-intellectualism and increasingly frequent resort to unfalsifiable bias claims against anyone
who happens to hold a position at university and produces findings that don't accord with current right-wing dogma.

[My response]
Most of us are not claiming mere bias in the HSPH articles. It is either massive incompetence or deliberate deception. And to correct your simplistic characterization of "the new right's...right wing dogma" please note that many researchers have started on the left/liberal/prohibitionist side of this issue and after they have conducted real, open-minded research, they recognized that the firearms owners were correct all along. Among these researchers are Prof. James Wright, Prof. Peter Rossi, Prof. Gary Kleck, Prof. William Van Alstyne, Prof. Sanford Levinson, and too many others to list. Incidentally, when you use epithets like "new right's anti-intellectualism" and "right wing dogma" to describe those who disagree with you, you reveal that you are a hard-line far leftist. It is no wonder you make excuses for HSPH.

[Pitchfork wrote]
Yet we seem the same kind of logic being raised to shoot down otherwise rigourous social scientific research.

[My response]
"Otherwise rigorous"? The entire experimental design is totally defective from the start when it is based on a proxy measure that is not realistic. This study is the worst type of "result oriented" research that is designed only to support a predetermined conclusion. But to make things worse, it isn't even internally consistent in its ability to explain the differences among states. As many people have mentioned already, there are other factors that produce a much better fit of the data.

[Pitchfork wrote]
The author of this study is at Harvard, Harvard is populated predominantly by liberals (although I can point to a number of rather conservative scholars), therefore this research is liberal crap.

[My response]
Hardly anyone has claimed that the study is "crap" because it is written by liberals, and it is irrelevant that you can point to the 5% of the Harvard faculty that is conservative, since they didn't conduct the study that is the topic of this thread. The study is crap because it is crap. GIGO.

[Pitchfork wrote]
Even if the liberal reseacher intends to be objective he/she will invariably skew the results. However, it does not follow that this researcher is liberal OR that all liberals fudge data any more than all conservatives would.

[My response]
Time for you to stop setting up straw men. No one was constructing a logical thread that stated that "because most researchers at Harvard are liberal, this researcher is liberal"--we know they are dishonest not because they are liberal, but because we have seen their other work. Nor did any of us make a specific claim that "all liberals fudge data any more than all conservatives would"--we are not saying anything about "all liberal", but we are saying that this particular group of anti-self-defense liberals has a very long history of deceptive and inaccurate reporting in this field.

[Pitchfork wrote]
The paper's results might not conform to your beliefs but that does not justify a claim of bais or incompetence.

[My response]
We are not using "conformity to our beliefs" as the test. Look at all the data that has been presented to you on this thread. Argue with the data if you dare.

[Pitchfork wrote]
The bias brush has a broad swath and each stroke
colors both the target, the discipline, and the process of scientific investigation.

[My response]
We are not talking about bias, but we are criticizing the "scientific investigation" competence and honesty of the HSPH. The really damaging approach is the one that you take when you defend the indefensible.

[Pitchfork wrote in previous message 142]
And sometimes the estimation techniques are controvertial but to simply reject out of hand qulity research because you don't like the result is sheer ignorace.

[My response]
It is not just that we reject "qulity research" out of hand because we do not like the results--many of us have studied this subject extensively over a span of many years or decades and we recognize bad research when we examine it. With all due respect, sir, you are the one acting out of ignorance. Have you actually read the body of research by HSPH? The estimation techniques in the current study are obviously unreliable and the conclusions are preordained--and in conflict with my point that all the jurisdictions with the highest murder rates are cities, which have low rates of legal firearms ownership, while many jurisdictions where firearms are almost universal have low murder rates.

[Pitchfork wrote in previous message 142]
The sheer ignorace of these postings is grating. You may not agree with the findings but unless you have some background or have read the article slinging around baseless accusations of shoddy or biased research is juvenile.

[My response]
Yes, some of the postings are based on ignorance, but the real point is that some of them are very substantive. Your complaints about some of the posters does not refute the points made by the others. You have not addressed any of those substantive points, so it appears that you cannot. I, for one, have read the Feb., 2002 sister report, which has the same authors, the same premise, the same rankings of states, and the same conclusions, but using "children" as the main subjects rather than the current article's "people". I have not yet read the current article, since it doesn't seem to be available for free, yet. But to reiterete, we are not criticizing the current study because we disagree with it, nor are we merely saying that it is biased or shoddy--we are showing you some real problems with their experimental design, with their premise, with their proxies, with their conflation, and with their conclusions. Most of us are being serious and realistic, not juvenile. It's time for you to stop being juvenile and time to stop the ad hominem attacks.

________________________

Randall N. Herrst, J.D.
President
The Center For The Study Of Crime
JOIN NOW! The Premier Resource for Innovative Activists!
http://www.studycrime.org

[challenger]

The point about firearms providing more safety is generally true. If you examine the time-series analysis conducted by Prof. John Lott, you will see that there is a distinct reduction in violent crime within a state after "shall issue" CCW/Self-Defense-Permit laws are enacted. This improvement in safety occurs no matter whether the state starts with a high or low rate of violent crime.

You are making the mistake of looking at violent crime rates in a one-time snapshot mode, rather than examining what happens when we change the number of firearms in society. For example, Florida used to have the highest violent crime rate of any state, by far. Their murder rate was more than 40% higher than the national average and their violent crime rate was almost double the national rate. After they enacted the "shall issue" concealed carry law in 1987, their murder rate fell dramatically, as did their violent crime rate. In 2001, they had a murder rate that had dropped so much that it is now 6% LESS than the national murder rate. Their violent crime rate was still the highest in the country, but just barely (Maryland, with strict gun control will probably take over that position next year), and it was now within 55% of the national average. These are huge improvements in public safety. Are Floridians now "safe"? No. Are the "safer"? Yes. So the direct answer to your inquiry is, "Yes, handguns make society safer, but not necessarily completely safe. Nothing can do that."

You miss the point regarding the press release about the study--they couldn't leave those states out because they wanted to use the bottom block of states (those with the highest rates of firearms ownership) in a comparison with the top block of states (those with the lowest rates of firearms ownership). Even their deception has limits.

__________________________

Randall N. Herrst, J.D.
President
The Center For The Study Of Crime
http://www.studycrime.org

[american_ranger]

Ice cream consumption causes drownings.

Every summer people eat more ice cream and there are also more drownings. Therfore icream consumption causes more drownings. It has nothing to do with people also boat and swim in the summer. It has everything to do with eating ice cream.

Look at me, I'm a researcher too.