From: coxb@netcom.com (Brian Cox) Subject: Univ of Maryland CCW Study (summary) To: firearms-alert@shell.portal.com Date: Wed, 15 Mar 1995 18:33:04 -0800 (PST) I have just been faxed a copy of the University of Maryland Violance Research Group's study "Easing Concealed Firearms Laws: Effects on Homicides in 3 States" by Dr's David McDowell, Colin Loftin, and Brian Wiersema from the University media Relations Office. [...] Anything in {} is direct comments relating to the material just summarized. If you don't want to read the study itself, a listing of what I see as primary problems is included afterward. I encourage you to read the summary as well though, even though it may be a little technical. (You should have tried reading the original 17 pages!) SUMMARY Introduction: "We find no support for the idea that the [CCW] laws reduced crime; instead, we find evidence of an increase in firearm murders." {This man is not on our side.} He mentions several other studies, although leaves out the famous Kellerman study. In the introduction he recognizes that CCW laws may cause in increase, decrease, or no change in homicide by firearm rates. Other Studies: The study mentions Kleck several times {probably an effort to ground this study in a firmly seated pro-gun basis} specifically in the matter of the 21% decline in Florida homicide rates between 1987 and 1992. "Although Florida's experience appears to support a deterrence effect [on crime when CCW laws are enacted], it suffers from several weaknesses. First these studies all use the UCR data compiled by the FBI. In 1988 the FBI did not publish crime counts for Florida. Evaluations must therefore ignore 1988 or use estimates of the 1988 total. This is important because 1988 was the first year after the law [was enacted]." {NOTE, as CCW laws to take some time to become effective (for a sizable percentage of the population to become armed), it is doubtful that the 1988 estimates actually pose any statistical flaw in previous study data. Also, leaving out 1988 does not change the fact that overal homicide declined between 1987 and 1992 significantly.} "Advocates of shall issue licensing often cite figures showing that few legal carriers misuse their guns. Yet greater tolerance of legal carriny may lead to higher levels of illegal carrying." {This is how he discredits the "law abiding citizens don't misuse guns" fact.} Research Methods: {Dr. Suter will be more able to assess this than I} The study covered Miami, Tampa, and Jacksonvile, FL, Portland, OR, and Jackson, MI. "Because there were relativelt few homicides in Portland, we combined data for [3 counties near Portland] in Oregon." {This may be a flaw. more detailed analysis is necessary.} The time period was January 1973 through December 1992 (240 months), except in Miami where "homicides increased sharply in May, 1980, following an influx of refugees from Cuba. Miami's homicide totals appeared to stablize by late 1982, and we thus analysed the period from January, 1983 through December, 1993." {Can we paint him as a racist for that comment?} An autoregrassive integrated moving average model was used, as well as an abrupt permenant change intervention model. {I don't know what either of those terms mean, but perhaps someone else can enlighten us.} Data: (duplicate of Table 1 on page 14) begin quote " BEFORE LAW CHANGE AFTER LAW #/month #/month S.E. % Firearms Homicide Miami, FL 25.88 +.079 1.09 +3% Jacksonvile, FL 6.24 +4.61 0.59 +74% Tampa, FL 4.91 +1.10 0.44 +22% Portland, OR 2.79 -0.34 0.35 -12% Jackson, MI 3.64 +1.57 0.47 +43% Mean change= +26.0% Inverse normal comb Z= -5.99, p< .0001 " end quote ----- Problems: I see several points on which this study could be attacked, and doubtless Dr. Suter will see others I have missed. First of all, the study uses the non-standard National Center for health Statistics for its data. I have never seen data from these pewople, there fore I have no way of knowing what type of information they are able to provide. It is likely that the author merely obtained all death certificates where the cause of death was a bullet wound. If so, he can be challenged on the same basis that Kellerman was challenged,as counting suicides as firearm homicides. Suicides usually make up a high portion of the number of deaths that would fit this criteria. Secondly, why did he combine the data in the Portland area? Perhaps he felt that the low numbers in portland were truly too small to be used in a study, but perhaps there was another reason. Maybe the Portland numbers showed some other characteristic that led him to believe they might not support his study. It should be noted that the city that he treated differently was the only one that came out with a decrease in firearm homicides. This is significant, although for which side I'm not sure. One glaring flaw absolutely jumped out at me when I read this report however. The study ran from 1973 through 1987 and took the mean as the pre-CCW-law firearms homicide statistic. Over this time period, firearms homicides were increasing at a consistent rate nationally, and I assume in his study areas (such as Florida) as well. Is it any wonder that data averaged from 1973 through 1987 then would be lower than date averaged after the 1987 date? In fact, if one merely compared the average of those 15 years with the 1987 figure, I'm sure the average would be much lower due to the gradual increase in firearms related homicides. If that explanation is a little foggy, perhaps Dr. Suter can explain it better. Finally, the study attempts to discredit the 21% decline in Florida by saying that the 1988 numbers were not included in that study. I find it unlikely that 1988 produced such a high murder rate as to account for this discrepency. Therefore, it is unlikely that the 21% figure (which has been quoted by the FBI on several occasiona in the UCR) is as inaccurate as this study claims. See above for more information on that. That concludes my summary, and if anyone can let us know whether autoregressive integrated moving average models and abrupt permanent change integration models are standard procedure, please do so. As I said above, I have no idea what either of them are. -- Brian D Cox | It is a conspiracy, the government | Finger for coxb@netcom.com | blanked out this space. | PGP public key __ Date: Wed, 15 Mar 95 10:20:00 EDT From: John Grossbohlin Subject: Re: MEDIA: Adverse CCW Study To: "WIENER, DAN x3708" On Tue, 14 Mar 95 16:09 you said: > > STUDY SHOWS HOMICIDES INCREASED AFTER CONCEALED-GUN LAWS RELAXED > . > . > McDowall said the results support other research showing that >policies to discourage firearms in public may help prevent violence. > > ### > > I know nothing about the biases or methodologies of these >researchers. However, this study has already been cited in an editorial >in today's Los Angeles Times, as part of their effort to rebut and >discredit the carry-reform movement. McDowall is associated with the folks who publish the anti-gun agenda support- ing "research" that appears in the NEJM. He was co-author on the study that found significant decreases in homicide and suicide in Washington DC after DC enacted its onerous gun laws in 1977 (Loftin, McDowall, Wiersema & Cotty (1991) NEJM). That study used raw numbers and not rates and ignored the population decrease in DC AND failed to note that the decreases in homicide and suicide started some TWO YEARS BEFORE the new law went into affect... McDowall, Lizotte & Wiersema (1991) CRIMINOLOGY examined the quasi-experimental data on the efficacy of defensive arms (Kennesaw, Orlando and others) and found that the implicit deterrant effect is not nearly what it appears to be. That is actually a pretty good study... and probably a function of having Lizotte involved (even if he does have a picture of a Porsche with "LIZOTTE" for a plate number on his office door ). Right off the bat I'd be concerned with the sample used in the latest "study" as it wouldn't be too difficult to find localities that fit the model they wished to test, i.e., guns are bad. Considering that the laws they referred to are state laws it seems they have also confounded their study by using city level data... i.e., they mixed up the unit of analysis. Alas, I too have to read the newest study before offering further comment. John John A. Grossbohlin GROSSBOJ@NPVM.NEWPALTZ.EDU Organizational Studies Ph.D. Program JG7831@UACSC2.ALBANY.EDU __ From: WIENER.GCS@pc.gcs.litton.com (WIENER, DAN x3708) To: firearms-alert@shell.portal.com (wiener) Subject: MEDIA: Adverse CCW Study Date: Tue, 14 Mar 95 16:09 For unknown reasons my early post was chopped off. Here it is again. The following are excerpts from an Associated Press story in today's local paper: STUDY SHOWS HOMICIDES INCREASED AFTER CONCEALED-GUN LAWS RELAXED . . The [University of Maryland's] Violence Research Group studied homicides by gun and other means before and after new, relaxed concealed gun laws took effect in Jacksonville, Miami and Tampa, Fla., Jackson, Miss., and Portland, Ore. Average monthly homicides by gun increased 74 percent in Jacksonville, 43 percent in Jackson, 22 percent in Tampa and 3 percent in Miami. Portland had a 12 percent decrease, the researchers announced Monday. They found that while homicides by gun increased after the less restrictive laws were adopted, homicides by other means remained steady. "While advocates of these relaxed laws argue that they will prevent crime, and suggest that they have reduced homicides in areas that adopted them, we strongly suggest caution," said University of Maryland criminologist David McDowall. "When states weaken limits on concealed weapons, they may be giving up a simple and effective method of preventing firearm deaths." . . McDowall said the results support other research showing that policies to discourage firearms in public may help prevent violence. ### I know nothing about the biases or methodologies of these researchers. However, this study has already been cited in an editorial in today's Los Angeles Times, as part of their effort to rebut and discredit the carry-reform movement. -- Dan Wiener (wiener@pc.gcs.litton.com) __ From: coxb@netcom.com (Brian Cox) Subject: Re: MEDIA: Adverse CCW study To: firearms-alert@shell.portal.com Date: Tue, 14 Mar 1995 16:02:53 -0800 (PST) [...] Dan, it seems some of your message was curtailed, so allow me to elaborate. This study, while I haven't seen it, deserves some serious attention based on the media coverage. It is from the University of Maryland Criminology Department, and is entitled "Easing Concealed Weapons Laws: Effects on homicides in 3 states" Apparently they studied 5 different communities and 4 of them showed an increase in homicide by firearm after CCW laws were liberalized. The states were Oregon, Massachusetts, and Florida. I doubt very much the integrity of this study, as I am personally familiar with both Oregon and Florida's murder rates before and after CCW. Both declined. Perhaps they are using straight numbers, not rates. This summary is from the LA Times editorial pages today (a major anti-CCW and pro Brady Bill editorial) so please take the numbers with a grain of salt. I would certainly not put selective reporting of statistics past the LA Times. I have been in touch with the University of Maryland to try to secure a copy of this study and find out where it was published. I will post a more complete summary when I have seen it, and Dr. Suter may be able to comment sooner than I. DIRPP tends to be able to get research papers fairly quickly. -- Brian D Cox | It is a conspiracy, the government | Finger for coxb@netcom.com | blanked out this space. | PGP public key