When gun control advocates argue for banning or severely restricting gun ownership, the comparisons drawn are usually the United States vs. Britain, Canada, or Japan. The argument presented is that availability of guns causes high crime rates. Occasionally, similar comparisons are made with different American states -- though usually such comparisons are made by their opponents, since state by state murder rate comparisons can be used (just as inaccurately) to "prove" that gun control laws increase crime rates.
That comparisons of such widely differing nations, cultures, and legal systems as Japan, Britain, and the United States are absurd should be obvious. But even ignoring these obvious differences, there is plenty of evidence that such comparisons are ignoring significant factors besides firearms availability. As an example, compare American and British rape rates. Unlike murder, rape seldom involves a gun. While 62% of murders in the U.S. in 1981 involved a firearm, only 7% of rapes did so.  Therefore, if crime rates in the U.S. and Britain can be fairly compared, we should find that British rape rates were equal to U.S. rape rates, minus the 7% of U.S. rapes committed with guns.
The 1984 British Crime Survey reported 2,288 rapes in England and Wales -- an area with a population of 49 million people! This gives 4.67 rapes per 100,000 people.  By comparison, America's rape rate for 1987 was 73 per 100,000 females  , or 36.5 per 100,000 people. Subtracting the 7% of U.S. rapes that are committed with firearms gives 34 rapes per 100,000 people -- far higher than Britain's rate. Britain's very low rape rate must be more than just the absence of firearms -- much more.
Similarly, there were 662 murders in England and Wales in 1984  . This gives 1.35 murders per 100,000 people. The U.S. murder rate in 1987 was 8.3 per 100,000 people  . Even if we assume that:
1. In the absence of firearms, not a single murderer using a firearm in the U.S. would have used another weapon to commit murder (very unlikely);
2. further assuming that not a single privately owned firearm was used to prevent a murder from happening in the U.S. (very unlikely);
3. assuming that not a single murder in Britain involved a firearm (not true);
subtracting out the 59% of murders committed with firearms in the U.S. in 1987  still gives a rate of 3.4 per 100,000 - - two and a half times higher than Britain. How valid is it to compare British and U.S. murder rates?
Comparisons of differing American states are somewhat fairer, since we share a basic set of criminal laws and culture, but even in this case, there are some dramatic differences that must be admitted in order to remain intellectually honest. New York is a highly urbanized (84.6% urban population  ), ethnically and culturally heterogeneous state. North Dakota, on the other hand, is rural (48.8% urban population  ), and much more homogeneous. Because of these differences, it would be absurd to compare New York and North Dakota crime rates, and draw the conclusion that New York's restrictive gun control laws cause its much higher murder rate.
To make a valid comparison, we need two states with comparable laws, populations (taking into account age, degree of urbanization, and ethnic composition), practical efficiency of justice systems, yet radically differing gun control laws. Unfortunately, such equivalent American states are not available. Using differing American states, and attempting to quantify various factors that contribute to crime is doomed to be unpersuasive, simply because of the highly statistical nature of such studies -- the average American knows that there are "lies, damn lies, and statistics". A persuasive yet non-technical case that lax firearms laws are, at most, a minor factor in murder rates, is not to be found by comparing different states.
There is a form of comparison that solves most of these problems -- intrastate crime comparisons where the gun ownership and transfer laws are uniform throughout the state. The particular set of data I chose to use was the FBI's 1987 Uniform Crime Reports for California. I chose these figures because:
1. California's laws regulating possession, carriage, and sale of firearms were uniform in 1987. California Government Code section 53071 prohibits cities and counties from passing more restrictive laws than the state in this area. (Unfortunately, the same statement can't be made for 1989, since some cities and counties passed restrictions on sale and possession of so-called "assault weapons". As the courts hear challenges to these local ordinances, they are being overturned -- but for 1989, it could be argued that a similar comparison might not be valid).
2. California has a great many cities over 10,000 (the minimum cutoff for Uniform Crime Reports). For the same reason that public opinion surveys poll a large number of people, to be sure that the sample is representative of a larger group, it is important that the number of cities, and the total number of people living in those cities, be large enough that the data reflects the average, and not a fluke.
3. I live in California, and knowing the dangerous cities to stay out of has a certain practical value for me and my family.
The Uniform Crime Reports compiled by the FBI provide detailed crime statistics for cities greater than 10,000 population; in California, this includes 278 cities, totalling 20,254,796 people, or 73% of the population of the state. These cities included 2,229 murders (11.0 murders per 100,000 population), or 76% of the state's total murders. While many of these cities are quite small, and it would therefore be questionable to draw conclusions about the safety of individual small cities from a single year's figures, studying multi-million person aggregates of these cities is statistically significant.
Everyone knows that there are dangerous places, and safe places -- but I was unprepared for how dramatic the difference really is. I was prepared to find some variation in murder rates in California, since it is obvious that big cities have more murder than small towns. I was also prepared to find quite a bit of murder almost everywhere in California; it is just common sense to assume that if guns are readily available, that there will be murders. But when I started working with the data, I found that this assumption was, in fact, quite wrong.
I divided California's 278 cities over 10,000 people into three groups. The first group contained 99 cities with individual murder rates below 2.5 per 100,000 people. (This is roughly the overall murder rate of Canada, a paragon of the virtues of gun control to the gun prohibitionists). A second group of 124 cities had murder rates between 2.5 and 10.0 per 100,000 population. The most dangerous 55 cities had murder rates above 10.0 per 100,000.
The first group of cities contains a total of 3,678,334 people in 99 cities. A total of 28 murders were committed during 1987, for a average rate of 0.76 per 100,000. In 78 of these 99 cities, there were no murders in 1987; the remaining cities had three murders or less during the year. If you lived in these safe California cities, the chances of being murdered are about half that in England and Wales -- and yet guns were readily available.
Not surprisingly, many of these cities are small rural or semi-rural communities like Napa, Gilroy, and my home, Rohnert Park. But not all fit in this category: Fullerton, with 111,499 people, and Torrance, with 138,997 people, are both in the Los Angeles megalopolis, and both on the safe list. Culver City, with 40,683 people, and within a few miles of some of the most dangerous parts of Los Angeles, is also on the safe list. Fremont, a few miles south of Oakland, with 157,462 people, and only one murder, is also among the safe 99.
The second group contains 8,260,007 people living in 124 cities, with 460 murders, for an average rate of 5.57 per 100,000 population. Here are a wide variety of cities, from small, rural communities like Turlock (34,818 people, one murder) to large cities like San Jose (730,079, 24 murders). Many are small to medium cities on the edge of megalopolises: Garden Grove, South Gate, Carson, Hawaiian Gardens are all part of the Los Angeles area. But proximity to Los Angeles isn't enough to explain their problems -- consider Culver City and Torrance, both in the safest group.
The third group, with a total of 55 cities, contains 8,316,455 people, with 1,741 murders, for an average rate of 20.93 per 100,000 -- roughly twice the California average. With a couple of notable exceptions, they are big cities, or suburbs of big cities: Los Angeles, Oakland, Santa Ana, Long Beach, Bell Gardens, Compton. What is really amazing, though, is how dangerous the worst cities in this group really are. Compton, for example, was the deadliest California city in the 1987 Uniform Crime Reports, with 95,894 people, and 80 murders -- 83.43 murders per 100,000 people -- more than twice the rate of Washington, DC. 
The disparities in murder rates are extreme -- yet the same firearms laws applied everywhere in California in 1987. There was no registration requirement for firearms, private sales were legal, and background checks were not required. The only background check system was for handguns purchased from dealers. If lax firearms laws are a contributing factor in murder rates, the enormous variation across California suggests that it is a very minor factor, or that some other factors must also be present for lax gun laws to cause a high murder rate.
The argument could be made that while the laws were the same, the rates of ownership of firearms differ dramatically. This might be a valid argument when comparing suburban Torrance to ghetto Compton -- after all, California doesn't have mandatory registration, so we don't know how many guns are legally present, per 100,000 population, in each place. (The overly demanding will point out that the number legally present and the number actually present aren't at all the same, no matter what the laws require). But even in the absence of valid measures of firearms ownership rates, no one familiar with California's cities is going to believe that the "cow towns" that make up most of the "safe list" are Handgun Control utopia -- quite the opposite!
The desire to reduce murder is both understandable and laudable. The desire to find one simple cause, which can then be eliminated, is also understandable, though clearly absurd. But the evidence shows that lax gun laws alone are not a major factor -- and the range of murder rates in California suggests that the major factors are many, many times more powerful than lax gun laws.
There are people out there, some of them reading this article, who have put tremendous energy and money into lobbying for restrictive gun laws, in the hopes that it will reduce the murder rate. I would submit that the evidence from California shows that they barking up the wrong tree. Relative to the more significant factors causing murder, restrictive gun laws are like trying to empty a lake with an eyedropper.
1. Bureau of Justice Statistics, Report to the Nation on Crime and Justice, 1st ed., (Washington, DC: Government Printing Office, 1983), p. 14.
2. "Still unsafe on the streets", The Economist, March 21, 1987, p. 56.
3. Federal Bureau of Investigation, Uniform Crime Reports for the United States 1987, (Washington, DC: Government Printing Office, 1988), p. 14.
4. "Still unsafe on the streets", The Economist, March 21, 1987, p. 56.
5. Federal Bureau of Investigation, Uniform Crime Reports for the United States 1988, (Washington, DC: Government Printing Office, 1989), p. 8.
6. Ibid., p. 12.
7. Mark S. Hoffman, ed., World Almanac and Book of Facts 1989, (New York, NY: Pharos Books, 1988), p. 615.
8. Ibid., p. 616.
9. Bureau of Justice Statistics, Report to the Nation on Crime and Justice, 2nd ed., (Washington, DC: Government Printing Office, 1987), p. 48.