For example, from the survey I linked to above:
http://ajl.sagepub.com/content/early/2011/02/01/1559827610396294.full.pdfI suggest reading the whole article, but here are some paragraphs about suicide and homicide:
Ecological studies try to explain varying
rates of suicide across different geographical
areas. Within the United States,
researchers have looked across US
regions,37-39 states,40-44 and urban areas.42,45
The studies using validated measures of
firearm ownership levels typically find
a strong significant positive association
between levels of gun ownership and
rates of suicide because of higher rates of
firearm suicide.
...
Many ecological studies link gun prevalence
with overall homicide rates because
gun prevalence is associated with high
gun homicide rates; there is typically no
association of gun prevalence with nongun
homicide.12,57 Articles include international
studies of high-income countries58
as well as US studies of regions,38,54
states,55,59-62 and counties.63,64
So now that we have the scientific point of view, let's see where your more informal analysis of the data goes wrong.
While the sale of firearms has skyrocketed in recent years, the violent crime rate and the rate of crimes involving firearms has dramatically decreased. This does not prove that more guns = less crime but does indicate that the number of firearms IS NOT the prime factor in determining crime rates.
We can agree that the number of firearms is not the prime factor in determining crime rates. Of course, nobody that I am aware of has ever claimed either of the following:
1) that the total number of firearms is the correct way measure of gun prevalence
2) that gun prevalence is the prime factor in determining crime rates
The argument you make is very common in the NRA bubble: gun sales are up, crime is down, thus no link. This is straight case of being uninformed, and it betrays a lack of knowledge regarding the scientific research on gun violence. Put another way, people familiar with the scientific literature would simply not say the things you are saying.
I've repeated the following several times in the last few weeks on this board, but here's why:
1) Virtually all serious scholars agree that the key statistic linking gun prevalence to crime is not the total number of guns in circulation, but the % of households and/or individuals with guns.
2) The ownership rates, for both households and individuals have in fact dropped over the last two decades (the plot you got from dailykos shows this, for example). Gun sales have increased, but most guns have gone to people or households who already own one or more guns. For obvious reasons, this has a much weaker effect on homicide and suicide rates than an increase in the number of households or individuals who have access to a firearm.
3) Most important, though, there are many factors that affect the crime rate,and thus nobody serious (on either side) attempts to base conclusions on the fact that homicide and gun ownership rates have dropped while the total number of guns has increased. This is the most significant point of all. Nobody is claiming that guns are the only factor or the main factor, so the whole game of trying to draw conclusions from general trends over the last two decades is flawed (despite the fact that, due to points 1 and 2, an informal look at the last few decades would actually support the scientific view rather than the NRA view).
One way to show that your methodology here is flawed is to show that the same type of coarse reasoning would lead you to reject other factors which essentially everyone agrees contribute to crime rates. For example:
1) Over the last two decades, the % of population in urban areas has increased, while crime has dropped. Despite this, there is virtual unanimity among social scientists that urbanization is a factor that contributes to crime.
2) Over the GW Bush decade, poverty increased, while crime dropped. Again, despite this fact, there is virtually unanimity that poverty is linked to higher crime.
So in both these cases, looking at changes in the entire country over the course of a decade or two leads to flawed conclusions. If some conservative tried to use the poverty statistic to argue that poverty and crime were unrelated, I'm sure you would object. And with good reason.
And so it is with guns. To tell what's going on, you need to look at the data more carefully and in more detail. Researchers have done so, and repeatedly found that higher gun prevalence leads to higher homicide and higher suicide rates.