Yesterday, the New York Times article, Share of Homes with Guns Shows 4 Decade Decline makes the claim that gun ownership has declined in the United States since the early 1970's. Their data comes from the General Social Survey regarding respondents that answered the question, Do you happen to have in your home (IF HOUSE: or garage) any guns or revolvers? Here is the survey data which, indeed, shows a decrease in households with guns. I would argue that a better question would be: Do you own a gun?" There is a difference. When you make assumptions about individuals based on household data, you may not see the whole picture. More than one person in a household may make the individual decision whether or not to purchase a firearm. All adults in the household may not be aware of each others decisions either.
However, one of the problems with surveys is that people aren't always honest in their response to survey questions. As you can see, 1.7% of respondents refused to answer this question in 2006, with the exception of a few years; this is a much higher refusal rate than most other years. I am not intending to do a sophisticated statistical analysis here, but this seems to indicate a reluctance do disclose this sort of information. How many simply said no to preserve their privacy? That is impossible to know. Even the New York Times article acknowledges
"Measuring the level of gun ownership can be a vexing problem, with various recent national polls reporting rates between 35 percent and 52 percent. Responses can vary because the survey designs and the wording of questions differ."
Despite varying conclusions of different analyses, the article accepts the one with the lower rate based on the title of the article.
Though there isn't a reliable way to measure gun ownership since data is scarce on an individual level, surveys are one of the few options. However, I am hesitant to give them too much weight for making the assumption that gun ownership is down. Nor do I believe that the number of gun owners should be the basis for supporting policy measures currently being proposed to restrict gun ownership.
Instead of relying on surveys that are often subject to error, let's look at data of which we do have definitive information. We do know that there are more guns in the United States. The New York Times article claims that this is largely due to the same gun owners purchasing multiple guns. Though it would not be any surprise that a gun owner owns more than one. For this rate of decline, wouldn't it be necessary for people that have previously owned guns to no longer own guns? That doesn't seem likely from most of the gun owners that I have known. I don't believe the difference can be wholly the result of demographic change (aging and death). Though there is certainly some of that.
Below, we can look at the firearms manufactured, exported, and imported. After subtracting out those exported, there has been an increase over time with the exception of a few years. There were more than double the number of firearms manufactured and imported in 2010 than there were in 1986. It seems a bit of a stretch to think with such a high volume being produced that they are being distributed to fewer people.
Furthermore, based on the FBI Uniform Crime Statistics, we can conclude that more guns do not result in an increase in crime. Regardless of what you want to attribute the decrease in crime to, there clearly has not been an increase in crime due to more guns since the crime rate has, in fact, gone down. John Lott makes a compelling argument that more concealed carry has reduced crime. I would encourage you to read his book, More Guns, Less Crime and see if you agree.