Chapter Three – Gun Ownership, Gun Laws, and the Data on Crime
To get to a well-reasoned study on the impact of gun-control laws, you need to have as accurate a picture as possible of gun ownership – what kinds of people own guns, and where they reside. That’s not exactly easy to find out, so, in this chapter, Dr. Lott describes how he’s tackled that difficult task.
WHO OWNS GUNS? (pages 37-47)
This section considers what types of Americans own guns and how that picture has been changing over time. It is difficult to collect precise data on gun ownership. Surveys have been a prime way used to calculate ownership rates, and exit polls after general elections every two years have been a main source of estimating gun ownership in each state. The set of relevant poll questions has grown over time so that it often includes items on:
- Background demographics – age, gender, race, income, place of residence, political views.
- Firearm ownership specifics – whether you own a gun, and whether your family owns a gun (i.e., a gun is present in the household, even if the person polled is not the owner).
Dr. Lott notes a significant increase in gun ownership from 1988 to 1996, both among voters (from 27.4% in 1988 to 37% in 1996) and in the general population (from 26% in 1988 to 39% in 1996). The Brady law went into effect in 1994, and Dr. Lott believes its requirement on background checks for gun purchase was responsible for some of the increase in gun sales around that time.
He points out several potential problems with exit poll surveys:
- The form of questions about gun ownership changed. Since the ways we ask questions can affect the answers we get, the meaning of the data shifted some in those survey years.
- It is still difficult to estimate the number of non-voters with guns.
- There was reason to believe that women under-report gun ownership.
- There is evidence that Republicans and conservatives (who are more likely to own guns) more often refuse to participate in exit polls, That skews the results toward underestimation of gun ownership in the overall population.
Here are some key findings from the 1988 and 1996 post-Presidential election poll data:
- Large numbers of Americans from all demographics own guns.
- The most typical profile for gun owners: rural, white male, middle-aged or older, conservative Republican, earning $30k to $75k per year.
- There was a larger growth rate in gun ownership during this period among women, not men.
- Individuals/families that are more likely to own guns include: high-income people, union members, rural dwellers.
- Gun ownership rates don’t explain the gap in crime rates between whites and blacks in the U.S.
The exit polls in 2004 used a different set of questions on gun ownership, and Republicans and conservatives refused to participate in higher rates. So, the 2004 data are difficult to compare with the findings of 1988 and 1996. Still, an estimated 41% of American households had guns in 2004 (up from 37% in 1996).
UNDERSTANDING DIFFERENT GUN LAWS AND CRIME RATE DATA (pages 48-55)
Violent crime (the category defined in the FBI Uniform Crime Reports as murder and non-negligent manslaughter, rape, aggravated assault, plus robbery) has risen significantly in recent decades. When More Guns, Less Crime was first published in 1998, violent crime statistics were 46% higher in 1995 than in 1976, and 240% higher in 1995 than in 1965. Though violent crime rates had declined some by the time Dr. Lott’s book was updated again in 2000, that was not enough to calm fears. He notes that by 2007, 39 states had adopted “shall-issue” concealed-handgun laws, meaning officials were required to issue a permit to every qualified applicant.
The total in 1985 for shall-issue states was only eight. The first edition of this book used the data available from 1977 through 1992 and during that time, 10 more states adopted shall-issue laws for concealed firearms. Between 1993 and 2007, 21 more states were added to that total. Dr. Lott presents in Table 3.3 the data to compare 1992 and 2007 for the seven kinds of FBI crime rates. For each of those benchmark years, he presents the crime rates per 100,000 people, divided between (1) states with nondiscretionary (shall-issue) concealed handgun laws and (2) all other states. He also shows the percentage of higher (or in a few cases, lower) crime rate in states without shall-issue laws.
Despite the problem with cross-sectional data, according to the data presented in the table for 1992, the difference is quite suggestive: violent crimes are 81 percent higher in states without nondiscretionary laws. For murder, states that ban the concealed carrying of guns have murder rates 127 percent higher than states with the most liberal concealed-carry laws. After almost all the states have adopted these laws in 2007, the difference is much smaller: just 25 percent for violent crime and 28 percent for murder. States with nondiscretionary laws have less violent crime, but the differences for property crimes are smaller and less consistent. (Page 49, emphasis added.)
Dr. Lott then describes the extensive process he used for compiling detailed county-by-county data in the following categories.
Crimes. FBI crime rates, conviction rates, sentence lengths, concealed-weapons permits.
Demographics. Bureau of Census data on population density per square mile, total county population, and the breakdown of racial, age, and gender categories.
Income Distribution and Unemployment Data. Per-capita personal income, unemployment insurance payments, income-maintenance payments, and retirement payments.
Other Legal Changes in Firearm Laws. Penalties for improper gun use, requirements for obtaining a permit to carry a concealed handgun, waiting periods required for purchasing handguns, etc.
This required a huge amount of investigative work to compile these data sets. But, there was no other way to start researching the possible connections among these factors on guns and crime, taken alongside data for states that had adopted shall-issue laws. And the carefully prepared data sets may be why Dr. Lott’s careful research methodology and conclusions are difficult to refute.