More Guns Less Crime by John Jr (ebook reader macos .txt) 📗

is ultimately allowed to obtain a concealed-handgun permit is up to the discretion of either the sheriff or judge who has the authority to grant the permit. The person applying for the permit must frequently show a "need" to carry the gun, though many rural jurisdictions automatically grant these requests.

Endogenous: A variable is endogenous when changes in the variable are assumed to caused by changes in other variables.

Exogenous: A variable is exogenous when its values are as given, and no attempt is made to explain how that variable's values change over time.

Externality: The costs of or benefits from one's actions may accrue to other people. External benefits occur when people cannot capture the beneficial effects that their actions produce. External costs arise when people are not made to bear the costs that their actions impose on others.

Nondiscretionary concealed-handgun law: Also known as a "shall-issue" or "do-issue" law; the term nondiscretionary means that once a person meets certain well-specified criteria for obtaining a concealed-handgun permit, no discretion is involved in granting the permit—it must be issued.

Pooled, cross-sectional, time-series data: Data that allow the researcher not only to compare differences across geographic areas, but also to see how these differences change across geographic areas over time.

APPENDIXTWO/ 251

Regression: A statistical technique that essentially lets us fit a line to a data set to determine the relationship between variables.

Statistical significance: A measure used to indicate how certain we can be that the impact of a variable is different from some value (usually whether it is different from zero).

Time-series data: Data that provide information about a particular place over time. For example, time-series data might examine the change in the crime rate for a city over many years.

ftppgndhr Thrgg

Description of the Data

This appendix provides a detailed discussion of the variables used in this study and their sources. The number of arrests and offenses for each crime in every county from 1977 to 1992 were provided by the FBI's Uniform Crime Reports (UCR). The UCR program is a nationwide, cooperative statistical effort by over 16,000 city, county, and state law-enforcement agencies to compile data on crimes that are reported to them. During 1993, law-enforcement agencies active in the UCR program represented over 245 million U.S. inhabitants, or 95 percent of the total population. The coverage amounted to 97 percent of the U.S. population living in Metropolitan Statistical Areas (MSAs) and 86 percent of the population in non-MSA cities and in rural counties. 1 The Supplementary Homicide Reports of the UCR supplied the data on the sex and race of victims and on whatever relationship might have existed between victim and offender. 2

The regressions report results from a subset of the UCR data set, though we also ran the regressions with the entire data set. The main differences were that the effect of concealed-handgun laws on murder was greater than what is reported in this study, and the effects on rape and aggravated assault were smaller. Observations were eliminated because of changes in reporting practices or definitions of crimes; see Crime in the United States for the years 1977 to 1992. For example, from 1985 to 1994, Illinois operated under a unique, "gender-neutral" definition of sex offenses. Another example involves Cook County, Illinois, from 1981 to 1984, which experienced a large jump in reported crime because of a change in the way officers were trained to report crime.

The additional observations that were either never provided or were dropped from the data set include those from Arizona (1980), Florida (1988), Georgia (1980), Kentucky (1988), and Iowa (1991). Data for counties containing the following cities were also eliminated for the crime rates listed: violent crime and aggravated assault for Steubenville, Ohio (1977— 89); violent crime and aggravated assault for Youngstown, Ohio (1977— 87); violent crime, aggravated assault, and burglary for Mobile, Alabama

APPENDIX THREE/253

(1977-85); violent crime and aggravated assault for Oakland, California (1977-90); violent crime and aggravated assault for Milwaukee, Wisconsin (1977-85); all crime categories for Glendale, Arizona (1977-84); violent crime and aggravated assault for Jackson, Mississippi (1977 and 1982); violent crime and aggravated assault for Aurora, Colorado (1977 and 1982); violent crime and aggravated assault for Beaumont, Texas (1977 and 1982); violent crime and aggravated assault for Corpus Christi, Texas (1977 and 1982); violent crime and rape for Macon, Georgia (1977-81); violent crime, property crime, robbery, and larceny for Cleveland, Ohio (1977—81); violent crime and aggravated assault for Omaha, Nebraska (1977—81); all crime categories for Eau Claire, Wisconsin (1977—78); all crime categories for Green Bay, Wisconsin (1977); and all crime categories for Little Rock, Arkansas (1977-79).

The original Uniform Crime Report data set did not have arrest data for Hawaii in 1982. These missing observations were supplied to us by the Hawaii UCR program. In the original data set several observations included two observations for the same county and year identifiers. The incorrect observations were deleted from the data.

For all of the different crime rates, if the true rate was zero, we added 0.1 before we took the natural log of those values. It is not possible to take the natural log of zero, because any change from zero is an infinite percentage change. For the accident rates and the supplementary homicide data, if the true rate was zero, we added 0.01 before we took the natural logs of those values. 3

The number of police in a state, the number of officers who have the power to make arrests, and police payrolls for each state by type of officer are available for 1982 to 1992 from the U.S. Department of Justice's Expenditure and Employment Data for the Criminal Justice System.

The data on age, sex, and racial distributions estimate the population in each county on July 1 of the respective years. The population is divided into