[2] See Cook (1982) and Zimring (1971) for these arguments.

[3] It is very easy to find people arguing that concealed handguns will have no deterrence effect. Uviller (1996, p. 95) writes that, "More handguns lawfully in civilian hands will not reduce deaths froom bullets and cannot stop the predators from enforcing their criminal demands and expressing their lethal purposes with the most effective tool they can get their hands on."

[4] Kleck and Gertz's survey (1995, pp. 182-3) of 10 other nationwide polls implies a range of 764,036 to 3,609,682 defensive uses of guns per year. Recent evidence confirms other numbers from Kleck and Gertz's (1995) study. For example, Annest et. al. (1995) estimate that 99,025 people sought medical treatment for nonfatal firearm woundings. When one considers that many criminals will not seek treatment for wounds and that not all wounds require medical treatment, Kleck and Gertz's estimates of 200,000 woundings seems somewhat plausible, though even Kleck and Gertz believe that this is undoubtedly too high given the very high level of marksmanship that this implies by those shooting the guns. Yet, even if the true number of times that criminals are wounded is much smaller, it still implies that criminals face a very real expected cost from attacking armed civilians. (See also Southwick (1995) for a discussion on the defensive uses of guns.)

⁵ Dawn Lewis of Texans Against Gun Violence provided a typical reaction from gun control advocates to the grand jury decision not to charge Gordon Hale. She said, "We are appalled. This law is doing what we expected, causing senseless death" (Potok, March 22, 1996, p. 3A). For a more recent evaluation of the Texas experience see Fort Worth Star-Telegram (July 16, 1996). By the end of June 1996, more than 82,000 permits had been issued in Texas.

⁶ In fact, police accidentally killed 330 innocent individuals in 1993, compared to the mere 30 innocent people accidentally killed by private citizens who mistakenly believed the victim was an intruder (Lott, 1994).

[7] Similarly, Multnomah County, Oregon issued 11,140 permits over the period January 1990 to October 1994 and experienced 5 permit holders being involved in shootings, 3 of which were considered justified by Grand juries. Out of the other two cases, one was fired in a domestic dispute and the other was an accident that occurred while an assult rifle was being unloaded (Barnhart, 1994).

[8] Wright and Rossi (1986, p. 151) interviewed felony prisoners in ten state correctional systems and found that 56 percent said that criminals would not attack a potential victim that was known to be armed. They also found evidence that criminals in those states with the highest levels of civilian gun ownership worried the most about armed victims.

Examples of stories where people successfully defend themselves from burglaries with guns are quite common (e.g., see "Burglar Puts 92-year-old in the Gun Closet and is Shot," New York Times, September 7, 1995, p. A16). Will (1993) discusses more generally the benefits produced from an armed citizenry.

In his paper on airplane hijacking, Landes (1978, p. 1) references a quote by Archie Bunker from the television show "All in the Family" that is quite relevant to the current discussion. Landes quotes Archie Bunker as saying "Well, I could stop hi-jacking tomorrow . . . if everyone was allowed to carry guns them hi-jackers wouldn't have no superiority. All you gotta do is arm all the passangers, then no hi-jacker would risk pullin' a rod."

⁹  These states were Alabama, Alaska, Arizona, Arkansas, Connecticut, Florida, Georgia, Idaho, Indiana, Maine, Mississippi, Montana, Nevada, New Hampshire, North Carolina, North Dakota, Oklahoma, Oregon, Pennsylvania, South Dakota, Tennessee, Texas, Utah, Vermont, Virginia, Washington, West Virginia and Wyoming.

¹⁰ These states were Alabama, Connecticut, Indiana, Maine, New Hampshire, North Dakota, South Dakota, Vermont, and Washington. Fourteen other states provide local discretion on whether to issue permits. California, Colorado, Delaware, Hawaii, Iowa, Louisiana, Maryland, Massachusetts, Michigan, Minnesota, New Jersey, New York, Rhode Island and South Carolina.

¹¹ All 22 gun control papers studied by Kleck (1995) use either cross-sectional state or city data or use time-series data for the entire US or a particular city.

¹² Equally damaging the authors appear to concede in a discussion that follows their piece that their results are highly sensitive to how they define the crimes that they study. Even with their strange sample selection techniques, total murders appear to fall after the passage of concealed weapon laws. Because the authors only examine murders committed with guns, there is no attempt to control for any substitution effects that may occur between different methods of murder. For an excellent discussion of the McDowall et. al. paper see Polsby (1995).

[13] Recent attempts to relate the crime rate to the prison population concern us (Levitt, 1996). Besides difficulties in relating the total prison population with any particular type of crime, we are also troubled by the ability to compare a stock (the prison population) with a flow (the crime rate).

[14] Arson was excluded because of a large number of inconsistencies in the data and the small number of counties reporting this measure.

¹⁵ Robbery includes street robbery, commercial robbery, service station robbery, convenience robbery, residence robbery, and bank robbery.

¹⁶ Larceny includes pick pockets, purse snatching, shoplifting, bike theft, theft from buildings, theft from coin machines, and theft from motor vehicles.

¹⁷ Among those who made this comment to us were: Bob Barnhardt, Manager of the Intelligence/Concealed Handgun Unite of Multinomah County, Oregon; Mike Woodward, with the Oregon Law Enforcemnt Data System; Joe Vincent with the Washington Department of Licensing Firearms Unit; Alan Krug who provided us with the Pennsylvania Permit data; and Susan Harrell with the Florida Department of State Concealed Weapons Division. Evidence for this point with respect to Virginia is obtained from (Lipton, 1995, p. A1) where it is noted that, "Analysts say the new law, which drops the requirement that prospective gun carriers show a 'demonstrated need' to be armed, likely won't make much of a difference in rural areas, where judges have long issued permits to most people who applied for them. But in urban areas such as Northern Virginia -- where judges granted few permits because few residents could justify a need for them -- the number of concealed weapon permits issued is expected to soar. In Fairfax, for example, a county of more than 850,000 people, only 10 now have permits." The Cramer and Kopel (1994) piece also raises this point with respect to California.

¹⁸ We rely on Cramer and Kopel (1994 and 1995) for this list of states. Some states known as "do issue" states are also included in Cramer and Kopel's list of "shall issue" states though these authors argue that for all practical purposes these two groups of states are identical.

¹⁹ The Oregon counties providing permit data were Benton, Clackamas, Coos, Curry, Deschutes, Douglas, Gilliam, Hood River, Jackson, Jefferson, Josephine, Klamath, Lane, Linn, Malheur, Marion, Morrow, Multnomah, Polk, Tillamook, Washington and Yamhill.

[20] However, the effect of an unusually large percentage of young males in the population may be mitigated because those most vulnerable to crime may be more likely to take actions to protect themselves. Depending upon how responsive victims are to these threats, it is possible that the coefficient for a variable like the percent of young males in the population could be zero even when the group in question poses a large criminal threat.

[21] For a discussion of the relationship between income and crime see Lott (1990a)

[22] A more detailed survey of the state laws is available from the authors, a brief survey of the laws excluding the permitting changes finds: Alabama: No significant changes in these laws during period. Connecticut: Law gradually changed in wording from criminal use to criminal possession from 1986 to 1994. Florida: Has the most extensive description of penalties. The same basic law (790.161) is found throughout the years. An additional law (790.07) is found only in 1986. Georgia: A law (16-11-106) that does not appear in the 1986 edition appears in the 1989 and 1994 issues. The law involves possession of a firearm during commission of a crime and specifies the penalties associated with it. Because of the possibility that this legal change might have occurred at the same time as the 198 changes in permitting rules, we used a Lexis search to check the legislative history of 16-11-106 and found that the laws were last changed in 1987, two years before the change in permitting rules (O.C.G.A. @ 16-11-106 (1996)). Idaho: There are no significant changes in Idaho over time. Indiana: No significant changes in these laws during period. Maine: No significant changes in these laws during period. Mississippi: Law 97-37-1 talks explicitly about penalties. It appears in the 1986 version, but not in the 1989 or the 1994 versions. Montana: Some changes in punishments related to unauthorized carrying of concealed weapons laws, but no changes in the punishment for using a weapon in a crime. New Hampshire: No significant changes in these laws during period. North Dakota: No significant changes in these laws during period. Oregon: No significant changes in these laws during period. Pennsylvania: No significant changes in these laws during period. South Dakota: Law 22-14-13, which specifies penalties for commission of a felony while armed appears in 1986, but not 1989. Vermont: Section 4005, which outlines the penalties for carrying a gun when committing a felony, appears in 1986, but not in 1989 or 1994. Virginia: No significant changes in these laws during period. Washington: No significant changes in these laws during period. West Virginia: Law 67-7-12 is on the books in 1994, but not the earlier versions.It involves punishment for endangerment with firearms. Removing Georgia from the sample, which was the only state that had gun laws changing near the year that the "Shall Issue" law went into affect, so that there is no chance that the other changes in guns laws might effect our results does not appreciably alter our results.

[23] Using Marvell and Moody's findings show that the closest time period between these sentencing enhancements and changes in concealed weapon laws is 7 years (Pennsylvania). 26 states passed their enhancement laws prior to the beginning of our sample period and only 4 states passed these types of laws after 1981. Maine which implemented its concealed handgun law in 1985 passed its sentencing enhancement laws in 1971.

[24] One possible concern with these initial results raises from our use of an aggregate public policy variable (state right-to-carry laws) on county level data (Greenwald, 1983 and Moulton, 1990). As Moulton (p. 334) writes: "If disturbances are correlated within the groupings that are used to merge aggregate with micro data, however, then even small levels of correlation can cause the standard errors from the ordinary least squares (OLS) to be seriously biased downward." Yet, this should not really be a concern here because of our use of dummy variables for all the counties, which is equivalent to using state dummies as well as county dummies for all but one of the counties within each state. Using these dummy variables thus allow us to control for any disturbances that are correlated within any individual state. The regressions discussed in fn. 26 rerun the specifications shown in Table 3 but also include state dummies that are interacted with a time trend. This should thus not only control for any disturbances that are correlated with the states, but also for any disturbances that are correlated within a state over time. Finally, while right-to-carry laws are almost always statewide laws, there is one exception. Pennsylvania exempted it largest county (Philadelphia) from the law when it was passed in 1989, and it remained exempt from the law during the rest of the sample period.

[25] However, the increase in the number of property crimes is larger than the drop in the number of robberies.

[26] Given the possible relationship between drug prices and crime, we reran the regressions in Table 3 by including an additional variable for cocaine prices. One argument linking drug prices and crime is that if the demand for drugs is inelastic and if people commit crimes in order to finance their habits, higher drug prices might lead to increased levels of crime. Using the Drug Enforcement Administration's STRIDE data set from 1977 to 1992 (with the exceptions of 1988 and 1989), Grossman et. al. (1996) estimate the price of cocaine as a function of its purity, weight, year dummies, year dummies interacted with eight regional dummies, and individual city dummies. There are two problems with this measure of predicted prices: 1) it removes observations during a couple of important years during which changes were occurring in concealed handgun laws and 2) the predicted values that we obtained from this ignored the city level observations. The reduced number of observations provides an important reason why we do not include this variable in the regressions shown in Table 3. However, the primary impact of including this new variable is to make the "shall issue" coefficients in the violent crime regressions even more negative and more significant (e.g., the coefficient for the violent crime regression is now -.075, -.10 for the murder regression, -.077 for rape, and -.11 for aggravated assault, with all of them significant at more than the .01 level). Only for the burglary regression does the "shall issue" coefficient change appreciably: it is now negative and insignificant. The variable for drug prices itself is negatively related to murders and rapes and positively and significantly related at least at the .01 level for a one-tailed t-test to all the other categories of crime. We would like to thank Michael Grossman for providing us with the original regressions on drug prices from his paper.

[27] By contrast, if the question had instead been what would the difference in crime rates have been between either have all states or no states adopting right-to-carry handgun, the case of all states adopting concealed handgun laws would have produced 2,020 fewer murders; 5,747 fewer rapes; 79,001 fewer aggravated assaults; and 14,862 fewer robberies. By contrast, property crimes would have risen by 336,409.

[28] We reran the specifications shown in Table 3 by also including state dummies which were each interacted with a time trend variable. In this case, all of the concealed handgun dummies were negative, though the coefficients were not statistically significant for aggravated assault and larceny. Under this specification, adopting concealed handgun laws in those states currently without them would have reduced 1992 murders by 1,839; rapes by 3,727; aggravated assaults by 10,990; robberies by 61,064; burglaries by 112,665; larcenies by 93,274; and auto thefts by 41,512. The total value of this reduction in crime in 1992 dollars would have been $7.02 billion. With the exceptions of aggravated assault and burglary, violent crimes still experienced larger drops from the adoption of concealed handgun laws than did property crimes. Rerunning the specifications in Table 3 without either the percentage of the populations that fall into the different sex, race, and age categories or without the measures of income tended to produce similar though somewhat more significant results with respect to concealed handgun laws. The estimated gains from passing concealed handgun laws were also larger.

[29] Other differences also arise in the other control variables such as those relating the percentage of the population of a certain race, sex and age. For example, the percent of black males in the population between 10 and 19 is no longer statistically significant.

[30] By contrast, if the question had instead been what would the difference in crime rates have been between either have all states or no states adopting right-to-carry handgun, the case of all states adopting concealed handgun laws would have produced 2,286 fewer murders; 9,630 fewer rapes; 50,353 fewer aggravated assaults; and 92,264 fewer robberies. Property crimes would also have fallen by 659,061.

[31] The Washington state data were obtained from Joe Vincent of the state Department of Licensing Firearms Unit in Olympia, Washington. The Oregon state data were obtained from Mike Woodward with the Law Enforcement Data System, Department of State Police, Salem, Oregon.

[32] Unpulished information obtained by Kleck and Gertz in their 1995 National Self-Defense Survey implies that women were as likely as men to use handguns in self-defense in or near their home (defined as in their yard, carport, apartment hall, stree adjacent to home, detached garage, etc.), but that women were less than half as likely to use a gun in self-defense away from home.

[33] See also McCormick and Tollision (1985) for an novel article testing the endogeniety of the "arrest rate" in the context of basketball fowls.

[34] We would like to thank Phil Cook for suggesting this addition to us. In a sense, this is similar to Ehrlich's (1973, p. 557) specification except that the current crime rate is broken down into its lagged value and the change between the current and previous periods.

[35] While county level data were provided in the Supplementary Homicide Report, matching these county observations with those used in the Uniform Crime Report proved unusually difficult. A unique county identifier was used in the Supplementary Homicide Report and it was not consistent across years. In addition, some caution is suggested in using both the Mortality Detail Records and the Supplementary Homicide Report since the murder rates reported in both sources have relatively low correlations of less than .7 with the murder rates reported in Uniform Crime Reports. This is especially surprising for the Supplementary Report which is derived from the UCR.

[36] Running the regressions for all Pennsylvania counties (and not just those over 200,000 population) produced similar coefficients signs for the change in concealed handgun permits coefficient, though the coefficients were no longer statistically significant for violent crimes, rape, and aggravated assault. Alan Krug, who provided us with the Pennsylvania handgun permit data, told us that one reason for the large increase in concealed handgun permits in some rural counties was because people used the guns for hunting. He told us that these low population rural counties tended to have their biggest increase in people obtaining permits in the fall around hunting season. If people were in fact getting a large number of permits in low population counties which already have extremely low crime rates for some reason other than crime, it will make it more difficult to pick up the deterrent effect on crime from concealed handguns that was occuring in the larger counties.

[37] We reran these regressions taking the natural logs of the arrest and conviction rates and it continued to produce statistically larger and even economically more important effects for the arrest rates than it did for the conviction rates.

[38] 182 million people lived in states without these laws in 1991 so the Tobit regressions would have also implied 9 more accidental handgun deaths in that year.

³⁹ Crime in the United States 1994.

[40] The ICPSR number for this data set was 6387 and the principle investigator was James Alan Fox of Northeastern University College of Criminal Justice.

[41] Dropping the zero crime values from the sample made the Shall Issue coefficients larger and more significant, but doing the same thing for the accident rate regressions did not alter those Shall Issue coefficients.

⁴² For further descriptions of the procedures for calculating intercensus estimates of population see ICPSR (8384): "Intercensal Estimates of the Population of Counties by Age, Sex and Race" (United States): 1970-1980. US Department of Commerce, Bureau of the Census. Winter 1985. ICPSR, Ann Arbor, MI 48106. Also, see "Intercensal Estimates of the Population of Counties by Age, Sex and Race: 1970-1980 Tape Technical Documentation." US Bureau of the Census, Current Population Reports, Series P-23, No. 103, "Methodology for Experimental Estimates of the Population of Counties by Age and Sex: July 1, 1975." US Bureau of the Census, Census of Population, 1980: "County Population by Age, Sex, Race and Spanish Origin" (Preliminary OMB-Consistent Modified Race).

⁴³ US Bureau of the Census, Current Population Reports, Series P-23, No. 103, "Methodology for Experimental Estimates of the Population of Counties by Age and Sex: July 1, 1975." US Bureau of the Census, Census of Population, 1980: "County Population by Age, Sex, Race and Spanish Origin" (Preliminary OMB-Consistent Modified Race), pp. 19-23.

⁴⁴ Statistical Abstract of the United States, 114th Edition, Table No. 746, page 487.

⁴⁵ Because of multiple arrests for a crime and because of the lags between when a crime occurs and an arrest takes place, the arrest rate for counties and states can be greater than one. This much more likely to occur for counties than for states.