Justification and refutation of the imperial model of J. Lott

In the United States, there are several models of legal regulation of the issuance of licenses to carry weapons at the state level.

The first model was the may-issue model, which existed until the late 1980s in many States. Decisions to issue a license were made by local authorities at the county level by county sheriffs, who had fairly broad discretion in issuing permits to carry weapons in the territories entrusted to them.

The situation was changed in 1987: a court in the state of Florida issued a decision that in the state, county sheriffs could no longer, at their discretion, restrict the right to bear arms of those citizens who seemed unreliable to them. From now on, local authorities were required to make their decisions about issuing licenses based on a set of formally defined rules. In other words, if a person does not have an outstanding criminal record or mental abnormalities, as well as several other formal criteria based on which the right of a citizen to carry a weapon can be restricted, the local authorities were obliged to issue this person a license. The Florida model, which greatly simplified the process for citizens to obtain a permit to carry a gun, was called the shall-issue and was later adopted by more than half of the states. The transition to a new model for issuing licenses in many states was actively lobbied by the National Rifle Association of America.

Thus, the transition from the may-issue model to the shall-issue model can be seen as a process of liberalizing legislation on firearms control, since the new rules have greatly facilitated the process of obtaining it for law-abiding citizens.

Changes in state-level gun regulations have drawn a lot of attention from researchers to the impact this has on the overall crime situation.

In American criminology, there is a whole research tradition of studying the relationship between the level of crime and the number of firearms in the hands of the population, with the first studies of this kind dating back to the 1930s. However, until the late 1980s, such studies did not address the question of how changes in firearms control legislation affect crime rates.

The first publications of this kind appeared almost immediately after the introduction of the shall-issue model in some states. D. McDowell and colleagues in a study based on data from Florida, Minnesota and Oregon concluded that the transition to the new system does not lead to a decrease in crime rates, according to representatives of the National Rifle Association. As a response to this study, J. Lott and D. Mustard wrote the article «Crime, intimidation, and legal firearms», and later the book «More guns, less crime». In their study, Lott and Mustard use criminal statistics collected at the county level from 10 States that adopted a new system for issuing licenses to carry weapons in the late 1980s and early 1990s. Using panel data from 1977 to 1992, they tried to assess how the shift to the shall-issue model affected crime rates. As dependent variables, they considered the number of crimes committed in different categories (murder, rape, theft, car theft, etc.). Based on these variables Lott and Mustard conducted two types of regression analysis.

In one of them (the dummy model), they construct a dummy variable that takes a positive value in the first full year after the state switches to the shall-issue model and stores it in the future.

In the other (trend model), they take into account the number of full years that have passed since the transition to the new system.

Thus, using different regression models, Lott and Mustard used different ways to represent the main independent variable in their panel data, namely the adoption of new rules for issuing permits to carry weapons following the shall-issue system. If the entire period from 1977 to 1992 is divided for each state into «before» and «after» the adoption of new rules when using a dummy variable, then the trend model also takes into account the number of years during which a particular state has a «shall-issue» model. The main control variable in the regression models was the number of arrests (arrest rate) at the county level, as well as 36 variables that take into account the demographic composition of the population (gender, skin colour, and age group). Lott and Mustard concluded that the introduction of new gun regulations has reduced crime rates in 10 States.

However, the impact was not the same for different types of crimes. They estimated that the shift to the shall-issue system helped reduce the number of violent crimes (murder and rape) by 4%, but at the same time, this reduction was accompanied by an increase in the number of property crimes (burglary or car theft) by 2%. As an explanatory model, Lott uses the theory of economic analysis of crimes proposed by Gary S. Becker.

In line with «economic imperialism» in the social sciences, Becker once suggested that the criminal be viewed as a rational individual who takes into account all of his benefits and potential costs of violating the law. If the offender does this, it is because their benefits from committing the offence exceed the potential costs associated with the threat of being punished.

Thus, the threat of being punished, on the one hand, and the potential benefits of the offence, on the other, are the main factors that influence the behaviour of the offender. The higher the costs and lower the benefits of violating the law, the lower the likelihood of committing crimes. Lott argues that simplifying the rules for obtaining permits to carry weapons leads to a reduction in the number of violent crimes by increasing the possible costs of their Commission for potential violators of the law. The increased availability of firearms to the public increases the likelihood that when committing a violent crime (murder or rape), the perpetrator will face armed resistance from the victim. This, in turn, means that switching to the shall-issue system increases the potential costs of violators of the law from committing violent crimes. According to Lott, the proliferation of weapons is the most important factor that deters criminal elements from committing violent crimes. The increased costs of committing violent crimes push potential criminals to commit illegal acts, in which the likelihood of encountering armed resistance from the victim is minimal. Property crimes (car theft or burglary) are a form of such criminal activity, since, as a rule, crimes against property are committed in the absence of the victim, when the probability of encountering armed resistance is extremely low.

Although the decline in violent crime was accompanied by an increase in property crimes when switching to the shall-issue system, Lott and Mustard still consider the introduction of this system advantageous from an economic point of view. They estimate that the introduction of new rules for issuing licenses to carry weapons in the 10 States analyzed led to the prevention of at least 1,414 murders and 4,177 rapes, and the combined annual gain from allowing firearms in 1992 was more than $ 5.74 billion. In their opinion, one additional permit to carry a weapon reduces the total losses of victims of violent crimes by $ 5,000. Thus, according to Lott, increasing the availability of weapons to the population is the most economically effective way to fight crime.

Almost immediately after the publication of the article and book, Lott and Mustard's research attracted the attention of both supporters and opponents of the policy of expanding the availability of weapons to the public.

Lott's main argument was also heavily criticized, according to which the presence of firearms in the victim's possession greatly reduces the predictability of the outcome of a violent crime, and therefore increasing the availability of weapons to the public should be seen as the best way to combat them. This thesis did not agree with the J. J. Donohue and S. Levitt, who developed a formal model according to which the increase in the number of firearms in the population increases the number of violent crimes that occur and consequently leads to an increase in the number of incidents involving weapons.

I. Ayres and J. J. Donohue tried to correct and Supplement Lott's research. They used more comprehensive county data and took into account internal trends at the state level. Among other things, they extended the period under review. The panel data they used covered the period from 1977 to 1997, i.e. 5 years more than Lott's one. The results of their study disproved the results of the regression analysis performed by Lott and Mustard. They estimate that most states that have adopted the shall-issue model have a statistically significant association with a higher crime rate, rather than a lower one, as Lott claims. And the incentives associated with switching to this system to commit more violent crimes tend to be particularly strong in those states that have adopted it in the last 10 years (in the 1990s). Ayres and Donohue also concluded that the transition of many States to the shall-issue system raised the total cost of committing crimes for criminals by $ 524 million. This small positive effect, however, is offset by an increase in the number of collisions with weapons (gun accidents), which in many ways makes the thesis «More guns, less crime» untenable. According to Ayres and Donohue, switching to the shall-issue system leads not to a reduction, but, on the contrary, to an increase in the number of violent crimes.

A comprehensive review of Lott's thesis was carried out by a team of researchers led by J. J. Donahue. In their article, they analyze data from 1979 to 2010 at both the county and state levels. They use the number of police officers (police) and the number of prison inmates (incarceration) as control variables. Also, for the period from 1999 to 2010, they take into account the dynamics of the «cocaine epidemic», so that the results of their analysis are more reliable than those of Lott. The developed model with a dummy variable (which takes into account the presence or absence of the shall-issue system in a particular state in a particular year) records a significant increase in the number of crimes in seven categories of crimes (murder, rape, aggravated assault, robbery, car theft, burglary, and theft). According to the proposed evaluation and transition to new rules for issuing gun, licenses is accompanied by growth in all seven categories of crime an average of 8%, except for murder which increased by 3%, but this figure is statistically insignificant. For the period from 1999 to 2010, their regression models record a statistically significant increase in the number of homicides by 1.5% annually.

Summarizing all the above-mentioned criticisms of the study by J. Lott, we can say that its main drawbacks are as follows. First, in a very limited number of control variables used (mainly only the number of arrests) and ignoring other important factors that affect the change in crime rates (the number of police officers and prisoners). Second, internal trends at the state and County levels are not taken into account. Third, including only 10 States that switched to the shall-issue system in the analysis and ignoring all the others made his research a victim of the selection bias effect, which makes his conclusions very sensitive to any manipulation of the number of cases (such as excluding Florida from the data under consideration).