Guns of Fortune:
State Gun Control Policies and the Laws of Supply and Demand
Michael Coates
MA Candidate, University of Rhode Island
Azhain@gmail.com
And
Shanna Pearson-Merkowitz
Assistant Professor, University of Rhode Island
shannapearsonuri@gmail.com
Abstract: Legislators face a compelling dilemma, how can they decrease the prevalence of guns
that may be used in deadly criminal activity? Cities and States around the United States have
laws intended to prevent violent criminals from acquiring and using weapons, but it remains
debatable whether these laws are effective. This study posits that guns are subject to the laws of
supply and demand and the variable gun laws in states across the country decreases the
effectiveness of local and state gun legislation. In short, guns are trafficked across state lines to
meet demand in states with stricter gun laws. Data for the study was collected from the Bureau
of Alcohol, Tobacco, Firearms and Explosives and the Brady Campaign to Prevent Gun
Violence. The analysis shows a strong positive correlation between the strength of a state’s
overall gun laws and the proportion of firearms it imports from other states. This analysis offers
compelling evidence that as states adopt laws to restrict in-state purchases of weapons,
innovative consumers will turn to other, less restrictive markets.
Introduction
The shared powers of the federal system in the United States are often seen as a net
benefit to the development of effective public policy. In a federal system, states serve as
“laboratories” where policies can be adopted to fit the varying needs and constraints of a given
state’s boundaries and policies can be “tested” before being adopted nationally.
Unfortunately, the states are imperfect laboratories. When scientists do experiments, they
do their best to create controlled environments in which they can isolate causes and effects.
Unlike controlled scientific laboratories states share unregulated borders with other
“laboratories” creating a system in which the population being experimented on can move back
and forth between “labs” and bring with them the exact causal agent the scientists were
attempting to control. As a result, in general, politicians try to claim credit for changes in social
problems that may be totally unrelated to any actions taken by the state government and policy
researchers working within the U.S. states have a hard time assigning causation to specific
policies and outcomes.
Using this logic, we argue that state policies often fail or have limited success because
they are unable to control the flow of goods across borders. Thus when the goal of a policy is to
control supply of a variable in which there is demand, the federal government is the only
government able to implement effective policy. To test our hypothesis we use the case of state
gun control laws and their seeming inability to achieve their desired outcome (e.g. Webster and
Vernick 2013; Kleck & Patterson, 1993; Kwon and Baack 2005) of decreasing the availability of
guns within their state. .
Using gun trafficking data from the Bureau of Alcohol, Tobacco, Firearms and
Explosives (ATF), and state gun regulations from The Brady Campaign to Prevent Gun
Violence, we investigate the relationship between the strictness of a state’s gun control
legislation and the proportion of guns used in crimes purchased out-of-state.
Our findings show that differing levels of gun trafficking experienced by states are
correlated with how strongly those states regulate the purchase of firearms. This data offers a
possible explanation as to why firearm regulations may be less effective on a state-by-state basis,
than under a more universal application.
The Case of Gun Control
No industrial country in the world comes close to America in firearm homicides; the
U.S.’s closest competitors enjoy a rate less than one-third the rate in the United States.1 Despite
this fact, gun ownership retains a unique political status in American culture. Gun manufacturers
have federally granted immunity against civil lawsuits as a result of crimes committed with their
products, an unusual and advantageous legal protection representing the strength of their industry
in public policy. Because of the unwillingness of the American federal government to engage
the topic of gun ownership, states are powerful vehicles of firearm regulation. Experimentation
in gun law has allowed substantial analysis to examine the effectiveness of individual regulatory
policies for firearms.
A quick read of the statistics would suggest that state gun control laws are working. In
2010, according to the CDC, the 25 states with the highest per capita firearm death rates in the
United States were also almost uniformly the 25 states with the weakest gun laws (Kaiser Family
1
"United Nations Office on Drugs and Crime. “Ninth United Nations Survey on Crime Trends and the Operations of
Criminal Justice Systems” Web. 12 May 2013.
Foundation 2010). However, the correlation between gun control legislation on the state and
local level and the rate of gun violence is mixed at best (Fleegler et al. 2013; Moorhouse and
Wanner 2006; Wintemute 2002; Kleck and Patterson 1993). Many studies to date show evidence
that state and local gun laws fail to achieve their desired effects (Kleck and Patterson 1993;
Kwon et al. 1998; Loftin et al. 1983; Callahan and Koepsell 1994; Jacobs and Potter 1995) and
that local efforts to curtail gun use are diluted by the inability to pass meaningful gun control at
the federal level (Miller 2008).
Kleck and Patterson (1993) conducted a meta-survey of 29 studies investigating the
effects of gun laws on gun crime. Of the studies in their survey, four found that gun control
lowers crime, eight had mixed results, and seventeen found no effect. Of the research that
investigated state-gun laws only two fond evidence that gun control laws reduced crime while
eleven were either inconclusive or found no evidence that gun laws predicted lower gun crime.
Indeed, Spitzer (1995) argues that state gun control policies are “more important for their
political consequences (i.e. the defeat of gun control opponents) than their policy impact.”
Studies in the past have examined the overall effectiveness of regulation on gun violence
and determined that the best predictors of gun violence are consistently socioeconomic indicators
like alcohol consumption, unemployment and poverty, rather than local gun laws (Kwon et al.
1997). That said, the most comprehensive study to date which employs a sophisticated
measurement of the totality of a state’s gun laws finds that multiple laws directed towards the
same end may be effective in reducing gun crime and gun stocks. Even then, however, demandside socioeconomic factors continue to be the ruling predictor of violence (Kwon and Baack
2005). The research to date indicates that it is individual level indicators that predict “demand”
for guns are far more effective than state policies that limit supply. While we do not in any way
argue that addressing demand would be more effective at reducing gun violence, our study seeks
to shed light on why states seem unable to control supply.
Decentralization Theory and Market Adaptation Theory
There are many criticisms about the functionality of federalism. In some cases, states
may have little incentive to create new, innovative policies that would serve their own
constituents and instead prefer to “free ride” off their neighbors inventions and avoid the risk of
innovating on a policy that could prove to be politically unpopular and/or may not work (RoseAckerman 1980). Galle and Leahy (2009, 1335) argue that “State and local governments can be
thought of as inventors without patents: because anyone can steal their new ideas, what incentive
have they ever had to invent?” Policies may create externalities in which states can benefit from
policies being passed in other states so states can “free ride” off the policies and programs being
funded in other states without expending the effort or funding themselves (Galley and Leahy
2009).
Peterson (1995) explains functional federalism theory which posits that the federal
government should control policies that require redistribution whereas smaller units of
government should control policies that are designed to spur economic development following
the Tiebolt thesis.2 However, federalism in the United States fails to work this way. Instead,
places that have better resources are better able to compete for economic growth and the federal
government fails to redistribute resources to offset shortfalls at the local level (Peterson 1995).
States may also engage in a “race to the bottom” in economic and social policy in which states
compete over attracting highly desired residents and try to push out needy populations (Peterson
2
According to the spatial competition model (e.g. Tiebout), policy is developed simultaneously in many
parallel jurisdictions, and citizens can choose which unit delivers services most efficaciously, leading to
more efficient government and greater match between citizen desires and local policies.
and Rom 1988; 1989). Moreover, while policy “experimentation” is seen by most students of
American government as advantageous, federalism can, “(1) promote duplication, inefficiency,
and a lack of clarity of responsibility; (2) leads to inequalities among citizens in different states;
(3) has contributed to struggles over legalized and de facto racial discrimination and (4) can
encourage state and federal opportunism” (Weissert 2013, 14-15).
Finally, by investigating gun control and crime policy Miller (2010) argues that
federalism has led to the prioritizing of interest group preferences from areas that do not
experience gun crime to the detriment of the interests of those that do. Policy as a result is
reflective of the desires of constituents unaffected by crime and leaves the needs of constituents
affected by crime unaddressed. The take away for students of American government, is that
federalism can promote the interests of some constituents at the expense of others and lead to
policy that does not reflect the reality of the problem.
We add to these arguments that there is another way in which federalism can affect state
policy efficacy. Our argument, in short, is that because states share unregulated borders, they are
unable to effectively regulate the supply-side causal mechanisms behind many issues they wish
to address. Our argument is not entirely different from Peterson’s (1981) argument about cities in
which he argues: “Too often cities are treated as if they were nation-states…On the contrary, by
comparison with national politics local politics is most limited. There are crucial kinds of public
policies that local governments simply cannot execute. They cannot make war or peace,” and
importantly for our study, “they cannot control imports or erect tariff walls” (3-4).
As previous scholars have argued about Peterson’s other work on cities (Hwang and Gray
1991), we argue that the argument Peterson makes above about states holds for state
governments as well. Nations can to some extent control the laws of supply and demand by
curtailing supply through regulations and tariffs. But although many states attempt to do the
same thing, given their fluid borders, their attempts to do so, we posit, should be met with
limited success at best. Instead the laws of supply and demand will continue to regulate the
market only in a quasi-illegal fashion.
Decentralization theory argues that local decision making is superior to a centralized
authority because local institutions hold a more specific understanding of local preferences and
needs (e.g. Oats 1972). Local legislative bodies are more acutely aware of specific problems
affecting their states and likewise are more aware of available solutions capable of solving those
problems. Or as Howard Dean put it, “It is unreasonable to apply laws that may be necessary in
California to rural states like Montana or Vermont.”3
Empirical evidence testing decentralized policies has not always bore out the assertion
that “local is better” though. There is no evidence that decentralization improves economic
growth (e.g. Treisman 2000) and regions tend to be specialized and decentralization is a barrier
to regional cooperation that ensures overall strength rather than local prosperity (Prud’homme
1995). This logic underpins Olson’s (1971) argument that individual actors in their attempt to
self-maximize will instead ensure the detriment of both themselves and their neighbors.
Our theory of the problem of decentralized policy making for gun control legislation rests
on a theory applied to the failure of state-level environmental legislation (Lowry 1992; Lester
1986)—that is, if one state regulates air pollution but the neighboring state does not, polluting
agencies will simply move to the neighboring state and the regulating state will therefore
experience no improvements to air quality but instead will simply lose the revenue these
agencies provide. Indeed, recent research by Monogan et al (2013) shows that states tend to
3
Concord Monitor / WashingtonPost.com on-line Q&A Nov 6, 2003
locate highly polluting industrial plants near their downwind borders, exporting most of the
environmental and health costs to neighboring states. In these cases, environmental regulation
will appear to have failed to achieve its intended goals.
For the purposes of this analysis, gun violence can be viewed as simply a disparity
between states. Because gun policy is so decentralized, states encounter a natural barrier to
addressing gun violence within their borders. Instead, firearm regulation, in that it only addresses
the supply of guns and not the demand for them only goes so far as to create an artificial supply
problem that can be easily overcome by crossing an unregulated border.
These assumptions about economic realities for gun control serve as a capable
explanation for why supply-side regulation has so far been unable to reliably and convincingly
decrease the supply of guns to those who wish to use them in crimes. Areas of high demand for
weapons may be a symptom of the various socioeconomic conditions that have long been
associated with gun violence (e.g. demand side), but gun laws attempt to address supply, not
demand. While gun laws are restrictive on the local level, the market reality is that the United
States no longer has what can be universally referred to as regional markets with capable barriers
to participation. Instead, in an era in which the drive from one state to another is only a matter of
gas costs, it is quite feasible for markets to simply shift as regulations make it easier to purchase
a gun outside the state’s borders than inside it.
Measurable Expectations
Our theory of state market adaptability assumes that demand for weapons should be a
constant force unaffected by restrictive access, and should therefore change very little as
associated with a local legislative agenda. However, the reality of participation in a national
market, as opposed to a local one, means that demand for weapons can be satiated regardless of
local regulation. Access is therefore not actually restricted and instead is manifested as an
artificial supply problem.
This means there should be a measurable movement of guns across state lines that are
highly correlated with the presence (or lack) of state regulations. Those states with stronger gun
laws should reliably “import” guns from states with weaker laws. This is driven by the existence
of an artificial supply problem which encourages savvy consumers to meet unaffected demand.
Such a reliable correlation would be significant evidence that local markets consistently adapt to
local regulations to obtain products for which there is demand.
To be clear, our hypothesis and our theory does not address whether supply-side gun
control legislation decreases gun-related crime or gun-related deaths, nor does our theory address
whether supply side legislation will decrease the prevalence of controlled guns in a state. As we
note above, there is significant research as of late that as part of a comprehensive approach to
decreasing gun-related crime/deaths, supply-side gun control laws can be one effective
component (Kwon and Baack 2005). But more importantly, that in the cases in which gun crime
does occur, whether or not it is less than it would be without the legislation, the guns used in
those crimes are more likely to have originally been purchased out of state.
Our hypothesis, to be clear, does not contradict the idea that supply side gun control laws
decrease the ease with which someone bent on committing a crime of passion can purchase a
firearm, and therefore may decrease gun-related crimes and gun-related death. For example,
“cooling off” laws are intended to help guard against impulsive acts of violence and suicide.
While someone in the heat of passion in Chicago, may be able to simply drive to Indiana to
purchase a gun to use to kill themselves or someone else in a passion crime, the drive time from
downtown Chicago to the closest gun retailer in Indiana, by our estimate (using Google maps) is
one hour in no traffic, meaning, a cooling off period law in Illinois, would still require a 2 hourminimum cooling off period for someone to get to the store, purchase a gun, drive back and find
their victim. The other choice, of course, would be for the person to purchase the gun on the
black market, again, this requires locating a black-market dealer and making an intentional
purchase of an illegal weapon. While both of these may be “viable” options, the increased
burden of either alternative method could prevent many crimes of passion even if it does not
prevent street-level gun-related violence. However, the logical implication of our theory is that to
truly reduce the availability of guns in the United States (and to drive down gun-related crime
through addressing availability) would take federal legislation and enforcement.
Formally stated, our hypothesis is as follows:
H1: As a state’s supply-side gun control laws increase, the proportion of guns used in crimes
originally purchased from out of state should increase.
Data and Methodology
Dependent Variable. To test our hypothesis we require an indicator of how many guns
move between states. Since 1968, the Bureau of Alcohol, Tobacco, Firearms, and Explosives
(ATF) has traced and kept records on firearms taken from crime scenes to their point of
purchase. Since 2006, the ATF has made publically available its reports on traces for each state,
including the District of Columbia and Puerto Rico.4 Using this data, we calculated our
dependent variable as the ratio of exports to imports of guns used in crimes. Succinctly, a state
4
Data for each state is available at the ATF website, here: http://www.atf.gov/statistics/index.html
with an Export Ratio of 3.0, had three guns found in crimes in other states for every one gun
used in a crime that was purchased in the state.
Despite the strength of the ATF’s reports as a unique data source, it does have
limitations. Unfortunately, from 2006 through 2011 the AFT data only includes the top 15-16
sources of firearms from each state, meaning some data is excluded and this skews the analysis
towards larger state sources of firearms which can reliably appear in the top 15 external
traffickers of weapons.5 However, in 2012 the ATF published trafficking data for the first time
which includes the trace results for every gun recovered from a crime scene. In order to harness
both the benefits of overtime data and the full data which is only available in one year, we run
our analysis on both types of data.
Independent Variables. To measure the strength of gun laws on a state-by-state basis,
we utilize data from the Brady Campaign to Prevent Gun Violence.6 Since 2007, the Brady
Campaign has collected a “scorecard” for each state based on the strength of the state’s gun laws.
This study utilizes those scores with some caveats. As the Brady Campaign has evolved, their
political agenda through their scorecard has changed. As a result, overtime different types of gun
laws are used in their score card. In addition, the BC weights their score based on an evolving
standard of which laws are considered more/less important. To deal with this, instead of using
5
Hawaii is excluded from the analysis because it enjoys significant barriers to trafficking that no other
state has, namely over 2000 miles of ocean. We also dropped the observation for North Dakota in 2007
from the model. In 2007, experienced an almost 500% spike in trafficking and then in 2008, its export
ratio returned to 2006 levels and continued at 2006 levels for the remainder of the time series. North
Dakota has a very low export ratio (averaging .55); because the ATF data from 2007 through 2012 only
includes top 15 or 16 supply states for each market state, in 2007 North Dakota was reported as a Top
supplier for California, meaning an additional 451 guns were reported in the data that don’t appear in
subsequent years. When this data is included in the statistical model it becomes unpredictable as the
report is anomalous.
6
Access to each year is limited to special request, but the Brady campaign current scorecard is available
here: http://bradycampaign.org/?q=programs/million-mom-march/state-gun-laws
the raw score provided by the Brady Campaign, each year of the scorecard was reweighted based
on the weighting scheme applied in 2011, and any law from the score which was not tracked
over the entirety of the period was dropped. As a result, while the original score was out of 100,
the reweighted scores have a maximum of 84. The intended result is that a state should have the
same score in 2011 as it had in 2007 if no changes in the law occurred. This method was
employed to protect the analysis from being skewed by changes in the scoring standard over time
instead of changes in the law. Also, our analysis for 2012 uses the same gun regulation score as
2011 because the Brady Campaign discontinued their report card format and so we do not have
information on the changes in the law between 2011 and 2012.
We also created a score where the strength of each law to equal 1, meaning for that score
no individual law is assumed to have a greater or lesser impact on gun trafficking than any other.
We ran our models using all three versions of the score (the original Brady Campaign score, our
preferred 2011-base weight score, and an unweighted score). The substantive findings of the
models did not depend on which form of the independent variable we employed.
We refer to this weighted score as the regulation scale, as it measures how regulated
firearms are within any given state during a given year. This scale should reliably gauge the
overall restriction placed on the firearm market by state officials, meaning that a consumer in a
state with a score of 50 should have a harder time legally acquiring a firearm of their choosing
than if they lived in a state which scores a 15. Table 1 gives the mean scores on the regulation
scale from 2007 through 2012.
[Table 1 about here]
The ATF also collects data on local firearm industries. We include the per capita number
of federal licenses granted to both firearm sales and manufacturing businesses to control for the
presence (or lack thereof) of local market availability. The data on market licenses, however, is
more limited than that of general trafficking information. At the time of this study, the ATF has
made licensing data publically available for the years 2010 through 2012. Thus, we report
models including this variable only in a model limited to 2010-2012.
A number of control variables are also included in the analysis. The violent crime rate
was taken directly from the Federal Bureau of Investigation’s (FBI) website.7 We use Census
data for the total population8 estimates for each state across the six years examined in the study,
as well as the level of urbanization, and overall population density.9 In addition, we include a
control for the size, in miles, of each state.10 We expect larger states to be less prone to importing
guns from other states as the cost of transportation and the likelihood of intervention will be
higher. Our expectations for urbanization and overall population density are less unidirectional.
On the one hand, we expect states with more densely populated urban areas to have higher
demand for guns but if those guns are originally purchased in state or out of state should be
reliant, to an extent, on the strength of the gun laws. Of course it is possible that the laws
themselves may just be stricter in places that experience higher levels of violence, as a reaction
to that violence. However, to not inflate the results based on demand for guns we include these
measures in our models.
Results
Bivariate results
7
Found here: http://www.fbi.gov/stats-services/crimestats
Population estimates are available here: http://www.census.gov/popest/index.html
9
The urbanization ratio was calculated by comparing the totaled population numbers from the urban
clusters page at the census (http://www.census.gov/newsroom/releases/archives/2010_census/cb1250.html) with the population listed by the census for 2010
10
Info is available at this link: http://www.census.gov/prod/cen2000/phc3-us-pt1.pdf
8
An anecdotal look at our data supports our hypothesis and gives a strong example of the
relationship we seek to understand. On its surface, there appears to be a relationship between the
strength of local gun laws and the rate of trafficking a state experiences.
Table 2 shows the top 5 gun exporting states and their 3 largest customers. First and
foremost, the expectation that states with weaker gun laws export more guns than those with
stronger ones has not been shattered. On a case by case basis, the flow of guns appears to be
from states with weak laws to those with stronger ones, with some exceptions. One of
Mississippi’s top ‘customer’ states is Louisiana which in fact scores lower on the regulation scale
than Mississippi. This may mean that states with similar levels of regulation have other,
unknown, variables which impacts trafficking. In addition, we see some regionalization.
Indiana’s top 3 customers are all states that share a border. However, Arizona’s top exports go to
three states, none of which share a border and two of which are a significant trek away.
[Table 2 about here]
A compelling example can be made of the trafficking relationship between Indiana and
Illinois: 63.3% of all the guns traced back to Indiana from other states in 2012 were recovered in
Illinois. Indiana is significantly less regulated than Illinois, scoring only a 4 on the regulation
scale to Illinois’ 26, and the two share a generous border. This has led to a veritable gun
highway, between the suburbs of Indiana into the city streets of Chicago.
This was illustrated very effectively when a 23 year-old college student named David
Lewisbey was arrested by FBI agents for trafficking 43 firearms to a government informant in
December of 2012. Of the incident, Assistant U.S. Attorney Christopher Parente said
“(Lewisbey) would go travel to Indiana, to these gun shows where he would load up literally a
duffel bag, go from table to table paying in cash, large amounts of cash … before returning right
into the worst neighborhoods of Chicago.” (Sweeney and Meisner, 2013)
While this type of trafficking may very well represent a very small percentage of the
overall trafficking figures, this particular example showcases the bases of our hypothesis:
underground suppliers will circumvent and exploit loopholes in the regulation net that controls
the local supply of firearms.11
Table 3 presents the top 5 “market” states and their top “suppliers” for 2012. Like the
previous tables, while there are exceptions, the expectation that states with stronger laws will
import guns from states with weaker ones once again holds out at first glance. Connecticut,
whose regulation scale scores a 55, is the 3rd strongest in the nation behind California (72) and
New Jersey (62). Connecticut imports more than 30 guns for every one that leaves its borders
and while its regulation score is only slightly lower than its number one supplier, its next two
suppliers, which are further away, both have significantly weaker gun laws.
[Table 3 about here]
Figure 1 presents the raw correlation of our regulation scale score and the export ratio for
each year. In each case, the correlation is very large (ranges from a low of .38 in 2008 to a high
of .52 in 2012) and highly statistically significant (p-value ranges from a low of .006 in 2008 to a
high of .0002 in 2012).
[Figure 1 about here]
Multivariate analysis
11
To be clear, it is unfortunate that we do not have the ability to control for gun shows, but public data
does not exist about the location and frequency by year of gun shows that we can identify.
The bivariate results above could easily be driven by other factors. In the following
section, we attempt to control for many of them. Table 4 presents a cross sectional time series
regression analysis for two models: Model 1 which ranges from 2007 to 2012, Model 2 which
ranges from 2010 to 2012 and includes a control for the number of fire arm licenses in the state
(both of these models include only the top 15-16 top exporters for each state). Both of the crosssectional time series analyses include a lagged dependent variable, random effects estimates, and
clustered standard errors by state. Model 3 includes the full population of guns traced for the
year 2012 and employs an OLS regression with robust standard errors. Each model includes data
on each state’s population, size, percentage of urbanization, rate of violent crime, and population
density. In each model, the variables have been recoded to range from 0 to 1 so that the
coefficients are comparable.
In Model 1 of Table 4, the only variables that show a statistically significant relationship
with the trafficking of firearms in the first two models are the strength of firearm regulations in
the receiving state, the size of the state, the population, and the lagged export ratio.
Not surprisingly, the largest coefficient is the lagged dependent variable. This is an
expected relationship because this variable captures established networks of gun trading.
Purchasers of weapons with the intent to resell them may utilize the same “routes.” For example,
gun shows consistently show up to the same places every year, and gun stores may be opened
along the border of strongly regulated states. Gun traffickers also likely have established
purchase locations and established buyers in receiving states. As a result, the previous year’s rate
of trafficking is very highly correlated with the next year.
Of paramount importance to this study’s hypothesis, when controlling against all of the
other variables present in the analysis, the strength of a state’s gun laws have a statistically
significant impact on the trafficking of firearms. The next two most important contributing
factors to the trafficking of guns between states, are the state’s size and population.
The size of the state being statistically significant suggests that one of the chief factors of
how guns are trafficked is convenience. Markets are still predominantly regional forces, and
guns are less likely to escape Texas than Indiana simply because Indiana is conveniently situated
and easier to get to. Also, a state’s population determines the size of its market. More people
indicates greater demand for guns than less people assuming all people who want guns want
generally the same number of them.
Model 2 introduces a control for the number of licensed firearm sellers in the state to
control for the convenience of sales. However, while it does improve the fit of the model
somewhat, this variable fails to meet statistical significance.
Model 3 includes the full data but is only for 2012, so it can not include a lagged
dependent variable. However, our analysis remains substantially unaltered. The coefficient is
much larger for the regulation scale than in Models 1 and 2. However, we posit that this finding
is largely due to the absence of the lagged dependent variable. The 2012 full data model (3) also
has several other deviations from the first two models. Population and Size are both not
significant but urbanization is. Again, we cannot determine how much of this is due to the lack
of a lagged dependent variable and how much of it is determined by the introduction of more
complete data. When we do not include the lag in Models 1 and 2, size and population remain
significant. However, without the lag, urbanization does become significant in the first two
models, although only at the .1 level for Model 2. Overall though, we find all three models to
strongly support our hypothesis.
[Table 4 about here]
The Export ratio and the regulation scale are, unfortunately, somewhat skewed. For this
reason, we ran the models with both variables (the scale and the export ratio) transformed into
their natural logs. The results of these models varied somewhat from those presented above. For
Model 1 and Model 3, the regulation scale score remains statistically significant, but in Model 2,
it does not reach standard levels of statistical significance (p<.106).
Discussion
The results of this analysis represent a strong bed of evidence to support the theory that
state borders are permeable, and therefore, relying strictly on supply-side laws passed at the state
level can only be partially successful. Additionally, the results show that overall, trafficking is
remarkably resistant to changes in the geopolitical landscape.
When thinking about how to generalize the results of this study, it is possible to consider
any public policy in which the states attempt to control the availability of goods. States that
make it more difficult to get an abortion may see those citizens with the financial means to travel
across borders, seek medical procedures prohibited locally. We have likewise seen similar
“consumer migrations” as a result of same-sex marriage becoming legalized one state at a time.
We can also hazard a guess at how the legalization of marijuana in Washington and Colorado
may impact the trafficking of drugs to states which prohibit them. Of course, the case of
fireworks, which are regularly sold on the border in states where fireworks are legal and states in
which they are not, serves as a prescient example. In each of these cases state-level supply-side
legislation would likely decrease the availability of the good but not for those with the means to
travel and purchase the good out of state.
A logical extension of this argument is that the size of the state, as our results suggest,
will matter a great deal to the extent to which the legislation has any measureable effect. In a
state the size of Texas, regulating the availability of a good will likely decrease the presence of
that good quite a bit, but in Rhode Island or Delaware where even the least savvy and or wealthy
individuals can easily drive just a few minutes into a neighboring state, the state governments
ability to pass meaningful supply-side regulation without the same regulations being
implemented regionally is unlikely to produce measurable effects.
References
Callahan, C. M., Rivara, F. P., & Koepsell, T. D. (1994). Money for Guns: Evaluation of the
Seattle Gun Buy-Back Program. Public Health Reports, 109(4), 472-477.
Fleegler, E. W., Lee, L. K., Monuteaux, M. C., Hemenway, D., & Mannix, R. (2013). Firearm
Legislation and Firearm-Related Fatalities in the United States. JAMA Internal
Medicine, 173(9), 732.
Galle, B. D., & Leahy, J. K. (2009). Laboratories of Democracy? Policy Innovation in
Decentralized Governments. Emory Law Journal, 58(6), 1335.
Hwang, S., & Gray, V. (1991). External Limits And Internal Determinants Of State Public
Policy. Political Research Quarterly, 44(2), 277-298.
Jacobs, J. B., & Potter, K. A. (1995). Keeping Guns out of the "Wrong" Hands: The Brady Law
and the Limits of Regulation. The Journal of Criminal Law and Criminology, 86(1), 93120.
Kaiser Family Foundation (2010) Number of Deaths Due to Injury by Firearms per 100,000
Population. Available at: http://kff.org/other/state-indicator/firearms-death-rate-per-
100000/#.
Kleck, G., & Patterson, E. B. (1993). The Impact Of Gun Control And Gun Ownership Levels
On Violence Rates. Journal of Quantitative Criminology, 9(3), 249-287.
Kwon, I., Scott, B., Safranski, S. R., & Bae, M. (1998). Rejoinder To Tomislav V. Kovandzic's
Comment On Our Article "The Effectiveness Of Gun Control Laws: Multivariate
Statistical Analysis". American Journal of Economics and Sociology, 57(3), 369-372.
Kwon, I. G., & Baack, D. W. (2005). The Effectiveness Of Legislation Controlling Gun Usage.
A Holistic Measure Of Gun Control Legislation. American Journal of Economics and
Sociology, 64(2), 533-547.
Kwon, I. G., Scott, B., Safranski, S. R., & Bae, M. (1997). The Effectiveness of Legislation
Controlling Gun Usage: A Holistic Measure of Gun Control Legislation. American
Journal of Economics and Sociology, 62(4), 533 - 547.
Lester, J. P. (1986). New Federalism and Environmental Policy. Publius, 16(1), 149-165.
Loftin, C., Heumann, M., & Mcdowall, D. (1983). Mandatory Sentencing and Firearms
Violence: Evaluating an Alternative to Gun Control. Law & Society Review, 17(2), 287.
Lowry, W. R. (1992). The dimensions of federalism: state governments and pollution control
policies. Durham: Duke University Press.
Miller, L. L. (2008). The perils of federalism: race, poverty, and the politics of crime control.
Oxford: Oxford University Press.
Monogan III, J. E., Konisky, D. M., & Woods, N. D. (2013). Strategic Placement of Air
Polluters: An Application of Point Pattern Models. Presented at the State Politics and
Policy Conference 2013. University of Iowa. Retrieved May 1, 2014, from
http://www.uiowa.edu/~stpols13/papers/Monogan_etal_airPollutersSPPC13.pdf
Moorhouse, J. C., & Wanner, B. (2006). Does Gun Control Reduce Crime or Does Crime
Increase Gun Control?. Cato Journal, 26(1), 103 - 124.
Number of Deaths Due to Injury by Firearms per 100,000Â Population. (n.d.). Number of Deaths
Due to Injury by Firearms per 100,000 Population. Retrieved April 23, 2014, from
http://kff.org/other/state-indicator/firearms-death-rate-per-100000/#
Oates, W. E. (1972). Fiscal federalism. New York: Harcourt Brace Jovanovich.
Olson, M. (1971) The Logic of Collective Action: Public Goods and the Theory of Groups.
Cambridge: Harvard University Press
Peterson, P.E. (1981) City Limits. Chicago: University of Chicago Press.
Peterson, P. E., & Rom, M. C. (1988). The Brookings Review. The Case for a National Welfare
Standard, 6(1), 24-32.
Peterson, P. E. (1995). The price of federalism. Washington, D.C.: Brookings Institution.
Prud'homme, R. (1995). The Dangers Of Decentralization. The World Bank Research Observer,
10(2), 201-220.
Rose-Ackerman, S. (1980). Risk Taking and Reelection: Does Federalism Promote Innovation?.
The Journal of Legal Studies, 9(3), 593-616.
Sweeny, A. and Meisner, J. (2013). Federal case outlines how guns made their way from Indiana
to Chicago. The Chicago Tribune. February 17, 2013. Available at:
http://articles.chicagotribune.com/2013-02-17/news/ct-met-guns-gangs20130217_1_levaine-tanksley-gun-laws-gun-trafficker/3. Accessed on 5/7/2014.
Treisman, D. (2000). Decentralization and Inflation: Commitment, Collective Action, or
Continuity. The American Political Science Review, 94(4), 837.
Webster, D. W. (2013). Reducing gun violence in America: informing policy with evidence and
analysis. Baltimore, MD: Johns Hopkins University Press.
Weissert, C. S. (2014). Chapter 1 - Federalism and Intergovernmental Relations. Guide to state
politics and policy (pp. 7-17). Thousand Oaks, California: CQ Press, an Imprint of
SAGE Publications.
Wintemute, G. (2002). Where the Guns Come from: The Gun Industry and Gun Commerce. The
Future of Children, 12(2), 54-71.
Figure 1. Correlation of Regulation Scale and
Export Ratio
0.6
0.5
0.4
0.3
0.2
0.1
0
2006
2007
2008
2009
2010
2011
Note: each correlation is statistically significant at the p<.01 level
2012
2013
Table 1. Mean Brady Campaign Regulation Scores
Year
Mean Score
Min Score
Max Score
2007
15.38
2
73
2008
15.04
2
73
2009
15.35
0
72
2010
14.64
0
72
2011
14.31
0
72
2012
14.31
0
72
2007 – 2012
14.84
0
73
Table 2. Top 5 Exporting States and Their ‘Customers’ for 2012
(Export Ratio) | {% of guns exported by the supply state} | [Brady weighted regulation scale]
1
2
3
4
5
Supply State
#1 ‘Customer’
#2 ‘Customer’
#3 ‘Customer’
Indiana
Illinois
Kentucky
Ohio
(3.74) [4]
{63.3%} [26]
{7.4%} [2]
{7.2%} [7]
Arizona
Tennessee
Illinois
Texas
(3.70) [4]
{21.4%} [8]
{20.9%} [26]
{18.4%} [4]
West Virginia
Maryland
Ohio
North Carolina
(3.11) [4]
{26.1%} [43]
{16.4%} [7]
{12.4%} [16]
Mississippi
Illinois
Tennessee
Louisiana
(3.08) [4]
{27.8%} [26]
{21.2%} [8]
{16.2%} [2]
Virginia
Maryland
New York
North Carolina
(2.95) [12]
{21.2%} [43]
{20.4%} [52]
{14.1%} [16]
Table 3. Top 5 Market States and Their ‘Suppliers’ for 2012*
(Export Ratio) | {% of local guns provided by supply state} | [Brady weighted regulation scale]
Market State
#1 ‘Supplier’
#2 ‘Supplier’
#3 ‘Supplier’
Connecticut
New York
{6.4%} [52]
North Carolina
{4.6%} [16]
Florida
New Jersey
Pennsylvania
North Carolina
Virginia
(.04) [62]
{13.2%} [26]
{9.4%} [16]
{9.2%} [12]
Wisconsin
Illinois
Indiana
Mississippi
(.10) [3]
{3.3%} [26]
{2.1%} [4]
{1.9%} [4]
Massachusetts
New Hampshire
Maine
Florida
(.13) [49]
{12.9%} [6]
{7.2%} [7]
{4.5%} [5]
New York
Virginia
Pennsylvania
Florida
(.13) [52]
{9.4%} [12]
{7.6%} [26]
{7.2%} [5]
(.03) [55]
{3.4%} [5]
*Hawaii is excluded because its geographical barriers preclude much trafficking from local
sources outside the state, District of Columbia is excluded because it is not a state. This makes
them difficult to compare against other states.
Table 4. The effect of state gun laws on gun trafficking.
2010 – 2012
Accurate 2012
2007 – 2012
With Licensing (robust linear
Data
regression)
Lagged Export
Ratio
.83**
(.05)
.81**
(.04)
N/A
Strength of Gun
Laws
-.08**
(.03)
-.10**
(.04)
-.26**
(.12)
Population
.05*
(.03)
.07*
(.03)
.03
(.09)
Size (in square
miles)
-.04**
(.02)
-.06**
(.02)
.44
(.11)
Urbanization
-.03
(.03)
-.01
(.02)
-.52**
(.13)
Violent Crime
Rate
.00
(.04)
-.05
(.04)
-.05
(.12)
Population
Density
.01
(.03)
.02
(.03)
.11
(.09)
Firearms
Licenses
Per capita
Constant
--
.01
(.03)
-.00
(.09)
.07**
(02)
.08**
(.03)
.67**
(.08)
N
244
147
49
Wald chi2
2838.52
2164.8
Prob>chi2
.00
.00
.00
R-squared
(overall)
.82
.84
.56
Note: Due to changes in how the ATF reports their data, the “Accurate 2012” contains
only one year of data and uses a robust linear regression model. “2007 – 2012” uses the
originally reported ATF data and includes the 2012 population converted to the original
reporting method. “2010 – 2012” contains licensing data which was not available from
2007 through 2009. Both the “2007 – 2012” and “2010 – 2012” use cross sectional time
series analysis. *=p<.1 **p<.05 *** P<.01