Interpreting the Data

К оглавлению
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 
68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 
85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 
102 103 104 105 106 107 108 109 110 111 112 113 114 115 

Braga et al. (2002) suggest that the three sources of data on illegal gun

markets are not directly comparable but broadly compatible. Each data

source has its own inherent limitations and, as such, it is difficult to credit

the insights provided by one source over another source.

None of the three sources of data contradicts the hypothesis that stolen

guns and informal transfers (as opposed to transfers from legitimate sources)

predominate in supplying criminals and juveniles with guns. However, they

also clearly suggest that licensed dealers play an important role and that the

illegal diversion of firearms from legitimate commerce is a problem. In their

review of these three sources of data, Braga and his colleagues (2002)

suggest that, in the parlance of environmental regulation, illegal gun markets

consist of both “point sources”—ongoing diversions through scofflaw

dealers and trafficking rings—and “diffuse sources”—acquisitions through

theft and informal voluntary sales. As in the case of pollution, both point

sources and diffuse sources are important (see also Cook and Braga, 2001).

Braga and his colleagues (2002) also speculate that the mix of point and

diffuse sources differs across jurisdictions depending on the density of gun

ownership and the strictness of gun controls.


General Model

Real interventions in gun markets tend to target particular types of

firearms or sources. If policy raises the difficulty (cost, time, risk) of obtaining

a particular type of gun or using a particular type of source, the effect

might be mitigated by criminals’ substitution across types of guns or sources.

The following framework is helpful for organizing what is known, and

what we would like to know, about whether access interventions can reduce

harms from criminal gun use.

There are many types of guns; the term “type” encompasses both the

literal firearm type (e.g., handguns versus long guns) and the source by

which it is acquired (e.g., retail purchase, private sale, theft, loan, and other

types of firearm transfers).3 Furthermore, there are many types of individuals

(legal possessors, juveniles, convicted felons and other persons prohibited

from legal gun possession). Restrictions aim at reducing firearm possession

or use by some of those groups. For analyzing the effects of these

restrictions, consumer demand theory provides a useful conceptual framework,

in which the use of each type of gun by each type of individual

depends on the total cost that individual incurs in acquiring or retaining

that gun. This generates a specific volume of use (possession or purchase)

by each type of individual for each type of weapon. When the difficulty

felons face in acquiring new guns rises, for example because of a targeted

intervention, we assume that new gun use will decline among felons;

whether that decline is substantial can be determined only empirically. Use

of other kinds of guns may rise.

We use the term “cost” as broader than the money required for purchase

of the item. Nonmonetary costs may be particularly important for

gun acquisition by offenders, compared with purchases of unregulated legal

goods; these costs include the time required to locate a reliable source or

obtain information about prices, the risk of arrest by police (and sanction

by a court), and the risk of violence by the seller. These are potentially

important in any illicit market and have received some attention in the

context of drug markets (Caulkins, 1998; Moore, 1973).

To make clear how this framework operates, consider an intervention

that raises the costs criminals face to obtain new guns. The direct or “own”

effect of this intervention is to reduce criminals’ demand for new guns. Yet

this is not the end of the story. The total effect of the policy intervention is

the sum of the “own effect” and a “cross-effect” reflecting criminals’ substitution

of used guns for new ones as new guns become more costly. Even

if the own effect is negative, the cross-effect might be sufficiently positive to

render the overall effect close to zero.

3For discussion purposes, we are dramatically simplifying the large variety of guns available

to consumers. Guns vary by type (revolvers, semiautomatic pistols, derringers, rifles, and

shotguns), caliber and gauge (e.g., .22, .38, 9mm, .45, 12 gauge, 20 gauge, and dozens of

other bullet calibers and shotgun gauges), and manufacturers (e.g., Smith & Wesson, Sturm

Ruger, Colt, Glock, Sig Sauer, Lorcin, Bryco, and hundreds of other manufacturers). There is

ample evidence suggesting that criminal consumers seem to prefer certain types of guns.

The patterns of substitution among sources may be different for different

types of potential buyers. Adults without felony convictions or other

disqualifications can presumably choose between buying new guns from

retailers and used guns from legal private sellers. Juveniles, by contrast,

cannot buy from retailers or law-abiding dealers in used guns. However,

they can conceivably substitute by obtaining guns from a number of sources

outside legal commerce, such as residential theft, informal transfers through

their social networks, and scofflaw dealers; as one source becomes more

difficult, youth may obtain more from another.

This framework is limited to an assessment of effects on the quantities

of guns owned, which is not the final outcome of interest. Rather, it is crime

or violence that ultimately interests policy makers. Whether changing the

number and characteristics of firearms in the hands of persons of a given

type increases harm is an additional question that requires different data

and is considered at the end of the chapter.

We classify potential market interventions in two dimensions: markettargeted

(primary or secondary) and supply or demand side programs. For

example, consider police undercover purchases from unlicensed dealers.

These aim to shift the supply curve in secondary markets by increasing the

perceived risk of sale; dealers will be less willing to sell to unknown buyers

and will charge a higher price when they do. Whether this has an influence

on criminal possession of guns depends on many factors, such as the share

of purchases that are made from nonintimate dealers and the price elasticity

of demand (i.e., how much an increase in the price affects the purchase and

retention of guns). Other interventions are focused on reducing demand,

for example, taxes on FFL sales (primary market) and increasing sentences

for purchasing from unlicensed dealers (secondary market).


What determines the demand for guns? Offenders acquire firearms for

a variety of reasons: self-protection, a means for generating income, a

source of esteem and self-respect, and a store of value. For example a rise in

violence in a specific city may shift the demand curve up because of the

increased return to self-protection. We assume that the demand for guns for

criminal purposes is negatively related to the price and other costs of acquisition;

there is no research on the elasticity with respect to either price or

any other cost component that would allow quantification of the importance

of this effect. Note that individuals make two kinds of acquisition

decisions, active and passive; passive refers to holding rather than selling a

valuable asset. Most market interventions aim only at the acquisition decision;

retention is affected only indirectly, in that an increase in the value of

a gun may lead to a greater willingness to sell to others.

Individual demand has an important time dimension to it, which makes

inconvenience of acquisition a potentially valuable goal for an intervention.

The value of a gun is partly situation-dependent; a perceived insult or

opportunity to retaliate against a rival may make a firearm much more

valuable if acquired now rather than in a few hours, when the opportunity

or the passion has passed. Analytically and empirically that is a substantial

complication; individuals are now characterized not only by their general

risk of using a firearm for criminal purposes but also by their time-specific

propensity of such use. This also allows for the possibility of positive effects

from interventions that merely reduce the fraction of time an offender has a