Availability and use of weapons in the neighbourhood as risk factors for criminal offending among prison inmates in Nigeria

The current study recruited participants from among the prison inmates in Nigeria to determine the relative impacts of availability and use of weapons in their respective communities prior to incarceration as risk factors for criminal offending. Eight hundred and twenty one participants made up of those awaiting trial and convicts, ranging in age from 16 to 65 years (M= 30.4, SD= 7.6) were recruited through opportunistic (non probability) sampling across ten medium and maximum security prisons in Nigeria to participate in the study. Adopting the quantitative analysis, the computed outcomes predict the effects of weapon availability as potential risk factors to criminal offending among this group. The implications of the findings for crime reduction policy in targeting offenders who are at risk of committing criminal offences due to weapon availability are emphasised.


Background
The impact of breaking the law and the subsequent arrest and incarceration of criminals may spaw n negativ e psychological consequences in offenders, w ho must rapidly come to terms w ith the shock of prison life and deal w ith the burden of know ing that their families may be suffering both emotional and financial losses because of them (Hagan & Dinov itzer, 1999).Giv en the effects of incarceration on the offender and ev eryone around them, it appears necessary to explore motiv ating factors to criminal offending in order to tackle or address the causes of crime than to opt for punishing offenders through incarceration.Of these motiv ating v ariables, the risk factors model is increasingly v isible and popular.Risk factors model assumes that there are multiple, and often ov erlapping, risk factors in an indiv idual"s background that interact w ith one another and consequently increases an indiv idual"s v ulnerability or propensity to engage in negativ e behav iour in the absence of protectiv e factors.I n other w ords, risk factors are those characteristics, v ariables, or hazards that, if present for a giv en indiv idual, make it more likely that this indiv idual, rather than someone selected from the general population, w ill dev elop problem behav iour (Mrazek & Haggerty, 1994).
One of the most reliable risk factors to criminal offending obtainable in offender researches is av ailability and use of w eapon in the neighbourhood (Kleck & Hogan, 1999;Turner, Simons, Berkow itz, & Frodi, 1977;Wells & H orney, 2002).I t has long been argued that w eaponry and firearms could giv e indiv iduals w ho are v ulnerable to inflicting injury on others the courage to attempt aggressiv e acts that they w ould otherw ise be afraid to attempt.I n particular, a w eapon may be especially important in facilitating attacks by armed robbers or other v iolent offenders against their v ictims.Psychologists hav e also argued that the sight of w eapons could stimulate aggression through classical conditioning processes resulting from the learned associations inv olv ing aggressiv e acts and w eapon use (for a review , see Turner, Simons, Berkow itz, & Frodi, 1977).I ndeed, the presence of aggressiv e env ironmental cues such as w eapons can increase the accessibility of hostile, aggressiv e thoughts and lead to more aggressiv e behav iour (for a rev iew , see Brennan & Moore, 2009).
Giv en the facilitativ e part played by w eapon as an important stimulus for dominance and aggression, it is unsurprising that armed robbers and other v iolent offenders are more likely to carry w eapons and firearms to perpetrate their v iolent acts.While it is appreciated that the use of w eapons and firearms by v iolent offenders is a global phenomenon (for a rev iew , see Brennan & Moore, 2009;I gbo, 2001;Kleck & Hogan, 1999;Turner, Simons, Berkow itz, & Frodi, 1977;Wells & Horney, 2002), the sources of firearms, such as rifles, pistols, and other dangerous, locally made w eapons used by v iolent offenders, in Nigeria has raised many of the usual questions among the general public.How ev er, it can be speculated that the country"s problems w ith small arms and w eaponry can be dated back to the 1967-70 civ il w ar, during w hich the southeast made a failed attempt to secede.During this period, most able bodied men enlisted-either v oluntarily or by conscription into the armed forces as fighting soldiers, especially on the rebel Biafran side.Those enlisted into the armed forces included the unemployed and underemployed, school leav ers, and drop-outs, as w ell as some members of the criminal population.These new recruits learned how to use rifles, machine guns, and other firearms against their opponents.At the end of the hostilities, it appears that many defeated and demobilised soldiers did not surrender their w eapons to the federal authorities.Some of these w eapons w ere even abandoned at the w ar fronts, w hile others w ere either buried underground or carefully concealed in bushes to make them easily retriev able should the need arise.Such a need could be for criminal purposes, as in the case of armed robbery.There is also the possibility that some serv ing policemen and military officers w ho returned to the barracks at the end of the w ar might hav e giv en out their officially assigned w eapons to persons w ith a criminal intent in exchange for cash (I gbo, 2001).The implication of all this is that there is a possibility that indiv iduals w ith access to w eapons may be tempted to resort to criminality.
Besides the civ il w ar risk phenomenon, it can be contemplated that cross border smuggling as a result of the civ il w ars in neighbouring countries like Sierra Leone and Liberia may hav e led to the proliferation of arms and w eaponry into Nigeria.This is facilitated by huge cross-border smuggling and mercenary activ ities (from Chad and Niger, for example) and the country"s long, porous borders that are poorly policed due to the inadequate resources and the lack of capacity of the security agencies (Ginifer & I smail, 2005).The three most notorious border posts for the illicit smuggling of small arms and w eaponry into the country, as reported by Ginifer and Ismail, are the I di-I roko and Seme (in the south-w estern States of Lagos and Ogun), Warri (in Delta State), and the border posts in the Adamaw a, Borno and Yobe states in the north-east.
Although v arious potential sources from w hich offenders in the country could obtain w eapons hav e been highlighted abov e, there is little or no empirical ev idence on the sources of the w eapons used by offenders in Nigeria, nor hav e there been many studies examining the probable contributions of w eapon av ailability in the neighbourhood as potential risk factors for criminal offending in the country.I n response to this, the current study recruited participants from among the prison inmates in order to determine the relativ e significance of the use of firearms to criminal offending among this group.The patterns of relationships betw een the use of firearms and the criminal history of the participants w ere also explored.I t is hoped that the findings from the current study w ill hav e the potential to inform the crime reduction policy in Nigeria, by prov iding an indication of the form and size of the problem under inv estigation, in addition to policing strategies aimed at stemming the supply and use of w eaponry to perpetrate criminal/v iolent offending.

Ethical considerations
The research receiv ed a fav ourable ethical opinion from the University of Surrey Ethics and Quality Committee.Permission to access prisoners w as obtained from the Comptroller General of Prisons in Nigeria.The Comptroller instructed the Assistant Comptroller of Prisons (Administration) to w rite an approv al letter for the researcher to be able to v isit prisons across the follow ing states for the purpose of data collection: Kano, Oyo, Edo, Delta, Abia, Lagos, and the Federal Capital Territory in Abuja (see table 1 & figure 1).
To ensure the confidentiality and informed consent of the participants, prison inmates recruited for the study w ere told - not to put their names or any of the pages of the questionnaire or put any marks that might identify them  that their participation in the research w as v oluntary  that the return of a completed questionnaire constituted informed consent to participate in the study, and  that the respondents should not discuss their responses w ith other inmates during the questionnaire administration session.

Methodology
Research population and sampling procedures At the time of collecting the data that inform the findings of the present study, there w ere 227 prisons across the country (including maximum and medium security, satellites prisons and 11 farm centres) holding approximately 46,000 inmates, comprising those aw aiting trial, conv icts, detainees and condemned prisoners.Of these prisons, the maximum security ones take into custody all classes of prisoner, including condemned conv icts, lifers, and those on long term sentences.The medium security prisons also take into custody both conv icts and remand inmates, but mostly inmates on short term sentences.The satellite prisons, on the other hand, are intermediate prison camps set up in areas w here the courts are far from the main prisons.They serv e the purpose of prov iding remand centres especially for those w hose cases are going to courts w ithin the areas.When conv icted, they could be easily mov ed to appropriate conv ict prisons at w hich to serve their jail terms.On the last note, the farm centres are agricultural prison camps that hav e been set up solely to train inmates in agricultural based v ocations.The rationale behind this is to equip the inmates w ith the agricultural based skills they w ill hav e to depend upon after completing their jail term.Of these prisons, the researcher w as only allow ed by the prison authorities to recruit participants from among the conv icts and those aw aiting trial in nine medium and one maximum security prisons (see table 1 & figure 1), that represent participants across the tribes, religions, and geo-political div ides in Nigeria.
How ev er, it should be highlighted that it has alw ays been the tradition in Nigerian prisons to select leaders among the inmates w ho liaise betw een the prison authorities and the other inmates.The leadership of each cell is selected by the prison w arder.I n the North of the country, the leadership of the cell is called Seriki, w hile they are mostly addressed as Prov ost in the Eastern and South Western parts of Nigeria.The Serikis/Prov osts are w ell recognised and respected among the inmates, and normally serv e as intermediaries betw een the prison authorities and the rest of the inmates w ith regard to the griev ances and other issues relating to the general w elfare of the prison inmates.The instructions that the Serikis/ Prov osts giv e to their fellow inmates are generally follow ed by other inmates, w ho see the Seriki/Prov ost as a superior inmate.Because of their influence, the researcher ensured that he established good rapport w ith the Serikis/Prov osts in order to facilitate the recruitment of the participants (inmates) for the study from a larger group (prison inmates) through an opportunistic sampling technique.The system of opportunistic sampling (i.e., non-probability technique) is justified in this type of research because the often-chaotic nature of booking facilities does not lend itself to systematic random sampling (for a rev iew , see Bennet, 1998;Wish & Gropper, 1990).
Originally in the ten prisons v isited for the administration of questionnaire, 979 inmates w ere approached and 821 respondents w ere considered for the analysis.
Questionnaires w ere discarded w hen they w ere largely incomplete, illegible, or contained similar answ er sets for all responses.The responses from the few female inmates w ere also discarded as they contributed a tiny number to the ov erall sample.The follow ing are the breakdow n of the prisons v isited and the patterns of the response from the respondents:

Measures
A self report standardised scale w as used to elicit a response from the participants.
While it appreciated that the use of self report is v ulnerable to manipulation and self presentation biases in the offender samples, due to the common belief that offenders are "cons" w ho should not be trusted and w ould not hesitate to lie or manipulate their responses to psychological measures, ev idence exists that selfreport questionnaires can be an accurate and robust instrument of data collection from the offender population (Kroner & Loza, 2001;Mills, Loza, & Kroner, 2003).What is important is that the questions or items on the self report questionnaire should be relev ant to the characteristics being measured.Hav ing ensured this, the self report scale used to elicit a response from the participants in the current study cov ered a range of topics, including personal demographic characteristics, criminal history, and w eapon av ailability and use prior to incarceration.
 Personal Demographic Characteristics: To ensure that the researcher has recruited a w ide v ariety of prison inmates, the participants w ere asked to state their age, gender, ethnicity, religion, highest educational achiev ement, occupation and marital status before arrest.
 The Criminal History Scale: This is a standardised scale dev eloped by the researcher to measure the respondents" prev ious contact w ith criminal justice system.I t contains questions on the arrest history, prison status, reason for admission, and the conv iction history of the respondents.A cumulativ e index of the criminal history scale w as constructed w ith a Cronbach alpha of 0.70.

 Measure of weapon availability and use:
The respondents w ere asked to respond to a number of questions about firearms/w eapon av ailability and their perception of crime.They w ere asked to indicate: (a) w hether they had ev er been injured by a gunshot in the past; (b) w hether they had shot at someone or attacked anyone w ith a w eapon in the past; (c) w hether they had carried a gun w ith them w hilst committing a crime, or had ev er used a gun to commit a crime; (d) w hether it w as important to hav e a gun in their neighbourhood; and (e), if so, w hat w as the reason for this.The participants w ere also asked to list any other w eapons that they had used in the past w hile committing a crime.The w eapon av ailability and use measure adopted in the current study also demonstrated good coefficient reliability, w ith a Cronbach alpha of 0.82.

Analytic strategies
Statistical Package for Social Sciences (SPSS) w as used for data analysis.A combination of univ ariate (frequency counts and odds ratio), biv ariate (chi square statistics), and multiv ariate (logistic regression) analyses w as employed to make statistical decisions from the data collected from the participants.I n particular, the univ ariate (frequency counts) analysis w as used to determine the demographic characteristics and lev el of the participants" accessibility to and use of w eapon and firearms before the present incarceration.The odds ratio statistics w as used to determine the risk estimate of w eapon av ailability and use by the prison inmates.
The difference betw een w eapon accessibility and use among different categories of prev ious offences and prison status of the participants w as determined by chi square statistics.The extent to w hich w eapon accessibility could contribute to criminal offending among the participants w as also established by logistic regression analysis.
The age range of the participants w as betw een 16 and 65 years, w ith a mean age of 30.34 (S.D. =7.6).Christians (60%) and Muslims (38.9%) dominated the religious faiths of the participants.There are also a few traditionalists (0.4%), and 0.7% did not declare their religious faith.Prior to their confinement, more than half (52.4%) w ere married, (45.9 %) single, and (1.7 %) div orced.Ov er half 65% of the sample had obtained a secondary education or less, w ith 28.9% hav ing receiv ed a diploma or univ ersity degree, and a small proportion of 3.9% hav ing an Arabic education.
Arabic education describes a process of sending children and w ards to Mallams (teachers) to study Quran, Hadith and other branches of I slamic know ledge.
Regarding the family background of the participants, more than half w ere from a polygamous background (59.8%), w hile the remaining 40.2% w ere from monogamous families.

Criminal history of the participants
The descriptiv e analysis of criminal history of the participants from self report indicates that the majority of the participants had a history of a prev ious arrest (73 %), and 27% w ere first-time offenders.With regards to the current prison status of the participants, almost tw o third w ere aw aiting trail (62.6%), and the remaining 37.4 % w ere conv icted.Their durations of admission into the prison v aried, w ith more than half of the participants hav ing been in custody for up to three years (48.2%).Others had been in custody betw een four and ten years or more.Various reasons w ere giv en for being in custody.Prominent among them w ere armed robbery (39 .7%),burglary and theft (17.3%), drug related offences (17.2%), assault (9.3%), manslaughter (5.8%), fraud (419), conspiracy to defraud (4.0%), and other miscellaneous offences (6.7%).Although it w ould hav e been appropriate to focus on offenders w ho had engaged in interpersonal crime or solely gun crime in order to obtain more precise results, practical reality is that it w as not feasible to group the participants according to their offences during the data collection due to the logistic problem associated w ith prison rules in the country.Nev ertheless, the response of participants w ho admitted to hav ing engaged in interpersonal v iolence among the inmates w as used to make predictions that inform the findings of the present study.

Firearms av ailability, possession and use
Out of the participants, 49.7% admitted to the accessibility of firearms in their neighbourhood; and a third (34.8%) admitted that they had access to a gun, and other w eapons.Various reasons w ere giv en for using firearms in their neighbourhood: protection and self defence, or for hunting or game expeditions.Out of the 34.8% of the participants w ho admitted hav ing access to a gun, 24.4% confessed that they had shot at someone.Similarly, 25.7% of them admitted to gun possession w hile committing crime.Of this particular, 25.7%, 9.6% w ere presently arrested for armed robbery or v iolent offences (see table 2).These responses w ere giv en for w hy the participants had firearms w hen committing the particular crime (i.e. in case they needed it, they alw ays carry a firearm, etc.) Almost a fifth of all the participants (18.1%) confessed to hav ing been prev iously shot by someone.They w ere probed further about w ho shot them, and table 3 summarises their v aried responses.The follow ing summarise the categories of prev ious conv icted offences w ithin participants w ith history of gun shot injury.Property and v iolent offenders w ere the most likely group to be shot and this is statistically significant (Chi square 65.9 p<.0001) Chi square analysis w as also employed to deter mine the differences betw een categories of prev ious conv icted offences w ithin participants w ho confirmed av ailability of firearms in their neighbourhood (see table 6).Note: *p<0.050 There was significant difference between cat egories of previous convicted offences and availability of firearms in the neighbourhood.
Using an odds ratio to determine the risk estimate of w eapon av ailability and use by the prison inmates, there w ere statistically significant effects of av ailability of and accessibility to firearms to criminal offending among the participants (see table 7).A small proportion of those w ho admitted to hav ing access to a gun hav e a history of gun shot injuries.The reasons giv en for previous gunshot injuries v ary, but tw o major facts can be deduced from the responses.Firstly, some admitted that they w ere shot by the police to force a confession during interrogation.While the police authority in Nigeria consistently denies the existence and use of lethal arms and torture to extract statements from suspects, ev idence abounds from the records of suspects and prison inmates of the extensiv e use of lethal firearms against suspects, beating and kicking, and the unnecessary use of restraints, such as handcuffs and leg chains (Etanibi & Chukw uma, 2000).I n research conducted by Etanibi and Chukw uma (2000) on police community v iolence in Nigeria, the findings rev ealed that 81%, 73.2% and 77.5% of the inmate respondents, respectiv ely, reported hav ing been beaten up by the police, threatened w ith w eapons and tortured in police cells.Further, 39.7% reported hav ing been burnt w ith hot objects, 33.3% receiv ing electric shocks and 50.8% being pierced by needles or sharp objects.
The other major reason for a prev ious gunshot injury reported by the participants in the current study w as that they w ere v ictims of the O"odua People"s Congress (OPC) and community v igilante groups.This may also be true because the members of these groups hav e cashed in on the grow ing sense of disenfranchisement among the Nigerian population in the face of the soaring armed robbery rate and ineffectiv e policing (Akinyele, 2001;Guichaoua, 2006).Although the use of ethnic militia, like OPC and v igilante groups, to enforce law and order has no basis in the Nigerian judicial system and constitution, their operations are only an attempt to complement the police efforts to combat crime, but they inadv ertently undermine the effectiveness of the police.The av ailable ev idence show s that members of the OPC, among other v igilante groups in Nigeria, are inv olv ed in extr a-judicial killings, torture, unlaw ful detention and serious abuses of the rights of alleged criminals (Akinyele, 2001;Guichaoua, 2006;Human Rights Watch, 2003).To further complement the unlaw ful activ ities of these v igilante organisations, some pow erful politicians and top gov ernment officials are know n to hav e hijacked some of these organisations and used them to threaten, intimidate and ev en kill their political opponents (I nternational Crisis Africa Report, 2007).
I t is also interesting to note that few respondents admitted that they had been shot during an armed robbery by their v ictims.This may be allow ed as a last resort for self defence by w ealthy indiv iduals w ith a license to hold firearms in Nigeria.But the other reasons mentioned for a prev ious gunshot injury, such as "an unknow n hunter", "shot by armed robbers", "during the secret cult clashes", "during the Biafra w ar", etc., need to be treated w ith caution, since the respondents may not be being completely honest in this regard.As mentioned earlier, most of them are aw aiting trail, and an admittance to firearms possession and use may implicate them and further compound their case in the law court.Nev ertheless, the computed outcomes from the odds ratio and regression analyses predicted that the av ailability of firearms in the neighbourhood are a potential risk factor and predictor of criminal offending among the participants w ho w ere detained or conv icted for violent and property offences.This finding can be interpreted from the perspectiv es of w eapon av ailability or possession could induce a psychological inclination to attack or the psychological strength in offenders to exert control ov er their victims by using the threat of harm or actual harm.This position is supported by Wells and Horney (2002) w ith regard to gun possession; more than any other w eapon, guns increase the possibility of attack because they empow er offenders or their users to inflict damage from a distance, w ithout endangering themselv es.
As a last w ord, it should be acknow ledged that the present study suffers from some limitations that must be addressed in future w ork.Firstly, the participants selected for the present study w ere male prison inmates.Although it is possible that male participants can be more easily approached, because the researcher w ho distributed the questionnaire w as of the same sex, the domination of the male sample may limit the extent to w hich w e can make an inference about w eapon accessibility and use among female prison inmates.I t w ould therefore be better if the future studies recruited an adequate number of female participants in order for such findings to be replicated to the entire prison population of Nigeria.Similarly, it w ould hav e been better if the participants recruited for the present study w ere mainly those arrested and detained for interpersonal crime or v eterans of the 1967-70 civ il w ar but, as noted earlier, the pragmatic reality of the circumstances during the data collection process does not allow this because the researcher w as constrained to comply w ith certain regulations in order not to jeopardise his security and that of inmates w ho participated in the study.I t w ould therefore be better if the future studies recruited a greater depth of subsamples among these groups for the better replication of the findings.The uncorroborated self report method of data collection and the extent to w hich the respondents underreported or ov er reported their inv olv ement in v arious activ ities and behav iours may as w ell not be truly determined.Although the researcher assured confidentiality of the participants" response, the research topic is sensitiv e and the admissions of the participants to prev ious weapon use are potentially compromising.I n this sense, social desirability factors cannot be ruled out, as participants may w ant to conceal certain information in order to prev ent themselv es from being implicated.All of the appropriate cav eats notw ithstanding, the findings of the current study hav e contributed to research know ledge that w ill be relev ant to researchers, practitioners, and policy makers in the criminal justice system in Nigeria to hav e a basic Availability and use of weapons in the neighbourhood 29 understanding of the impact of w eapon av ailability and use in the neighbourhood as probable risk factors to criminal offending among prison inmates in the country.

Concluding thoughts
The findings of the current study hav e established the relativ e contributions of w eapon av ailability as a potential risk factor to criminal offending among prison inmates in Nigeria.To mediate the effects of w eapon av ailability as a risk factor for criminal offending, there needs to be a stringent law and enforcement of the law controlling w eapon (firearms) possession and use in the country.While it is appreciated that some local dangerous w eapons, such as machetes, cutlasses, and axes, may be difficult to control because of their proliferation in the country, in addition to the fact that firearms are mostly obtained through v arious means such as theft from armouries and seizures from security officials during robberies, the gov ernment should intensify its efforts in tracking the license, possession, and use of small arms, w hich are largely concentrated in the hands of armed groups, criminal gangs, and elites.Any law enforcement agents, be they serv ing or retired, w ho lend out firearms to offenders for them to perpetrate criminal activ ities should be adequately sanctioned according to the law .The law guarding or protecting unlaw ful firearms possession and collaboration to posses should make no exception for anyone, if crime induced by w eapon av ailability has to be reduced in the country.I t may as w ell be mentioned that a clearly conceiv ed w elfare policy should be put in place by the Nigerian gov ernment for demobilised soldiers, militia men, as w ell as disadv antaged populations.This must be the central feature of crime control policy necessitated by risk of w eapon av ailability.Without this, the prospect of ensuring crime free society aimed at promoting the safety of ev eryone in the community in a meaningful and sustainable fashion w ill remain an illusion.

Figure 1 :
Figure 1: Map of Nigeria

Table 1 :
Response rates from each prison v isited

Table 3 :
History of gunshot injury

Table 5 :
Prev ious conv iction by history of gun shot injury

Table 6 :
Categories of prev ious conv icted offences w ithin participants w ho confirmed av ailability of firearms in their neighbourhood

Table 7 :
Risk estimate of accessibility to firearms and criminal offending Not es: * indicates statistically significant effect at the 0.05 level