Penalization and multidimensional poverty: improving our understanding of poverty amongst offenders in Ciudad Juarez, Mexico

Using a multidimensional approach to poverty measurement, this article aims to contribute to an improved understanding on the main aspects of deprivation experienced by former participants of organized crime in Ciudad Juarez, Mexico, a city that became the epicenter of violence due to the ‘war on drugs’ declared in 2006. A sample of 180 surveys and 20 in-depth interviews were implemented to evaluate multidimensional poverty amongst young men serving a prison sentence for a series of crimes related to organized criminal activity (aged 12 to 29) in Ciudad Juarez, Chihuahua. An equal number of surveys were implemented to a group of non-offenders who had no criminal record and resided in a marginalized area of the same city. The research finds that while offenders fared worse in several non-income dimensions of poverty compared to non-offenders, qualitative evidence revealed that experiences of poverty of offenders were not homogenous, as suggested by existing studies. One of the key findings to emerge from the research is that participation in organized crime decreases income poverty; however, participation did not constitute an effective nor sustained pathway out of poverty nor did it decrease deprivation in other dimensions due to a highly skewed distribution of income in the illegal economy and the use of gains from illegal activity to fuel conspicuous consumption, findings that are similar in the established Western literature on youth gangs.


Introduction
The phenomenon of organized crime in Mexico has been an important subject of headlines since the declaration of the 'war on drugs' launched under the Felipe Calderón administration (2006)(2007)(2008)(2009)(2010)(2011)(2012). Throughout this period, there has been a notorious increase in the penitentiary population 1 and a much publicized herculean effort on behalf of the state to punish those responsible for crime. In the years spanning from 1995 to 2012, which marked the end of the Calderón administration, the penitentiary population had more than doubled (Secretaría de Seguridad Pública Federal and Secretaría de Gobernación in [2]).
The expansion of the prison population did not begin with the 'war on drugs'. It was part of a structural change that started with the dismantling of the welfare state in Latin America in the 1980s and the introduction of neoliberal policies, which reached their peak in the mid-90s. These policies were accompanied by what the renowned sociologist Loïc Wacquant described as Bthe criminalization and penalization of poverty^ [3], in other words, incarceration as a way to for the state to manage poverty. According to Hathazy and Müller, B…imprisonment expansion-in addition to, and accompanied by punitive policing-serves to manage and regulate the most unruly sectors of the informal-criminal economy while reproducing the social hierarchies that emanate from the triangulation of formal, informal and criminal economic structures^( [4]: p. 16).
There is existing empirical evidence in Mexico that those being punished are not high-level criminals, but small-time offenders that dangle at the bottom of the organized criminal hierarchy. Unfortunately, knowledge of the prison population in Mexico is almost non-existent and does not feature prominently in the academic literature. Reasons include the exclusion of prison population from census data, the reticence of Mexican authorities to make data on offenders available to the public, the inherent administrative difficulty in prison access to work with offenders, and the safety concerns for researchers working in prison settings. Due to situations of auto-gobierno, or self-government, in some penitentiary institutions of the country -where powerful criminals exert some degree of control over prison officialsprison access and work with offenders can constitute a hazardous affair. As a consequence, not enough work has been carried out on prison populations in Mexico, nor has this work targeted former participants or organized crime exclusively. Academic work to understand the conditions of poverty under which former participants of organized crime in Mexico lived prior to their incarceration is to my knowledge, inexistent. Exceptions of scholarly work that have assessed socioeconomic conditions of offenders in the country have looked at penitentiary populations as a whole and have not focused exclusively on organized crime participants. Moreover, these poverty assessments have conflated poverty with income and level of education. Given the significant increase in homicide rates experienced in Mexico perpetrated by organized criminal groups since the declaration of the war on drugs, 2 and given that insecurity and crime constituted the prime concern amongst citizens in the country in 2014 [6], the absence of this specific population from the academic literature is striking. As a consequence, our understanding of deprivations formerly experienced by the penitentiary population is very limited, as the current evidence lacks the use of a more comprehensive definition of poverty and lacks qualitative data. This paper will attempt to fill that gap by exploring the background of this understudied but crucially important population group.
Using Loïc Wacquant's argument on the penalization of poverty and providing evidence from the case of Ciudad Juarez, Mexico, this article focuses on former participants of organized crime that were serving a prison sentence at the time of the fieldwork (October 2014-April 2015) involved in this research. A multidimensional approach to poverty was used to gain a more nuanced understanding of poverty experienced by offenders. Through material derived from surveys and interviews, these dimensions are used to compare the findings of this study with other qualitative research that has taken place in Western settings. To my knowledge, no work on penitentiary populations in the country has yet incorporated a multidimensional poverty framework. This article will specifically contribute to an improved understanding on the main aspects of deprivation experienced by former participants of organized crime in Ciudad Juarez, and will more broadly feed into a larger discussion on how findings derived from Western settings compare to evidence from a developing country setting.
This paper is structured as follows: It first presents a contextual section describing the effects of neoliberalization on imprisonment in the United States and Mexico, followed by a brief overview of existing work on prison populations in Mexico. The next section provides a description of multidimensional poverty and the methodology implemented in this research. An analysis and discussion section follows, where multidimensional poverty indicators are used as a framework to describe offenders' experiences of poverty. The article concludes with a summary of the findings.

Context Neoliberalization and imprisonment
The implementation of neoliberal policies -which reached their Bgolden age^at the end of the 1980's under the leadership of Margaret Thatcher and Ronald Reagan, led to a visible retrenchment of the state in which fiscal austerity, drastic reductions in the budget of social programs, and privatization dominated the socio-political scene. 3 The latter, coupled with a media frenzy that emphasized the idea that insecurity and crime were rampant, led to feelings of insecurity amongst the middle and high echelons of society [7]. To address this anxiety, a drastic increase in incarceration rates occurred in the United States, a trend which was not necessarily a response to a rise in crime rates, but instead, as Wacquant explains, due more to: Bthe extension of recourse to confinement for a range of street crimes and misdemeanors that did not previously lead to custodial sanction, especially minor drug infractions and behaviors described as public disorders and nuisances...^( [3]: p.125) The heightened incarceration rates led to an exponential growth of the prison population in the US, which rose from 380 thousand in 1975 to half a million by 1980. In 1990, the prison population stood at 1 million and by 1998, it was close to 2 million [3]. At the present time, the United States has the highest prison population in the world in absolute numbers [8,9].
Incarceration rates did not affect everyone in the US equally, however. There was a class, but also a racial component, in the new security policies that were becoming the norm in some cities of the country. A prime example was New York City, where under the auspices of Rudolph Giuliani, mayor from 1994 to 2001, and William Bratton, commissioner of the NYPD from 1994 to 1996, policies of Bzero tolerance^became the norm. These policies were based on the Bbroken windows theory^, which states that in order to deter large-scale urban pathologies, crime has to be stopped at its lowest level [10]. Under the logic Bwhoever wants to steal an egg also wants to steal an ox^an assumption for which there is no substantial empirical evidencethe duo began a crack-down on crime that saw an unprecedented number of arrests and an apparent decrease in crime [3].
Giuliani and Bratton were praised all over the world for their success in the reduction of crime in the Big Apple. Across the Atlantic and down south in Latin America, zero tolerance began to creep into the political discourse of decision-makers in the region, mirroring the Manichean view that George H.W. Bush (1989) espoused at height of the so-called Bwar on drugs^in the US, when he declared that in order to start building a safer society, it was necessary first to agree that: B…society itself doesn't cause the crime -criminals cause the crime^(as cited in [11]). And so, a trend in which a heightened sense of insecurity, coupled with individual responsibility, took precedence in the discourse, side-lining the idea that environmental factors (which the champions of neoliberal ideology called Bsociological excuses^) contribute to explaining individual behavior [3]. Under this prevailing structure of thought, poverty and inequality cannot be blamed for influencing an individual's personal choice to commit delinquent acts.
It did not take long for Latin America to adopt this discourse. In a setting where inequality was and still is rampant (the 10 countries with the highest Gini index in 2012 were all countries from the Latin American region), 4 institutions were (and still are) flailing, and democracy is still relatively young, we took ownership of these ideasciting the argument that zero tolerance policies, or mano dura/iron fist policies were a Bnecessary evil^. Mexico was not an exception. After the country experienced a devastating financial crisis in 1994, it became the recipient of a significant and much needed bailout by international institutions. The bailout required to regain stability came at a price and the Mexican government implemented the Brecommendations^of the Washington Consensus to the letter. At the time of the signing of the North American Free Trade Agreement (NAFTA), which of course, boded well with the introduction of neoliberal polices, crime in the Federal District of Mexico increased exponentially, from a rate of 1700 crimes per 100 thousand population in 1993 to 2835 in 1995 [12]. Likewise, crimes accompanied by violence increased by 500% in the period from 1990 to 1996 [13].
In tandem with the adoption of neoliberal policies in the economic front, the increase in violence and crime witnessed in Mexico was tackled with the adoption of iron fist policies. 5 As in the case of the US, these translated into higher incarceration rates throughout the decade of the 1990s, a period in which the Mexican prison population increased at an average annual rate of 9%, with the most pronounced growth taking place after 1996 [15]. According to the secretariat of state and the secretariat of federal public security of Mexico, in the period from 1995 to 2012, the incarcerated population more than doubled, from 93.6 thousand to 239.9 thousand (as cited in [2]). Today, Mexico is the country in Latin America with the second largest prison population in absolute terms, surpassed only by Brazil [9].
One of the significant findings of academic work with offenders in Mexico is that Binmates are generally not the most dangerous criminals, but they are the poorest^ [15]. According to the results from a sample of more than 800 inmates serving time in eight federal rehabilitation centers 6 -the term used for prisons in Mexico -Azaola and Pérez-Correa [16] found that almost three fourths of the sample had a secondary school education or less and only 2.5% of the sample had completed a bachelor's degree. The highest percentage (13.6% of the sample) was self-employed, possibly in the informal sector, followed by taxi driver and trader (comerciante), with 12.4% of the sample falling into these last two employment areas. The monthly income for 75% of males serving time in these federal detention centers was between 240 and 12,000 MXN (9.44 and 472.44 GBP) [16]. 7 Any amount below 1242.61 MXN (48.92 GBP) per month on a per capita basis constitutes extreme poverty, which means that income is insufficient to buy a basic nutritional basket [17].
An earlier study conducted in 2002 in 15 prisons in the states of Mexico, Morelos and the Federal District also found evidence that it is the poorest members of society that end up serving time in prison. For drug-related crimes, for example, the composition of individuals found in prison were Bnot major drug traffickers, but actually small-time pushers or possibly, consumers who were arrested with small quantities that barely exceeded the amount tolerated for personal consumption^ [15]. The authors also found that almost one quarter of those serving a sentence for theft had stolen less than 1000 MXN (45.8 GBP) and almost half had stolen less than 6000 MXN (275.2 GBP), leading them to conclude that the Bjustice system concentrates primarily on punishing property crimes perpetrated by small-time criminals^ [15].
Although this work has been illustrative in providing an overview of the socioeconomic background of inmates in several areas of the country, these efforts have been 5 One of the prime examples of the glorification of these policies is evidenced in the 4.3 million USD paid to Giuliani and associates for a consultancy to evaluate and issue recommendations for the reduction of crime in Mexico City. The resulting document is essentially an extrapolation of the policies applied in New York to manage crime in the most populous city in the world. However, according to Juarez, the implementation of BZero Tolerance^policies was faced with important obstacles, starting with the lack of data to the deficient law enforcement structure that persists and which is evident in the poor selection, training and salaries of the police force, as well as rampant impunity and corruption (2007). As an example, the starting salary of a police officer in Mexico City is equivalent to $6 thousand USD. The starting salary for a New York Police Department officer is $34 thousand USD. See [14]. 6 Federal centres house individuals who have been sentenced or are in the process of being sentenced for federal crimes. 7 Study cited originally mentioned weekly income, which was between 60 and 3000 MXN (2.36 and 118.10 GBP).
scarce. The first study, which constitutes the first systematic effort to survey the prison population in federal penitentiary centres [16], is not representative of local penitentiary centers, which concentrate the largest percentage of prison population in the country: In 2012, there were 420 penitentiary centers, of which 405 were local and only 15 were federal [1]. The generalisations made from this study are therefore relevant to federal penitentiary centres; which means that the background of offenders in local institutions is largely unknown. In addition, the studies mentioned do not measure poverty according to the criteria established by the National Council for the Evaluation of Social Development Policy of Mexico (CONEVAL), the official body tasked with poverty measurement in the country and in charge of producing the Multidimensional Poverty Index (MPI), which provides a more comprehensive picture of poverty by measuring six deprivation indicators in addition to income. Instead, these studies have provided indicators on educational achievement and previous employment and have lacked a qualitative component that would provide a deeper understanding of experienced deprivations of offenders prior to incarceration. In addition, the work cited above does not compare the poverty indicators with other population groups, which makes it difficult to conclusively determine that it is the poorest members of society who end up in the confines of the prison system. The next section will briefly describe how poverty is measured in Mexico as well as the methodology used for MPI measurement implemented in this study.

Defining poverty
This paper strongly argues for conceiving poverty not solely as income-based, but rather as a multidimensional construct. Currently, Mexico is one of the few countries in Latin America that produces their own national multidimensional poverty index (MPI) [18], 8 currently calculated by CONEVAL. According to this agency, Mexico currently has a population of over 120 million inhabitants of which 55.3 million, equivalent to 46.2% of the population, were considered poor and 9.5% were considered extremely poor in 2014 [19]. Poverty also varies regionally, with the north of Mexico being more prosperous and industrial, while the south is home to some of the poorest states in the country, such as Oaxaca and Guerrero, which have poverty rates exceeding 60% [17].
For its calculation, poverty is measured using two income lines: the minimum wellbeing line and the well-being line. The first measure considers the per capita monthly cost of a food basket per person, whilst the second measure considers the per capita monthly cost of food and non-food baskets. 9 The cost of the baskets varies between the urban and the rural areas. 10 Because CONEVAL recognizes that poverty is a multidimensional 8 The other countries in Latin America implementing national multidimensional poverty measurements are Colombia (2011), El Salvador, Chile and Costa Rica (2015). 9 The non-food basket items include but are not limited to the cost of public transportation, cost of items to clean the household, cost of items for personal care, cost of clothes, shoes and accessories, among others. 10 For example, an individual is considered poor if her/his income is below 2542.13 MXN (100.08 GBP) per month, which is the cost of the food and non-food baskets per month in the urban area. phenomenon, six additional dimensions other than income are considered for the classification of individuals into several categories of poverty. These are: 1) Educational lagging, 2) access to health services, 3) access to social security, 4) access to food, 5) quality and spaces of the household, and 6) access to basic services in the household. Using the six indicators on social deprivation and the corresponding well-being lines, individuals are grouped into the following categories of poverty, shown in Table 1:

Methodology
The present study involved the administration of 180 surveys containing questions to evaluate multidimensional poverty among a sample of young male offenders between the ages of 12 and 29 (definition of youth according to the Mexican Youth Institute, IMJUVE), 11 who were serving a prison sentence in a local 12 penitentiary centre in Ciudad Juárez, located in the state of Chihuahua (in the north of Mexico), at the time of the research. Ciudad Juárez became the epicenter of the 'war on drugs' declared by Calderon and in 2009 was the city with the highest homicide rate in the world due to a turf war waged between two rival organized criminal groups seeking control of the lucrative territory, which sits directly at the border with the United States (see Fig. 1). The sample of inmates had been sentenced for a variety of crimes related to organized criminal activity. 13 Twenty in-depth semi-structured interviews were also conducted amongst a sub-sample of survey participants to gain a more nuanced understanding of their experiences of poverty.
The sampling in prison used a combination of purposive sampling coupled with random selection. The aim of the study was to work with young men who had participated in organized crime exclusively (purposive sampling). A list of offenders who had been sentenced for offenses related to organized crime was provided by the director of the correctional facility in which the fieldwork was carried out. Every fifth person from this list was selected for survey participation. The surveys were administered face to face to those who consented to participate by three research assistants (RAs) and myself. We recorded the answers respondents provided on printed surveys. Only two candidates were unwilling to participate in the survey.
All 180 survey participants were asked at the end of survey administration whether they would be willing to participate in a follow-up interview (prison access was restricted to 2 weeks: in the first week, we conducted surveys, and in the second week I personally conducted all interviews). From those who consented, RAs and myself chose 20 participants for in-depth, semi-structured interviews. I instructed RAs to suggest candidates based on their survey experiences. Each day that surveys were 11 Age group considered youth by the Mexican institute of youth (Instituto Mexicano de la Juventud, IMJUVE). This study was limited to including young males, the demographic group that perpetrates a high proportion of crime in Mexico. 12 The shortcomings of research in federal institutions have been previously addressed. 13 Law against organized crime (Ley federal contra la delincuencia organizada) in Mexico considers that when three or more individuals have organized to execute -in a permanent or systematic way -a series of crimes, they will be sanctioned as members of organized delinquency. The crimes considered in this research were: 1) crimes against public health -which included 1.1) transport and possession of drugs, 1.2) distribution, 1.3) consumption and/or 1.4) cultivation, among others; 2) crimes related to arms-trafficking; 3) others form of organized delinquency including human trafficking; and 4) theft. The study also considered individuals sentenced for crimes such as: 5) kidnapping, 6) extortion; and 7) homicide related to organized crime. conducted in prison in the first week, RAs and I would discuss the candidates that they had deemed suitable for interview and the reasons for their suitability. We discussed criteria such as openness in the survey and skimmed the surveys we had administered to selected candidates. Unfortunately, because prison access was a time-sensitive issue, I was unable to input all of the data from the surveys and carry out preliminary analyses that would have allowed for a more sophisticated interview selection process. Based on the recommendations of RAs, final candidates that fit some of the study's assumptions (deprivation was apparent along multiple dimensions) as well as those that were Population with an income below the minimum well-being linemeaning that income is insufficient to purchase the food basket, and deprived in three or more of the social indicators just described.

Moderate multidimensional poverty
Population with an income below the minimum well-being line, meaning that they cannot purchase the food basket, and deprived in one or two of the social indicators.
Multidimensional poverty Population with an income below the well-being line -meaning that income is insufficient to purchase both the food and non-food basket items, and deprived in at least one of the six social deprivation indicators

Vulnerable due to deprivation in social indicators
Population that presents deprivation in one or more of the social indicators but has an income above the well-being line.
Vulnerable due to income Population that presents no deprivation in social indicators but whose income is below or equivalent to the well-being line.
Not poor and not vulnerable The population whose income is above the well-being line and who also show no deprivation in terms of the social indicators.
Source: Elaborated by author based on information from CONEVAL outliers (those who did not show deprivation) were selected. I conducted all interviews face to face in the selected prison. Each interview lasted between 45 min to an hour. An equal number of surveys (n = 180) were also administered to a comparison group of young men in the same age group (12 to 29) who had no self-reported prior criminal record and who resided in a marginalized area of Ciudad Juarez (Felipe Angeles). Access to this population was obtained by a trusted member of the community who was a football coach for young men in the area of the city where data was collected. He facilitated access to candidates willing to participate in the study. Because he had a close relationship to most of the young men who participated in the survey, he could corroborate that participants had not engaged in delinquent activity.
For the evaluation of multidimensional poverty, I sought to maintain the same questions included in the national survey on income and expenditures (Encuesta Nacional de Ingresos y Gastos del Hogar, ENIGH), which is used to generate the multidimensional poverty index in Mexico. For income, delinquent respondents were asked about the number of members in the household where they resided prior to incarceration and to indicate an income range for the household 14 ; non-delinquents were asked about the household where they resided and income range at the time of the data collection. This figure was divided into the number of members to determine whether the individual was poor or extremely poor by income standards. The other six indicators, included in the MPI measurement -educational lagging, access to health services, access to social security, quality and spaces of the household, access to food, and access to basic services in the household, were assessed according to questions drawn from the ENIGH survey (see Table 2 for criteria to establish deprivation below). Survey questions for the non-delinquent sample were the same; the only difference was that delinquents were asked about deprivation experienced prior to incarceration and non-delinquents about deprivation at the time of the data collection. For the interviews, respondents were asked to elaborate on their experienced deprivations.

Analysis of data and discussion
Comparison of data indicates that the offender sample fared worse along five of the dimensions considered in the multi-dimensional poverty analysis, finding that those in the confines of the prison system were more deprived in access to health services, access to social security, quality of the spaces of the household, access basic services, and had a higher share of educational lagging. However, they were less deprived in income and in access to food. The following graph summarizes the proportion of deprivation in the delinquent population sampled along the dimensions previously discussed, comparing results to the sample of non-offenders (Fig. 2).
Despite these deprivations amongst offenders, qualitative evidence from the in-depth interviews finds that deprivation was not experienced homogeneously across delinquents. For example, while at least three out of the 20 interview participants described experiences of extreme poverty, at least six participants indicated that they did not feel 14 Income ranges were calculated using data on income deciles for households according to the ENIGH survey. This method to determine income has been used in other national surveys. For example, see the national youth survey, Instituto Mexicano de la Juventud (IMJUVE), Encuesta Nacional de Juventud 2012. they grew up deprived. One respondent who grew up in extreme deprivation belonged to an indigenous group residing in the Tarahumara mountains. Gabriel, 15 who lived with his nine siblings, described having no access to basic services and struggling at times to pay for foodstuffs prior to his incarceration. He described that when working as a day laborer, for example, he was earning a daily wage of MXN $120 (GBP £4.76) for picking fruits and vegetables. When asked whether he had access to basic services, he summed up his situation as follows: 15 All names have been changed for the protection of study participants.

Educational lagging
Participants did not complete secondary school or were not enrolled in any educational programme at the time of the study (non-delinquents) or prior to incarceration (delinquents) 3. Access to health services No access prior to incarceration (delinquents) or at the time of the study (non-delinquents) via any of the providing institutions mentioned.
4. Access to social security If they failed to answer positively to at least one of the three questions concerning access to social security prior to incarceration (delinquents) or at the time of the study (non-delinquents).

Quality and spaces of the dwelling
The dwelling where respondents resided prior to incarceration (delinquents) or at the time of the study (non-delinquents) had any of the following characteristics: -Main material of the floor is earth -Roof made from waste, cardboard, metallic or asbestos plates or straw -Material of the walls made from clay; bamboo, palm or reed; carton, metal or asbestos plate; or waste -more than 2.5 occupants per room in the household 6. Access to food* There was a positive answer for the items indicating moderate and severe food insecurity.
7. Access to basic services and household appliances* The dwelling where respondents resided prior to incarceration (delinquents) or at the time of the study (non-delinquents) presented at least one of the following characteristics: -Water obtained from a well, river, lake, stream, or tubed water brought to the dwelling from another house, public fountain or hydrant -The dwelling did not have drainage or the drainage is connected to a tube that leads to a river, lake, ocean or crevice -The dwelling did not have electricity -The fuel used to cook or heat food items was firewood or carbon -Household did not have access to either a refrigerator, washing machine, or both *Indicates those that have been altered from the original indicators developed by CONEVAL Source: Elaboration by author Well, we were always humble. We never had anything. We were…didn't I tell you that we had a hole in the ground outside for a bathroom? There was no drainage, there was no light. Gradually, we have built our house.
Another respondent, Alfredo, who had worked as a hired assassin, remembered growing up in abject poverty, not having money for buying a pair of shoes. He described that he and his siblings would walk around barefooted, without sandals, for months. At times, they would not even wear shirts, just some sort of underwear garment. Although these cases highlight extreme poverty, not all interview participants shared these experiences. For example, when asked about his economic situation and whether he grew up feeling deprived, Jorge described his situation as follows: Not really, I told you. Deprived? No. On my father's side of the family, well, thank God they are well off economically. And because I am the favorite grandson, I was given everything…and my mother too, well, she was practically killing herself to provide for us…when I was young, thank God, I never lacked shoes, or clothes… In another interview, Pedro mentioned that he abandoned school not due to income constraints (the main reason why school abandonment occurs in Mexico, see [20]), but rather because he wanted to have a stable income for the independence that it represented. Juan José also mentioned that although he did not feel he grew up deprived, although he did mention that his parents would refuse to give him money to buy 'lujos', or luxury goods.
Classifications of respondents into the different categories of poverty confirms that poverty was not homogenously experienced: Most individuals in the delinquent sample were categorized as 'multidimensionally poor', meaning that prior to their incarceration they had an income below the well-being line (income was insufficient to purchase both the Source: Author's elaboration from own dataset and data from CONEVAL.  food and non-food basket of basic items) and were deprived in at least one of the six deprivation indicators. However, there were also individuals belonging to other categories of poverty: More than one quarter of respondents were 'vulnerable due to deprivation in social indicators' (26.5%), meaning that they were deprived in one or more of the social indicators but had an income above the well-being line, and almost one fifth were in 'extreme multidimensional poverty' (18.9%), meaning that they could not afford to purchase a basic food basket and were deprived in three or more of the indicators. There was also a small percentage who were found to be 'not poor and not vulnerable' (4.6%), who had an income above the well-being line and who showed no deprivation in terms of the indicators (see Fig. 3).

Income deprivation and conspicuous consumption
One of the often cited explanations of the appeal of participating in organize crime in Mexico is economic [21]. The comparison of income deprivation across the delinquent and non-delinquent samples reveals that delinquents are in fact less income deprived than non-delinquents, suggesting that participation in organized crime is reducing income poverty. While 72% of delinquents were poor by income standards, for the nondelinquent sample, the figure was 86.9% (see Fig. 2).
In interviews, participants who had participated in organised crime confirmed obtaining significantly higher amounts in the illegal economy in comparison to their earnings in legal employment. Data in Table 3 summarises the proceeds of organised criminal activity and compares them to formal and informal employment: Guillermo, for example, mentioned that he would make a daily wage of between MXN $2000.00 and $3000.00 (GBP £79.33 and £119.00) through participation in delinquent activities. 16 In comparison, when he held a job in public transportation, he was making a daily wage of only MXN $83.33 and $100.00 (GBP £3.30 and £3.97), which means that the proceeds of crime were between twenty and thirty times greater than the daily wage in his formal job. Rodrigo mentioned making between MXN $5000.00 and $10,000.00 (GBP £198.33 and £396.67) on a weekly basis in organised crime related activity, while in a construction job he made one-tenth or less than that amount.
Despite the disparity in earnings between formal employment and organised crime related activity shown in Table 3, it is important to bear in mind the following observations. First, income gains made from participation in the illicit economy are insufficient to have an impact on income distribution. Despite suffering from less income deprivation, a large proportion of the delinquent population still belongs in the income deciles that control a very small proportion of total income. Figure 4 illustrates the proportion of the total national income received by each decile of the Mexican population, where the first decile is the poorest and the tenth decile is the richest. Data from the ENIGH survey reveals that at the national level, the two highest income deciles receive more than 50% of total income 17 and the two lowest income deciles only receive less than 5% of total 16 The exchange rates mentioned throughout the thesis correspond to the exchange rate between the Mexican Peso (MXN) and the British Pound (GBP) at the time of the writing, which spanned from 2014 to 2017. Throughout this period, the exchange rate has fluctuated between a low of MXN $22.92 = GBP £1.00 and a high of MXN $25.2 = GBP £1.00. 17 According to data from the ENIGH survey, total income for 2012 was estimated at 1199245100 MXN. This was divided by the income estimated for each decile, to come up with the proportion of total income (%) that is controlled by each income decile.
income. Of the delinquents sampled, almost half (41.5%) belong to the lowest two income deciles.
Second, participation in this type of crime is neither stable nor constant. While a significant amount of money is made from illicit activity (e.g. through kidnapping), it is unlikely that this happens frequently enough to represent a stable income. Second, there is evidence that incomes earned depend on the individual's position in the hierarchy of organised crime, meaning that there are significant disparities in the income gains for those at the top of the hierarchy in comparison to those at the bottom. While no data is available on the distribution of profits in organised criminal groups in Mexico, Muehlmann's ethnography of Mexican narco-culture on the US border documented evidence of a 'routine exploitation to which the lower-level participants in the drug trade are subjected to' ( [23]: p. 80). According to Azaola and Bergman [15], the literature on offending populations in Mexico illustrates that those who end up behind bars are usually small-time drug peddlers, not the most dangerous or prosperous dealers. Although part of the reason is rooted in the unfairness of the justice system in Mexico -meaning that those who can't buy justice, seldom obtain it, it is also the case that most participants of organized crime do not manage to reach top-level positions in the organized crime hierarchy that would allow them to escape poverty.
Valuable insights from research on youth gangs in other contexts provide further evidence that skewed distributions in profits corresponding to position in the hierarchy of the criminal structure are common. In a research on youth gangs in London, Densley [24] found that Bincome variation between elders and youngers (terms used to denote hierarchy within the gang) appears 'highly skewed' B (p.57). According to his study, one elder gang member, who occupied a more 'senior' position in the gang, indicated that those below him would get a weekly wage of approximately GBP £500.00, while the real weekly proceeds of the drug sale would amount to around £5000.00 or £6000.00, meaning that those in the lowest tier of the gang were only getting a '10% cut' (p.57).
Densley compares gangs to multi-level marketing, which he explains is Bintrinsically flawed, yet it appeals because it sells the dream of material wealth and     [22] independence and appears to be outside the mainstream of business as usual^( [24]: p.80). This is very similar to the 'sales pitch' that street gangs and organised crime use when they are recruiting youngsters to participate in the illegal economy (p.80).
Densley points out that Bwhat both groups fail to advertise is that the system benefits the few over the many. For every gang younger who makes a decent living or even a decent supplementary income from drug dealing, there are countless others who do little more than break even^(p.80). If the illegal drug economy in Mexico does generate an estimated yearly revenue of between 30 billion [23] to 50 billion USD [25], constituting between 3 to 4% of the country's GDP and employing more than three times the number of people than the nationally-owned petroleum giant, Petroleos de Mexico, PEMEX [26], then the findings from Densley's study can only ring as true in the Mexican setting. Let's not forget that more than 70% of delinquents who had participated in organized crime in this sample were still income poor, a high share of delinquents belonged to the lowest income deciles and more than a quarter were vulnerable due to deprivation in social indicators. This finding corroborates well the reality of organized crime participation: BEso de que todo aquel que entra al narco se hace rico, es nomás un pinchi mito^[the idea that all who participate in narcotrafficking become wealthy, is only a *** myth] [27]. In an interview conducted by a Mexican journalist, a participant of organized crime from the city of Culiacán -one of the prime 'hot-spots' of organized criminal activity, revealed that 70% of those who participate in this type of criminal activity are very poor, and most of the money they make from the trade is spent on drugs and alcohol. This perhaps explains why despite less income deprivation, more than one quarter of respondents were vulnerable due to social indictors, meaning that although they had an income above the well-being line, they were still Source: Author's elaboration. Note: Proportion calculated based on information from ENIGH survey, households and current total trimestral income.  Note: Proportion calculated based on information from ENIGH survey, households and current total trimestral income deprived in one or more of the social indicators. This shows that the gains in income from illicit activity were not used to decrease poverty in other dimensions, for example, to access health care or to decrease educational lagging. Evidence from the interviews in this study confirmed that spending the proceeds derived from illicit activity on material items was prevalent. For example, Juan José described how Balcohol consumption, partying, wanting to be well dressed, wanting to have a car, having, I don't know, jewelry…you know, looking good,^were motivating factors for criminal participation, findings that fall in line with other studies of marginalized urban youth. MacLeod's study of young, poor, men residing in the projects of Massachusetts pointed out that ( [28]: p.48) access to things like Bbeer, sneakers and joints^were important and in his ethnography of youth gangs, Densley recalled witnessing 'bedrooms adorned with large flatscreen televisions and stacked ceiling-high with boxes of Nike shoes', [24]. For Anderson [29], the fashion imperative of the Philadelphia street youth were Timberland boots and gold chains. According to Shover and Honaker the intense concern of offenders with outward appearances, with 'looking good', is derived from a strong attachment to the values of street culture, values that place great emphasis on the 'ostentatious enjoyment and display of luxury items' ( [30]: p. 283). In his study of street culture in Philadelphia, Anderson has referred to fashionable clothes as sources of esteem and respect [29].
For the marginalised youth of Ciudad Juarez, the coveted luxury items are constituted by the apparel worn by the elite, the middle and high-class Mexicans. These clothes, referred to as 'ropa fresa' [preppy clothes] in some interviews, are best exemplified with clothes bearing the Polo, Abercrombie & Fitch, and Hollister logos. However, these are not representative of 'street' culture; instead, subjects are subscribing to the fashion statements set out by the affluent class of Mexico. According to the Mexican National Survey on Discrimination (ENADIS), lacking economic resources was the main reason cited for which individuals had felt their rights had not been respected, followed by physical appearance [31]. In Juarez, the imitation of consumption patterns of the elite, a class that commands respect, represents a mechanism to prevent discrimination based on appearance. As Raphael [32] explains, in Mexico, wearing certain labels constitutes a Bpassport^to enter a particular social group. In this case, it symbolically reduces the distance between the class that is discriminated (the poor) and the class that commands respect (elite). It is not surprising to see that one of the most notorious drug leaders was wearing a green shirt with the distinctive logo of the Polo brand when he was arrested (see Fig. 5). According to Veblen [33], there are many ways to show that one is wealthy, but clothes and accessories, in contrast to other material items, can be shown off on any occasion and it is difficult for them not to be noticed by observers.
Along with the emphasis on outward appearances, spending the proceeds of illicit crime on drugs was also not uncommon, a finding that has been documented widely in other work with young gangs in Western contexts, including Bourgouis' work with crack dealers in Harlem [34], Densely's work with youth gang members in London, the 'street' youth of Philadelphia that feature in Anderson's work, as well as the armed robbers who star in Decker and Wright's ethnography of armed robbers. In this study, several interview participants in Juarez mentioned using drugs frequently and a couple of others admitted having a strong addiction to several substances. Family members who used drugs or were themselves involved in the drug trade was also a common phenomenon. In an interview with Milán, he described his father had a cocaine addiction and used on a daily basis, and eventually he and his brother also picked up the habit. When he was involved in crime, he used some of the proceeds of crime to buy alcohol and cocaine. Jorge for example described that his family was involved with drug trafficking. He started to use heroin, cocaine, and Ball types of drugs^. Initially, he described that consumption was sporadic but eventually became an addiction. He was doing drugs on a daily basis. Manuel, described that although his family -he lived with his grandparents, was not in a difficult situation economically, he felt the need to work to sustain his addictions. BI worked only for my drugs and to be clothed, 'andar bien' […]^BI started to consume marihuana and alcohol, tobacco…and well, whatever my family gave me, it wasn't enough […]^Although the levels of drug use in Mexico and the US are drastically different -with prevalence rates estimated at only 2.3% nationally in 2001 [35], there is evidence that consumption among youth and especially young people who live close the US border, have a higher prevalence of drug use ( [36]).
The motivation behind using the income derived from illicit activities to foster conspicuous consumption and to live a fast life is succinctly summarized by what a participant of organized crime in Mexico expressed to a journalist in an interview in 2010: 'prefiero vivir dos años como rey que toda la vida como un wey' [I would rather live 2 years like a royal than my whole life like a fool]. As Wright & Decker point out in the criminal universe, Bthere is no reputational mileage to be gained through deferred gratification^( [37]: p.37). That organized crime does not constitute a pathway out of poverty or an overall improvement in other dimensions of deprivation is however an important message to communicate: in personal communication with a specialist on education in Mexico, there is evidence from rural schools in states where organized crime is rampant that young students openly aspire to be narco-traffickers when they grow up (

Deprivation in non-income indicators
Exploring other indicators beyond income, the data indicates that there is a significant disparity in educational lagging when comparing both groups: almost half of the delinquent sample (47.2%) were lagging in education, meaning that they had not completed secondary school and were not enrolled in any educational programme prior to incarceration. This figure is more than three times the proportion for the non-delinquent sample, where educational lagging was present in only 15.6% of respondents. Furthermore, five out of eight delinquent respondents in the age group 12 to 15 were not enrolled in school prior to incarceration and for those aged between 16 and 29, almost half (46.8%) indicated that they had not completed secondary school. Although the most recent report on school abandonment cites income constraints [40,41] as the main motivation for abandoning school, as was previously pointed out, income constraints were higher amongst non-delinquents, meaning that economic constraints cannot explain the disparity in educational lagging across both groups.
The literature on school abandonment in Western contexts, largely the United States, indicates that the reasons why students drop out of school may have a lot to say about their engagement in crime. While dropping out of school to assume adult roles (employment, marriage and/or parenthood) predicts a lower likelihood of participation in crime, poor educational performance has been associated to school abandonment and delinquency [42,43], meaning that those individuals who drop out of school due to poor performance in school are more at risk of participating in delinquent acts.
The present study first found that almost half of offenders sampled already had parental responsibilities, and some had started families since the age of 14, which contrasts with the finding that assuming parental roles decreases the likelihood of criminal engagement. Rather, the pressures to provide in a culture where men are still expected to be the main, if not the sole breadwinners, can act as a catalyst for criminal participation. In addition, it was found that the respondents who were serving time in prison had performed well in school (only 24% of respondents expressed that they Bhighly agreed^and Bagreed^with the statement Bmy grades in school were usually low^), and hence low educational achievement was not the reason that they were dropping out, nor could poor performance in school explain their engagement in delinquency. A recent study of school abandonment at the secondary level in Mexico found that the second most important reason for dropping out amongst students surveyed in the country was a strong dislike for school ([20]).
In additional analyses, I found that those who agreed more with the statement Bthe things I was taught in school were not very useful^had a higher probability not only of abandoning school, but also of participating in delinquent activity [44], providing evidence that a dislike for school and not finding the material taught useful may be more decisive factors in explaining dropping out and engagement in delinquency than academic achievement. This was further corroborated in the interviews, where over half of delinquents pointed out that they dropped out of school due to an inclination for 'andar en la vagancia', or a preference for hanging out with their peers. In an interview with Milán, he recalls his experience at school in the following way (CC denotes author, P is for interview participant): In another interview, Ricardo is asked what he disliked about school: BIt's just that it was a long time. It had been 6 years (he dropped out after sixth grade), one starts to grow up and likes the street more than school.^[On the street] Bno one is there watching out for what you are doing or anything, you make your own law [...] B. Manuel also confirmed that he quit school because Bhe didn't like it anymore^and because he preferred to be out on the street, Bandar de vago^, an often cited phrase by interview participants. Jorge mentioned that he dropped out of school due to a drug habit and Juan José mentioned that he Bdidn't study because he didn't want to, because he never liked school^, recalling that he spent the money his parents gave him for tuition in night clubs and alcohol. Other respondents mentioned having discipline problems in school and being expelled. For example, Pedro mentioned that he performed well in school and had Bgood grades^, but he liked trouble and had disciplinary issues. He was expelled from three different schools. As Bourgouis explains, the formal institutional vacuum that is created by truancy [and some cases, being expelled from the school system] is filled by a 'proto-criminal youth crew, a gang' in the context of Harlem, by 'street' people in the context Philadelphia [29], and by 'vagos' in the case of Juarez. What all groups share is their immersion in urban poverty and a generalized attitude of dislike for school.
It was surprising that given the regional differences in poverty between the north and the south of Mexico -with the north (where the state of Chihuaha and Ciudad Juarez are located) being significantly more prosperous, close to one half of respondents in the delinquent sample of this study had experienced deprivation in access to food: more than 40% of respondents had been in a condition of moderate or severe food insecurity prior to their incarceration; for non-delinquents the share of deprivation in access to food exceeded 50%. It is interesting to note that the protagonists of the Western literature relied on social welfare to access basic goods. Not one qualitative study reviewed made mention on critical deprivations in access to food among participants. In Mexico, instead, social programs promoted by the government, such as the program BSin Hambre,^have been ineffective in fulfilling their purpose -to eliminate hunger amongst people in extreme multidimensional poverty through an adequate nutrition. Although data for the state of Chihuahua, where Ciudad Juarez is located, indicates that 127 communitarian dining halls were set up under the program [45], there is no data to effectively evaluate the impact of the program in reducing the number of individuals in a situation of moderate or severe food insecurity.
In contrast to discouraging indicators in terms of income and education, delinquent and non-delinquent sample offenders were found to have lower shares of deprivation in access to social security (38.5 and 20.4% are deprived, respectively), when compared to deprivation at the national level (which was almost 60% in 2014). One of the plausible explanations for high enrolment of both delinquents and non-delinquents in the Mexican Institute of Social Security (IMSS) -which employers are not forced to provideis that a high percentage of the inhabitants of Ciudad Juárez work in maquilas, or assembly plants. According to Ávila, the industry employed 272,702 workers in 2015, in a city of roughly 1.3 million inhabitants [46].
Additional questions on employment before participation in illicit activities found that one of the most common jobs carried out by offenders was precisely in the maquila sector. Around 32% of the delinquent sample signalled that they had worked at some point in the assembly plants, most of them in the role of operadores. Access to social security provided by employment in the maquila is, however, insufficient: the weekly salary for the lowest level of employment in the assembly plant -as an operador -is estimated at a meagre 500 to 800 MXN (19.6 GBP to 31.5 GBP) [44]. According to the national poverty line, an individual is considered poor if his income in an urban area is below the cost of a basket of food items and a basket of non-food items, roughly equivalent to a monthly amount of $2542.13 MXN (116 GBP) per capita. A weekly salary of under 600 MXN, the median salary of the maquila employee fulfilling the role of operador, falls below the poverty line. In contrast, the minimum wage in the US is currently 7 USD per hour. Despite these stark differences, some of the protagonists of Wright & Decker's work agreed that they did not see themselves holding down a job, nor did they want to: some of the armed robbers resented receiving orders from a boss as well as the low wages accompanying the low skilled jobs they could perform. In Bourgouis' work, Harlem crack dealers had held jobs in office environments where they were subjected to constant discrimination, eventually giving up those jobs entirely. While discrimination and low wages constitute important contributing factors to work outside the legal economy, it is clear that the constraints in the Mexican context are and will continue to be far more severe. Despite that, all delinquent respondents had occupied at least one form of employment prior to incarceration or engaged in illicit activity while holding down legal jobs; not one interview participant mentioned not wanting to work. Furthermore, it is possible that Mexicans who under these economic constraints migrate to the US illegally sometimes fill those very jobs that the protagonists of these authors' work are reticent to take up.
The findings on Mexico reveal that indeed offenders in this study were poor along multiple dimensions, especially so in the educational dimension. The findings from the study also point to important contrasts with some findings established in the Western literature. First, leaving school to assume adult roles (becoming parents) does not decrease the likelihood of criminal engagement; rather what is seen amongst offenders is a high incidence of adolescent pregnancy that results in young men assuming adult roles since an early age. In this context, starting a family can constitute a catalyst for criminal participation, partly due to a culture that still places a heavy burden on men who are traditionally considered to be the main breadwinners. Secondly, the conspicuous consumption driven form the extra gains of the illicit economy are not used in the purchase of a 'street' look, as illustrated in Anderson, Densley and Wright & Decker's work. Instead, in a context of glaring inequality such as Mexico (not to say that the US does not suffer from high inequality, it is just a question of degree and Gini indices clearly show that levels of inequality are higher in Latin America), young marginalized men seek to imitate consumption patterns of the wealthy with the hope of reducing the distance between themselves and the elite, or ultimately, to gain access to the social group of elites that commands respect. The way that one earns respect across these contexts is therefore inherently different: while the protagonists of the Western literature earn respect developing their own sense of style (the 'street' look), in Ciudad Juarez this respect is instead drawn from mimicking the look of the elite. Lastly, access to social welfare and higher wages in low-skilled employment was more readily available in Western contexts, but there was reticence on behalf of some of the study participants in this literature to engage in the legal economy. In Mexico, respondents in this study had no access to a regular welfare program that provided food stamps, resulting in a high proportion of food deprivation that Western counterparts did not largely face. Participants in the Mexican context also faced significantly more obstacles in legal employment due to low wages that were insufficient for meeting basic needs.

Limitations
In their ethnography of armed robbers, Wright and Decker criticize researchers who have refused to work with offenders in their 'natural habitat' (outside the prison), pointing out that there are biases to data collected in correctional settings. First, it is possible that study participants may recall their life prior to incarceration in a more positive light and secondly, respondents who had spent more time in jail may have had more difficulty in recalling their circumstances prior to incarceration. The evidence collected for this study did not indicate, however, that when asked about deprivations, respondents recalled their experiences more positively, nor were there high rates of non-response in the evaluation of poverty indicators that would have pointed to difficulty in respondents' ability to recall prior conditions.
In addition, there are several shortcomings with Wright and Decker's criticism concerning researchers' reticence to work with offenders outside institutional settings. In the first place, the authors produced an ethnography focusing exclusively on armed robbers. In contrast, the range of crimes of participants in this study largely included homicide; hence, my reticence to work with this population in their natural environment. They also fail to consider gender in their criticism overlooking that the risks for women who decide to do research outside of an institutional setting, in offenders' 'natural environment', are different. Furthermore, as I have argued before, working in prison in a developing country cannot be compared to working in this same environment in a developed country. In instances of autogobierno, or in prisons where officials are mired in deeply corrupt practices, as is the case in some prisons in Mexico, the lines between offenders and officials become blurry and prisons come a step closer to offenders' natural habitats.

Conclusions
The objective of this paper was to test Wacquant's argument on the penalization of poverty in the specific case of Ciudad Juarez, Mexico, one of the cities most affected by the recent 'war on drugs' declared by Felipe Calderon. The present analysis of multidimensional poverty used in this study, based on the methodology that has been used for the measurement of poverty in Mexico at the national, state and municipal levels since 2010 [18] but that has traditionally excluded institutionalized populations, confirms that offenders are deprived on several dimensions. Because our understanding of poverty has exclusively focused on income measurement, there is a tendency to assume that experiences of poverty of people in prison are homogenous, when this is not necessarily the case. The analysis showed that while offenders fared worse off in several dimensions of poverty when compared to non-delinquents, there were also individuals who belong to other categories of poverty and interviews revealed that experiences of poverty varied.
In addition, while the most common explanation for participation in organized crime is economic in nature, the study reveals that while delinquents were less deprived in the income dimension, participation in activities of this nature did not constitute an effective nor sustained pathway out of poverty, as most respondents were still considered income poor and an important share were still vulnerable due to deprivation in social indicators. The failure to escape poverty is due to a highly skewed distribution of income in the illegal economy depending on the position in the organized criminal hierarchy and the use of gains from illegal activity to fuel conspicuous consumption, findings that are similar in the established Western literature. However, there are important differences in the consumer goods purchased when compared across contexts: while in Juarez, imitation of consumption patterns of the elite to earn respect is visible, qualitative work in Western settings reveals that respect is earned through the creation of a 'street' style that stands in direct contrast to the middle and high classes.
This analysis also found that educational lagging is much more pervasive amongst delinquents when compared to non-delinquents. Amongst delinquent sample respondents, a dislike for school constitutes an important motivating factor for dropping out, despite evidence that respondents performed well academically. This stands in contrast to the established literature in the US which finds that those who drop out of school due to poor academic achievement are more likely to participate in delinquent acts. On a more positive note, the study finds evidence of lower shares of deprivation in access to social security amongst both delinquent and non-delinquent respondents in comparison to deprivation at the national level, which can be attributed to the pervasive character of assembly plant employment which permeates the economic structure of Ciudad Juarez, and where access to social security is guaranteed by employers despite meager wages. However, in contrast to the Western literature, where protagonists have access to welfare in the form of food stamps and higher remunerated low-skilled employment, this study finds high degrees of deprivation in access to food amongst participants and wages in the legal economy that do not meet basic living costs. These initial findings show that multidimensional poverty is a useful framework to link penalization and poverty and to gain a more nuanced understanding of offenders' experienced deprivations. In so doing, they point to the relevance of building an evidence base around work in prisons in developing country settings and particularly in Mexico, as well as to question established assumptions that have been based almost exclusively on evidence derived from Western settings.