Double Expectations: Law Enforcement Workers and Dilemmas on Handling Drug Use at the Street Level

: This paper analyses law enforcement workers’ experiences when transforming policies for crack cocaine and heroin into practice. It focuses on dilemmas workers have and choices they make when approaching drug users. Grounded theory principles and ethnographic techniques were used to gather and analyse in depth interviews and extensive participant observations of 20 workers across the cities of Amsterdam (Netherlands) and Porto Alegre (Brazil). The comparative analysis enables to explore the impact of different socio-economic and political environments on workers’ practices around drug use. Three areas in which law enforcement workers reported to have dilemmas concern workers’ choices on how to deal with violence; choices between being tough or friendly towards users; and on defining their role on curbing public nuisance. If a first sight shows differences between the cities, in a closer look interesting similarities appear. When deciding upon approaching drug users, law enforcement workers drift between order and care approaches. At the street level, different contexts produce ambiguities and workers’ experiences question the etic reductionism in simple dichotomies of care vs. order approaches. From this closer perspective, similarities appear more clearly, as well as counterpoising perspectives to the stereotypical views of Dutch and Brazilian workers.

In the field of drug policies, an ongoing debate concerns which direction policies for illicit drug use should take. Historically, many governments have supported a 'law enforcement' or 'public order' approach towards drug use, which involves enforcement of prohibitionist laws to achieve its goal of completely eradicating drugs from society (Marlatt 1998). Drug use is seen as a safety issue and treated with punishment and repression. The last 30 years have seen a shift in drug policies of many countries, towards the inclusion of a 'public health' (or a 'harm reduction') approach. This approach considers drug use a public health issue, rather than a safety one, and is concerned with the health of people who use drugs. It focuses on reducing the harms caused by drug use and trade rather than expecting to completely ban them (Inciardi, Harrison 2000). Public order and public health approaches have often been combined in countries' policies. Combining their different aims and priorities, however, have led to contradictions in the establishment of national/local policy statements and the way policy happens in practice.
One main problem in integrating these approaches is how to be repressive against illicit trade and use of drugs and worried about the health of drug users at the same time. For some (e.g. Hunter, McSweeney et al. 2005), contradictions are related to the debate about *Address correspondence to this author at the Utrecht University, The Netherlands; Tel: +31302531179; E-mail: R.dequadrosrigoni@uu.nl who should have the leading role regarding drug use: the law enforcement or the health sector. In this context, law enforcement workers' activities are mostly portrayed as focused solely on repression, and as being harmful to public health activities carried out by care workers and services. Examples include crackdowns and intensive policing in areas where health services are offered to drug users, which compromise the functioning of these programs (Davis, Burris et al. 2005, Small, Kerr et al. 2006, induce drug use in riskier circumstances, and hinders users' access to care by driving them underground (Aitken, Moore et al. 2002, Cooper, Moore et al. 2005. Other examples include the lack of compliance of street level police workers towards new harm reduction laws that allow syringe possession and purchase: workers tend to seize syringes given by health programs and use its possession as legal ground for searching and arresting injection drug users (Beletsky, Macalino et al. 2005, Small, Kerr et al. 2006. This, again, harms public health activities with this population. A limitation of these studies, however, is to focus on care workers' interpretations of law enforcement workers' activities only. Little is known about law enforcement workers' own point of view, and the possible reasons leading them to decide for strategies focused on order rather than care. Besides, although the above-mentioned activities, indeed, contrast with a public health oriented approach, this may not represent the full picture of dilemmas and choices law enforcement workers face daily. Some studies found, for instance, that lack of material resources and services for drug users can make collaboration between police workers and health services difficult (Connolly 2006, Vermeulen, Walburg 1998. Others, focusing on police workers, found that these workers have to deal with a 'double' expectation of being repressive towards drug use but also of collaborating with harm reduction programs (Bull 2005, Lister, Wincup et al. 2007, Beyer, Crofts et al. 2002. Police workers do not represent a homogeneous way of dealing with drug use; workers may have more dilemmas regarding how repressive or caring one should be than what is usually imagined. To understand law enforcement workers' choices, a more complex account of the several nuances of law enforcement meanings and activities is needed. This is what this study aims at. Based on principles of a grounded theory method (Urquhart 2013), it builds a view from law enforcement front-line workers' level, and derives an understanding of policy in the place it happens daily: the streets. The ways in which policy happens in practice are usually different from the ways in which it was stated in official documents. From official guidelines to the ground, rules, goals and regulations of any written policy must undergo a set of steps, enter a world of institutions and resources, different contexts and a set of workers who translate rules into practice. Lipsky (2010), founder of street level bureaucracy approach, contends that front-line workers must find ways to cope with the gaps between rules and expectations and the reality they find in the streets. In doing this, they end up transforming the ways in which policy happens. This is what street level bureaucracy scholars call discretion: a freedom in exercising one' work role (Evans 2010).
The present study analyses law enforcement workers' use of discretion when transforming drug policy into practice, and choosing for different approaches towards drug users. It focuses on the socalled problem drugs -crack cocaine and heroin -and compares two cities in very different settings -Amsterdam, in the Netherlands, and Porto Alegre, in Brazil. The first, in a developing country, with a history of a military-dictatorship and strict policies towards drugs, but with growing tendency towards a more open drug policy. The second, in a developed country with a historically liberal, but recent and growing tendency towards a stricter drug policy. The comparative analysis of cities with such a diverse socio-economic and political context enables to explore the impact of different environments on law enforcement workers' discretion.

METHODOLOGY
This paper draws its data from a wider study (Rigoni 2015b), which analysed dilemmas and strategies not only for law enforcement workers but also for social and health workers from Amsterdam and Porto Alegre when putting drug policies into practice. The research benefitted from a qualitative and ethnographic approach based on grounded theory principles (Urquhart 2013). Amsterdam and Porto Alegre are treated as two case studies (Yin 2000). Within each case, two districts were chosen to cluster the observations: the city centre and a neighbourhood known for being one of the most problematic for drug related issues-namely the Bijlmer in Amsterdam, and the North Zone in Porto Alegre.
Fieldwork was conducted over 14 months: in Amsterdam from February to July 2010 and in Porto Alegre from August 2010 until March 2011. The researcher was in the respective cities during these fieldwork periods and, moreover, lived in Porto Alegre from 1994 until 2008 and in Amsterdam since 2011. For the main study, 80 workers were in-depth interviewed and 800 hours of their activities were done. These were balanced between the three studied sectors and the two cities and districts. The present paper focuses on the data of law enforcement workers only. These comprised of 20 participants, being 14 street level workers and 6 key-informants. Street level workers -focus of the study-had as a participation criteria having a work contract in the service and contact with drug users as an important part of their daily work. Key-informants were also interviewed due to their vast experience in the drug policy field, mostly related to street or middle management level. Different semi-structured questionnaires guided street level workers and key-informants' interviews, but both lasted around 1 hour, and were voice recorded, transcribed and then analysed with Atlas.ti software 1 . Interviews were done in English in Amsterdam and in Brazilian Portuguese in Porto Alegre. Interviewees signed an informed consent assuring their secrecy, anonymity and right to withdraw the study. Observations of street level workers' activities included, in general, a minimum of 2 shifts (around 8 hours) in each service. They included the district context, services available, work conditions, workers' activities, and users' presence and treatment. Like interviews, observations were typed into field notes and analysed with Atlas.ti.
Following grounded theory principles (Urquhart 2013), sampling, data gathering and analysis were understood as part of the same process: they occurred sequentially, with analysis guiding places and people to sample for data collection and in turn being modified by the data. Given the variation of services and experiences of workers in each city, a sample was drawn to choose those participating in this study based on the different types of experiences emerging as categories in theoretical sampling (Bryant, Charmaz 2010). From the six informants interviewed, three of them occupied medium management positions in the  Temporary contract-1 (tutelary counsellor) police, one a local coordination position, and one a representative position in a Drug Council. The profile of street level law enforcement workers participating in the study is described on Table 1.
In the Netherlands, law enforcement workers getting more frequently in touch with drug users at the street level are the community police officers (buurtregisseurs), the patrolling police workers, and the probation officers. While community police are responsible for contacting community and community services along with curbing crime, patrolling police focuses on curbing crime only. Probation officers follow drug users who have committed crimes into reintegration programs. In Brazil, law enforcement workers getting more frequently in touch with drug users at the street level are the military police, civil police, and tutelary counsellors. While military police curbs crimes on the streets, civil police workers are responsible for crime investigation (drug dealing, drug possession, and drug use in doubtful cases where dealing might be involved). Since the early 90's, a few military police workers participate in the PROERD (Programa de Resistência às Drogas), a drug prevention program based on the American D.A.R.E. (Drug Abuse Resistance Education). In this program, trained police workers offer lectures to children from public primary schools based on the idea of abstinence as both prevention and treatment. Tutelary counsellors get involved in cases where drug use affects children, youth, and/or their parents 2 .
Street level workers' sample was balanced for sex, age, and experience working with drug users, since the literature contends that these factors impact workers' beliefs around drug use and their activities with users (Forman, Bovasso et al. 2001, Humphreys, Noke et al. 1996. Emerging categories for theoretical sampling were also added to form the sample: type of service; type of work contract; main approach adopted regarding drug use; being considered liberal or strict by colleagues and drug users; type of relationship with colleagues and users. Amsterdam and Porto Alegre law enforcement workers presented similarities in terms of their levels of formal education, number of jobs (just one), and type of contract they had with their services (mostly stable). Main differences were related to a higher monthly income for Amsterdam workers, and a higher average of working days and working 2 Created with a social function of protecting children's right, Tutelary Councils ended up having an image and use related to their law enforcement power. hours a week for Porto Alegre law enforcement workers.

CONTEXT
When thinking about tensions at the street level regarding law enforcement of drug use, both Brazil and The Netherlands are interesting cases to consider. The cities are similar cases in terms of being medium-sized cities with a cosmopolitan culture, having a history of hard drug use and dealing in public places, with a drug problem concentrated in certain neighbourhoods and among people at the margin of society. Both also, share process of evolution in drug policies towards harm reduction. These cities, however, contrast in their histories and types of hard drug use, the resources available to street level workers in the two cities, and the histories of their national politico-bureaucratic regimes.
Worldwide and in the European Union (EU), The Netherlands has been seen as a leading country in adopting and defending a public health approach. It is in the drug policies of the Netherlands that harm reduction approach has its roots (Inciardi, Harrison 2000), with the notion of drug use as a social-health problem rather than a crime (VWS 2003). In South America, Brazil is a leading country claiming to adopt public health strategies (Bueno 2007), being considered an example among developing countries on harm reduction policies implementation (Mesquita 2006). Even though both Brazil and the Netherlands are considered pro-harm reduction, these countries' different background and context provide an interesting comparison of street level workers' responses in practice.
In the Netherlands, a market division between hard and soft drugs operates in the country since the Opium Act in 1976: trade of hard drugs is penalized, but sale of soft drugs is tolerated under certain rules (van der Gouwe, Ehrlich et al. 2009). Although drugs are illicit, drug use is not. The advent of an HIV/Aids epidemic among injection drug users, in the mid 80's, played an important role in developing a harm reduction approach (Hedrich, Pirona et al. 2008, Korf, Riper et al. 1999. Nowadays, syringe exchange programs, opiate substitution therapy, heroin prescription, consumption rooms and shelters were drug use is allowed, are some of the harm reduction measures carried out by the care sector and supported by law enforcement. For Dutch police, priority is put on curbing drug trade rather than punishing users. Large-scale dealing and production of drugs is prosecuted, while possession of drugs for personal use (in general 5 grams of soft drugs and 0.5 grams for hard drugs) has low priority (van der Gouwe, Ehrlich et al. 2009). Both on official policy statements and on the ground, the Netherlands is considered to have achieved a good balance between 'tolerance' and 'repression'. An integrated approach is taken between different ministries involved in drug policy 3 ; collaboration between care and law enforcement professionals is seen as an absolute necessity, and has been happening for some decades (de Kort, Cramer 1999).
In Brazil, a harm reduction approach started about a decade later, also stimulated by an HIV/Aids epidemic among injection drug users in the 1990's. Different from the Netherlands, however, harm reduction in Brazil encountered very repressive policies towards drug use. These were inherited from the Military Dictatorship period  in the country, when Brazil imported the 'war on drugs' from the US (Carvalho 2006). Brazilian 'antidrug law', from 1976, penalized dealers and users, ordering coercive treatment for the later (Brazil 1976). In 2003, Brazilian Health Ministry reformed prevention and treatment policy regarding alcohol and other drugs, officially stating national political support for harm reduction strategies for the first time (Brazil 2003). In 2006, a drug policy reform established that drug use is not a reason for arrest, even though one can be penalized with optional treatment, counselling and/or communitarian work (Brazil 2006). However, since the law does not establish clear quantities allowed for possession, it is at police discretion to define a given case as drug use or trade. The law forbids both sale and trade of any illicit drugs. Different from Amsterdam, public health and law enforcement professionals have a clear preference for not working together, and perceive their roles as hindering each other's (Rigoni 2015a). How would these different contexts interact with dilemmas and choices law enforcement workers have around approaches towards drug use?

DILEMMAS AND CHOICES WHEN FACING DRUG USERS
Study participants reported a variety of dilemmas when translating official drug policies into practice. These were clustered into three areas, described in the 3 The Ministry of Health, Welfare and Sports (who coordinates drug policies), the Ministry of Justice, and the Ministry of the Interior and Kingdom Relations.
following sub-sections: deciding on how to deal with violence; choosing between being tough or friendly towards users; and defining their role as law enforcement workers on curbing public nuisance 4 .

How to Deal with and when to Apply Violence
Deciding on how to deal with (potentially) violent situations is part of law enforcement workers' job at the street level. Given the illegality of drugs such as crack and heroin, its use scenes are strongly connected to a social image of violence, aggressiveness, and danger. The actual level of violence involved in these drug use scenes, however, is very much dependent on contextual factors.
Brazil occupies the fourth place among the most violent when compared to other Latin American countries (Waiselfisz 2012). Sadly, drug trafficking and police workers' conflicts are thought to account for a large part of the violence, mainly in the big cities (Rodrigues 2006). In the city of Porto Alegre, the registered homicide rate for 2015 was 39,5 for each 100 thousand inhabitants, which places the city among the most violent in the world (IGARAPE 2017). The Netherlands, on the other hand, has one of the lowest levels of criminal offences for drug use in the EU (EMCDDA 2017). Property crime related to drug use as well as public nuisance is said to have declined in recent years, despite the claimed increases in organized crime (van Donk, Boekhoud et al. 2009). In the year of 2010 Amsterdam police workers registered 18 homicides (Kranenburg, Vugts 2012), which represents an approximate rate of 2 per 100 thousand inhabitants. Not surprisingly, law enforcement workers' dilemmas and choices regarding violence presented differences across the cities.
In Porto Alegre, physical violence, armed conflicts, life threats, and community aggression towards police, were part of the work experience of many law enforcement workers. During informal conversations, but not during voice-recorded interviews, some participants explicitly mentioned to make use of verbal aggression, lower leg kicks, arm torsions, and slaps in the face as approach techniques with drug users regarded as uncooperative or violent. 4 The main study found yet another dilemma related to differentiating drug users from drug dealers. This, however, involved mostly law enforcement workers from Porto Alegre and therefore, is not analysed in this paper. In Amsterdam, small scale drug dealing is not a priority for the police. Distinguishing users from dealers, however, may become a problem inside care facilities where drug use is allowed, becoming a dilemma for social and health workers. For more information, please consult Rigoni (2015b). BR03 [police worker] tells me that, when approaching a user/suspect, the technique is to act fast: say you are a cop, tell them to put the hands on the wall, move the legs away from each other, and do the body search [to check for drugs]. It can be that they resist and then police has to use the force. Active resistance, he explains, is to flee or attack the police, while passive is to take his hand off the handcuff, not open his legs as ordered by the police, or not to put his hands on the head when ordered. The police worker tells me that if the guy takes his hands from the handcuff, for instance, and he does not react violently, soon after that guy will try to take his gun. By active resistance, more strength is necessary. In case of fleeing, for instance, shooting one leg can be considered. He affirms: "One has to show who's the boss, who's the authority". (Porto Alegre, fieldwork notes, 27 October 2010).
Police workers in Brazil are constantly portrayed as violating human rights. Brazilian police is recognized by its selective and violent approach, inherited by the days in which a military dictatorship ruled the country (Coimbra 2001). Far from justifying or diminishing the gravity of these acts, a closer analysis of their work context brings some clarity into law enforcement workers' reasoning for choosing violence. In this study, workers mentioned to have learned violence techniques from more experienced colleagues in the corporation. The support for this type of approach among chiefs and colleagues make the choice for (violent) punishment more appealing. Performing violence could be a way of being recognized and getting rewards inside the corporation.
The criminal context was also seen by workers from Porto Alegre as "asking for" violence. Especially the police working in the streets felt the constant need to protect themselves against a hostile environment.
BR10: There is no gun for everybody. If everyone takes a gun when they are out of service, it will be missing for others. So, if you're out of service, that's where you use your private weapon. You buy your private gun with your own money.
Researcher: And why do you need a private weapon? BR10: Yeah, it's your choice. But you must buy, I advise you to buy. Every policeman, when you first enter the corporation, the moment you are taking the course, you already put the money together so when you graduate you buy a gun. Several times you get out of your job ... you go and arrest the guy, okay? You take him to the judiciary, they don't judge it as flagrant [small-scale dealing case]. They release him. Then you come here ... you're doing your shift. Then you leave the service you are walking in the street, you go catch the bus and you see that guy. But now you're unprotected, you're unarmed, and he's going to kill you. Now if you have a weapon too, you can defend yourself. (Porto Alegre, military police worker).
Fear, perceived lack of support from their organizations, and lack of trust in the judicial system made many law enforcement workers from Porto Alegre to find personal solutions for their safety. Many times, the strategy chosen for self-protection could be one of escalating (perceived) violence, in an attempt to show power and generate fear on the "enemy" side. In these workers' narrative, violence becomes the only approach perceived as feasible to deal with a lifethreatening context.
A contrasting narrative comes from community police workers from Amsterdam, who reported to feel very supported by their organizations and use violence de-escalation methods whenever possible. In their stories, beatings and guns where virtually absent. Not only violence was perceived as not necessary, but as harming their professional activity. Dutch law enforcement has been considered to have a social and non-violent approach, having developed a community policing style for the last 30 years (Punch, Kees van et al. 2002). Law enforcement workers in this city declared to know most of the drug users in their policing areas (many times by name). Especially community police workers frequently visit health care centres in which drug users are assisted. A closer relationship with community (and users) brought forward the idea of mutual respect as central to good and effective policing. Different from Porto Alegre, in Amsterdam police has to be trusted (not feared) to be respected.
When certain boundaries are disrespected, however, and violence is more visible, fear may also appear as a feeling leading Dutch workers' experiences. An example are cases when police is confronted with a person who is under drug effect and having a violent crisis episode. In Amsterdam, police specially mentioned the case of some drug tourists, who were regarded as turning violent due to their inexperience with drugs. NL06: We have big problems with the guys that are using everything. They are coming [to Amsterdam] and they are drinking too much, they are blowing too much, they are eating mushrooms… and they are using heroin, or cocaine, or LSD, or ecstasy. Some guys are making a real mess, taking a very dangerous cocktail. [..] And the mushrooms they eat it, but you must wait 45 minutes before it will start to work. And they are thinking: "hey, I ate it, but it didn't work", so they drink, and eat a second box, and drink, and eat… and then it will work very, very heavy. Those are the really crazy people, and they will fight with everything, with glass, with blood. […] It is very dangerous, because they don't know what they are doing. They can put a knife in your neck, and they are thinking they are having a bath or something. It's really scary! They have such eyes… you can't talk to them. (Amsterdam, community police worker). When studying discretion among street level police workers, Maynard-Moody and Musheno (2009) affirm that a significant part of workers' discretion involves assessing whether the person being confronted poses a threat to the worker's safety. Feelings of fear and danger would tend to lead workers to dehumanize that person, classifying them into stigmatized identity categories. For these, deemed as " unworthy" citizens, punishments are more likely to be applied than care. As the community police worker says, a conversation is not possible and arresting could be an option. Amsterdam law enforcement workers, however, can also count with other options.
In the program Vangnet & Advies (Safety net and Advice), police can call a social psychiatric nurse to help handling cases of drug users who are being aggressive, making nuisance in the streets while intoxicated, or having problems with neighbours. The service, developed in the last decades, works 24/7. Vangnet & Advies is one of the programs between law enforcement and care which were built specially for the population with mental illnesses. Even if not specifically directed to drug users only, many of these programs attendees are users.
Porto Alegre police, unfortunately, does not count with the same options. In this city, a weak partnership with health care along with less resources makes police feel demanded to "cover holes" from other state services, instead. At the same time they felt as having to perform activities which were not police-related, they felt judged by the community for not handling these situations in an adequate way. BR10: ... when they use drugs and they start breaking everything inside the house and then we go there, we talk with the family and we call [the ambulance]. But sometimes they don't come, and then we must do everything to conduct the user. But sometimes is not easy because they are aggressive [...] and then the police must be very careful, because he is not… a thief, you have to be careful with him, you know? And maybe is his mother or father who is there, and they can turn to be against the police. I went already to a house where the boy had an axe in his hands, breaking everything inside the house, you understand? And then?
Researcher: and then? BR10: and then you must use the force! You understand? But then you do that and you drop the person [on the floor] and then the parents are against the police! So, you have to be careful to contain the guy, he must be on handcuffs and sometimes it hurts him… (Porto Alegre, law enforcement worker).
As the police points out, it might be expected that a health worker will have the techniques to handle with care a person during a violent crisis under effects of drugs; but the same cannot be assumed from a police worker.
A different socio-cultural context, different culture and trainings inside the police force, as well as a differentiated level of availability of services to support law enforcement work, create different dilemmas for workers in the streets, and may make it easier for them to choose for different postures regarding violence. What is considered violence, and which law enforcement responses are pondered as acceptable towards (perceived) violent behaviour also vary according to these factors.

To be tough or to be Friendly Towards Users
From the experiences and choices regarding violence in Amsterdam and Porto Alegre, one could expect Amsterdam law enforcement workers to be mostly friendly towards users, while Porto Alegre workers mostly strict. Even though this stereotypical view may portrait well the situation in a first sight, a further analysis proves that workers' experiences cannot be generalized or differentiated in such a black and white manner. In Porto Alegre, military police workers involved in drug prevention in schools (PROERD), tutelary counsellors, and some civil police workers reported dilemmas regarding strict-only attitudes. These workers perceive a contradiction between their beliefs and practices towards a friendlier behaviour with users, and an organizational culture favouring tough behaviours and punishing the ones considered 'too soft'. BR07: Honestly, I think that in the police, if you show yourself as very sensitive, very understanding with the accused, if you try to act like that, like a human being, this is frowned upon by your colleagues. […] [Then] You shape yourself so people like you, you know? [But] I always get very much into conflict with colleagues who think that kicking butts goes for everything. If I'm going to give a glass of water to the accused [They say]: 'Oh, you'll give a glass of water to the accused? What's that? Tomorrow they'll be back wanting something else ...'. But I cannot be any different here than I am at home, you know? (Porto Alegre, civil police worker) The dictatorship period in Brazil left a military culture as heritage: torture techniques, punishments and rewards used at that time are still perpetuated and valued by the corporation (Coimbra 2001). This contrasts with a community policing culture, prevention programs, and caring relationships with users. Brazilian police organization has been trying to develop community policing and support prevention programs in the last decade. Since 2009, Porto Alegre workers get lectures on community policing in the welcome training, and the yearly recycling courses include a five-hour lecture on this subject. However, according to police officers, only 20 police workers from Porto Alegre were trained in community policing by 2011, and even these were not yet attached to specific communities. Despite the movement towards a new more caring perspective, police workers considered it was far from enough to change current military mentality.
Study participants perceived that the main organizational culture is that efficient police workers must be tough, fearless, and collect many arrests. When a worker escapes this pattern, they become targets of prejudicial jokes by their colleagues. An example of workers suffering from this type of prejudice are the military police workers joining the drug use prevention program PROERD. Despite the program being based on a very repressive approach towards drugs, with abstinence-only guiding both ideas on prevention and treatment, being part of PROERD could be judged as being too soft and fearing "real" police activity.
BR22: But there are comments, right? [They say] That is one person less in the streets to work, to combat crime […] 'Ah, you go there play with children; we are here arresting, while you are there playing'. And no, that's not it, it's a matter of prevention. Often, is that student there that will not cause a problem later on (Porto Alegre, military police worker).
Police from PROERD also felt misunderstood by their military police colleagues, to whom fighting drug use meant performing punishing activities. For them, on the other hand, prevention played a big role in the fight against drugs, and this could be done by having a friendly attitude to achieve trust and credibility, so to be able to influence children's (future) choices.
As questionable as a "just say no" approach towards drugs can be from a social and health perspective, police workers joining the PROERD were more inclined to see drug as a public health issue rather than a criminal one.
PROERD workers also perceived contradictory signals about the program inside the police organization. Since 1998, Porto Alegre's military police officially recognizes the program, which has been transformed into a state policy (Estado do Rio Grande do Sul 2010). Practical organizational support, however, is perceived as still missing: lecturing for the program is a voluntary choice, and workers have to assume the costs of training and trips this might require. Above and beyond, it depends on managerial discretion on priorities to decide if workers can lecture during their working hours, or only in their free time. If police force supported new policing practices verbally and on official policies, street level workers did not feel it provided enough incentives for them to perform these practices on the ground.
In Amsterdam, on the other hand, police gets frequently in touch with social and health workers as part of their prescribed function and within their working hours. Community policing has been developed for over 30 years, and the idea of community trust is important inside the corporation. For most police workers from Amsterdam, being friendly towards users can help them to perform their repressive role in the streets. By acquiring users' respect, wrongdoings can be prevented or solved in an easier way than by using purely repressive techniques. Some workers perceived that, when on duty, less wrongdoings would happen in the streets, since users were ashamed of getting a fine from them. The mutual respect included police help when possible, and in return, more collaboration regarding rules from the users' side. Friendly police workers, however, perceived that not always their approach was admired by their colleagues. NL11: There is respect from both sides. My side towards them, because I treat them like humans, not like junkies. That's why they respect me, I think. But it's…sometimes it's difficult. I have to explain to my colleagues, because they don't understand how can you do things like this. They think it's too soft (Amsterdam, community police worker).
Different visions within Amsterdam police force exist, related both to personal preferences of workers, as well as to the shifts in the development of police strategy along the years. The friendliness of Dutch police has been both praised by its ideas of normalization and equality, and condemned by its perceived lack of effectiveness in curbing crime. Shifting paradigms on responses towards security brought different discourses within police force, where also tougher actions were claimed for (Das, Huberts et al. 2007).
For those workers who were closer to a caring approach, more dilemmas appeared on the boundaries between friendliness and repression. In this city, the presence of community police workers inside care facilities frequented by drug users produced a closer relationship between police and users. This proximity could evolve to a personal concern over users' problems with justice, family situation or health. Law enforcement workers in these situations mentioned to have dilemmas related to the perceived mismatch between "being human" and the type of role expected from them. For me it's more difficult than for them. Because they see a policeman... Okay, today I came with chocolates to give them. One day I give you chocolate, but on the streets, I give you a fine. It's strange for me, but also strange for them. But for me it would be easier to be a policeman, straight, always. So that black is black, and white is white. Because I have black, grey, and white. Grey area, is this area. It's easier to think black and white because you don't think about the guys. You do your work and it's over. Sometimes if there's a case, even if it is not a police case, I try to look at it, because I'm not only a policeman, I'm also a human being. (Amsterdam, community police worker).
Even tough with different nuances, both Amsterdam and Porto Alegre law enforcement workers faced dilemmas on how friendly or tough to be towards drug users they encounter daily. The more involved with users, the more they would perceive them as people in need of help, and the more dilemmas would arise. As Punch, van der Vijver et al. (2002) say when studying policing development in the Netherlands, community policing cannot solve the hard-soft dilemma in police. When black and white ways of thinking and performing their profession become mixed, or grey, it is harder for workers to decide on which approach to take, and to defend it before their colleagues and society at large.

Criminal or Help Seeker? Role on Public Nuisance and Care
Freeing the city from illegal and unwanted activities and behaviors is part of police role. Recently, public nuisance has increasingly become one of such happenings which is thought of as requiring police intervention. Important facts here are what type of behavior comes to be considered a nuisance to others, as well as who are the citizens considered to bother and who are the ones who cannot be troubled. The state activity of deliberately hiding the ones considered unworthy from the ones who (claim to) have right to the city has been called "clearing policies". These refer to clearing the city from unwanted activities and behaviors, but also, from unwanted people. In the field of drug policies, this includes supressing or decreasing the presence of identifiable drug use, users' gathering (and related noise) in open public spaces, and other drug-related 'visual nuisance', such as not being clean or not wearing proper clothes. Clearing policies have been criticized for its disrespect to the rights of the most vulnerable population, besides its inefficiency in terms of solving social problems such as drug use (Wandekoken, Quintanilha et al. 2015, Varanda, Adorno 2004.
The ways in which police workers from both cities interpret their role around public nuisance and clearing policies, however, showed surprising features. Interestingly, while law enforcement workers from Porto Alegre held a criticism towards displacing drug users from the "good citizens" view, their Amsterdam colleagues were very much at ease with it.
NL33: […] I always think, when I'm walking here with my children, they don't have to see drug use; they don't have to see addicted persons. The streets are not from them [users], the streets are from everyone. They have a problem, that's not very nice for them. Not everybody can do something about that problem, but their problem must not become our problem. Let them go into the user room, don't do that on the streets; not everybody has to see what is happening with them. (Amsterdam, patrol police worker) BR34: Drug use is not a police matter, but it ends up in the police. I like to underline this, so people get to know our view on this. Our goal is not to go out hunting, as we pejoratively say, to go out hunting pot heads. We have more to do than that, but eventually it falls into our lap. We are passing by, we see the guy ... It's kind of demoralizing for us to pretend not to see. The community holds us accountable: "look, the police passed by, saw the guys smoking and did not take action". For that person, the resident, they don't want to have people smoking in their square. (Porto Alegre, military police worker).
The key difference between workers' position regarding clearing policies and punishment relates to the context of care services availability and the presence (or lack of) collaboration with social and health care professionals. In both cities, using drugs in public are supposed to be sanctioned with an administrative punishment: those are fines in Amsterdam, and signing a written register in Porto Alegre. In both cities also, clearing policies are into place, with police workers being expected to keep unwanted people away from the rest of the population. However, while in Amsterdam this clearing-punishment was understood as pushing users into care, in Porto Alegre it was seen as a useless, given its lack of efficacy both to solve public nuisance and the (perceived) drug problem. BR10: We guide them, but then you take them from one's front door and put them on someone else's front door. Then the community says "Ah, the police is seeing the homeless users, all laying there on the street, and does nothing." But where am I going to take them, tell me? Will I take them to the police battalion? Take them to my place? Am I going to take them to a shelter if all shelters are full? If you had shelters, more shelters, more houses... But you see, the shelter is open, you can get in and out whenever you want! Then what… I'll cuff him and I'll arrest him inside the shelter...? They are not hurting anyone, they are just there, quiet, lying down, sleeping. (Porto Alegre, military police worker).
Lack of vacancies in shelters or other care services to refer users who are taken off the streets, made the task of clearing the city from drug users to be perceived as just moving people around from more to less visible places. Besides, for many police workers in Porto Alegre, conducting a drug user to a police station to sign a written register was seen as a "useless work". While they would spend at least 1 hour in this process, the streets would be uncovered and someone could be killed or assaulted. Besides, in police workers' experience, the user signs the register and there is no follow up. Most of the times, the user will not be called by the judiciary, since they have more important processes to judge. For these police workers from Porto Alegre, thus, to follow the prescribed policy procedures was perceived as a waste of their time and of public money.
Despite disagreeing with clearing policies, police workers felt pushed to perform it due to community pressure and lack of better alternatives. Sometimes, besides fearing to lose the respect from society, not acting could also be seen by police workers as not fulfilling their professional role. Even with criticisms towards clearing policies and punishment alone, police sometimes felt that by not acting in a strict way, they could be giving the wrong impression that "everything goes". While believing to lack other alternatives, punishment judged to be the best response to these workers.
In Amsterdam, on the other hand, law enforcement workers believed that the best response was to push users to care. Different from Porto Alegre, in Amsterdam low-threshold facilities are widely available. For police workers in the last city, investing in care facilities such as walk in centres, drug consumption rooms, shelters, and opiate substitution treatment, served the purpose of securing society's safety, improving drug users lives, and facilitating law enforcement work.
NL28: … their lives have improved a lot since they are in care. Then they don't want to go out, most of the times. Sometimes they are fighting inside [care services] because one of them has taken a little bit of drugs of another one, or something like that. But, most of the time it's inside and we don't even hear about it. Only when there are consequences we come. But we try to put them into care, because as long as they are in there we don't have anything to do with them. If we didn't cooperate with the helping institutions, we would have much more work to do. It's not like you don't want to do work, but when we don't see any result of it that's not a good feeling. (Amsterdam, community police worker) In Amsterdam, a financial fine can be translated into days in prison in case users do not have money to pay or decide not to spend money on a fine. Alternatives to prison, however, are usually offered in care institutions. For law enforcement workers, users inside shelters, walk in centres and user rooms, and having assured their basic needs, are seen as having no need for committing crimes or making nuisance. Clearing policies are, therefore, both considered to help police to fulfill their role on curbing public nuisance, as well as helping drug users. Although indeed pushing to care do help drug users to better their life quality, a possible danger here is to perform a selective role by targeting the most vulnerable population and suppressing their right to the city. All this, in name of a public health approach towards drugs.

LAW ENFORCEMENT WORKERS AND MEANINGS OF CARE AND ORDER
When looking from a general perspective and focusing on the (stereo)typical behavior of police workers, one finds mostly differences when comparing Amsterdam and Porto Alegre. From this general point of view, one could label Porto Alegre's police workers as violent and focused solely on punishment, and Amsterdam's workers as friendly and focused solely on care for drug users. There are, indeed, clear contextual reasons for these differences in terms of: the history of more authoritarian, less accountable policing in Brazil under military rule which has long duration persistence, notably in the military police brigades; and the great difference in the resources available both for police officers and the services to which police officers can direct people who use drugs. Understanding drug policies, however, needs a deeper consideration of work processes and its nuances. In this study, a qualitative approach inspired in ethnographic and grounded theory principles allowed to capture workers' voices and experiences to show a more complex picture of law enforcement practices in the studied cities.
In general lines, escalating violence seems to be the general choice of Porto Alegre street-level workers for dealing with a perceived life-threatening context. Perceived lack of support from their organizations, as well as community pressure, may lead law enforcement workers to find personal solutions to safeguard themselves and fulfil their professional role of keeping society safe. In Amsterdam, differently, violence is generally a non-acceptable behaviour, and deescalation methods are the preferred choice. When violence becomes closer, however, workers might have the tendency -similar to those in Porto Alegre -of building stigmatizing identities for drug users.
Even with the obvious differences, in a closer look it was possible to see that both in Amsterdam and in Porto Alegre law enforcement workers believe that drug use is not a police matter. There are more important wrongdoings for them to supress. Both also agree that punishment is not the best solution for someone who use drugs, at least, arresting is not going to solve drugs problems. This, however, does not immediately make law enforcement workers from Porto Alegre to follow their Amsterdam colleagues into pushing drug users into care or into adopting a public health approach towards drugs.
In the case of Porto Alegre, some law enforcement workers feel entrapped between a military culture and their beliefs on benefits of a more caring approach towards users. Social and health services are understood as lacking, which leads workers to position themselves away from pushing users into care. In the case of Amsterdam, pushing users to care is understood as a fundamental practice for law enforcement action. Care, however, has a strong linkage with enhancing public order by taking users (and their potential criminal and nuisance behaviors) off the streets. The main concern, in this sense, is not necessarily with the individual user, but with the undesirable effects s/he may cause for society at large.
Through debating workers' dilemmas into three interrelated areas, the paper showed the careconcerned side of Brazilian law enforcement workers and the core role of public order in Dutch workers' approach towards care. Overall, law enforcement workers' activities and beliefs are, most of the times, far from being solely concerned or aimed at repression, as portrayed by most studies. When deciding upon how to deal with encounters with people who use drugs, law enforcement workers drift between care and order. From this closer perspective, similarities between Dutch and Brazilian law enforcement workers appear more clearly, as well as counterpoising perspectives to the stereotypical views of police workers in these places.