Proximity police and its impact on the decrease of crime in Ecuador: An analysis of the period between 2009-2015

Abstract In Ecuador, according to reports, most of robbery modalities constantly decreased between 2009 and 2015. Homicide rate rose from 1999 to 2009 but decreased from 2010 to 2017 around 65%, from 18.74 to 5.78. It was the only case of this kind in the Americas. We described the nature of the studied proximity strategy and evaluated it before-after through exhaustive control groups. For so, we measured correlations between crimes decreasing and Euclidean radius around Communitarian police headquarters during the period when policies were implemented. We found that police proximity and related strategies are visibly associated with crime decreasing, but other contexts could also be involved. We suggest studying the Ecuadorian experience in depth as an important example of crime reduction. ResumenEn Ecuador, la mayoría de las modalidades de robo disminuyeron constantemente entre 2009 y 2015, según denuncias administrativas, y aunque la tasa de homicidios aumentó de 1999 a 2009, disminuyó de manera progresiva y persistente de 2010 a 2017 en torno al 65%, de 18.74 a 5.78. Fue el único caso de este tipo en las Américas. Describimos la naturaleza de la estrategia de proximidad policial y la evaluamos antes-después. Para ello, medimos las correlaciones entre los crímenes decrecientes y el radio euclidiano alrededor de las sedes de la Policía Comunitaria en el período mismo de la implementación. Se encontró que la proximidad policial está visiblemente asociada a la disminución de la delincuencia, pero también podrían estar involucradas otras políticas y contextos. Sugerimos estudiar la experiencia ecuatoriana en profundidad como un ejemplo importante de reducción del crimen.


PUBLIC INTEREST STATEMENT
It is unusual to find reliable crime statistics on the Third World, and much less usual on crime decreasing. Due to the fact that Latin America is one of the most violent regions in the world, mostly among drug trafficking-related countries, this article is particularly useful. In Ecuador, between Colombia drug production and Peru and Bolivia biggest coca cultures, most of the robbery modalities constantly decreased between 2009 and 2015. Also, the homicide rate rose from 1999 to 2009 but decreased from 2010 to 2017 around 65%, from 18.74 to 5.78. This was the only case of this kind in the Americas at the time, and it might be the only one. We described how police proximity strategies were clearly related to these results. We showed Euclidean radius around Communitarian Police headquarters in the period when policies were implemented, in several subperiods, clearly pointing to the policy's impact on crime reduction inside each headquarter. evaluamos antes-después. Para ello, medimos las correlaciones entre los crímenes decrecientes y el radio euclidiano alrededor de las sedes de la Policía Comunitaria en el período mismo de la implementación. Se encontró que la proximidad policial está visiblemente asociada a la disminución de la delincuencia, pero también podrían estar involucradas otras políticas y contextos. Sugerimos estudiar la experiencia ecuatoriana en profundidad como un ejemplo importante de reducción del crimen.

Purpose and theoretical reference
This article aims to show how Ecuador, implementing proximity police, inside a preventionoriented frame, obtained concrete results on the decrease in some of their crime indicators.
Proximity police correspond to an extensive discussion in contemporary criminology. It is generally homologated to communitarian police, and it is possible to trace its origins to the very anglosaxon concept of policing. However, its concept is evasive. Jerome Ferret (2004, pp. 184-185) explains this conceptual condition because the political pressure that constantly is exerted over this kind of since the concept of proximity was coined in France. Political pressure makes proximity Police perceived as an ethical issue, related to citizen rights, prevention (in opposition to undesirable reaction) and a desirable alternative space for solving conflicts; this pressure also made the concept change according to political conjunctures. Because these reasons, in general, proximity is very difficult to evaluate. In order to know how to evaluate proximity Police, Ferret finds some common denominators. Here we highlight three aspects, all of them related to administration of a territory, and ethics: "a) The proximity police was based on territorialization b) The local police should be in permanent contact with the population, which manifested itself, especially, in an active association with everybody actors of security, association in which the local contract of security it was the reference by excellence.
(. . .) e) the service offered to the population and its improvements should guide the organization of the service, as well as the missions that had to be fulfilled before the inhabitants: it was especially necessary to ensure a better treatment for the victims and to the vulnerable population or in difficulty" (Ferret, 2004, p. 186) The ethical and political pressure over the proximity police is easy to find in specialized literature and throughout the narrative of politicians. Philosopher Vicente Herrera even finds in proximity police an expression of applied ethics. To him, in practice, proximity Police represents (. . .) the relationship with the community, because an ethical model of police action and strategy can never be designed with the backs of the citizenry. Prevention, because citizens rather value prevention before the resolution of crimes. The resolution of problems, because a 21st century policeman cannot simply comply with the law, but has to look for imaginative ways to solve problems. And trust, as this value is reinforced as a driving factor of an ethical and moral regeneration of our society (Herrera Arando & Moratalla Domingo, 2018) Latin American criminological tradition among others explicitly recommends adopting proximity police, associating it with professionalization and reforming it in order to create a police service nearer to citizenship (Frühling, 2003). The UNDP citizen security handbook for Latin America says: "the models of proximity police or communitarian must be developed into the preventive frame" (PNUD, 2013, p. 134). The ethical and political issues on this subject are clear in many other texts. In a certain way, it is expected from proximity police to be a local expression of administration of justice, which means a sort of government in small spaces, more than just a security service. In the regional literature, it is expressed by what Guillén Lasierra calls "Governmental Police," in his texts about Police Model classification (Guillén Lasierra, 2016; 2020), which tends not to establish fight against delinquency as the model goal. From this point of view, delinquency existence can be used as an alibi for producing political order against national threats.
Because of these perceptions, much of the regional debate sees the substitution of the governmental model by the proximity one, whatever it works or not in the reality, as well accepted from a Human Rights and pro-democratic point of view, but: Is police proximity useful to reduce crime?
Even if it was politically difficult to accept, Ecuadorian native specialized literature finally accepted the decrease, arguing that it was the effect of the judiciary reform and not of the police reform (Pontón Cevallos, Rivera Vélez & Amores Leime, 2020). 1 Some well-referenced criminologists even explicitly argue that crime reduction is not itself a function or objective of police, or crime decreasing by police is too complex to be measured (Gabaldón, 2007, pp. 255-262, Jérome, 2004. This article aims to evaluate national proximity police experience in Ecuador not because its ethical condition, nor relative to the perception of security, nor as component of democratic transformation, but because its ability to reduce crime and therefore, to protect social peace (Cano, 2002, p. 5;Mohor, 2007;Tudela Poblete, 2011). In other words, we will define proximity as Ferret did, but evaluate police through its most important objective, according to the 9 th Peelian principle of policing: to reduce crime. 2 Sherman 3 directed a referential report in Criminology, requested by the U.S. Congress in 1996, assessed the extent of more than 500 Crime Prevention Projects, to identify whether public funds then allocated by the State, were justifiable. The report stated that some police strategies were demonstrably more efficient than others. This report is an important piece that helped to build new developments on Problem Oriented Policies [POP], Geometry of Crime and more recently, Criminology of Place, among other related theories. Among its conclusions, more efficient polices were as follows: 1. Increase in the number of Police Servants, 2. Arrests aimed at high-risk to crime persons 3. Development of the Police legitimacy conditions through the door-to-door relationship.
A convened principle behind POP, Geometry of Crime and Criminology of Place, which is important to the purpose of this article, is: "A strong minority, a small proportion of all units of criminal conduct, causes the most damage to most types of crime," and because of this: "efficiency in crime prevention may increase when resources are concentrated in strong minority units, identified by the use of patterns from past behavior" (Sherman, 2012, p. 8;Sherman & Sherman, 1996). This perception is supported by the relatively recent situational crime theories, which highlight the "proven" efficiency of a criminal policy, fundamentally preventive, based on micro spaces, that is, where a small group of people, actors from small territories, are responsible for producing most of the crimes of an urban conglomerate.
Also, it has been related to the efficiency of proximity police in the fight against terrorism. Early reports on suspicious activities and a healthy relation between proximity police seems to be one of the most important tools for preventing terrorism, as it was studied recently in Nigeria (Tarela, 2018).
In this sense, micro-territorial crimes are not a "minor" evil. Statistically expressed, they affect in an extended way to almost all the population of the country, above all, in the corresponding urban conglomerates. In fact, policing on these micro-territories has "proved" efficiently that allows to reduce by approximately 40% the total crimes in several cities of the world, leading to the declaration of a "Law of Concentration of Crime" (Weisburd, 2015, pp. 133-157;Weisburd et al., 2012).
Given the experimental nature of situational theories in criminology or anthropology (Geertz, 1973, p. 23;Clarke, 1995) it is not possible to automatically project the logic of a criminal opportunity from one situation to another. For this reason, it is important not to generalize crime policies in large territories if they cannot be previously tested by rigorous tools. It implies that criminal situation diagnoses also have to be made from small territories, where proximity police would have a better incidence because of its nature.
From this perspective, proximity police would constitute an important tool for applying effective prevention strategies from the situational perspective. It is possible to presume that a welloriented police in small territories would have, at least, some dissuasive effect in reducing crime. This is the angle we want to highlight in this article.
For doing so, we applied some public policies basic methods, as conglomerate control groups and before-after data compilation (Howlett et al., 2009;Laws & Hajer, 2008;Aguilar Villanueva, 2007a;2007b). The used and available data were an exhaustive body of all the sub-circuits and proximity police units of Ecuador.

The new model of police management
Proximity police in Ecuador is not new, but it is possible to show that it could improve its efficiency since the New Model of Police Management (here in after, NMPM) promoted by Ministry of Interior of Ecuador and National Police since 2009. This is partly based on: 1. Strategies that induce Police legitimacy among communities, 2. Administrative decentralization of Policing, re-directed to small territorial units, 3. Geo-referential computerization of criminal dynamics in order to prevent and to react, 4. A police educated to improve negotiation abilities with the community, 5. Directed arrests of people at high risk of crime, and 6. To increase the Police staff in order to enhance deterrence in small places. According to the literature indicated above, these Ecuadorian policies somehow match with the indicated specialized recommendations (Sherman, 2012).
Such policies, although they were started in 2010, took shape through three 2012 concepts: The "Police Decentralization" (Desconcentración Policial, in Spanish), the "New Management Model of the National Police" and, at the programmatic level, the "Citizen Security and Social Peaceful Coexistence". Such impact is recognizable by the national extent of territorial penetration, which is exclusive to these policies in that country.

Territorial decentralization and communitarian policing
In 2009, police zoning was described like this (. . .) at the Metropolitan District of Quito, police zoning conforms to the administrative zoning of the municipality. In the other cities of the provinces, there is an incompatibility between the police and the municipal zoning (. . .) Thus, there is not, from the standpoint of administrative organizational visions, affinity between Police and Municipalities. (Pontón, 2009, p. 105) However, as a result of the Organic Code of Territorial Organization, Autonomy and Decentralization, the Ministry of Planning and Development in 2012 established a new division based on new units of administrative planning: National/Zonal/District and Circuits (Asamblea Nacional, 2010;SENPLADES, 2012).
In the same year, the Ministry of Interior created the model of decentralization of the National Police by adding a smaller instance for the purpose of generating a greater proximity to Police: the sub-circuits (Ministerio Del Interior, 2012). This administrative division of Communitarian Watching Units (hereinafter, UVC) was implemented at the district level and at the Communitarian Police Units (hereinafter, UPC) at the sub-circuit level. The provision of material support to some of the old units were also improved; specific patrolling programs in small territories were activated, as the socalled Transversal Programs: Safe Neighborhood, Citizen's Contact, Local Security, Safe School, Safe Public Space, Security Buttons, Security Alarm and Citizen's Assemblies. Sub-circuits finally allowed to national security policies to penetrate the territory, where municipalities were not operating or just operated in separated focus.
A sub-circuit is calculated according to the linear distance traveled by a patrol, whose average speed is 13.67 km/h, in a Police shift, i.e.: 8 hours. So, speed x time, it covers about 109 linear km. If a circuit has 300 km., linear route will produce 2.75 areas, that is, 3 sub-circuits which, in practice, will be approximated to 1 km2 each. The definition of sub-circuit in real space is also subject to a) road splits which constitute fundamental road frames of a nominable territory, b) a density between 5,000 and 10,000 inhabitants per Sub-Circuit, c) the "problem of Coexistence and Public Security", estimated by crime rates, d) topography, and e) police resources already set in place (Policía Comunitaria, 2014a) Thus, the map of the Police's Service of Ecuador was radically transformed from the years 2011-2012, dividing the 140 traditional administrative districts in 1134 circuits about 5 km 2 and 50,000 hab. Approx. each one, and, only for the purposes of policing, dividing those Circuits into 1880 subcircuit about 1 km 2 , from 5,000 to 10,000 hab. approx. each one. The new territorial units serve as operating units of 1829 headquarters of the Communitarian National Police, under the forms of UPC (Communitarian Police Units) and UVC (Communitarian Surveillance Units), spread all over the country (Policía Comunitaria, 2014a). Sub-circuits are the most important decentralization expression for prevention purposes because they support the main proximity strategies of the NMPM and the Communitarian Police.
The NMPM, in addition to distributing the Communitarian Police on sub-circuits, also improved its infrastructure and inputs, improved police servant salaries, reformed the educational model, the way new agents were selected, increased the police staff and trained the new Police in implementing Transversal Programs to include citizen participation in security strategies at the subcircuits level. the scenario that led to a political schism: the so-called 30-S-The facts and interpretations of this historical process are too complex to be explained here, but it is important to state that it allowed the government to replace influential chiefs of Police and to facilitate the Police Reform in 2010. This last process was also called the depuration. In 2012, the achievements were clearer, the government deepened this strategy in several aspects and formally called it, in Spanish, el Nuevo Modelo de Gestión Policial (NMPM). Figure 1 illustrates this process The reform of communitarian policing meant, among others, two simultaneous changes: First, territorial and Police responsibilities were distributed in 2,028 sub-circuits, and secondly, 39% of the 45,015 agents of the National Police were directed to Communitarian Police, in order to activate the   (2014)), which has also been instructed in some way in the Transversal Prevention Programs, but they don't fully implement these programs, do not yet have the patrols, nor the final technological equipment and infrastructure planned for the new model. So that, for purposes of statistical measurement to be made at a later section, the UPC and UVC who are not in full operations, will serve as a control group against those that do fully correspond to the New Model of Police Management, in order to determine whether they represent a weighty change in the policy impact.
Finally, Police Circuits at the UVC level have also served as a bridge to bring specialized services of the Penal System, such as Crime Labs, Prosecutors, Police Reports Collecting Centers, Peaceful Mediation Services and Specialized Police on Domestic Violence, Women and Children, including others.

The police servers increase
The other relevant aspect of Police Reform is having increased members and their quantitative distribution in the territory. Ecuador aims to achieve the rate of 1 Police/250 inhabitants, established by the UNDP. In 2014, it declared having reached 1 police/330 inhabitants (CTI/EL CIUDADANO., 2017; PNUD-AECID, 2009, p. 31) that is, including special and decentralized police. Sub-circuits distribution improves the vicinity of the Police to the whole community, because it places at that level, assuming optimal conditions, between 16 and 22 Police servers, divided into three shifts of 8 hours each, along the 1,714 UPCs.
By October 2013 it was estimated that there was a shortfall of about 17,000 officers to do their jobs effectively and to meet international standards (EL CIUDADANO, 2013). Because of that, the government planned to increase from 39,000 agents in 2011 to 69,000 Preventive Police agents in 2017, and also to increase the number of agents of the various specialized branches (TERRA

Transversal programs
The UPCs and UVCs are responsible for the implementation of nine key prevention programs, regardless of their own patrol functions, conflict-resolution, arrests and police reports: Citizen Contact, Safe Neighborhood, Citizen Training, Local Security, Safe School, Safe Public Spaces, Security Buttons, Communitarian Alarm and Communitarian Assemblies (Dirección Nacional de Policía Comunitaria, 2014). The following 5 transversal programs are fundamental to understand the played role by Police as an actual proximity police.
Citizen Contact program consists in police visits to residents or space users (as merchants) door to door; police delivers the phone numbers of the unit, and fulfills basic and strategic information about the community in a system; the residents are explained about how the UPC works and how to act in several situations. Then, the data is projected on the Police's map and in the respective tabs. During 2014, 1,191,426 Citizen Contacts were reported throughout the country until August. For the Metropolitan District of Quito (DMQ), the rate of contacts was approx. 6,500 new citizens  per month. For the Metropolitan District of Guayaquil (DMG) was 7,500 people per month. In the available statistics, monthly results looked stable enough to say that the instruction was rigorously and methodically implemented. Citizen Contact is the first important strategy that a new policeman must implement and, because of that, it is the basis of other future strategies.
Citizen training consists in identifying members of the community in order to invite them to workshops and conferences on general issues related to security. In the Metropolitan District of Quito community members were trained at a rate of approximately 3,500 people per month, and in the Metropolitan District of Guayaquil, 5,500 people per month.
The Safe Neighborhood/Security Assembly program consists on to organize a permanent assembly between the police and its community in order to design local security strategies. According to statistics, only in 2014, police settled 1,755 Safe Neighborhoods in Ecuador and 126,129 "participating leaders engaged in Public Safety".
Communitarian Alert and Safety Buttons. There are two types: the safety button, which can be fired from the mobile phone to alert the police; and the communitarian alert, which can be a kind  Safe Public Space consists to diagnose the territorial gaps in neighborhood public spaces through statistics, surveys, contacts, and meetings with neighbors. Subsequently, the Police and the community have to plan an intervention schedule that includes inter-sectoral coordination.

David system. Qualitative and geo-referenced statistical information
Since a presidential request, the Police and the Ministry of Interior created an Office of Information Analysis of Crime (OAID, after changed to DAID in Spanish). Actually, the DAID is a team of about 140 Police and civil professionals spread on the country, mainly geographers, statisticians, and engineers in systems.
To date, the David System is the main product of the DAID (Castro-Aniyar et al., 2015). It consists on a geo-referenced CompStat to identify qualitative and quantitative patterns of crime, from subcircuits to national level. From 2014, it decided to include Composed Cognitive Maps, an innovative victimization tool in order to produce qualitative and larger information based on situations called Spatial Dynamics of Crime (Castro-Aniyar, 2018). The David System compiles and relates Prosecutor's reports, judiciary information, data from the National Institute of Statistics of Ecuador, independent statistics, ECU-911 reports, qualitative approaches, and geo-referenced information, among others. The georeferenced mapping of this information allows to offer instruments to proximity police servers in sub-circuits and microterritories (Núñez, 2018).   (c) Social Leadership: designing strategies and intervention plans based on assertive community relations;

The new educational curriculum of the national police
(d) Guidance to the Community: to be part of an institutional culture of service and evaluation of community satisfaction; and (e) Teamwork: promoting participation of the members, evaluating the different experiences and share information (Dirección Nacional De Educación, 2017;

Overview of the crime drop in Ecuador
The Ecuadorian government launched a set of new policies to reduce crime and build Peaceful Social Coexistence at least between 2010 and 2015. Such policies were designed and implemented with various public and civilian agents of Ecuador, such as civil society organizations (Safety Committees, Neighborhood Brigades, Associations, Citizens' Assemblies, Chambers of Commerce and Industry, etc.), the Autonomous Governments, the Judiciary Council, the General Attorney, the Ministry of Justice, the Coordination of Security Ministry and, mainly, the Ministry of Interior.
At the end of 2014, it was established that Ecuador had improved in some its most important crime indicators (Castro-Aniyar et al., 2015). Not only its achievements in the fight against crime deepened through the years, surpassing the state of epidemic violence as defined by UNDP, but its Homicide Rate dropped to less than half in a persistent, single curve in Latin America, and today approaches to low-violence countries as Uruguay and Chile (Gaibor, 2015). No country in the region presented these indicators in a progressive and sustained way. Pointed droppings in other countries in the region, such as Colombia, Guatemala, Mexico and El Salvador, are linked to truces or struggles between cartels, which does not correspond to the situation in Ecuador. The Ecuador's abruptly homicides falling reached in 2015 less than half of the 2009 level, giving an important validity to its criminal policies.
From the programmatic point of view, Government's National Plan for Good Living envisaged to reduce this indicator to 8 points by 2017, which means that in 2014, 3 years before, the country having reached 7.2 incidents/100,000 inhabitants, already exceeded in 0.8 his own goal (SENPLADES, 2009).
The relevant indicators obtained from the David System, INEC (National Institute of Statistics Ecuador) and the National Police, will be shown next. Since computerization and validation of the data also corresponds to the process initiated in 2010, validated, and standardized available information will be placed in the graphics according to the available periods. Some graphs were made by us, only with an illustrative purpose. Other graphics focus on periods of interest to clearly show the purpose of the article.
The decline in the homicide rate is visible in the following charts: (b) The decline in the rate and the absolute number is constant over time.
As can be detached from the graph below, in absolute terms, this decline implied, statistically speaking, more than 1335 lives saved in 2014 compared to 2010, inside the dropping period: Indicators of several modalities of Robbery have also declined between 2012 and 2014, during the validated and standardized information period at the time this article is written. In Robbery to People, to Stores, to Cars and to Homes, measured by Police and General Prosecutor reports, a significant improvement is reflected, as shown in the following examplesFigure 5,6 This decrease would point to the existence of a deterrent presence in the public spaces, like patrolling and proximity police. However, these records can be weakened because of the fact that Ecuador, as many other countries in the world, has a strong under-registration, which means that non-territorialized national reports could not reflect reality.
Since it is possible to argue that declines in Robbery are the result of declines of reports, and not of decreases in commission of crime, it is important to triangulate the provided information.
Criminological test conventionally used to counteract the effect of the dark numbers of crime behind the formal complaints is the Victimization Survey (Damnert et al., 2010;Dijk et al., [2008] 2014; IIDH-República de China, 1999)   The graph shows, first, that Victimization Rate (in blue) fell from 7.14% to more than a half, and stabilized near the first quarter of 2015, reaching a rate of 2.42%. This means that citizens declared that crimes against them have fallen by 66% for the period in which most of the policies were implemented and, also, that such phenomena stabilized since December 2012 to March 2015 -surpassing traditionally associated stations with increasing of crime. On the other hand, the curve of the reports rate made by the same people after being victimized (in red), although it is unstable, tends to increase throughout the period of the graph. So, it is not possible to say that the increase in complaints corresponds to an increase in victimization, but, on the contrary, it increases while crime decreases, which is associated with a growing legitimacy of the complaint as resource in society. By this measure, the decline in indicators of crime is accompanied by a more determined attitude of the population to make a report.

Relationship between the decrease of robbery and homicide, and the new model of police management
Next, there will be information collected from the David Systemin relation to the commission of crimes to 200 mts. radius around the UPCs/UVCs, between 2010 and 2014, it means, the period when proximity police was implemented and fully implemented through (OAID, 2014;OAID, 2015;Sistema David, 2016;Policía Comunitaria 2014b) NMPM Figures 8, 9, 10, 11, 12 and 13 will allow to understand crime behaviour in this period.
The information will be separated on the type of Unit, and before/after the fully implemented model. We argue that proximity police efficacy has a strong relation with police servers increasing, full infrastructure presence and transversal programs, which gives proximity to the actual ability to operate in micro-territories, as indicated by international handbooks and criminologists. In Ecuador, the process of readjustment and construction of new headquarters had three stages: From 2008 to 2010 there was the Safe Zones Pilot Plan, mainly in Quito and Guayaquil. They were the first antecedents of the NMPM, consisting on multiplying and connecting self-defense and prevention civil initiatives, called Brigadas Barriales (Neighborhood's Brigades). From 2010 to 2012 the old headquarters were readjusted, received new equipment and police servers were trained for the community context. In some ways, the National Police was distributed on sub-circuits (without being formally constituted) but it not received yet the Transversal Programs. From 2012 to 2013 the first new headquarters were built, replacing most of the time to the above, the sub-circuits map was formally completed as the administrative units. Transversal Programs, for the first time, were induced in all venues of all models. From 2013 to 2014 the construction of new headquarters was expanded and induced again, but more broadly, as the Transversal Programs. Such extension of the program involved a greater number of Servers and Communitarian Alarms, Security Buttons, Inter-Agential Security, Citizen Contact and Communitarian Assemblies.
For this reason, the productivity of Police Units (UPC and UVC) for the full implementation of the NMPM, that is, the sum of the three waves (rehabilitation of the structure, construction of new headquarters with Transversal Programs and expansion in depth of the model) must be clear for reading and understanding the maps. In order to separate the old model that still functioning in Ecuador, the blue curve will represent it as a control group, and the full implementation will be read as "Nuevo Modelo de Gestión Policial" (NMPM) corresponding to the red curve.
The graphs also will show the units that have reduced crime in their territories, divided by the number of total units in each year, in order to observe their productivity in declining of crime, in relation to other units.
It will be visible in most graphics that crimes increased from the beginning of the studied period, 2010: Because this year appeared the David System, a new Prosecution system, efficiency of the proximity Police service increased. Such change augmented the available information, generating more reports and less loss of information. Therefore, the graphs indicate steepest ascending curve in 2010, before the descending curves.

Analysis of results and conclusions
By isolating the implementation of the NMPM by years, all displayed graphics show a positive effect of proximity police on the decrease in Robbery Rates. This decrease is visibly more pronounced and more consistent in the Communitarian Police Units where the new model was fully implemented. This difference is more visible after the expansion of the new headquarters, servers increasing and the deepening of the Transversal Programs that began in 2012; in this case, the curves clearly precipitated from 2012 to 2014.
In the case of robbery, the curves show a rise from the beginning of the implemented model in 2010, which is related to an improvement in the collection of reports and, consequently, less loss of information. Then curves responded to the intensification of the model in 2012, associating this result to the common variables for all units in the analysis: Transversal Programs, new infrastructural presence, and more police servers in small territories. The actual evidence points to the basic formulation that prevention must be focused on "the units of strong minorities" and small places (Sherman, 2012;Sherman & Sherman, 1996). This approach was also developed by a strong criminological tradition based on the situation/ opportunity relationship (Scott & Stuart, 2012;Ponton 2009) and focus on micro territories (Weisburd et al., 2012).
Robbery graphics express results better than Homicide graphs, since they are produced from larger databases. In other words, on Homicide variable, although the presence of a corpse is a much stronger event and, therefore, it is less susceptible to underreport, the measurement corresponds to smaller databases; therefore, its weight is weaker. This is the reason why Homicides were calculated as 1000 mts. radius from the UPC. This could affect the measurement of UPC productivity for this crime. However, the chart shows a particularly pronounced and more stable Homicide descending curve than the Robbery ones, which also can be related to a territorial policy impact.
The fact that the homicide curve does not look sensible to the intensification of Model in 2012 suggests that the decline in homicides was triggered by policies or previous factors since 2009-2010. However, since the deterrent factor to Robbery looks clear from 2012, there are reasons to think that the implementation of NMPM in time also facilitated the decline in the Homicide Rate. In other words, the Ecuadorian proximity police model, close to Ferret and Sherman descriptions of the service, has a significant effect on crime decrease during the studied period.
Moreover, in order to rule out the possibility that the cause on Robbery decreasing was the mere proximity of the UPC/UVC, and not the proximity police model, expressed through the NMPM itself, it is important to note that in the period 2009-2011 there were 1276 UPC in operation, which were spread through a proximity strategy throughout Ecuador. The distribution of the Units on the map is very similar to the 2012-2014 period. In fact, since 2012 there were already 1382 UPC/UVC, which means that only 7.6% more units were added. In other words, the presence of proximity police not only can be explained by its simple presence, but to associated crime policies oriented to prevention in small and micro territories, as described above.
So that, by contrasting the criminal behavior around the units where the model had not been fully implemented-as a control group for the period 2010-2011, to a number and similar distribution, it indicates that the curve breaking effect appeared in 2012 for both curves, far from the simple proximity factor. This new factor is the NMPM, even if it is not fully implemented. It has been already found that communitarian presence of the police, as the transversal programs look to consolidate in sub-circuits, can affect perception of security and generate deterrence (Vidales Rodríguez, 2012, pp. 478-488).
Homicide rate behavior seems to show other playing factors in the whole crime reduction phenomenon (Vinueza, 2018). Even if proximity police shows a positive effect on studied variables, it is important to deepen the analysis considering other angles and to compare to available data.
From the model analysis itself, even if it is not possible to understand how precisely transversal programs (i.e., direct arrests or door to door visits) affected the decreasing, the model shows a positive relation to crime reduction as a whole. Because of the specific behavior of the policies, it is possible to stablish that mere proximity does not impact the same way the statistics, as the period when transversal programs, quantitative-qualitative methods as MCC and computerization of crime indicators were applied. These results point to think an interesting relationship between this Latin American experience and the now a day criminological literature about the importance of small territories, as expressed above.
From the political point of view, paying attention to the Latin American mainstream narrative about the democratic need of the proximity model, it was not possible to measure how it could empirically improve, or not, civil rights. For the case of democracy indicators increasing, the NMPM does not focus strongly on citizen political participation, but mostly on citizen organization, which would imply a coordination with policemen in small territories. The NMPM is far from the implemented examples where policing and Human Rights are directed by the same community (Aniyar de Castro, 2010) or even the ancient Ecuadorian system of citizen security brigades (Brigadas Barriales, in Spanish). It gives to the police the ability to develop professional criminometric techniques on territory as computerization and MCC, without depending on the community political leadership. Because of this, it is not possible to relate any deepening of democratic security tools with the Ecuadorian crime reduction, even if democracy indicators were available. But, in the same time, if policemen do not professionalize or acquire the social abilities to measure and acting in the community context, the model will not work either, as it is pointed in the figures.
The introduced conclusions allow policymakers in general to understand that the proximity model of police, oriented to give the policemen an active role in the situational and crime opportunity context can be related to robbery and, at least indirectly, homicide decreasing. Even if, in the one hand, the real context of civil rights and democracy improvements under this model still remain theoretical, these results show that, on the other hand, crime decreasing is empirically possible.
Finally, these results could also have an impact in the way we use to imagine crime justice. If we share the critical criminology standing point, stating that crime policy and crime justice include all the hegemonic apparatus, as school, church, or police (Zaffaroni, 2011, p.3); Cancio Meliá & Pérez Manzano, 2019) the Ecuadorian experience can fit, even if unconsciously, into an imagined major system of crime justice, according to a referential author on the subject: the State model-civil/mediation variant (Delmas-Marty, 1992, pp 137-151). From this perspective, in the studied period, citizens, in a solitary way, provide intelligence data insights to professional police assistance, which nourish a designed system where mediation, retributive justice, and prevention can be developed. In this sense, it is discussable in the near future, if the implemented policy efficiency can be projected into the imagination of a different kind of a crime policy system, compared to the existing one.