The exercise of strength is a foundational element of the modern State. There is not one constitutional State that does not include the presumption of the risk of attack in its founding text, the Constitution, indicating that safeguarding public order is of utmost importance to the modern State (Sajó 2006: 2282–2283).

Increasing demands for improved public safety have occurred in parallel with a demand for a solution to crime. In Mexico, public safety is a primary concern for 57 per cent of the national population (INEGI 2013), even more so than unemployment, according to the National Survey on Victimisation and Perception of Public Security (ENVIPE) 2013 (ENVIPE 2013). Mexican citizens also demand a more effective system to investigate crimes and administer justice.

The security and justice constitutional reform of 2008 sought to address these two legitimate interests. The reform firstly sought to respond to the institutional and public concern about the increase of crime and the perception of insecurity, and secondly, it acknowledged the need for a justice system that would put an end to impunity and overturn the practices of the traditional system, which violate fundamental rights.

Discourse surrounding the reform justified the need to improve both aspects without distinguishing them as two independent phenomena. Under this premise, the resultant constitutional framework contained conflicting ideals: while the framework was based on the highest standards of human rights, it also contained exceptional measures that severely restricted rights, such as mandatory pre-trial detention, extra-judicial detention and forfeiture, provisions which were based on arguments of ‘effectiveness’ and were intended to improve public safety (Fondevila and Mejía: 2011).

The reform did not meet its goals, however. The crime rate did not decrease in the following years and the rate of unreported crimes of impunity changed little. ENVIPE 2013 reported an increase in the rate of victims of crime from 24,327 victims per 100,000 inhabitants in 2011 to 27,337 in 2012. The crime rate also increased from 29,200 crimes per 100,000 inhabitants in 2011 to 35,139 in 2012, while the rate of unreported crimes was 92.1 per cent (INEGI 2013).

Crucially, the public safety regulations that restrict fundamental rights are costly from a constitutional point of view and from the perspective of democratic legitimacy. When these move beyond ‘norms’ they become emergency measures that do not require a formal declaration – as an emergency situation would require. The arguments of ‘effectiveness’ allow, from the beginning, a strong presumption against its permissibility. Additionally, when administrative practices are opaque, the risk of impacting innocent people that cannot be protected through normal channels is very high (Sajó 2006: 2270 and 2290).

To assume that the criminal justice system alone can solve the problem of crime is a demand that exceeds its capacity. It is also true that other mechanisms can contribute to public safety, such as the Office of Public Prosecution, which typically assumes responsibility for the issue within their institutional objectives.

In terms of prevention, the Public Prosecutor can coordinate with the police if, during investigations, they can identify criminogenic spaces. Also, properly attending to victims counters perceptions of impunity and reduces feelings of fear.

The response to a demand for public safety cannot be abstract, as this can result in the hardening of the justice system, greater restrictions on rights, and increased penalties. Thus, the clear identification of a specific problem allows for rational, concrete responses.

Thus far, the responses to improve public safety have not necessarily been reflected in prison sentences. This highlights the need to utilise more productive responses of the system towards crime, such as detoxification programmes, effective reintegration, and supervised parole.

Accordingly, this paper will focus solely on the study of pre-trial detention and other precautionary measures supported by the new criminal justice system, how they relate to the principle of presumption of innocence, and the tension between it and the punitive demands for increased imprisonment. In addition, this study will analyse an updated technical solution, pre-trial services, which seek to balance the presumption of innocence and the right to personal liberty with public safety.

A new justice system and the demand for public safety

The ‘Rule of Law Index 2014’ prepared by the World Justice Project reports that, with a rating of 0.25/1.00 in the area of criminal justice, Mexico ranks 97th among 99 countries surveyed globally, and 14th of 16 regional countries (WJP 2013: 117). These results confirm the need for an improved criminal justice system that corresponds to a democratic rule of law.

The constitutional reform of 2008 enshrined potential alternatives to pre-trial detention (Article 19, Paragraph 2). These bail measures are designed to ensure the presence in court of an accused, that he or she does not hinder the criminal process, and to guarantee the safety of complainants, witnesses, and the community. In this sense, bail measures are purely procedural mechanisms that provide balance between the presumption of innocence and public safety.

Controversially, the same paragraph in the constitutional reform contemplates restrictions on the application of bail measures when a person is being tried for a separate offence or has a criminal record. It also establishes a catalogue of offences that do not permit pre-trial release on bail. On the contrary, the Constitution requires judges to order mandatory pre-trial detention when any of the following situations apply: organised crime, intentional homicide, rape, kidnapping, human trafficking, crimes committed by violent means such as with the use of weapons and explosives, and serious crimes as defined by national security, free personal development, or health legislation.

Founded on the principle of presumption of innocence, the new constitutional standard regulates a series of bail measures under the accusatorial and adversarial system. Unfortunately, the practical application of these new mechanisms has generated perceptions of impunity on the grounds of a ‘revolving door’ argument, which refers to the opinion that ‘the police arrest criminals and the system lets them out’. However, there is no evidence linking the implementation of such bail measures with the rate of unreported offences, much less with increasing crime rates and violence.

In this manner, when public safety is of higher importance than the rights of detained and accused persons, the existence of the presumption of innocence is questionable. Defending public safety over the rights of these individuals ‘hinders’ the execution of adopted criminal policies and promotes myths about pre-trial detention; for example, that it counteracts insecurity, reduces the incidence of crime, or is a tool to stop ‘dangerous people’ (Zepeda 2010).

In spite of criminal policy, pre-trial detention figures have not changed significantly in recent years. According to figures from the Administrative Office for Prevention and Rehabilitation (OADPRS) of the National Security Committee (CNS), as of June 2014, Mexico had a total prison population of 254,641 people, of which 110,116 were on remand. That is, 43.24 per cent of people in prison have not been sentenced.

Similarly, OADPRS reports that from 2005 to June 2014, the prison population grew by approximately 20 per cent, while the remand population remained at levels between 40 and 43 per cent of the total. Table 1 shows these indicators divided by jurisdiction.

Table 1

Number and percentage of people in pre-trial detention by jurisdiction.


FEDERAL 23,391 26,647 53.25%
LOCAL 121,134 83,469 68.9%

Note: These figures are from the OADPRS 2014 report.

These numbers challenge the aforementioned assumptions of pre-trial detention. If these myths were true, the victimisation and crime rates would be inversely proportional. Also, pre-trial detention is not a determining factor that responds to the demand for public safety, and on the contrary, it carries high social and economic costs (Zepeda 2009).

The attempt to transform criminal policy that favours the deprivation of liberty demands reflection of the presumption of innocence, and why its effective implementation should be tied to the call for public safety.

Presumption of innocence

The presumption of innocence is a complex concept that manifests itself in two concrete meanings. Firstly, it is a probative rule that is reflected in the ‘in dubio pro reo’ principle and secondly, in the rule that determines that the accusing party —the State— carries the burden of proof.

As the ‘in dubio pro reo’ principle prescribes, presumption of innocence is ‘the requirement that the sentence, and thus, the application of a penalty can only be found on the certainty of the court that finds as to the existence of a punishable offence…’ (Maier 2004: 495). Certainty, not probability.

As the burden of proof, it imposes the obligation of proving guilt on those who accuse. This involves a series of pre-sentence procedural steps that override the presumption of innocence until they achieve the conviction of the court, beyond reasonable doubt, that the person is guilty. Such preliminaries support a positive probability in relation to the charge, such as the precautionary measures that allow for the restriction of freedom of accused persons as they face trial, including pre-trial detention (Maier 2004: 496–497).

With respect to the second meaning, the presumption of innocence is also understood as a rule of treatment. In this sense, it is a fundamental right associated with the right to a defence. The Inter-American Court of Human Rights stated in the case of Ricardo Canese v. Paraguay, ‘[…] the right to the presumption of innocence is essential to the effective realisation of the right to a defence and accompanies the accused throughout the trial proceedings until a judgment determining his guilt is secured.’

The second meaning also gives it an extra-procedural perspective. That is, it obliges third parties, not just authorities, to comply, as recognised by the Supreme Court of Justice (SCJN) (SCJN 2007). In this regard, the Human Rights Committee in its General Comment no. 32 imposes a ban on officials making public statements that violate this right. It also establishes an obligation on third parties, such as the media or experts, to respect the presumption of innocence. In both cases there is no direct state responsibility for this conduct, but the principles apply to criminal proceedings requiring every judge, as arbiter and guarantor of the rights of the victim and the accused person, to intervene when this right is violated.

Presumption of innocence and reform of the criminal justice system

The 2008 reform established a differentiated rights regime depending on whether one dealt with the inquisitorial criminal justice system or the oral adversarial system. Within the former, it would be understood that the presumption of innocence applied jurisprudentially, and not specifically recognised in the Constitution. In the adversarial system, the reform expressly included the presumption of innocence in the list of rights of an accused person in Article 20.

With the human rights constitutional reform of 2011 and later, the judgment issued by the SCJN through the contradiction thesis 293/2011, the catalogue of fundamental rights, includes all those set out in the Constitution and in international human rights treaties. The presumption of innocence, therefore, is present in the Mexican legal system and has the consequence that a person charged ‘enjoys the same legal status as an innocent person.’ It is indeed a political starting point that the law of criminal procedure assumes - or should assume - in a Rule of Law state…’ (Maier 2004: 491).

The presumption of innocence is a right with an essential practical significance that protects those accused of a crime during a criminal trial, and acquires even greater relevance in a context where arbitrary detention and torture remain common practices of national police agencies. This was the conclusion of the United Nations Special Rapporteur on torture and other cruel, inhumane or degrading treatment during his last visit to Mexico (Montalvo 2014).

These practices that violate human rights, by being carried out systematically, instil fear and distrust of the justice and public safety systems in citizens. They thus lose legitimacy, which makes it essential to give meaning to principles such as ‘due process,’ and to provide certainty in meeting the elements that make up these rights through laws, judicial practices, and approaches that effectively link authorities and third parties (Maier 2004: 540).

This is essential to understanding why the tension between the presumption of innocence and public safety is a false dilemma. From a comparative perspective, it is not uncommon for legislatures to establish rules denoted ‘exceptional’ or ‘extraordinary measures’ on issues related to organised crime and terrorism. These tend to reverse the burden of proof onto the accused.

Courts that have studied these cases recognise that even without the reversal being complete, it is possible to impose such burdens on accused individuals whenever guided by criteria of necessity and proportionality (Gupta 2002: 177 and Hamilton 2011: 4).

A concrete example of this type of exception is found in the case of forfeiture: a typical regulation of extraordinary character that requires the verification of the legal origin of a good owned. The need for this is because the alleged owner is best placed to check, without involving the State and its prosecution arm, in the performance of all activities necessary required to prove the illicit origin.1

The European Court of Human Rights has ruled on these presumptions of guilt and considered that they should be subject to limits consistent with the rights of the defence in criminal proceedings. In the Court’s opinion, it is wrong to give a ‘blank check’ to the legislature to contemplate them.2

In Mexico, the laws establishing criminal policy, based on efficient public safety arguments - from which derives mandatory pre-trial detention - have not found a legitimate justification from the point of view of human rights. The mere fact of constitutional regulation nullifies the effectiveness of any legal remedy available for the protection of fundamental rights violated by them (Patrón 2012: 16–19).

By constitutionalising such measures, it is necessary to count on all possible safeguards that protect freedom, integrity and personal security, and due process. When this does not occur, everyone becomes vulnerable. Therefore, our perception of insecurity may increase.

Jurisdictional function and restrictions on liberty

One of the major criticisms of pre-trial remand described is based on the interference of this figure with the judicial function. That is, by dispensing with the obligation of judges to order pre-trial detention in the case of offences listed in Article 19, the principle of separation of powers is violated.

The role of a judge is critical since it determines the guilt of a person once the parties have been heard. In other countries judicial precedents have always regarded pre-trial release as the rule, even in high-profile cases such as murder. The liberty of a person while facing criminal proceedings was mandatory for cases in which the death penalty did not apply. In those in which it did, the judge decided discretionally according to the particular circumstances of the case. The judge understood that ‘…to deny freedom to a person who may be later acquitted, was much worse than the risk to the community by releasing the defendant.’ (Baradaran 2011: 728–729).

Following the human rights constitutional reforms of 2011, it was thought that there might be an alternative response to the constitutional restrictions based on an interpretation using the pro homine principle. According to the analysis undertaken by the SCJN in the Radilla case, many optimistically greeted the possibility that judges could better protect citizens from irrational restrictions, or alternatively, could decide on a reasonable restriction when necessary and on a case by case.

Accordingly, the SCJN, by resolving the matter ‘Several 912/2010,’3 gave content to the constitutional parameter whereby the rest of the norms must be construed. Judges are obliged to take into account the human rights contained in the Constitution and the jurisprudence of the court, human rights contained in international treaties ratified by Mexico, and binding criteria of the Inter-American Court of Human Rights, regarding the judgments in which the Mexican state was a party – and guiding criteria if Mexico was not a party to the case.

Regarding convention compliance, the Court recalled the decision of the Inter-American Court of Human Rights in the case of Cabrera García and Montiel Flores v. Mexico, according to which:

Judges and bodies involved in the administration of justice at all levels are obliged to exercise ex officio control of “conventionality” between internal standards and the American Convention, clearly within their respective competencies and corresponding procedural regulations. In this task, judges and bodies involved in the administration of justice must take into account not only the treaty, but also the interpretation thereof made by the Court, the ultimate interpreter of the American Convention.4

With the decision 293/2011 that resolved a jurisprudential contradiction thesis, the Supreme Court gave effective validity to an extended catalogue of human rights. That is, one that integrates not only those fundamental rights enshrined in the Constitution, but also those delineated in international human rights treaties. The Court also considered that the judgments of the Inter-American Court of Human Rights are binding on the Mexican state, whether it was a party or not.

However, the SCJN also maintained the validity of the constitutional restrictions against rights and the interpretation of those foreseen under international treaties. That is, regardless of international precedents in certain matters, if the Constitution says otherwise, the latter holds precedence.

In principle, this would be the case when the nature of the crime mandates pre-trial detention. There are international precedents that oppose this. For example, in Bayarri v. Argentina, the Inter-American Court noted two important things. One, that pre-trial detention cannot be extended when the need for caution disappears; and two, that the need to deprive a person of liberty must provisionally be required to comply with the procedural purposes described above.5

Following this and other cases,6 the Court confirms that pre-trial precautionary measures have purely procedural purposes. Furthermore, the severity of the crime, according to the Inter-American jurisprudence, is not a sufficient argument for the automatic imposition of pre-trial detention. In López Álvarez v. Honduras, the Inter-American Court held that ‘[…] The personal characteristics of the alleged perpetrator and the seriousness of the crime imputed are not, by themselves, sufficient justification for pre-trial detention. Pre-trial detention is a precautionary and non-punitive’ measure.7

A provision such as Article 19 ignores the court’s discretionary power to decide on the most appropriate precautionary measure, according to the circumstances of the case, regardless of the conduct in question.

Courts such as the Brazilian Federal Supreme Court have ruled that legislative, or non-constitutional, provisions that require judges to order pre-trial detention in drug cases violate the presumption of innocence and the right to a fair trial. In the majority opinion, preventing the grant of bail usually means denying the judge the opportunity, in this case, ‘to analyse hypotheses about the need for pre-trial detention in anticipation of it, going against several constitutional provisions.’8

Other international provisions such as Article 9 of the Covenant on Civil and Political Rights, General Comment no. 8 of the Human Rights Committee, or the United Nations Standard Minimum Rules for Non-custodial Measures (The Tokyo Rules) confirm the view that remand is for procedural purposes only and is a measure of last resort to guarantee them. Moreover, it is a measure that should be limited to a reasonable period of time.

Guidelines for the rationalisation of pre-trial detention

The ‘Report on the use of pre-trial detention in the Americas’ (hereinafter, ‘the Report’) issued by the Inter-American Commission on Human Rights in 2013 condensed in one of its sections guidelines for the rationalisation of pre-trial detention in accordance with inter-American and comparative jurisprudence (CIDH 2013: 56–88).

The Report recognises that the presumption of innocence and the principle of exceptionality are the first standard to take into account in legislation, application and determination of pre-trial custody. The Report lists a number of conditions that must be taken into account in making a decision on pre-trial detention.

Among others, there must be legitimate grounds or causes of origin in its application. According to the Inter-American system there are just two grounds: 1) the risk of interfering with the criminal process; and 2) the risk of absconding by the person charged with a crime. Mexico recognises the additional risk to the safety of the victim and witnesses. The Report also refers to the need for judicial review of pre-trial detention, to periodic review, effective legal defence to counter it, and above all, the requirement to use criteria of necessity, reasonableness, and proportionality in determining the imposition of said measure.

In addition, and related to the discussion raised by this article, there are causes of provenance that are unjustified if they are taken into account individually, such as pre-trial detention solely based on the type of crime, the possible sentence to be imposed if the person is convicted, or the existence of a criminal record. All of these elements must be considered together with the particular circumstances of each case.

According to the Inter-American Court, the application of pre-trial detention that only considers the type of offence precludes a test of proportionality That is, if the objective pursued by the measure justifies the restriction of personal liberty and the presumption of innocence. If there is no proportionality, the measure is arbitrary. Therefore, in this line of argument, mandatory remand is arbitrary.

Pre-trial Services: A possible answer to balance liberty and public safety

The systematic abuse of pre-trial detention coupled with the implementation of an adversarial criminal justice system that includes a number of alternative bail measures opened the door to proposals such as pre-trial services (hereinafter, ‘PTS’), based on an American model. In fact, the Report cited in the previous section indicates that the existence of alternatives to remand is vital to its rationalisation. Of equal importance is the existence of mechanisms to make these alternatives operative and rational. (CIDH 2013: 123).

The PTS reflect those mechanisms that the Inter-American Court has They are administrative offices with two main functions. Firstly, ‘procedural risk assessment’ generates accurate and objective information about the specific social conditions of accused persons so that the judge may issue an appropriate release or detention measure accordingly. Secondly, the area of ‘supervision’ monitors compliance with court-imposed bail conditions in order to fulfil the procedural objectives identified (Aguilar and Carrasco 2014).

The evaluation phase informs the monitoring phase because regardless of the bail measure imposed, the evaluation allows the procedural risk level of the accused to be determined. Thus, the stage of supervision designs its monitoring programs on this basis - without losing sight of the obligation that the judicial ruling represents.

The PTS is also an auxiliary service to the administration of justice. In addition to providing reliable information on the appropriate precautionary measure for each person concerned, it allows judges to be certain that their decisions will be properly discharged.

Among other things, these mechanisms also seek to identify whether alternatives to detention meet their objectives, and to elucidate the benefits that proper implementation can have for the life of the accused, his or her family, and society.

PTS also indirectly support the public safety system as the information they generate can identify areas of vulnerability of accused persons that could potentially be offset by prevention policies. For example, which aspects of their social environment, such as labour and education, civic coexistence programs, could be improved? For this to be possible, proper maintenance of records and statistics is essential.

In addition, partnerships between the Services and civil society organizations (CSOs) assist in supervision of defendants. This type of initiative allows citizens to inspect the activities of the pre-trial services offices. PTS, then, can help build trust in, and legitimacy of, the justice system and public safety.

Currently, there are four PTS programs in the country. Three of them have demonstrated positive performance results,9 as reflected in the number of accused persons under supervision who have complied with the bail measures that were imposed on them, as opposed to those declared as having absconded. Figure 1 shows the compliance rates reported by state bail measures, as according to information obtained through the government transparency portals.

Figure 1 

Percentage of compliance with bail conditions by state. Source: Response to access to information petitions serial numbers: 00259114 (Puebla), 00370814 (Morelos) and 141321 (Baja California).

The above graph shows a high level of compliance with bail measures in the states of Morelos and Puebla, while the level of compliance marginally decreased in Baja California.

The PTS model was first implemented in the state of Morelos in the juvenile justice system. The success was such that the state supported the creation of an office within the adult system. Based on this experience, Puebla, Tabasco, and Baja California did the same.

From the beginning, the PTS have provided feedback for legislative and public policy decisions in the criminal justice and public safety systems. In this sense, the PTS aim to identify cost-effective administrative mechanisms to implement bail supervision measures, in contrast to the high cost of keeping a person in custody.

The state spends approximately US$3 million to maintain the prison population per day. Besides the amount spent by the state, roughly US$1.5 million are spent each day by the families of those detained on defence costs and to cover the basic needs of inmates, corruption, and the lost income of days not worked (Zepeda 2013).

At the state level, there is empirical evidence indicating that the deprivation of personal liberty is linked to poverty. Most detainees are accused of minor offences and find themselves in difficult situations (Zepeda 2013). In general, poor people are unable to pay for adequate defence and are therefore at a greater risk of spending a prolonged time in detention, impacting their rights as well as the prison system.

According to information provided by the directors of the Adult PTS Unit of Morelos, the state spends about US$15 per day on a person in pre-trial detention, while supervising a person on bail costs less than US$1 per day.

As has been widely documented, abuse of pre-trial detention negatively impacts social reintegration as it causes overcrowding in prisons, where capacity to care for the needs and rights of persons deprived of liberty decreases significantly. The Inter-American Commission documented that persons deprived of liberty pre-trial are ‘equally exposed to the riots, escapes, violence, drug abuse, murders and the prevailing self-government in prisons’ (CIDH 2013: 31; México Evalúa: 2013: 4–6).

This point is even more relevant considering the administration of bail measures has legislatively depended on state public safety authorities, often through social reintegration systems. Those in charge of the administration of remand centres or jails have become the natural allies of alternative measures to pre-trial detention, as they know they can directly counteract overpopulation rates, and thus enable better control within prisons.

According to the Citizen Observatory of the Justice System (OCSJ) in the state of Morelos - an entity that applies the adversarial system throughout its entire territory - in 2012, 67 per cent of people were remanded in custody pre-trial, while from January to June 2013, the same measure was 64.6 per cent of accused persons subject to criminal proceedings. Similarly, in the district of Mexicali in Baja California, pre-trial detention was ordered for 68.2 per cent of persons charged in 2012; while from January to June 2013, the percentage was 61.65 per cent (OCSJ 2014).

Although this indicator is not the only non-exceptional use of pre-trial detention, it depicts the tendency to request and order pre-trial detention as a general rule. This depends largely on those who prosecute and decide precautionary measures. It is mandatory for judges to meet the guidelines for rationalisation of pre-trial detention in order to facilitate the work of PTS.

Despite the current legislation and practice but thanks to the low rates of failure to appear PTS were specifically regulated by the uniform National Criminal Procedure Code (CNPP). Although the CNPP recognises the importance of creating public policy conducive to pre-trial release, rather than detention, the Code itself fails to offer alternatives and maintains the constitutional pre-trial detention regime.

PTS aim to generate best practices in the management of bail measures. In doing so, they may support, in the long term, relaxing conditions for ordering pre-trial detention and the rationalisation of its use, benefitting the presumption of innocence without neglecting public safety.


Pre-trial detention is the most restrictive precautionary measure within a criminal proceeding. It is likely to result in violations of the human rights of accused persons by unreasonably restricting personal liberty and due process, particularly regarding the presumption of innocence.

Moreover, it provokes social and economic costs for those on remand, their families, and the public. Pre-trial detention should be used in exceptional cases, and it is therefore essential that entities distance themselves from decisions based on arguments of public safety when determining criminal policy. Such arguments made on the grounds of public safety alter the guidelines for the rationalisation of pre-trial detention, such as occurs with the crime catalogue that mandates pre-trial detention.

Therefore, it is the obligation of governments to conduct appropriate assessments that lead to the creation of public policies that respond to the needs of the justice system and public safety of a country - such as pre-trial services - beyond political speeches or electoral interests.