Assessing the Extent of Human Trafficking: Inherent Difficulties and Gradual Progress

This article explores some of the key current research and statistical evidence available on the global scale of trafficking in human beings, and considers the assumption that the occurrence of trafficking is increasing. The value and limitations of this statistical data is identified, as is the relationship between the research base and the resulting legal and policy responses. This allows an assessment of whether there is a connection between the perceived problem and the responses to trafficking victims’ circumstances. It questions whether assumptions, generalisations and policies can be based around the available data and the responsibilities of individual countries, including the UK and the wider international community, in relation to the improvement of data collection. The article also considers signs of progress in terms of data collection and suggests further future improvements that need to be made to the approach taken.


Introduction
This article considers the inherent difficulties of data collection and statistical information on human trafficking. This statistical information shapes and informs, in some part, legal and policy responses to this phenomenon and therefore it is vital that steps are taken to overcome these difficulties and ultimately improve the quality of the data collected. This is not a new problem and is one that has been recognised as being in need of attention. As the data has the potential to result in policies where there is little correlation between the perceived problem and the reality of trafficking victims' circumstances it is an issue that must stay high on the priorities of all countries throughout the world. An overview of the key statistical information gathered, both globally and within the UK will be provided with the aim of highlighting some of the known problems with this data. The use and reliability of the available data is discussed and it is questioned whether assumptions, generalisations and policies can be based around this data with any confidence. The article also considers some of the progress made in the responses aimed at eradicating human trafficking and examines evidence of changing approaches and attempts taken to further improve this data and ultimately the international community's knowledge. The UK is used, as a case study, to demonstrate the difficulties and assess progress taken by one individual country in the attempt to deal with the problem.

The Extent of the Global Problem of Human Trafficking: Statistical Evidence
Human trafficking has been reported to be a major global problem, with the number of victims increasing at a rapid rate. It is acknowledged however, that the precise global scale of human trafficking is uncertain as the nature of the phenomenon makes it notoriously difficult to quantify (Stoecker, 2000). With this reported increase in the occurrence of human trafficking came recognition that action had to be taken at an international level aimed at dealing with the problem. In 2000 the first international consensus was reached on the definition of human trafficking in Article 3 of the United Nations Protocol to Prevent, Suppress and Punish Trafficking in Persons, especially Women and Children (United Nations, 2000b) which was attached to the Convention against Transnational Organised Crime (United Nations, 2000a). Article 3(a) of the Protocol states: "Trafficking in persons" shall mean the recruitment, transportation, transfer, harbouring or receipt of persons, by means of the threat or use of force or other forms of coercion, of abduction, of fraud, of deception, of the abuse of power or of a position of vulnerability or of the giving or receiving of payments or benefits to achieve the consent of a person having control over another person, for the purpose of exploitation. Exploitation shall include, at a minimum, the exploitation of the prostitution of others or other forms of sexual exploitation, forced labour or services, slavery or practices similar to slavery, servitude or the removal of organs. This definition has been adopted and used as the basis for defining trafficking by both the Council of Europe in the Trafficking Convention on Action against Trafficking in 2005 and EU Directive 2011/36 on Preventing and Combating Trafficking in Human Beings and Protecting its Victims. Article 19 of Directive 2011/36/EU stresses the importance of gathering statistics, stating that: Member States shall take the necessary measures to establish national rapporteurs or equivalent mechanisms. The tasks of such mechanisms shall include the carrying out of assessments of trends in trafficking in human beings, the measuring of results of antitrafficking actions, including the gathering of statistics in close cooperation with relevant civil society organisations active in this field, and reporting.
Internationally, there have been attempts at collection of data and statistics aimed at increasing the international communities' understanding and knowledge of human trafficking and to assist in the development of policies and legislation to deal with it. There are statistics from the US Department of State Annual Trafficking Reports that have often been cited within academic literature in an attempt to demonstrate the extent of the problem (for example, Bokhari, 2008;Garrard, 2006;Ray, 2006). These reports contain information gathered from US embassies and consulates throughout the world. Each embassy report is compiled following discussions with the host government, local nongovernmental organizations (NGOs), immigration officials, police, journalists, and victims, in addition to reviews of government, press, and NGO reports. Information is also gathered from the United Nations Children's Fund (UNICEF), the Office of the United Nations High Commissioner for Refugees (UNHCR), the International Organization for Migration (IOM), Human Rights Watch, Amnesty International, the Protection Project, and media reports (US Report, 2001).
The first US human trafficking annual report in 2001 stated that at least 700,000 women and children were trafficked internationally each year. This figure was said to be based on "reliable estimates" but no further information regarding the source of the information was provided. In the 2004 report the figures quoted were between 600,000 and 800,000 people trafficked across international borders, in addition to an unknown number of people trafficked within their own country for the purpose of exploitation. These figures were restated in the 2008 report which also reported that out of the total number of people trafficked internationally 80% are female and up to 50% of the total are thought to be minors. The 2008 figures were based on US Government sponsored research which was published in 2006(US Report, 2008. The first EU level report on statistics in trafficking in human beings was published by Eurostat in 2013 with data covering the period 2008-2010 collected from 27 Member States. The EU Commission has recognised that the issue of collecting comparable and reliable statistics is a key part in the response to tackle human trafficking and it is this that the Eurostat report attempts to provide. However, there are still doubts as to the reliability of the data and warnings that the data contained within the Eurostat report should be used and interpreted with some care. This stems, it is argued, for a number of reasons including both the overreporting of presumed cases of trafficking and the under-reporting of undetected cases (Vogel, 2014). This is in fact something that the report itself acknowledges. Although there are increases in the number of both presumed and identified victims over the 3 year period the report focuses on, it is recognised that this increase could be due to a number of factors, such as new legislation, improved identification processes or a higher priority of the problem within particular countries (Eurostat, 2013). Despite this, there is a strong statement within the report that claims that "the full implementation of the EU Directive, the EU Strategy and Action Orientated Paper provide the necessary legislative and policy tools to end trafficking in human beings". This, it could be argued, is more an aspirational aim rather than a known fact.
The subsequent and most recent Eurostat Report in 2014 covers the period 2010-2012 with data being provided by all 28 Member States and a number of candidate countries and provides information on the number of victims, suspects, prosecutions and convictions recorded by the authorities in each Member State. Although it is claimed that there have been improvements in data collection, it is also acknowledged that this needs further attention due to the fact that not all Member States have provided information on all aspects of human trafficking requested by the EU. This results in data that is still not comparable between Member States and it is important that no generalisations about the extent of human trafficking within the EU are made. Any changes in the reported level of human trafficking remain questionable with varying identification procedures, new legislation and data collection processes still effecting the data collected (Eurostat, 2014). Despite the limitations of the data, the 2014 Eurostat report does provide some detail about victims' gender dependant on the type of exploitation, finding that a majority of women are being sexually exploited compared to men who were the majority of those exploited for their labour. However, Vogel (2014) commenting on the Eurostat 2013 report, finds that the level of under-reporting for male-dominated forms of exploitation is much higher than that where women are the dominate group within a particular form of exploitation.
In terms of the suspected traffickers, these were reported to be predominately male (Eurostat, 2014) a point that the UNODC Global Report (2014) also highlights. The UNODC 2014 report also claims that there has been a noticeable increase in the number of child victims of trafficking, although this is dependent on the region. For example in Africa and the Middle East children have been found to be the majority of victims that have been detected, whereas in Europe and Central Asia there are clearly more adult victims detected. There is little doubt, in the next UNODC Global Report in 2016, that there are going to be further increases in detected cases if, the 9 countries who currently have no legislation to criminalise trafficking and the 18 that only criminalise some forms of exploitation rectify this situation and introduce appropriate legislation to deal with the problem.
There are occasions where trafficking statistics are subsumed within figures of those who are in forced labour. For example, the US annual report of 2010 stated that globally there were 12.3 million adults and children in forced labour, bonded labour and prostitution. This figure was taken from the figures cited by the International Labour Organisation (2009) rather than being based on primary research. Similarly, all forms of modern slavery, which includes trafficking, are reported as one figure, as seen for example, by Australian Walk Free Foundation's 2014 Global Slavery Index. By dealing with the statistics and collating data in this way, it hinders the assessment of the true extent of the trafficking problem because not all those in forced labour or modern slavery are there as a result of trafficking. The Global Slavery Index has been criticised for "a mysterious, inconsistently applied methodology, a raft of unverified assumptions and multiple, critical errors of fact and logic". Gallagher (2014) states that "Even the basic unit of measurement of 'modern slavery' is flawed: the definition is self-created and, bizarrely, changes from one year to the next". This she states "creates an almost irresistible temptation to make a silk purse out of a very tattered sow's ear: to harness the power of statistics and numbers to create an illusion of concreteness that masks the slipperiness of what we are counting". The criticisms highlight the importance of providing clear legal definitions of the various terms involved, for example there are legally defined differences between trafficking, forced labour, modern slavery and smuggling, yet data collected on one may often include another. There is also a problem with the inconsistent use of the definition of trafficking within national contexts which compounds the problem of comparable data collection.
Similar criticisms were made of the US Annual reports which gave details of the number of victims of trafficking who were identified worldwide as well as the number of successful trafficking prosecutions, although it is not entirely clear how these figures were arrived at. 1 In each of the US annual reports the figures were said to be based on "reliable estimates" (for example US Report, 2001). However, the US Government Accountability Office (2006) viewed these figures as "questionable" because of the methodology adopted as the basis of the research. This was due to the fact that the figures were based on one piece of research that was undocumented and unlikely to be capable of being replicated. The discrepancy between the high number of estimated victims and the number of victims officially recorded was also queried. Inaccurate reporting of data may, it is argued, lead to policies that fail to account for all victims in various type of exploitation experienced and policies which are over simplistic (Vogel, 2014). 1 The actual numbers stated in the report were as follows: Adults and children in forced labour, bonded labour, and forced prostitution around the world: 12.3 million; Successful trafficking prosecutions in 2009: 4,166; Successful prosecutions related to forced labour: 335; Victims identified: 49,105; Ratio of convicted offenders to victims identified, as a percentage: 8.5; Ratio of victims identified to estimated victims, as a percentage: 0.4; Countries that have yet to convict a trafficker under laws in compliance with the Palermo Protocol: 62; Countries without laws, policies, or regulations to prevent victims' deportation: 104; Prevalence of trafficking victims in the world: 1.8 per 1,000 inhabitants; Prevalence of trafficking victims in Asia and the Pacific: 3 per 1,000 inhabitants.

Gathering the Data
The difficulty obtaining reliable statistics stems from the fact that human trafficking is predominantly a hidden phenomenon, coupled with the lack of consistent central point for the collection of data nationally, regionally and internationally (Bokhari, 2008). Data has been either unavailable or unreliable (Van Reisen & Stefanovic, 2004). O'Connell Davidson (2006) shared concerns over the discrepancy in relation to the perceived increase in human trafficking. She questioned the scale of the problem and referred, as an example, to the number of women who were discovered following raids in 2003 by the London Metropolitan Police in massage parlours. Out of 295 women, only four or five were identified as victims of trafficking, whilst the rest were found to be in breach of UK immigration law. This does not necessarily indicate that there is very little trafficking. It may instead be an indication that trafficking victims are not being discovered or are being incorrectly identified, which results in the statistics not showing these individuals as trafficking victims due to the wrongly recorded classification of their circumstances. The difficulty with small scale operations and research is the problem of extrapolating any generalisations from the findings. The discrepancies in the figures quoted above can only serve to illustrate the inability to make such generalisations, which do nothing to indicate the true severity of the global problem. Another issue exacerbating the collection of reliable data is the inconsistent use of definitions of trafficking "from country to country and scholar to scholar" (Aromaa, 2007). This results with incomparable data which does not assist in providing a true picture of the extent of the problem of trafficking.
Different bodies and organisations such as the US Government, the International Labour Organisation (ILO), the International Organisation for Migration (IOM) and the United Nations Office on Drugs and Crime (UNODC) all collect information and research trafficking. However, each has a different lens through which to view human trafficking as well as a different geographical interest and this undoubtedly affects the statistics available and the research undertaken. For example, both the US Government and the IOM focus on estimating the global number of victims, whilst the UNODC focuses on tracing the international trafficking routes and the country patterns of international trafficking.
The different mandates of international organisations and national bodies determine what information is collected by each of these bodies. The ILO develops international labour standards which set out the minimum standards of basic labour rights through the use of Conventions and Recommendations (see Cullen, 2007). Its main aims are to promote rights at work, encourage decent employment opportunities, enhance social protection and strengthen dialogue in handling work-related issues (ILO). In relation to trafficking, the ILO has promulgated numerous Conventions which are relevant to both trafficking and forced labour, for example the ILO Worst Forms of Child Labour Convention 1999, The Forced Labour Convention 1930 and the Protocol of 2014 to the Forced Labour Convention 1930. The ILO is also concerned with children's rights in the context of employment and labour relationships. The ILO's International Programme on the Elimination of Child Labour (IPEC,2013) aims to eradicate child labour with a priority given to the worst forms of child labour, which includes trafficking of children as contained in Article 3(a) WFCL Convention (ILO, 1999). There are several ways that IPEC aims to achieve this goal, including through the support of research, campaigning to raise awareness and change attitudes towards child labour. IPEC (2007) also provides encouragement and support to States to enable them to ratify ILO Conventions.
The focus of the International Organisation for Migration is the number of actual victims assisted by the organisation in the countries where they have a presence and concerns itself with trafficking as an issue of migration and migrants' rights. A comprehensive approach is taken to trafficking within the wider context of managing migration and IOM's activities are implemented in partnership with governmental institutions, NGOs and international organizations. Research has been funded by the IOM to examine different elements of trafficking in human beings. For example, Anderson and O'Connell Davidson (2003) conducted research on behalf of the IOM examining whether trafficking was being driven by demand in the sex and domestic work sectors with pilot studies in Sweden, Italy, Thailand and India. The emphasis of IOM's work is on the prevention of trafficking and the protection of victims, and the research specifically considers children within the broader context of migration (Touzenis, 2008).
In contrast, the United Nations Office on Drugs and Crime focuses on drugs, crime and terrorism and all transnational crime including trafficking in human beings. The Trafficking Protocol is directly relevant to the work of the UNODC. The UNODC Global Programme against Trafficking in Human Beings (GPAT) was set up in March 2007 in conjunction with the United Nations Interregional Crime and Justice Research Institute (UNICRI). The overall aims of GPAT are to assist States in understanding the causes and methods of both smuggling and human trafficking and to encourage effective responses. It was also set up with a view to raising awareness, increasing knowledge about trafficking in human beings and it collects and analyses data for this purpose. GPAT aims to strengthen existing partnerships and establish new ones in order to coordinate efforts across Member States (UNODC). In conjunction with this assistance GPAT provides support to States in anti-trafficking efforts, training and provision of improved victim support. They also assist countries in their efforts to combat this crime (UNODC, 2011).
Each of the organisations above gathers data according to their mandate, and this may result in individuals being counted in more than one dataset. The extent of the data collected also varies according to the resources available to each organisation. It is claimed that these varying approaches are hampering the collection of reliable statistics, as Munro (2005) states: Statistics vary widely between the various agencies involved (police, immigration, sex workers and abolitionists, human rights activists and social service providers), reflecting both a pervasive lack of reliable intelligence and a potentially divisive conflict of interests among relevant stakeholders.
The available statistics are therefore held by various organisations that have collected them in different ways, at different times and for different purposes. This means that the statistics are not generally comparable, either within the country the statistics have been gathered from, or between countries (Eurostat, 2014;Salt, 2000).
It may be argued that the reliability of any of the statistics discussed above is therefore dubious, and should be viewed with a degree of caution. Although there has been an increase in the literature on human trafficking, many of the studies undertaken to date lack an extensive research base, both in terms of the type of exploitation researched, the selected group of people (male, female, girls or boys) and the geographical area of interest (Laczko & Gramegna, 2003). Conversely however, it could be argued that the broad range of perspectives actually provides a more realistic view of the problem and assists in furthering current knowledge and understanding.

Examining the Data: Reliability, Methodologies and Mandates
Before considering signs of progress that are emerging in terms of data collection, it is useful to consider what the data discussed above actually tells us about human trafficking. At first glance these statistics suggest that trafficking in human beings is in fact an extensive global problem with the number of victims at alarmingly high levels. There is a general perception that trafficking in human beings has dramatically increased as observed by Javate de Dios (2002): No one now disputes that trafficking today has reached alarming proportions, the magnitude of which affects many countries as countries of origin, transit and destination points. This sentiment is shared by many academics and NGOs and the media has assisted in the creation of this perception of the phenomenon of trafficking in human be-ings. Often the figures cited by the media are merely a reiteration of the statistics discussed above. Various concepts, such as trafficking, smuggling, forced labour and modern slavery are also often confused and inaccurately reported by the media which adds to the confusion of the number of victims involved (Dasgupta & Murthy, 2009). Vogel (2014) questions the media reports of "alarming trends and a predominance of women and children among the victims" and states that the claims being made, cannot in fact be supported by the data and that there is what he terms an "exaggeration presentation policy". The media have been accused of sensationalising trafficking in their reports of various incidents: In the field of human trafficking, sensationalism, stereotypes and assumptions on the profile of victims and circumstances of their exploitation unfortunately often prevail over in-depth investigative journalism. Journalists and editors often cannot tell the difference between illegal migration, prostitution and trafficking in women and are often not too concerned about accurate definitions. The linkages between issues such as globalization, migration policies, marginalization, discrimination as well as economic inequality and exploitation are rarely explored. (UN. GIFT, 2008) On closer examination of the figures cited, there are actually no completely reliable, conclusive and comparable statistics. Even if it is accepted that human trafficking is increasing (Egan, 2008), the precise scale of the problem is still unclear with significant inconsistencies in the reported number of trafficking victims. There are reports of particular incidents of trafficking in human beings (for example, Duncan, 2014;Potts, 2003), although these cases are extremely important, they can only provide an indication rather than indisputable evidence that trafficking is increasing. Askola (2007) regards both estimates of human trafficking and suggestions that trafficking is increasing as nothing more than "guesstimates" due to the clandestine nature of trafficking which results in the lack of reliable data collected. The UNODC identify particular challenges to data collection such as the range of interpretation and approaches taken to the Trafficking Protocol, the result of which is: that countries may not refer to some practices when referring to human trafficking cases, thereby limiting the scope and effectiveness of the response and hindering international cooperation.
This has an impact on data collection which prevents effective implementation of policies and strategies (UNODC, 2012). Similarly where countries have action plans in place to tackle human trafficking there is often little evaluation of the effectiveness of these plans, which are then subsequently updated but without clear evidence to support the developments of these plans (IOM, 2008). Where countries are carrying out evaluations of research or projects, they are not consistently looking for long-term, significant and lasting impacts and are instead more likely to be looking at the more immediate output of each project. The longer term view needs to be developed to enable antitrafficking projects to have a lasting effect and for any good practices to be shared and developed.

Can Any Conclusions Be Drawn from the Data?
The statistical data collected by the various organisations discussed above reveal some shared characteristics. For example, according to all the data collected more women are trafficked than men and sexual exploitation is the primary reason for which women are trafficked. There are however, some significant variations, notably the percentage of trafficking victims who are children with figures ranging from 13%-50% (US Accountability Office, 2006). Some information can be drawn from these various pieces of research. The first is that countries throughout the world are affected by trafficking either as countries of origin, transit or destination, or as a combination of these (Melrose & Barrett, 2006). Secondly, there is some indication that trafficking in human beings is increasing, with some countries reporting an increase in the numbers of suspected cases of trafficking (Eurostat, 2014;Laczko & Gramegna (2003). 2 This does not mean that all trafficking in every country has necessarily increased or increased at the same rate and these reports are only statistically relevant for the individual country concerned. It may also mean that the effectiveness of reporting of trafficking has improved as awareness of the problem has increased.
The reported growth in trafficking in human beings has resulted in a higher priority on introducing policies to deal with the issue. As stated earlier, data relating to trafficking and smuggling are often combined (Laczko & Gramegna, 2003), despite the fact that these are separate concepts which require different legal responses. This contributes to the distortion of the true extent of the trafficking problem. It has been claimed that the difficulty in providing accurate reliable statistics has the potential to impact negatively on the policy response to trafficking in human beings. Dottridge (2003) argues that inaccurate estimates are likely to lead to the proposal of inappropriate remedies and responses. Responding to trafficking of an estimated 10,000 people would require a very different response where the estimated number of victims is 100,000. This is a view supported by Tyldum and Brunovskis (2005) who suggest that the overestimation of the extent of the problem may have equally negative consequences as underestimation. They consider the uncritical use of research findings as potentially leading to mis-information which in turn would "…hinder the creation of relevant policies and appropriate programmes". They argue for the need for research to be based on clear conceptual and practical identification of the target group, accompanied by indications of the appropriate end users of the data. They also regard misleading data as worse than no data at all. Encouraging and supporting a more robust system of data collection resulting in more clearly defined and "useable" data is understandable. It is however, questionable whether having no data at all would be preferable to misleading data.

An Example of Data Collection in the UK: Difficulties, Limitations and Progress
An example of the difficulties and limitations in data collection discussed above can be seen in the approach taken by the UK. Small scale operations on a national basis within particular geographical areas focusing on particular "types" of trafficking are often carried out and statistics are collected without any reference to other countries or even a consistent specific focus on all individual groups of victims. Even with the overall lack of reliable statistical information, it has been claimed that "…there is sufficient evidence to indicate that, at an absolute minimum, hundreds of people are being trafficked into the country [UK] for sexual or labour exploitation each year" (Skrivankova, 2006). This statement is not supported with reference to identifiable research and therefore its validity is still questionable.
In the UK, there has been a combination of police operations and an attempt at data collection in order to increase the knowledge of trafficking into and within the country. The United Kingdom Human Trafficking Centre (UKHTC), established in October 2006, is central to this aim and has been involved with police operations such as Pentameter 1 in 2006 and 2 in 2007. These were nationwide operations involving 55 police forces across the UK aimed at assessing the extent of trafficking, rescuing and protecting victims, and prosecuting traffickers. The ultimate aim was to tackle the problem of trafficking for the purpose of sexual exploitation. Pentameter 1 resulted in the discovery of 84 victims of trafficking who had been sexually exploited, including twelve children, with victims originating from twenty different countries. This operation involved visits to approximately 10% of the estimated number of sex establishments in the UK. As a result of this operation there were a total of 232 arrests, which led to 134 people being charged with a variety of offences (Gloucestershire Constabulary, 2006). The following year Pentameter 2 resulted in 528 criminals being arrested with 164 victims of trafficking for sexual exploitation reportedly discovered, which again included thirteen children. Assets of over £500,000 were confiscated and over 822 premises were visited (Police Oracle, 2008). Although both these operations appear to be very successful in the number of criminals arrested and victims discovered, there has been some criticism of the reported success of the operations, compared with their actual success. For example, Gilbert and Moore (2010) identified that out of the initial reported 528 arrested during Pentameter 2, only 93 were arrested on suspicion of trafficking and, of those, only 67 were charged, resulting in fifteen individuals (ten males, five females) being convicted. The initial reports present the results of the operations in extremely positive terms which are misleading because they do not actually mirror the reality of the actual conviction rates for trafficking offences.
The UKHTC has also produced Statistical Quarterly Reports covering the period from 1 st April 2008 to 31 st March 2009 to assist in extending the existing knowledge of human trafficking into and within the UK (UKHTC, 2009a). The data provided on the number of identified victims is disaggregated, detailing victims' nationality, gender, age, number of victims of EU or non-EU origin and the type of exploitation for which they were trafficked. There is similar, but not as extensive disaggregation in relation to the defendants in prosecution of trafficking offences. Details include the number of defendants dealt with under sections 57-59 Sexual Offences Act 2003, which concerns the trafficking of individuals for the purpose of sexual exploitation, on the gender of the defendants and the eventual outcome of the cases. Due to the nature of the phenomenon, coupled with the fairly newly developed data gathering mechanism, the collection of this data from a number of different sources occurred in a fairly unsystematic manner.
The UK ratified the Council of Europe Convention against Trafficking in Human Beings 2005 and one of the practical effects of the UK's ratification was the introduction from 1 st April 2009 of a National Referral Mechanism for the identification of victims in the UK, as required by Article 10 of the Convention (UKHTC, 2009b). Each party to the Convention is required to establish a "competent authority" with the responsibility for the identification of trafficking victims and to manage all the information relating to human trafficking. In the UK this responsibility is shared between the UKHTC, which deals with referrals from the police, local authorities and NGOs and the Home Office Immigration and Visas, dealing with referrals identified as part of the immigration process for example where trafficking may be an issue as part of an asylum claim. Information concerning confirmed victims of trafficking is reported to UKHTC, resulting in a central point where more robust and systematic data is collected. From April 2009 this data was reported in National Referral Mechanism Reports (NRM) rather than the previous Statistical Quarterly Reports.
During the first year of data collection under the new NRM system there was more than a 200% increase in the number of potential victims identified when compared with the Statistical Quarterly Reports for the previous year. Of these 520 were female, compared to 186 males, and over three times as many individuals from non-EU countries as EU countries. The first annual NRM report in March 2010 also saw more than a four-fold increase in the number of referrals made for those exploited in domestic servitude (106) as compared to the total number for 2008-2009 (23). In the Annual NRM report in 2012 there was a reported 1186 potential victims referred which represented a 25% increase in the number referred in 2011. In the most recent annual NRM Report (2013) there was a reported sharp increase in the number with 1,746 potential victims referred to the NRM which represented a 47 per cent increase on 2012. The UK National Crime Agency also carries out a Strategic Assessment of the extent of human trafficking and found that there were 2744 potential victims during 2013, 3 and Silverman (2014), attempting to estimate the extent of modern slavery in the UK in 2013, found there were between 10,000 and 13,000 potential victims. To obtain this figure he carried out an exploratory analysis of the Strategic Assessment data using a statistical technique of Multiple Systems Estimation, but does acknowledge the data should be viewed cautiously due to the limitations involved in the method used.
The apparent increases in the number of victims could be explained by the introduction of a more effective and robust referral system, with referrals being received from key bodies and agencies such as Police Services, Local Authorities and NGOs. Clarity over the responsibility for referrals and data collection will undoubtedly lead to increased reporting. It could therefore be argued that this is a fairly rapid practical effect of the ratification of the Council of Europe Convention, rather than an indisputable rise in trafficking into and within the UK. Longitudinal data collection should provide a more detailed picture, although it must be remembered that these are still only the cases where victims have been identified. There are undoubtedly many more adults and children who remain undetected. This is particularly true due to the hidden nature of trafficking.
Data collected by the UKHTC specifically relating to child victims revealed some significant information. Child referrals increased from eighteen in 2008-2009 to 179 in 2009-2010 with children representing 25.4% of the total number of referrals. This increase may once again be the result of a clearer reporting and referral mechanism. Alternatively it may be due to an increased awareness of the various forms of child trafficking rather than an actual rise in the number of child victims. The division between male and female victims, as well as adults and children have been fairly consistent throughout the reporting periods 2010-2013 with approximately 67% of victims being female, 33% male and of the total number of victims, the percentage of children under the age of 18 ranged between 25% and 31%.
Prior to May 2009 and the introduction of the NRM, child victims were grouped into one category, of children aged between ten and seventeen, with a total of eighteen children being recorded as victims of trafficking between April 2008 and March 2009. In contrast in the first year under the NRM (April 2009 to March 2010) the age categories of children were disaggregated into sub-categories. There were 95 children identified aged 16-17, whilst 61 were aged between 12-15. There were six children aged between ten and eleven and seventeen under the age of ten. Of these children, 60% were female and 40% were male. Again in the 2013 NRM statistics, the majority of children referred were between the age of 16-17 (57) with 34 between 12-15 with a nearly identical split between the gender of girls (59%) and boys (41%) as the earlier reports. Similarities in the type of exploitation experienced by these children were also identifiable in the 2010 and 2013 NRM. This indicates that there are some patterns emerging and the disaggregation of the statistics is starting to provide a clearer picture of the extent of the problem for UK authorities.
In relation to the implementation of UK trafficking legislation, the statistics reveal that the Sexual Offences Act 2003 is used on a much more frequent basis than section 4, Asylum and Immigration (Treatment of Claimants, etc.) Act 2004 with more prosecutions for trafficking which results in individuals being sexually exploited than in forced labour. For the period 2009-2011, the UK Government Report (2012) stated that there had been 311 prosecutions and 41 convictions for trafficking for sexual exploitation, in comparison to only 78 prosecutions and 8 convictions for labour exploitation. This difference may be an indication that more attention is given to the prosecution of those involved in sexually exploiting victims. Alternatively it may be as a result of a lack of sufficient evidence or it being more difficult to prove that trafficking has taken place where the type of exploitation involves other forms of exploitation. In these circumstances defendants may be charged with other offences, such as assisting unlawful immigration or holding someone in forced labour, which may not involve the same difficulty in establishing the necessary proof required as trafficking.
Although there is a clear disparity between the convictions for sexual exploitation and labour exploitation, there has been gradually more attention being given to trafficking which results in exploitation of an individual's labour. In November 2008, Operation Ruby, which involved the East Midlands Foreign National Crime Team, Northamptonshire Police, the UK Human Trafficking Centre, the UK Border Agency (since abolished), the Serious Organised Crime Agency (now part of the National Crime Agency) and the Gangmasters Licensing Authority, resulted in over 60 people aged 15-67 being found picking leeks in a field in Holbeach, Lincolnshire. Eight people were subsequently arrested on suspicion of trafficking (Williams, 2008). Similarly, a husband and wife were charged with trafficking adult workers into the UK and exploiting their labour working in two care homes in Worthing (Worthing Herald, 2009a). They were charged under section 4, Asylum and Immigration Act 2004 and sentenced in July 2009 to two years' imprisonment and ordered to pay £25,000 in legal costs (Worthing Herald, 2009b). The individuals exploited in the care homes were not permitted to see a doctor whilst working for the couple and their illegal status in the UK was employed as a deterrent for them to go to the police to report their exploitative situation.
Operation Ruby was clearly focused on organised crime. Similarly, the trafficking and exploitation of individuals in care homes involved an element of organised crime because employment agencies made the referrals. Although there is some progress being made in tackling labour exploitation and similar operations have taken place, the focus still remains primarily on organised crime. This is not only evident through Operation Ruby, but other Police Operations in the UK such as Operation Maxim (2008), Operation Glover (2009), Operation Acumen (2010, and Operation Pentameter 1 and 2. This results in the neglect of reporting in the statistics the more informal trafficking arrangements, where individuals who are not part of organised criminal groups are responsible for trafficking. The situations described above involved exploitation of individuals in public settings rather than private households or hidden environments and although the exploitation was still carried out in a clandestine way, did require individuals to be in public view whilst being exploited. This is not the case for all trafficking victims where they can be hidden out of sight of the public eye. This exacerbates the problem of obtaining an accurate perspective, of the various forms that trafficking takes, through statistics alone. The collection of data by the UKHTC has provided some insight into the various forms of trafficking. The National Referral Mechanism (SOCA, 2012) appears to be having a positive effect on the collection of data relating to trafficking victims within the UK. There are some elements that could further improve the data. The current categories of exploitation could benefit from further disaggregation, for example, the category of forced labour could detail the specific exploitation experienced such as agricultural, factory work or cannabis cultivation. Details of locations within the UK where victims are discovered would assist in targeting particular forms of trafficking in specific geographical areas. Data detailing the nationality of defendants could be collected in order to assess whether there is a direct link between nationalities of victims and traffickers. This in turn may provide opportunities to identify particular countries of origin and facilitate more opportunities than are currently available for effective cooperation in preventative measures. If given time to develop, the National Referral Mechanism and the corresponding reports have the potential to provide a much more detailed picture of the current trafficking problem in the UK and eventually identify more concrete trends, information and knowledge.

Signs of Progress and Future Steps to Improve Data Collection
Some attempts are being made to improve data collection such as the change of approach which is evident in the 2014 US Report. It moves away from providing generalised figures to focusing on individuals and their experiences and provides extensive useful insights that statistics alone cannot provide. Apart from the US annual reports, the International Labour Organisation's (ILO) figures are the other statistics frequently cited. In relation to trafficking, the ILO stated in 2005 that 12.3 million people are in forced labour throughout the world including, at a minimum, 2.45 million people trafficked into forced labour each year. 4 In the latest ILO report (2012b) those in forced labour, which includes those trafficked into this situation was 20.9 million, however unlike the 2005 report there is no separate breakdown for those who are in forced labour as a result of trafficking. It is important to note at this point that the ILO 2005 and 2012 statistics are not comparable due to the difference in methodology used to collect the data, this is something the ILO themselves recognise. There is however, in some cases, more evidence of and attempts to collect data which is disaggregated in terms of both victims and traffickers by gender, age and form of exploitation as well as citizenship and in some cases the type of assistance and protection provided to victims (for example, Eurostat,4 The figure of 2.45 million is the number of those in forced labour as a result of trafficking as reported by ILO (2005ILO ( ). 2014). This level of disaggregation does need to be consistently applied across all data collection in order for a more holistic picture of human trafficking to emerge. The UNODC (2014) sought and received advice from a number of experts with the ultimate aim of creating "a methodologically sound estimate of the global number of trafficking victims", which it was concluded would require "significant resources and a long-term perspective". In addition, the experts encouraged UNODC to …try to take advantage of existing data collection vehicles by making efforts to have relevant trafficking in persons-related questions included. This is particularly relevant for industrialized countries as these often carry out various national surveys with some regularity. There are also some United Nations-led surveys that could be used similarly.
The ILO also aims to produce the most up to date statistical information as part of the ILO strategy 2012-2015. This involves setting up a Global Slavery Observatory which would act as "a clearing house" for all data on forced labour, slavery and trafficking. In turn this could, argue the ILO (2012a), be used to "inform investment decisions and to measure the impact of action against forced labour by monitoring the change in prevalence across countries and regions" and would "…enable evidence-based policy making at country level".
In terms of signs of progress within the EU, this in some part can be seen with the Strategy towards the Eradication of Trafficking in Human Beings (2012) covering the period 2012-2016 which was adopted by the Commission 19 June 2012. A particular aim is for the development of an EU wide system for data collection and publication of this data with sufficient disaggregation. As highlighted in the Eurostat (2014) report, the EU Commission will work together with the Member States of the EU as well as national rapporteurs or equivalent mechanisms with the aim of the collection of comparable and reliable data.
In terms of the UK, the government have recently published the UK Modern Slavery Strategy (2014) which sets out the approach the UK will take over the coming years to ultimately significantly reduce the rate of modern slavery in the UK and to strengthen international collaboration. The Strategy has four main parts: (1) Prosecuting and disrupting individuals and groups responsible for modern slavery (Pursue); (2) Preventing people from engaging in modern slavery crime (Prevent); (3) Strengthening safeguards against modern slavery by protecting vulnerable people from exploitation and increasing awareness and resilience against this crime (Protect); (4) Reducing the harm caused by modern slavery through improved victim identification and enhanced support and protection (Prepare).
In terms of data collection, the NRM remains the method the UK will continue to use and the UKHTC as a competent authority will coordinate the collection of the data available. As well as the Modern Slavery Strategy, the UK currently has the Modern Slavery Bill 2014-2015 progressing through Parliament which, if enacted, will consolidate the UK's current legislation relating to trafficking and slavery offences; create two new civil orders to prevent modern slavery and make provision for the protection of modern slavery victims.
All these indications of progress are promising signs that the issue of human trafficking will remain high on the political agenda; however it is not guaranteed to ultimately improve the quality and comparability of data collected. With the ILO, UNODC, the EU and national authorities all intending to collect data on the phenomenon, the disparate mandates and aims of this data collection still remain, as discussed earlier. If improvements are to take place and a concerted effort made to ultimately provide truly comparable data, there needs to be discussions taking place with all key actors at all levels. This discussion needs to find agreed parameters of what data is being collected, with consistently agreed and applied definitions of each of the terms involved as well as a uniform methodology being used at each level of data collection. Currently with each of these bodies intending to collect data in the future, there is still the problem that this could lead to sets of data which include instances of victims being double-counted or trafficking being still being subsumed into forced labour data, which will result in similarly flawed data as is currently available.

Conclusions
This article has illustrated that the generalisations made about the extent of human trafficking are unconvincing. There is no clear evidence that determines the extent of the problem, but this is unsurprising given the inherently hidden nature of trafficking in human beings. Despite the problems with data collection, there are attempts being made to improve the process, reliability and comparability of the data, however more needs to be done.
The definition contained within the Trafficking Protocol and the various terms contained within it need clarification and consistent application, as do the indicators used to collect data. There needs to be detailed descriptions of how the data is gathered and any limitations of the method employed needs to be acknowledged and clearly explained in order for more effective policies to be implemented. Key actors at all levels must work towards creating clear parameters and levels of disaggregation for data collection and only by working together and being brutally honest about the limitations of the data can a more effective system be developed which is consistently applied in all countries throughout the world. Although it may take some time to reach a consensus and apply it in practice, it may eventually result in data which is internationally comparable, which in turn would allow more appropriate legal and policy responses to reflect the trends exposed within each country. If a system of data collection is implemented correctly this may also reveal a more diverse set of trafficking situations and implementation of policies that more accurately reflect the varied experience of all trafficking victims. Coupled with this, is the need to improve the monitoring and evaluation of both data collection and the impact of responses taken, to enable countries to know whether the policies and activities they are undertaking are actually effective. By working closely together to strive to develop one methodologically sound approach to data collection it may enhance the opportunities to share best practices within the international community. Any country that has not already done so should ensure that legislation is in place to deal with all variations of trafficking that can occur, as doing so will ultimately improve the possibility of collecting comparable data.
Each Member State of the EU should guarantee that the obligations under the Council of Europe Convention are being met to ensure a central monitoring unit exists to record all incidences of trafficking in human beings in each country. Collection of data at regional, national and international level needs to be systematically collected with one centralised body at each level taking responsibility. Without accurate statistics, disaggregated by age, gender, country of origin, nationality, type of exploitation, etc., policies are formed which are not necessarily going to most effectively deal with this problem and capture the full range of the experiences of trafficking victims. If managed effectively and sufficient disaggregation of the information is undertaken, this is an opportunity to provide in the longer-term, a more accurate picture of human trafficking. At UK level, the activities of the National Referral Mechanism and its data collection role has the potential to develop understanding of human trafficking, identifying trends and patterns which over time could assist in the creation of better informed policy-making. Human trafficking is such a complex phenomenon involving so many different elements and forms of exploitation that it is unrealistic to expect that statistics alone can ever completely and accurately inform policy responses. What is important is that data continues to be collected to assist in building the knowledge base and providing as detailed a picture as possible. Only when a consensus is reached, by the International community, in terms of exactly what data should be collected and the extent of disaggregation necessary, with each country taking the same unified and methodologically sound approach to data collection, can a true comparison be made between the various countries throughout the world to develop a more targeted and cohesive international response to human trafficking.