Abstract
Because both corruption and trafficking are very complex phenomena, a multivariate approach was used to analyse the relationship between corruption and trafficking. The results of the analysis show that, in Brazil, corruption is a causal factor in human trafficking. The paper offers a number of suggestions to help researchers and policy makers better understand where, how and when corruption facilitates trafficking in human beings in order to combat both crimes more effectively.
Keywords
Correlation, Corruption, Trafficking in Human Beings (THB), Triangulation.
The technological advances of the past century have brought about unprecedented levels of organized crime (Savona and William 1996). In a world no longer restricted by geography, transnational crime is on the rise. With the advent of new and more sophisticated technologies, the reach of criminal networks has expanded while domestic legal systems have struggled to keep pace (ibid.). One of the key types of organized crime that researchers have been focusing on is trafficking in human beings (THB) (European Commission 2001). According to a research report sponsored by the United States government, every year some 800,000 people are trafficked across national borders and several million more are trafficked within countries (US Department of State 2008). This makes THB one of the most lucrative crimes as well as one of the most difficult to combat.
Social scientists and experts such as Lyday (2001) and Bales (2005) have labelled this type of criminality the ‘new slave trade' because it commodifies human beings into mere merchandise that is being transported across borders, traded for currency and recycled through the domestic economy (see Hughes 1999). Whereas other sorts of contraband are easily identified and have proven methods of detection (e.g. dogs trained to sniff out drugs and explosives), THB is much more difficult to identify and prosecute (see Ribando 2005). In addition, the exponential rise in global traffic has added pressure on national borders that are already swamped by unprecedented levels of trade and human movement.
Criminality is also becoming increasingly endemic among government officials. As noted in the United States Victims of Trafficking and Violence Protection Act of 2000 (section 102(b)(16)): ‘In some countries, enforcement against traffickers is also hindered by official indifference, by corruption, and sometimes even by official participation in trafficking.' Worse still is the involvement of judges and politicians in the facilitation of sexual services provided by trafficked victims (see CPMI 2004). When different agents of the criminal justice system participate in this type of criminality, it undermines the basic structure of government and the moral responsibility of its officials to uphold and protect the integrity of the law. According to Lyday (2001: 8): ‘few other issues so well reflect not only government corruption, but the larger issues of poverty and the low status of women and children.' As such, trafficking has its roots in the much broader conditions of state failure.
In Brazil, both corruption and human trafficking have increased greatly over the past few years (PESTRAF 2002; Ribando 2005) owing to the high rates of poverty, income differentials, illiteracy, gendered cultural practices, discrimination and homelessness, all of which have been described by Bales (2005) as critical ‘push' factors. Focusing on these elements, this article provides an analysis showing the linkage between corruption and THB in Brazil, which is considered a source country for trafficked victims. The 2006 report by the United Nations Office on Drugs and Crime (UNODC), Trafficking in Persons: Global Patterns, describes Brazil as the third-largest source country of THB in the western hemisphere, after Mexico and Colombia. Tsutsumi and Honda (2005: 35) classify Brazil as a ‘leading source country of victims moved to Europe on the Entertainer visa, and ... leads by far the number of Entertainer visas issued to Japan'. Based on these findings and other available scientific evidence (see Lezertua 2003), the hypothesis of this study is that THB is dependent on corruption in Brazil.
Confronting the lack of reliable data
THB is difficult to classify because it encompasses a range of crimes, from the falsification of documents to the sexual exploitation of trafficked victims. For this reason, THB is commonly referred to as an ‘irregular phenomenon' - a criminal process that is in a constant state of change, making it difficult to recognize and measure (see Laczko 2005; Tyldum and Brunovskis 2005). Furthermore, it is difficult to determine the ways in which other criminal activities mask THB. The inability to distinguish THB from its associated criminal activities not only inhibits the process of data-gathering but also distracts law enforcement, social organizations, non-governmental organizations (NGOs) and the criminal justice system from accurately focusing their attention on those areas where it would be most effective (CPMI 2004). For example, in underdeveloped countries, where crime rates tend to be higher and resources are scarce, gathering sound data is especially difficult. As Mahmoud and Trebesch (2009: 3) note: ‘despite frequent calls for empirical research on the topic THB, there is still very little knowledge on human trafficking as an economic phenomenon. One major reason for this is the great difficulty in gathering reliable and representative data.'
Although the limitations of data-gathering vary from country to country, basic controversies remain over how to define and classify reported crimes (see Laczko 2005).
Criminal activities associated with THB may not be included in THB case files as a result of differences in classification or of misidentification owing to the failure to complete the full investigative process into crimes. This is usually because they are deemed less significant on their own, even though they may be connected to a larger criminal enterprise. In many case files, THB is identified only at the end of the investigative process, especially when sexual offences are the primary focus. Any single case involving sexual abuse or sexual exploitation may be a manifestation of THB (Laczko 2005).
Considering the potential for one individual criminal offence to betray the occurrence of another, law enforcement agents must be well trained to identify the clues that point to THB. For this reason, a database exclusively on THB is insufficient for identifying the real scope of the phenomenon. Without an adequate database in which all potential related factors can be isolated, identifying every instance of THB is almost impossible, especially when only isolated sets of data are available. Laczko correctly noted (2005: 15): ‘fragmentary datasets cannot be collated into national figures or compared at international level.' In Brazil, the problem of data collection is significant. The core issue for Brazil is the lack of a clear, nationwide system for data collection. The absence of a unified and efficient national database has a direct impact on the ability of governments at every level to formulate specific public policies targeting THB. The primary explanation for this is budgetary constraints.
Beyond the technical difficulties of collecting data, Brazil, as South America's most geographically extensive country and with over 184 million inhabitants, faces a multiplicity of ideological, political, moral and cultural issues, many of which are regionally defined. Differences in regional attitudes influence both the way data are recorded and the way data are evaluated. Likewise, cultural factors play a significant role in the way data are gathered and collated (CPMI 2004). When sexual offences are involved, especially those involving public officials, a level of discretion is observed that tends to hinder the recording of cases. These issues are relevant to understanding THB in Brazil, as well as the corrupt practices that enable it.
Data on corruption itself are especially difficult to obtain because no official database exists and no attempt has been made to distinguish and classify the different types of corrupt practices involved. Lambsdorff observed that determining precise definitions of the different types of corruption-related crimes in different countries is very difficult since ‘the statistical methodology of counting and aggregating used in each national agency can differ considerably from that used elsewhere' (2001: 1). In Brazil, researchers attempting to compare the data recorded in different regions or federal states have discovered how difficult it is to perform even interregional comparisons. For example, to compare conviction rates among states would require a lengthy period of field research - an endeavour with uncertain results because records are likely to reflect not actual rates of corrupt practices but rather the relative efficiency of various public agencies (Ades and Di Tella 1997).
Recent literature on the link between THB and corruption
Two studies on the relationship between THB and corruption have been conducted in recent years, one by Bales (2005) and the other by Lyday (2001). Both studies used a cross-national approach. Lyday correlated the 2001 Corruption Perceptions Index (CPI) of Transparency International (TI)1 with tiers that classify countries according to their compliance with certain anti-trafficking standards. He divided the CPI into three groups and then compared the CPI groups with the three tier categories established by the US Department of State's Trafficking in Persons Report. Lyday's analysis revealed a strong relationship between perceptions of corruption and the level of government response to THB. He then established a positive relationship between corruption and THB at the cross-country level.
Meanwhile, drawing on his field research data, Bales (2005) correlated TI's CPI index with estimations of THB. Bales performed tests using indicators for poverty, corruption, sexism and other variables to determine which might be the causal factors (explanatory variables) in human trafficking. The multivariate analysis performed by Bales showed that government corruption was the most important factor predicting trafficking from a source country. Although each study's exploration of the interrelationship is limited by their reliance on secondary data, I incorporate Lyday's cross-country level of analysis and Bales's statistical method of analysis to form the theoretical and analytical bases for this article. The main purpose is to demonstrate that THB in Brazil depends on corruption and, hence, both criminal phenomena are correlated.
For the purposes of this article, the term ‘trafficking in human beings' (THB)2 will follow the definition described in Article 3(a) of the United Nations Protocol to Prevent, Suppress and Punish Trafficking in Persons, especially Women and Children (2000). The definition of corruption follows the 2003 United Nations Convention against Corruption (Article 15) and the broader definition laid out by Transparency International (‘abuse of public office for private gain').3 Therefore, in this article, corruption refers only to official corruption, which follows the definition established by Article 2 of the Convention.
Data sources and methodology
In order to follow a multi-modal approach, this study used five different sources of data. The first source is a survey I conducted in 2006 among law enforcement agents and federal prosecutors in Brazil on the interrelationship between THB and corruption. The second source is the report of a Joint Parliamentary Commission of Inquiry in Brazil - the Comissão Parlamentar Mista de Inquérito (CPMI) report of 2004. The third source is Transparency International's Corruption Perceptions Index (CPI), and the fourth source is data from the Brazilian Federal Police Official Statistics between January 1990 and March 2006. Finally, the fifth source is a report on Municipal Corruption by the Brazilian federal government between 2003 and 2005.
The concepts of THB and corruption were operationalized so as to enable an examination of their possible interrelationship. In the survey, following the TI's CPI index, the operationalization procedure applied was measurement. In the CPMI report, the Federal Police Official Statistics and the Municipal Corruption report, the operationalization procedure applied was counting, since the property of the operationalized concept is discrete (unique and expresses individual incidents of THB). In the survey, THB-related corruption was measured via survey respondents' perceptions of the relationship between both variables. In the CPMI report, THB was quantified by the number of cases/occurrences involving domestic and international THB of female minors; corruption, in the same data set, was quantified by the number of incidents of THB-related corruption found in THB case files. In the Federal Police Official Statistics, THB was quantified by the number of cases/occurrences of international THB. In the report on Municipal Corruption, corruption was quantified by the number of incidents of corrupt behaviour by public officials, on duty, in municipalities (fraud related to procurements, simulated payments, etc.).
The variables were measured and described in both absolute and relative numbers. The initial units of analysis were individual incidents of THB, corruption and individual perceptions of THB-related corruption. The data were then grouped by federal state and region for comparability. The data had a cross-sectional character (27 Brazilian federal states) and were analysed using descriptive statistics as well as multivariate techniques. Each individual data set, together with its methodology, is explained below.
Survey on perceptions of the linkage between THB and corruption
The survey was conducted in the form of a self-administered questionnaire, which was distributed via e-mail. All responses were received between June and November 2006. After the questionnaire was sent to the pool of potential respondents, follow-up reminders were sent via e-mail and telephone. Consistent with observations by Heidenheimer (2004), the survey was designed to explore perceptions of corruption among law enforcement and court officials. The questions were intended to provide an insight into the respondent's knowledge and opinions of the present state of THB and corruption in Brazil. The questions sought opinions, evaluations, judgements and sensitivity to criminal phenomena in general, and were designed to assess the perceived interrelationship between THB and corruption among respondents.
Out of a potential pool of approximately 1000 state employees, only 109 (10.9 percent) completed and submitted a questionnaire and, of those surveys completed, 36 respondents were subsequently dismissed as ineligible owing to contradictions in their responses, leaving the final number of respondents at only 73. The final sample consisted of 59 criminal prosecutors, 10 federal police officers, 2 members of government institutions, 1 expert from an NGO, and 1 judge. Although the sample is small, the respondents represent a select group with informed knowledge of THB. Their responses reflect considerable experience with investigations, interrogations, victims' statements and evidence disclosed during trials. In addition, almost half (47 percent) had worked directly with THB cases.
The CPMI report
The CPMI report is a set of cross-sectional and longitudinal data that include information on incidents of domestic and international trafficking in and from Brazil. The report originated from an investigation conducted by a Joint Parliamentary Commission of Inquiry whose purpose was to identify cases of sexual abuse and sexual exploitation of minors throughout Brazil. The Commission analysed cases under investigation, as well as ongoing court proceedings, in order to better understand the criminal justice system's response mechanisms. Between May 2003 and July 2004, the Commission investigated 561 case files of sexual offences and analysed factors related to the environment in which these cases occurred. All of the cases referred to crimes committed between 1994 and 2003.
The Commission identified a number of root causes of the sexual exploitation of minors in Brazil, tracing the effects of globalization, social exclusion, a history of physical or sexual abuse, lack of familial support and/or strong social networks, vulnerability associated with age, especially vis-à-vis culturally rooted practices, and the negligence of those responsible for the protection of minors, among many other factors, to their end result in ‘modern day slavery'. These findings reflect those of several other foreign studies (see Phinney 2001; Hughes 1999) that THB is rooted in poverty and poverty-related phenomena.
In an effort to better address the situation in Brazil, the parliamentary commissioners highlighted several key factors specific to Brazil that have a direct impact on the data-gathering process, especially when sexual offences are involved. The major factors identified ranged from an institutional bias against victims of sexual offences to a lack of institutional concern towards THB and sex-related crimes, with particular concern for the involvement of public officials and persons of influence (e.g. business persons) in the trafficking process.
The CPMI report places special emphasis on the impunity of sexual abusers and exploiters, mainly because of the involvement of public officials and persons of influence in such crimes (Figueiredo and Hazeu 2006). The low likelihood of indictments being sought against public officials makes law enforcement and criminal justice agents as reluctant to investigate sex crimes as victims are to report them. Similarly, even when a case is presented, prosecutors are unlikely to achieve convictions owing to corruption within the courts. In fact, an indictment does not guarantee a conviction, and conviction itself does not guarantee a formal sentence. As mandated by due process, many judicial appeals are possible. Furthermore, people of influence, who possess the means to either exhaust the entire appeals process or prolong it until an amenable solution is found, can unduly exploit the system to postpone punishment, sometimes indefinitely (see CPMI 2004).
Therefore, one of the objectives of the study was to re-analyse the data of the CPMI report and to isolate THB case files from the data set in order to test the possible linkage between THB and corruption in Brazil.
Federal Police Official Statistics and TI's CPI
Unlike the CPMI report, which deals with only the THB of under-age females, the Federal Police statistics cover the international trafficking of all women, regardless of age. This data set covers all Brazilian federal states and lists the annual number of investigations in each state between January 1990 and March 2006. The total number of incidents investigated within this period (480) refers to the number of investigations brought to the attention of prosecutors. It does not, however, include the actual number of cases reported or the number of investigations where a ‘positive conclusion', such as a conviction, was reached.
The CPI was used to compare CPI data with the data provided by the Federal Police in order to test the general relationship between both concepts (THB and corruption) longitudinally.
Corruption in municipalities
The Municipal Corruption report consists of cross-sectional and longitudinal data, in which multiple incidents of corrupt behaviour were observed in each federal state between 2003 and 2005. This report was obtained from the federal government anti-corruption programme initiated by Brazil's Office of the Comptroller General in April 2003 (Controladoria-Geral da União, CGU). Employing the random auditing of municipal government expenditures, the programme had two primary objectives: to use the threat of future audits to discourage public administrators from abusing public funds, and to ensure the transparency of public expending, thereby encouraging an active interest among the general population in fiscal/financial matters (see Klitgaard et al. 1998). The programme seeks to inhibit corrupt practices, to strengthen institutions and to engage the public in the monitoring of the government's decision-making process, as well as its allocation of public funds. By restricting opportunities for corrupt practices, the programme seeks to ensure the regular and appropriate distribution of public funds.
The data set covers the period from 3 April 2003 to 27 September 2005, and provides data from 921 audits. The data for 2003 and 2004 comprise seven random audits; the data for 2005 comprise four random audits. The audits required that municipalities have a population under 450,000 inhabitants to ensure the audits were completed in a timely fashion. Auditors uncovered a wide range of abuses of public trust between 2003 and 2005, including: improper spending, undisclosed bank transfers, illegible record keeping, and phantom purchases. The incidents recorded do not distinguish the degree of corruption present, nor do they specify the post of the public officers in question.
This data set is important because it presents an objective measure of the corrupt practices of public officials in a cross-sectional comparison between the 27 federal states. As direct measures of corruption are rather scarce in anti-corruption research, the results of this data set have an added scientific value. The data will be used, along with the Federal Police statistics, in the multiple regression analysis in which four independent variables will be tested against human trafficking as the dependent variable.
Limitations of the data sources
The data sources used have limitations. The main purpose of the questionnaire was to test the hypothesized link between corruption and THB through the views of law enforcement agents towards THB-related corruption. The survey respondents do not come from all federal states, which makes cross-sectional comparisons impossible. Nevertheless, all five regions were represented even though the heavy workload of public officials in some federal states explains the relatively low response rate.4 The other sources of secondary data - the CPMI report, the official Federal Police statistics and the ‘Corruption in Municipalities' report - present limitations regarding the methodology. Specifically, the data collection and primary analysis performed were deficient. To overcome these limitations, I re-analysed the data, adjusting the results into comparable units, which enabled a cross-sectional analysis.
Findings
Survey
The relationship between corruption and THB was measured by asking respondents whether they considered corruption and THB to be related: 89 percent of respondents perceived THB to be related to corruption and 81 percent of respondents perceived THB to be increasing in Brazil. The follow-up question asked respondents about the perceived strength of the relationship. The majority of respondents described the interrelationship as either ‘medium' (38 percent) or ‘strong' (37 percent), resulting in a total of 75 percent (n = 54) of all respondents estimating the relationship as being of medium strength or greater. In addition, nominal variables were used to determine the points in the trafficking chain perceived as being the most vulnerable to corruption. The analysis of the most vulnerable points for corruption in the human trafficking chain sheds additional light on the connection between both variables.
In the domestic trafficking chain, the preparation of documents and the control of trafficked victims were the two most frequently selected points of risk (26 percent and 29 percent, respectively). A further 13 percent of respondents identified recruitment of trafficked victims as a stage vulnerable to corruption. The percentage of ‘unable to answer' was also quite high (23 percent), which suggests that, even when respondents claimed to be aware of the correlation between corruption and THB, many were reluctant to pinpoint the most vulnerable points. The results suggest that there has been no systematic attempt to either record or coordinate responses to various types of crimes inextricably linked with trafficking. In failing to obtain concrete information, government offices demonstrate the weakness of the information-gathering process, as well as the absence of a systematic method for registering Brazilian judicial procedures.
In relation to the international trafficking chain, the ‘logistics' phase was ranked by more respondents (38.35 percent) as a point of vulnerability than any other, presumably because of the direct and indirect involvement of border control officers in the screening and processing of persons entering the country (PACO 2002). Respondents found the process of preparing documentation to be the second most vulnerable point (23.0 percent), and another 23.0 percent gave non-attitude responses or did not answer at all. A further 11.0 percent identified recruitment. The control and exploitation of trafficked victims were not considered owing to the fact that Brazil is generally a country of origin rather than the destination for the trafficked victims. As such, no comparative analyses could be completed for this factor between the domestic and international trafficking chains. These results demonstrate that, according to those surveyed, human trafficking is strongly aided by corruption and therefore is an explanatory variable of THB.
Findings from the CPMI report
Out of the total of 561 cases of sexual offences analysed between 1994 and 2003, the CPMI report indicated that 89 were investigated as THB case files. The investigated cases of THB involved victims trafficked either within the territory of Brazil or from Brazil to foreign destinations. Of the 89 THB incidents, 63 involved corrupt behaviour (71 percent) -15 (24 percent) were related to domestic trafficking, and 48 (76 percent) were related to international trafficking. Of the latter, there were 10 occurrences in which domestic trafficking was also present (21 percent), suggesting that many criminal networks traffic both domestically and internationally, blurring the commonly held distinctions between the two. This finding points to the need for further research into the THB networks operating in different federal states, because they may differ in size, complexity (of structure) and modus operandi.
The data suggest that corruption is more frequent in international trafficking (48 out of 63 cases) than in domestic trafficking (15 cases). However, this finding must be analysed with caution because all cases involving international THB are considered to involve at least one form of corrupt behaviour. Because minors cannot exit the country without documentation proving that they are adults, especially when travelling without someone responsible for them by law or otherwise authorized, either documents must be falsified or public officials must ‘look the other way' when minors cross borders (Shelley 2003: 80).
Table 1. Transparency International's CPI scores and the number of incidents of international THB in Brazil: 1999-2005
Year Brazil's CPI scores Total number of THB investigations
1999 4.1 20
2000 3.9 35
2001 4.0 48
2002 4.0 39
2003 3.9 56
2004 3.9 72
2005 3.7 119
Source: Author's extrapolation from Transparency International's CPI and Brazilian Federal Police data.
Note: Only data from full calendar years were used.
Federal Police Official Statistics and the CPI
Using seven annual indexes of perceived corruption and the total number of investigated THB cases in Brazil between 1999 and 2005, the analysis suggested that THB and corruption are strongly interrelated. Table 1 compares Brazil's scores on the CPI from 1999 to 2005 with the total number of investigations in the same period of the Federal Police data set. Brazil has scored poorly in recent years, experiencing ‘a significant worsening in perceived levels of corruption' (Transparency International 2006: 1).
The yearly CPI scores show a strong negative correlation with the annual number of investigations into THB (the coefficient of the correlation is -.91). The inverse correlation suggests that the number of investigations into THB may be related to perceived levels of corruption. Although it is not possible to establish the direction of causation, whichever way causation runs the public perception of corruption has increased significantly during the same periods that suspected cases of THB have, according to the official Federal Police statistics. This finding is reinforced by the findings of the survey, in which respondents perceive both phenomena to be interrelated and increasing in Brazil. The data from the Municipal Corruption report also indicated that the overall number of corrupt incidents increased over a consecutive three-year period (2003, 2004 and 2005). Although fewer incidents were recorded when the number of random audits was reduced in 2005, the number of corruption incidents nevertheless shows a tendency towards growth. This result confirms TI's finding that Brazil's general corruption score has been falling in the past few years. Figure 1 shows both growth patterns of THB and corruption. Figure 1 suggests that THB and corruption are interrelated phenomena. This assertion was subsequently tested using the multiple regression technique.
Multiple regression analysis
So far, all the empirical findings of the previously discussed data sources have been analysed using descriptive statistical methods, providing simple correlations between the variables. The results were presented mainly in terms of rates and percentages. However, a simple correlation between variables does not reveal the truth about the relationship between them. In order to identify the correct relationship between THB and corruption it is necessary to fix or control for the effect of other relevant variables by conducting multiple regression analysis.
In order to calculate the predictive potential of the relationship on THB (dependency), multiple regression was conducted using two different (independent) data sets (the Federal Police Official Statistics and the Municipal Corruption report). This allowed for the application of more advanced statistical methods in order to stipulate a positive causal effect of one variable on another (predictive effect), while simultaneously controlling for the effects of other potential predictors and their interrelationship (multiple inter-correlations). The classic predictors for THB include poverty, unemployment, official corruption and illiteracy (see CPMI 2004; PESTRAF 2002; UN Protocol to Suppress, Prevent and Punish Trafficking in Persons 2000). The variable to be explained in the regression equation is THB for the purpose of sexual exploitation. The explanatory variables (causal factors), based on previous research (see Bales 2005; Hughes 1999; Hojman 2004; Morris 2004; United Nations Commission on Human Rights 2000), are:
•household income5
•Gini coefficient6
•illiteracy rate per 100,000 persons7
•municipal corruption incidents
All the explanatory variables (except corruption and THB) were measured by data drawn from all the Brazilian federal states and were obtained from the Brazilian Institute of Applied Economic Research (IPEA). The data on THB were drawn from the Federal Police Official Statistics and the data on corruption from the Municipal Corruption report. The Municipal Corruption report is the only available source of data on Brazilian corruption incidents with a cross-sectional and longitudinal character. As an independent measure of corruption, it furthermore fulfils the necessary criterion for comparability in multiple regression analysis.
Potential predictors such as poverty (measured by the poverty rate, see Bales 2005), infrastructure (measured by the extent of paved roads and number of exit points), population density and regional economic success (measured by federal state income) were all tested without showing any significant effect on THB and were therefore excluded from further analysis. Other economic variables, such as GDP or GNP per capita, measuring the wealth/poverty of each federal state were not available for all federal states and therefore could not be used.
Following the guiding hypothesis that THB for the purpose of sexual exploitation depends on corruption in Brazil, a positive partial causal effect of corruption on THB is predicted. The regression model uses combined cross-sectional and time-series data from three consecutive years, 2003-2005. Each variable was sequenced by the 27 federal states and by year. As a result, 81 observations were obtained. Logarithmic transformation was then applied in order to reduce multicollinearity. Two time dummy variables and one constant variable were used to control for time variations.
Based on the model, random and fixed effects were estimated. In order to test whether the fixed effects or random effects model should be used, the classical Hausman (1978) specification test was applied. Since no significant correlation between unobserved, person-specific random effects and the regressors could be found, the Hausman specification test suggests the random effects model is most suitable (Yaffee 2003; Hausman 1978).
Two variables predicted THB: corruption and the Gini coefficient. The analysis demonstrated that the coefficient of corruption is statistically significant (p ≤ .01) and has the predicted positive sign. In other words, the higher the number of incidents of corruption, the higher the number of incidents of THB. At first glance, it seems that the Gini coefficient is also a statistically significant predictor at any reasonable level. However, the value of the Gini coefficient is very high (8.27). One possible explanation is that the Gini coefficient is a compound variable, which means that it is itself dependent on several different factors, each influencing the others (Agenor 2004). This internal interdependency may have had a negative impact on the outcome of the analysis. This can be seen in the lower statistical significance of the Gini coefficient, which is half that of the level of corruption (z-value of 2.39 versus 4.60). The other remaining variables - household income and the illiteracy rate - were not significant predictors of THB.
The relationship between corruption and THB appears to be linear (the estimated parameter is 1.08), indicating that a 1 percent increase in corrupt practices will lead to an increase of 1.08 percent in occurrences of THB. This result substantiates the guiding hypothesis of this analysis, that THB strongly depends on corruption in Brazil, reinforcing the results drawn from related research findings (see Bales 2005).
Conclusion
The findings in this study demonstrate that THB in Brazil is not only related to levels of official corruption but also dependent on official corruption. Respondents to the survey questionnaire perceive corruption to be correlated with THB. The CPMI report reveals that corruption is involved in 71 percent of all examined cases. This finding demonstrates a direct linkage between THB and corruption. The trend analysis performed on the Federal Police Official Statistics and Transparency International's CPI suggests that THB and corruption have both increased in recent years.
The data in the Municipal Corruption report also showed an increase in occurrences of corruption between 2003 and 2005. These findings were supported by the results from the survey questionnaire and the trend in Brazil's scores on the TI's CPI. Overall, the findings show that THB is highly dependent on corruption. The main inference from the multiple regression analysis postulates that, to combat THB effectively, Brazil must also tackle corruption. Effective anti-corruption measures are likely to have a demonstrable impact on actual levels of THB in every region in Brazil. The results of the regression analysis imply that, when corruption rates fall, THB rates are expected to drop accordingly. By lowering the overall rate of corruption and strengthening the integrity of public officials, the occurrence of THB will decline.
Notes
1. Transparency International is a leading NGO on corruption issues. The CPI, released annually by TI, is a composite index of cross-country polls based on annual surveys of attitudes to corruption-related crimes carried out by a variety of independent and reputable institutions (Lambsdorff 2005: 5-6). The index ranks countries around the world according to their perceived levels of corruption. The ranking ranges from 10.0 (‘highly clean') to 0.0 (‘highly corrupt').
2. Brazilian Penal Code (1940), Articles 231 and 231-A, follows the UN Convention and the Protocol to Prevent, Suppress and Punish Trafficking in Persons, and establishes both domestic and international trafficking as crimes.
3. See https://www.transparency.org/news_room/faq/corruption_faq (accessed 22 September 2009).
4. This reluctance to cooperate is unfortunately a common phenomenon in Brazilian scientific research (see CPMI 2004).5. The variable ‘household income' refers to the calculated average income of all persons sharing a single residence.
6. The ‘Gini' coefficient variable measures the degree of inequality in the distribution of persons relative to per capita household income. Its value ranges from 0, when there is no inequality (the salaries of all persons have the same value), to 1, when inequality is at its highest. ‘Brazil is one of the world's most unequal countries' (see Hinton 2005).
7. The variable used for illiteracy measures the illiteracy of all persons aged 15 years or older.
References
Ades, A. and Di Tella, R. (1997). The new economics of corruption: A survey and some new results. Political Studies 45, 496-515.
Agenor, P. (2004). The economics of adjustment and growth. Cambridge, MA: Harvard University Press.
Bales, K. (2005). Understanding global slavery. Berkeley: University of California Press.
Brazilian Penal Code (1940). URL (accessed 9 September 2009): https://www010.dataprev.gov.br/sislex/paginas/16/1940/2848.htm.
CGU [Controladoria-Geral da União] (2003). Programa de Fiscalização a partir de Sorteios Públicos. URL (accessed 9 September 2009): https://www.cgu.gov.br/CGU/index.asp and https://www.cgu.gov.br/Imprensa/Noticias/2003/noticia06903.asp.
CPMI [Comissão Parlamentar Mista de Inquérito - Joint Parliamentary Commission of Inquiry] (2004). Câmara dos Deputados e Senado Federal no Brazil internal report.
European Commission (2001). Trafficking in women. The misery behind the fantasy: from poverty to sex slavery. A comprehensive European strategy. URL (accessed 9 September 2009): https://ec.europa.eu/justice_home/news/8mars_en.htm#a1.
Figueiredo, D. and Hazeu, M. (2006). Migração e tráfico de seres humanos para Suriname & Holanda, Projeto Jepiara - Programa de enfrentamento do tráfico de seres humanos e exploração sexual em Belém. URL (accessed 9 September 2009): https://www.faor.org.br/CD/download/4_trafico_seres_humanos.pdf.
Hausman, J. A. (1978). Specification tests in econometrics. Econometrica 46, 1251-71.
Heidenheimer, A. J. (2004). Disjunctions between corruption and democracy? A qualitative exploration. Crime, Law & Social Change 42, 99-109.
Hinton, M. S. (2005). A distant reality: Democratic policing in Argentina and Brazil. Criminal Justice 5, 75-100.
Hojman, D. E. (2004). Inequality, unemployment and crime in Latin American cities. Crime, Law & Social Change 41, 33-51.
Hughes, D. M. (1999). Pimps and predators on the internet: Globalizing the sexual exploitation of women and children. Coalition against Trafficking in Women (CATW).
IPEA [Instituto de Pesquisa Econômica Aplicada - Brazilian Institute of Applied Economic Research]. Ministério do Planejamento, Orçamento e Gestão. URL (accessed 9 September 2009): https://www.ipea.gov.br.
Klitgaard, R., MacLean-Abaroa, R. and Parris, H. L. (1998). A practical approach to dealing with municipal malfeasance. Working Paper No. 7, UNDP/UNCHS/World Bank-UMP. First printing, May 1996.
Laczko, F. (2005). Data and research on human trafficking. International Migration 43, 5-16.Lambsdorff, J. (2001). Background paper to the 2001 Corruption Perceptions Index. Framework document, Transparency International and Göttingen University, June.
Lambsdorff, J. G. (2005). The methodology of the 2005 Corruption Perceptions Index. Transparency International and University of Passau. URL (accessed 9 September 2009): https://www.transparency.org/policy_research/surveys_indices/cpi/2005/methodology.
Lezertua, M. (2003). Opportunities for corruption in the trafficking of human beings. Paper presented at Transparency International's 11th International Anti-Corruption Conference, Seoul, Korea, 25-28 May.
Lyday, C. B. (2001). The shadow market in human beings: An anti-corruption perspective. Paper presented at Transparency International's 10th International Anti-Corruption Conference, Prague, Czech Republic,7-10 October.
Mahmoud, T. O. and Trebesch, C. (2009). The economic drivers of human trafficking: Micro-evidence from five Eastern European countries. Working Paper No. 1480, Kiel Institute for the World Economy.
Morris, S. D. (2004). Corruption in Latin America: An empirical overview. SECOLAS Annals 36, 74-92.
PACO [Programme against Corruption and Organised Crime in South-Eastern Europe] (2002). Trafficking in human beings and corruption. Council of Europe. Report on the regional seminar, Portoroz, Slovenia, 19-22 June.
PESTRAF (2002). Pesquisa sobre tráfico de mulheres, crianças e adolescentes para fins de exploração sexual comercial no Brasil [Study on trafficking in women, children and adolescents for commercial sexual exploitation in Brazil]. National Report, edited by M. L. Leal and M. F. Leal. Brasília: CECRIA.
Phinney, A. (2001). Trafficking of women and children for sexual exploitation in the Americas. The Inter-American Commission of Women, Organization of American States. URL (accessed 9 September 2009): https://www.paho.org/english/hdp/hdw/TraffickingPaper.pdf.
Ribando, C. (2005). Trafficking in persons in Latin America and the Caribbean. CRS Report for Congress. Congressional Research Service, The Library of Congress. URL (accessed 9 September 2009): https://www.oas.org/atip/Latin%20America/CRS%20Dec%202005.pdf.
Savona, E. U. and William, P. (1996). The United Nations and transnational organized crime. Boca Raton, FL: Taylor & Francis.
Shelley, L. I. (2003). Human trafficking: Transnational crime and links with terrorism. Testimony before the US House Committee on International Relations, Subcommittee on International Terrorism, Nonproliferation and Human Rights.
Transparency International (2001). Corruption Perceptions Index 2001. URL (accessed 9 September 2009): https://www.transparency.org/policy_research/surveys_indices/cpi/2001.
Transparency International (2006). Corruption Perceptions Index 2006. URL (accessed 9 September 2009): https://www.transparency.org/policy_research/surveys_indices/cpi/2006.
Tsutsumi, K. and Honda, S. (2005). OAS Rapid Assessment Report: Trafficking in persons from the Latin American and Caribbean (LAC) region to Japan. URL (accessed 9 September 2009): https://www.oas.org/atip/PDFs/Rapid%20Assessment%20(English).pdf.
Tyldum, G. and Brunovskis, A. (2005). Describing the unobserved: Methodological challenges in empirical studies on human trafficking. International Migration 43, 17-34.
United Nations Commission on Human Rights (2000). Integration of the Human Rights of Women and the Gender Perspective: Violence against Women. Economic and Social Council, E/CN.4/2000/68, 29 February.United Nations Convention against Corruption (2003). URL (accessed 9 September 2009): https://www.unodc.org/pdf/crime/convention_corruption/signing/Convention-e.pdf.
United Nations Protocol to Prevent, Suppress and Punish Trafficking in Persons, especially Women and Children, supplementing the United Nations Convention against Transnational Organized Crime (2000). URL (accessed 9 September 2009): https://www.uncjin.org/Documents/Conventions/dcatoc/final_documents_2/convention_%20traff_eng.pdf.
UNODC [United Nations Office on Drugs and Crime] (2006). Trafficking in Persons: Global Patterns. URL (accessed 9 September 2009): https://www.unodc.org/documents/human-trafficking/HT-globalpatterns-en.pdf.
US Department of State (2008). Trafficking in Persons Report 2008. Washington, DC: US Department of State. URL (accessed 9 September 2009): https://www.state.gov/g/tip/rls/tiprpt/2008/.
Victims of Trafficking and Violence Protection Act of 2000. United States Department of Justice, Civil Rights Division, Public Law 106-386-Oct. 28, 2000. URL (accessed 9 September 2009): https://www.state.gov/documents/organization/10492.pdf.
Yaffee, R. (2003). A primer for panel data analysis. Social Sciences, Statistics and Mapping Group of ITS' Academic Computing Services, New York University. URL (accessed 9 September 2009): https://www.nyu.edu/its/pubs/connect/fall03/yaffee_primer.html.
Author Biography
Andrea Cirineo Sacco Studnicka is a Criminal Prosecutor in the Ministério Público do Distrito Federal e Territórios of Brasília, Brazil. She has a PhD in Criminology from the Catholic University of Milan, Italy. Her specializations are juvenile delinquency, domestic violence, organized crime, the special task of controlling the police, corruption, and trafficking in human beings.
European Journal of Criminology