Crime Law Soc Change https://doi.org/10.1007/s10611-018-9780-0

Compliance with anti-human trafficking policies: the mediating effect of corruption Cassandra DiRienzo 1

# Springer Science+Business Media B.V., part of Springer Nature 2018

Abstract Human Trafficking is an atrocious crime that represents a gross assault on human rights and the United Nations states that it is among the fast growing types of criminal activity. Recognizing the need for counteractive measures, in 2000, the United Nations General Assembly adopted the Convention against Transnational Organized Crime and its Protocol to Prevent, Suppress and Punish Trafficking in Persons, especially Women and Children (Protocol). Using measures of country compliance with the Protocol, past research offers empirical evidence that corruption is a primary deterrent to compliance. Further, previous field studies and surveys suggest that a greater share of women in government should positively contribute to country compliance; however, this result is largely not borne out in empirical studies. It is hypothesized that the effect of the share of women in government on compliance is fully mediated by corruption, indicating that there is an indirect effect of women in government on compliance, rather than a direct effect. This hypothesis is empirically tested using a mediation model and the results indicate that the indirect effect is statistically significant. The empirical results presented suggest that a greater percentage of women in government reduces country corruption, which in turn increases country compliance with the Protocol. The policy implications of these findings are discussed and include suggestions to enhance female participation in government.

Introduction Human trafficking represents a gross assault of fundamental human rights as victims are bought and sold like commodities and subjected to such horrors as forced labor, prostitution, bonded labor, domestic servitude, pornography, organized begging, and organ harvesting [1–3]. Although human trafficking is generally acknowledged as an atrocious crime, the United Nations [4] reports that it is among the fastest growing

* Cassandra DiRienzo [email protected]

1

Elon University, 2075 Campus Box, Elon, NC 27244, USA

C. DiRienzo

types of criminal activity and, in the next ten years, is predicted to surpass both drug and arms trafficking to become the largest form of organized crime in its incidence [5–7]. The United Nations Office on Drugs and Crime (UNODC) states that, at any one time, a conservative estimate of the number of human trafficking victims is 2.5 million and that human trafficking affects every country in every region of the globe [8]. As Duong ([9], p. 48) notes, B…no other crime has such a high prevalence of victims like human trafficking.^ With such incidence rates, it should not be surprising that human trafficking has been classified as the third most profitable global crime by the United Nations [10], which according to the UNODC, generates tens of billions of dollars in annual profit for criminals [7, 8, 11]. Recognizing the need for counteractive measures, in 2000, the United Nations General Assembly adopted the Convention against Transnational Organized Crime and its Protocol to Prevent, Suppress and Punish Trafficking in Persons, especially Women and Children (Protocol). The Protocol outlines a set of three broad anti-human trafficking policies; the prevention of the crime, the protection of victims, and the prosecution of traffickers. The United Nations Treaty Collection reports that over 180 countries have ratified the Protocol and, although states are not legally bound by the Protocol, states that ratify the document commit themselves to undertake a series of antihuman trafficking actions [7]. Considering the significant estimated number of human trafficking victims and the crime’s projected growth, the question arises as to the effectiveness of the Protocol. As Cho and Vadlamannati [12] note, there are surprisingly few empirical studies assessing the effectiveness of the Protocol, despite the number of years it has been in existence. To this end, Cho et al. [13] created the 3P compliance index that measures the quality of countries’ anti-trafficking policies and their level of compliance on the three main policy dimensions of the Protocol; prosecution, protection, and prevention. The 3P index was originally created in 2012 and assessed countries over the 2000–2010 period, but is updated annually and the index is currently available for 189 countries and has been used by several researchers as a proxy for country compliance with the three policy dimensions of the Protocol [8, 12, 14, 15]. Using the 3P index, several researchers have explored the factors that make countries more or less likely to comply with anti-human trafficking policies. Briefly, using cross-country panel data, Cho et al. [13] find that overall country compliance with the three policy dimensions of the Protocol significantly decreases with higher levels of corruption and significantly increases in countries that respect the rights of women. In an empirical, cross-country study, Potrafke [8] finds that countries with a greater majority of Christians implement stricter anti-human trafficking policies compared to Muslim majority countries and that democracies also have stronger policies compared to dictatorships. Cho and Vadlamannati [12], Cho [16], and Neumayer [17] also identify democracies as being more likely to be in policy compliance. Further, in an extensive, cross-country analysis Heller et al. [15] conclude that countries with greater economic freedoms are more likely to adopt anti-human trafficking policies. Finally, Cho and Vadlamannati [12] find that countries with a greater share of female legislators tend to have a greater level of compliance in regard to the protection policy dimension of the Protocol, but relationship is found to be insignificant with the prosecution and prevention policy dimensions. When exploring the determinants of country compliance with the Protocol, it is important to discuss the root causes and drivers of human trafficking as many of these

Compliance with anti-human trafficking policies: the mediating...

factors are intrinsically linked to the strictness of a country’s anti-human trafficking policies and to compliance. These factors have been divided into two categories; the supply side or ‘push’ factors and the demand side or ‘pull’ factors. The ‘push’ factors, or the reasons that people seek employment opportunities elsewhere, and in that process fall victim to human traffickers, have largely been identified as poverty, organized crime, corruption, institutional quality, information, and conflict [1, 18–21]. In reference to ‘pull’ factors, Cho [19] provides empirical evidence that wealthier countries are more likely to have the demand for low-skilled labor and the means to pay the traffickers. Although Cho’s [19] analysis uses country level data, it is important to note that 72% of victims are trafficked either domestically or within their own region [10]. Thus, in more general terms, it is the conditions of poverty, and weak and unstable institutions and environments - conditions that can be exacerbated by corruption - that drive people to seek a better life elsewhere and corruption and organized crime facilitate traffickers into ensnaring their victims and relocating them into wealthier areas – areas that can be within country, region or cross-country. While corruption has been identified as a significant factor in the human trafficking crime and as a hindrance to country compliance with anti-human trafficking policies [1, 2, 13, 18, 21, 22], it is argued that corruption plays a key role in every stage of human trafficking; from weakening institutions and creating greater income inequality, to facilitating the crime, and finally to creating disincentives for government officials to comply with anti-human trafficking policies. In regard to its definition, corruption is commonly defined as the ‘misuse of public power for private or political gain’ [23–26]. The World Bank [27] defines a corrupt practice as the ‘offering, giving, receiving or soliciting, directly or indirectly anything of value to influence improperly the actions of another party’. More specifically, behaviors or practices such as bribery, nepotism, collusion, fraud, embezzlement, self-dealing, trading in influence, and misappropriation of public resources have been defined as corrupt [25, 28–31]. In reference to ethics and corruption, Tanzi [32] describes corrupt public sector practices as those that are considered to be dishonest and/or unethical. In the case of human trafficking, it is argued that corruption can serve an important intervening role. Specifically, it is hypothesized that corruption can, from a statistical perspective, fully mediate the effect of factors that should lead to greater compliance with anti-trafficking policies such as having a greater share of women representation in government. Considering that the majority of persons trafficked are women and children, many researchers have pointed to the important role of gender related issues to human trafficking; gender discrimination, share of women in government, female literacy, years of schooling, female labor force participation, and measures of female economic and social rights [19, 33–35]. Nonetheless, the empirical results largely do not support this relationship both in terms of a push or pull factor or in terms of country compliance with the Protocol, although field studies and surveys indicate that such factors should be relevant to human trafficking [12, 19]. It is theorized here that when regression analyses include both corruption and measures of gender equality, especially the share of women in government, corruption fully mediates the effect of gender in government. In the case of full mediation, the mediating variable, corruption, is found to be significant, but the independent variable, share of women in government, is found to be insignificant. The primary objective of this study is to empirically demonstrate the full mediation effect of corruption on the share of women in government in reference to

C. DiRienzo

a country’s compliance with the Protocol. In other words, the effect of the share of women in government does not have a direct effect on country compliance with the Protocol, rather it has an indirect effect that is dependent on corruption.

Human trafficking, corruption and the role of women in government Human trafficking The Protocol defines human trafficking as Bthe 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^. Victims of human trafficking tend to be from vulnerable populations such as those living in high poverty, children, and young women, many of whom are single mothers. These individuals are typically in search of low-skilled work and opportunities to increase their economic status. Traffickers target these vulnerable individuals using a variety of deceptive ploys. In many cases, traffickers promise these individuals a better life elsewhere with low-skilled, but reputable, higher paying work in positions such as a child care provider or nanny [36]. Victims are tricked into believing they are migrating to a better life and once traffickers have ensnared and relocated them, the traffickers typically confiscate victim’s identification and/or passports [33, 34]. The victims are then forced into such atrocities as prostitution, domestic servitude, pornography, or other forms of forced labor. Often facing language barriers and typically without identification, or a cultural or legal understanding of their rights, traffickers are able to control these victims through threats, coercion, and violence [37, 38]. Human trafficking: the role of corruption Corruption has been named as both a root cause and means to facilitate human trafficking and has been identified in many qualitative and empirical studies as a significant contributor to the crime [39]. In a study of human trafficking in Nigeria, Agbu [1] concludes that the crime is worse in highly corrupt societies and any efforts to fight human trafficking must also address the issues with corruption in both the private and public sectors. Using a multi-modal approach, Studnicka [2] finds that corruption is not only a causal factor to human trafficking in Brazil, but that human trafficking is, in fact, dependent on corruption in Brazil. In a cross-country, multivariate study, Bales [40] identified government corruption as the most critical factor in predicting human trafficking from an origin country. Further, in a cross-country analysis, Lyday [41] finds that there is a strong relationship between perceived corruption and the quality of government action against human trafficking. This research represents a few of the many studies linking corruption to human trafficking and corruption has been named both as a root cause and a facilitator to the crime.

Compliance with anti-human trafficking policies: the mediating...

In reference to corruption as a root cause, the World Bank [42] states that corruption harms economic activity and growth, contributes to higher-order crimes, and exacerbates income inequality as the poor have been found to pay the highest percentage of their income in bribes. Corruption weakens institutional foundations, distorts rule of law, creates distrust among citizens, and causes instability in both the private and public sectors [43]. In sum, corruption plays an important role in creating the kind of environment that the poor and vulnerable desire to leave in search of a better life. As Studnicka [2] states, Btrafficking has its roots in much broader conditions of state failure^. Corruption is also a key facilitator to the crime, which in turn, forms disincentives for government officials to comply with anti-human trafficking policies such as the Protocol. As McCarthy [22], Surtees [21], and UNGIFT [39] describe, corrupt officials can be paid bribes to falsify travel documents, allow illegal border crossings, and turn a blind eye to prostitution venues where trafficking victims are likely to be found. Perhaps even more egregious, investigators, prosecutors, and judges can be bribed into compromising or even dismissing criminal investigations or cases against traffickers [21, 22, 39] and Studnicka [2] states that even corrupt government officials have been found to participate in human trafficking. Surtees [21] states that the corruption of a variety of government officials from politicians, state functionaries, law enforcement, to foreign service staff has allowed for trafficking to continue originate and move throughout South-Eastern Europe. As an additional complication, Surtees [21] notes that the presence of corruption makes victims distrust law enforcement and can even reduce the likelihood that a victim would accept assistance from an agent. Finally, corruption can also play a role in the re-victimization of trafficked persons. Holmes [44] describes how corrupt officials can also exploit suspected human trafficked victims through prostitution, domestic servitude, or other forms of forced labor. As Uddin [45] notes, corrupt government officials can treat victims of human trafficking as criminals, effectively re-victimizing those who have survived the atrocities of being trafficked. Holmes [44] refers to this re-victimization as ‘triple victimization’. In such corrupt societies, there are disincentives to comply with anti-human trafficking policies. Government agents receiving bribes or participating in the crime do not have an economic incentive to actively work against the crime. Further, as Cho and Vadlamannati [12] discuss, compliance with the Protocol is costly as it often requires funding for new policy programs and amendments of national law. In sum, in corrupt societies where officials are either participating in human trafficking are being paid substantial bribes, there is not an economic incentive to discontinue their current practices and work to combat the crime, especially if compliance is costly and there is risk that they will be exposed as criminals. Not surprisingly, using the 3P index, Cho et al. [13] offer empirical evidence that corruption is a primary determinant of noncompliance with the Protocol. Human trafficking: the role of women in government As Winterdyk and Reichel [7] and Cho [46] discuss, field studies and surveys indicate that the majority of human trafficking victims are female, especially

C. DiRienzo

young women, single mothers, and young girls. As a result of the greater share of female victims, Cho [46] notes that human trafficking is a form of gender based violence. Nonetheless, it is interesting to note that the share of female offenders is greater in human trafficking compared to most other crimes [10]. Considering human trafficking is at least in part a gender based crime, a society’s attitude toward women and the extent to which women share in governance becomes a very relevant issue in anti-human trafficking efforts. Specifically, studies such as Chattopadhyay and Duflo [47] and Bartilow [48] have shown that gender representation in government affects policy making. The general conclusion of these studies is that when females hold a greater representation as legislators and/or political representatives, there is more investment and policy creation that benefits women’s interests. Further, Chattopadhyay and Duflo [47] provide empirical evidence that a politician’s gender does influence policy decisions. Cho and Vadlamannati [12] summarize that female officials tend to be more concerned about women’s issues, making countries with a higher percentage of females in government more likely to comply with anti-human trafficking policies such as the Protocol. Given the above findings, several studies exploring the drivers of human trafficking and/or compliance with the Protocol have included the measures of the proportion of women in government [12, 19]. Considering both the theoretical arguments and the analytical evidence indicating the gender representation in government affects policy outcomes, it is surprising that there is very little empirical evidence that countries with a greater share of women in government are more likely to comply with the Protocol. In an extensive cross-country panel data using the 3P index as a measure of country compliance with the Protocol, Cho et al. [13] find that the share of women legislators in parliament does not significantly affect county compliance. Further, using the subindices of the 3P index - the individual measures of compliance with prevention, protection, and prosecution - Cho and Vadlamannati [12] do offer empirical evidence that the share of female legislators has a significant positive effect on the protection subindex, but no significant effect on the other two indices of prevention and prosecution. More broadly in reference to the outflow and inflow of human trafficking victims at the country level, Cho [19] notes the controversial findings of her results indicating no significant relationship between human trafficking and a variety of gender discrimination measures. It is argued here that gender, specifically the share of women in government, does significantly affect country compliance with the Protocol; however, when empirical studies include measures of corruption and share of women in government in regression analyses, corruption fully mediates the effect of women in government. In other words, the effect of the share of women in government on country compliance is not direct, rather it is indirect such that the effect of the share of women in government passes through, or is mediated by, the level of corruption. In this situation, when both corruption and share of women in government are included in a regression analysis using compliance with the Protocol as the dependent variable, the effect of corruption is found to be statistically significant, while the effect of women in government is insignificant.

Compliance with anti-human trafficking policies: the mediating...

Mediating effects: corruption and share of women in government Mediating models Regression models that specify a mediating effect hypothesize a causal chain in which the influence of an independent variable on a dependent variable is mediated through a third variable [49]. The basic statistical mediation model is illustrated Fig. 1. Testing for mediation is a multi-step process in which three regression models are estimated. Specifically, to test the effect of the independent variable on the dependent and the potential mediating effect, estimates from the following regression models are needed [50]: Y ¼ β0 þ β1 X þ ε

ð1Þ

M ¼ α 0 þ α1 X þ ζ

ð2Þ

Y ¼ λ0 þ λ1 X þ λ2 M þ ς

ð3Þ

In Model 1, it is necessary for the independent variable to significantly affect the dependent variable. From Fig. 1, Model 2 tests path a, or the effect of the independent variable on the mediator variable. Model 3 tests path b from Fig. 1. In the third model, either partial or full mediation is supported if the effect of the mediator variable remains significant after controlling for the independent variable. If the independent variable remains significant after the mediating variable is controlled, but the effect is less than in Model 1, there is evidence to support partial mediation; however, if the independent variable becomes insignificant when the mediating variable is controlled for, there is support for full or perfect mediation [49–52]. The direct effect of the independent variable on the dependent, or path c from Fig. 1, ^ 1. The indirect effect of independent variable on the dependent is measured by λ variable through the mediator variable can be found by substituting Model 2 into ^ 2 . The total effect of the independent variable on the ^1 * λ Model 3 and is equal to α ^1 + α ^2. ^1 * λ dependent variable is the sum of the direct and indirect effects which is: λ Testing the significance of the indirect effect or the estimated product of the coefficients ^ 2 is rather complex and past research has typically used either a bootstrapping ^1 * λ α method that is nonparametric in nature, or a parametric Monte Carlo method. Recently, Biesanz et al. [49] offer a new inferential method for testing the significance of a

a

Mediator Variable

Independent Variable

Dependent Variable c

Fig. 1 Illustration of mediation model

b

C. DiRienzo

mediating effect and provide evidence that the new method outperforms previous techniques both in regard to preserving Type 1 error rates while maintaining power. Corruption as a mediator As discussed above, there are strong theoretical arguments for a higher share of women in government to have a positive effect on country compliance with the Protocol; however, this result is largely not borne out analytically. Nonetheless, it is possible for corruption to have a full mediating effect on the share of women in government and when both are included in the same regression analysis, the effect of the share of women in government is found to be insignificant. In this case, the underlying assumption of the mediating effect is that there is a causal relationship between the share of women in government and corruption. In other words, a greater share of women in government reduces corruption levels; a result that has been identified in empirical studies. Dollar et al. [53] summarize numerous behavioral studies that find women to be comparatively more honest and take stronger stances on ethical behavior. Dollar et al. [53] hypothesize that the results of these studies suggest that women are less likely to engage in activities that sacrifice the common good for personal gain; in other words, women are less likely to engage in corrupt practices. Dollar et al. [53] offer robust empirical evidence of this relationship using a large cross-section of countries and conclude that there is a strong, negative and significant relationship between the share of women in government and the level of corruption within the country. Further, in cross-country studies using several independent data sets, Swamy et al. [54] also finds that corruption is less severe in countries where women hold a greater share of parliamentary seats and senior positions in government and, in another analysis using micro-level data, Swamy et al. [54] show that women are less likely to take bribes or otherwise be involved in bribery. While there are several theoretical, behavioral, and empirical studies identifying the causal relationship between a greater share of women in government and lower corruption levels, other research suggests that while this relationship can exist, it is likely to be contextual. Specifically, Esarey and Chirillo [55] summarize this literature by stating that the women’s attitudes and willingness to participate in corrupt practices is dependent on institutional and cultural contexts. In cross-country empirical analyses Sung [56] and Sung [57] offer empirical evidence that the share of women in government does not significantly affect the prevalence of corruption, but rather it is the strength of the liberal institutions that drives both the prevalence of corruption and its trend. Recognizing that the relationship between women in government and corruption can be contextual, the general consensus of much of the literature is that, all else equal, a greater share of women in government is associated with lower corruption levels. In order for the share of women in government to be mediated by corruption, it is necessary to assume that the relationship is causal. Figure 2 illustrates the hypothesized relationships. To test these relationships, the three regression models are estimated where share of women in government serves as the independent variable X, corruption is the mediating variable M, and compliance with the Protocol is the dependent variable Y. Further, the indirect effect of share of women in government on country compliance is estimated

Compliance with anti-human trafficking policies: the mediating... Corrupon b

a

Share of Women in Government

c

Compliance with the Protocol

Fig. 2 Illustration of corruption mediation model

and tested for significance, in addition to estimating the direct and total effects. A description of the data used to test these relationships follows.

Data The most recently updated 3P index published by Cho et al. [13] is 2015 index, which is used to measure country compliance with the Protocol. The 3P index represents an aggregate measure country compliance on the three main policy dimensions of the Protocol; prosecution, protection, and prevention. As described in Cho et al. [13], the prosecution dimension assesses the effort level of a country’s government to punish and prosecute traffickers, the protection dimension measures the effort level of a country’s government to protect and assist the victims of human trafficking, and the prevention dimension assesses the effort level of a country’s government to prevent and combat human trafficking. To create the three policy dimension scores, Cho et al. [13] use raw data from the United States State Department’s Annual Reports of Trafficking in Persons and the UNODC’s Reports on Trafficking in Persons. Cho et al. [13] consider the Protocol’s key requirements for each of the policy dimensions and a country’s compliance with each of the requirements is evaluated using a minimum of two trained coders. Cho et al. [13] note that the coders were given clear coding instructions and decision rules and each coded country compliance independently. For each policy dimension, Cho et al. [13] assigns countries a score on a discrete one to five scale such that a score of five represents the greatest level of compliance and a score of one represents complete non-compliance. To create the 3P index, Cho et al. [13] take the un-weighted sum of the three sub-indices scores, yielding a range of scores from three to 15. Countries used in this analysis receiving the lowest (2015) 3P value of three (least compliance) are Eritrea, Libya, North Korea, and Syria, which were followed by Equatorial Guinea, Iran, Kiribati, Russia, and Yemen, each with a score of four. At the other end of the compliance spectrum, Austria, Spain, and the United Kingdom received the highest possible score of 15 and Armenia, Belgium, Philippines, and South Korea followed, each with a score of 14. Examples of countries that fell more in the middle range with a score of six are Algeria, Benin, Democratic Republic of Congo, Mali, Mauritania, Morocco, Papua New Guinea, and Seychelles. The Corruption Perception Index (CPI) created by Transparency International is used as a proxy measure for the degree of corruption across countries. Transparency International describes the CPI as estimating the degree to which officials and politicians are perceived to accept bribes or illicit payments, embezzle public funds, or participate in other practices considered to be corrupt. The CPI ranges from zero (allpervasive corruption) to 100 (no corruption).

C. DiRienzo

Berg [58] notes that the CPI is likely to be the most well-known measure of country corruption; however, other measures of country corruption such as Business International, International country Risk Guide, World Bank index, and the World Competitiveness Report are available and the CPI is not without its criticisms. Specifically, the CPI is a perceptions-based measure of corruption and Sampford et al. [59] discuss the over-reliability of the data over time, which can be linked to the use of a small number of corruption perceptions surveys that do not vary significantly on an annual basis. Further, Jones et al. [60] state that the CPI is unlikely to capture the perceptions of the poor, women, or informal businesses as the CPI is largely based on the assessments of experts, which do not necessarily capture household or business experiences. Nonetheless, it is important to note that other measures of country corruption have been criticized for relying on individual sources [61, 62]. Recognizing the concerns regarding CPI and acknowledging that any quantitative measure of a generally clandestine activity will be subject to measurement error, the CPI is used in this analysis to proxy country corruption as Berg [58] and Serra [62] describe the CPI as the most comprehensive and robust measure of country corruption. Finally, the World Bank’s percentage of seats held by women in national parliaments (WP) is used to proxy the share of women in government. For Model 1, paired country data is available for 174 countries and for Models 2 and 3, 165 country observations are available. Table 1 provides a list of all of the countries included in this analysis and Table 2 shows the descriptive statistics for the each of the three variables; 3P, CPI, and WP (across all countries included) and Table 3 provides the estimated correlations. In cross-country economic, socio-economic, and political analyses, it is common to lag the independent variables by at least one year as their effect on the dependent variable cannot be expected to occur simultaneously [63–65]. In the analyses presented, the CPI and WP variables are lagged to 2013, or by two years; however, the entire analysis was also conducted with just a one year lag to 2014 and there are no significant differences in any of the results presented.

Analysis and results The estimated regression results from the three models are shown in Table 4. The estimated results from Model 1 establish that the percentage of seats held by women in national parliaments (the independent variable) positively and significantly affects country compliance with the Protocol as measured by the 3P index (the dependent variable). As described in Baron and Kenny [50], establishing a statistically significant relationship between the independent and dependent variables is a necessary condition for the mediational model. Second, the estimated results for Model 2 show that the percentage of seats held by women in national parliaments (the independent variable) positively and significantly affects the level of corruption as proxied by CPI (the mediating variable). Recalling that greater CPI values indicate a reduction in perceived corruption, these results indicate that the independent variable significantly affects the mediator variable with the correct sign, which is the second condition for the mediation model outlined in Baron and Kenny [50]. Finally, the results of Model 3 provide statistical evidence that corruption fully, or perfectly, mediates the percentage of seats held by women in national parliaments as when both the independent and mediator

Compliance with anti-human trafficking policies: the mediating... Table 1 Countries included in analyses Afghanistan

Czech Republic

Latvia

Poland

United Arab Emirates

Albania

Denmark

Lebanon

Portugal

United Kingdom

Algeria

Djibouti

Lesotho

Qatar

Uruguay

Angola

Dominican Republic

Liberia

Romania

USA

Antigua and Barbuda*

Ecuador

Libya

Russia

Uzbekistan

Armenia

El Salvador

Lithuania

Rwanda

Venezuela

Australia

Equatorial Guinea Luxembourg

Saudi Arabia

Vietnam

Austria

Eritrea

Senegal

Yemen

Macedonia

Azerbaijan

Estonia

Madagascar

Serbia

Zambia

Bahamas

Ethiopia

Malawi

Seychelles

Zimbabwe

Bahrain

Finland

Malaysia

Sierra Leone

Bangladesh

France

Maldives*

Singapore

Barbados

Gabon

Mali

Slovak Republic

Belarus

Gambia

Malta

Slovenia

Belgium

Georgia

Marshall Islands*

Solomon Islands*

Belize*

Germany

Mauritania

Somalia

Benin

Ghana

Mauritius

South Africa

Bhutan

Greece

Mexico

South Korea

Bolivia

Guatemala

Micronesia*

South Sudan

Bosnia and Herzegovina

Guinea-Bissau

Moldova

Spain

Botswana

Guyana

Mongolia

Sri Lanka

Brazil

Haiti

Montenegro

St. Lucia

Bulgaria

Honduras

Morocco

St. Vincent and the Grenadines

Burkina Faso

Hungary

Mozambique

Sudan

Burma/Myanmar

Iceland

Namibia

Suriname

Burundi

India

Nepal

Swaziland

Cambodia

Indonesia

Netherlands

Sweden

Cameroon

Iran

New Zealand

Switzerland

Canada

Iraq

Nicaragua

Syria

Chad

Ireland

Niger

Tajikistan

Chile

Israel

Nigeria

Tanzania

China

Italy

North Korea

Thailand

Colombia

Jamaica

Norway

Timor-Leste

Comoros

Japan

Oman

Togo

Congo, Democratic Republic of

Jordan

Pakistan

Tonga*

Congo, Republic of

Kazakhstan

Pala*

Trinidad and Tobago

Costa Rica

Kenya

Panama

Tunisia

Cote d’Ivoire

Kiribati*

Papua New Guinea

Turkey

C. DiRienzo Table 1 (continued) Croatia

Kuwait

Paraguay

Turkmenistan

Cuba

Kyrgyz Republic

Peru

Uganda

Cyprus

Laos

Philippines

Ukraine

*Missing 2013 CPI score

variables are included in the regression analysis, the independent variable becomes insignificant while the mediating variable remains positive and significant. The estimated direct effect of percentage of seats held by women in national ^ 1 , which is equal to 0.0259 from Model 3. The parliaments on 3P compliance is λ direct effect is insignificant; a result necessary for full mediation. The estimated indirect effect, or the effect that the percentage of seats held by women in national parliaments ^ 2 , which is equal to ^1 * λ has on 3P compliance through the mediating effect of CPI is α (0.4654)*(0.0568) = 0.0264. The indirect effect represents the change in 3P compliance for every one unit increase in the percentage of seats held by women in national parliaments that is mediated by CPI. Finally, the total effect of the percentage of seats ^ 2 , or ^1 + α ^1 * λ held by women in national parliaments on 3P compliance is equal to λ 0.0259 + 0.0264 = 0.0523. While the estimated regression results offer support for the full mediation effect of corruption, it is important to test the significance of the indirect effect, which represents the path of a to b in Fig. 1 or the causal relationship between the independent and mediating variables. Using the new inferential method presented in Biesanz et al. [49], the null hypothesis of no indirect effect is rejected with a p-value less than 0.0001. Thus, there is significant empirical evidence to support the mediating effect of corruption on the share of women in government in regard to county compliance with the Protocol. It is important to note that mediation models are designed to test the causal relationships illustrated in Fig. 1. Although the empirical results presented support the relationships between women in government, corruption, and country compliance with the Protocol, other potential intervening factors such as culture and the quality of a country’s institutions and government are not included in this analysis. The focus here is on the primary relationship between these three variables, as illustrated in Fig. 2, and the results should be interpreted in this light.

Table 2 Descriptive statistics Analysis variables 3P

CPI

WP

Mean

9.08

42.27

19.96

St. Deviation

2.53

20.01

11.54

Min

3

8

0

Max

15

91

63.8

n

174

165

165

Compliance with anti-human trafficking policies: the mediating...

Table 3 Correlation matrix

3P 3P

CPI

WP

1

CPI

0.4878

1

WP

0.2956

0.2592

1

Concluding remarks Human Trafficking is a dreadful crime that has left no country or region of the world untouched. The United Nations [66] states that it is among the fastest growing types of criminal activity and as Potrafke [8] discusses, it generates tens of billions of dollars in profit for criminals each year. With the creation of the Protocol, the United Nations recognized the need for an international commitment of the global community to fight to crime. Considering that the Protocol has been in existence for over a decade, it is time, as Cho et al. [13] note, to measure the effectiveness of the Protocol, which is tied very closely to country compliance with the three anti-human trafficking policy dimensions. Empirical analyses exploring the factors that facilitate or hinder countries’ likelihood of compliance are important to the anti-human trafficking effort. Knowledge of these driving factors are important to policy makers who are developing new antihuman trafficking policy, or are working to facilitate compliance with the Protocol. In this vein, empirical support for the relationship between the share of women in government and country compliance is offered. Field studies and surveys on human trafficking, in addition to previous research by Chattopadhyay and Duflo [47] and Bartilow [48], suggest that the gender violence aspect of the crime should result in Table 4 Estimated regressions Dependent Variables 3P Model 1

CPI Model 2

3P Model 3

Intercept coefficient estimate Standard Deviation p-value

7.788***

32.584***

6.293***

0.3675

3.208

0.4571

<0.0001

<0.0001

<0.0001

WP coefficient estimate

0.0647***

0.4654***

0.0259

Standard Deviation

0.0160

0.1358

0.0157

<0.0001

0.0008

0.1010

p-value CPI

0.0568***

coefficient estimate Standard Deviation

0.0087

p-value R2a n

<0.0001 0.0821 174

0.0615 165

0.2412 165

C. DiRienzo

countries with a greater share of women in government to be in greater compliance with the Protocol. Nonetheless, previous research has found little empirical support for this relationship. It is argued here that these past findings can be a result of the mediating effect of corruption, which in the analyses presented in this study, fully mediates the effect of women in government on country compliance. It is concluded here that the share of women in government does have a significant impact on compliance; however, that effect is indirect and is dependent on corruption. In reference to the implications of this study, the primary contribution reconciles the findings of field studies and survey that conclude that a greater percentage of women in government should have a positive effect on anti-human trafficking efforts and the empirical analyses that do not support this relationship. More specifically, the analyses presented here indicate that both the percentage of women in government and corruption play a significant role in the strength of country compliance with the Protocol. The Protocol is arguably the most significant international anti-human trafficking effort and considering the number of years the majority of the countries have ratified the document, knowledge of the factors that aid and hinder country compliance is critical as the international community continues its battle against human trafficking. Thus, government officials and policy makers need to consider both anti-corruption policy and increased female participation in government as two important anti-human trafficking tools. Policies to increase the share of women in government will obviously be dependent on numerous country characteristics such as culture and a variety of socio-economic factors. Such policies could range from developing educational programs designed to support and encourage women who are interested in politics to large-scale government campaigns that promote the image of women in government. Although this study offers insights into the relationship between women in government, corruption, and country compliance with the Protocol, it is not without limitations. All quantitative measures of largely qualitative variables such as corruption and compliance, must be recognized as having inherent imperfections and must be acknowledged as proxy measures. While the measures used in this study have been used extensively by other researchers, it is noted that the results of this study are dependent on the quality of the data measures used. Further, this analysis concentrated on the primary relationships illustrated in Fig. 2 and did not consider other potentially intervening variables. The results presented here should be viewed in this light. Finally, future research could consider other factors that theoretically should affect anti-human trafficking efforts, but the empirical evidence supporting the relationships is lacking. These studies provide more information for policy makers regarding the relevant drivers of compliance; information that is critical to not only new anti-trafficking policy development, but also for those working on compliance efforts with the Protocol. As Agbu [1] states, human trafficking is the ‘commercialization of humanity’ and if current forecasts are correct, human trafficking will become the largest form of international crime in the next decade. Considering that the Protocol was established over a decade ago, it is critical to the anti-human trafficking effort to understand the factors that drive and hinder compliance as non-compliance enables the crime to continue to grow.

Compliance with anti-human trafficking policies: the mediating...

References 1. Agbu, O. (2003). Corruption and human trafficking: The Nigerian case. West Africa Review, 4(1), 1–13. 2. Studnicka, A. C. S. (2010). Corruption and human trafficking in Brazil: Findings from a multi-modal approach. European Journal of Criminology, 7(1), 29–43. 3. United States Department of State. (2016). Trafficking in Persons Report 2016. https://www.state. gov/documents/organization/258876.pdf Accessed January 11, 2018. 4. United Nations. (2012). Heinous, fast-growing crimes of human, drug trafficking will continue to ravage world’s economies without coordinated global action, Third Committee http://www.un.org/press/en/2012 /gashc4039.doc.htm Accessed March 5, 2017. 5. Dijk, J. V., & Mierlo, F. K.-V. (2014). Quantitative indices for anti-human trafficking policies: Based on reports of the U.S. Department and the Council of Europe. Crime, Law and Social Change, 61, 229–250. 6. Wheaton, E. M., Schauer, E. J., & Galli, T. V. (2010). Economics of human trafficking. International Migration, 48(4), 114–141. 7. Winterdyk, J., & Reichel, P. (2010). Introduction to special issue human trafficking: Issues and perspectives. European Journal of Criminology, 7(1), 5–10. 8. Potrafke, N. (2013). Policies against human trafficking: The role of religion and political insitutions. CESifo Working Paper, No. 4278. 9. Duong, K. A. (2012). Human trafficking in a globalized world: Gender aspects of the issue and antitrafficking politics. Journal of Research in Gender Studies, 2(1), 48. 10. United Nations Office of Drugs and Crime (UNODC). (2014). Global report on trafficking in persons 2014. Retrieved from https://www.unodc.org/documents/data-and-analysis/glotip/GLOTIP_2014_full_ report.pdf. 11. Fitchtelberg, A. (2008). Crime without borders: An introduction to international criminal justice. Upper Saddle River: Pearson. 12. Cho, S-Y & Vadlamannati, K.C. (2010). Compliance for big brothers: An empirical analysis on the impact of the anti-trafficking protocol. Discussion Papers, Center for European Governance and Economic Development Research, No. 118. 13. Cho, S.-Y., Dreher, A., & Neumayer, E. (2014). Determinants of anti-trafficking policies: Evidence from a new index. Scandinavian Journal of Economics, 116(2), 429–454. 14. Cho, S-Y. (2017). The economics of human trafficking. http://www.economics-human-trafficking. org/anti-trafficking-3p.html Accessed March 20, 2017. 15. Heller, L. R., Lawson, R. A., Murphy, R. H., & Williamson, C. R. (2016). Is human trafficking the dark side of economic freedom? Defence and Peace Economics, (2016), 1–28. https://doi.org/10.1080 /10242694.2016.1160604. 16. Cho, S-Y. (2010). International human rights treaty to change social patterns – the Convention on the elimination of all forms of discrimination against women. Cege. Discussion Paper Series No.93. GeorgAugust University of Goettingen, Germany. 17. Neumayer, E. (2005). Do international human rights treaties improve respect for human rights? Journal of Conflict Resolution, 49(6), 925–953. 18. Bales, K. (2007). What predicts human trafficking. International Journal of Comparative and Applied Criminal Justice, 31(2), 269–279. 19. Cho, S.-Y. (2015). Modeling for determinants of human trafficking: An empirical analysis. Social Inclusion, 3(1), 2–21. 20. Rao, S., & Presenti, C. (2012). Understanding human trafficking origin: A cross-country empirical analysis. Feminist Economics, 18(2), 231–263. 21. Surtees, R. (2008). Traffickers and trafficking in southern and Eastern Europe. European Criminology, 5(1), 39–68. 22. McCarthy, L. A. (2010). Beyond corruption: An assessment of Russian law enforcement’s fight against human trafficking. Demokratizatsiya, 18(1), 5–27. 23. Uslaner, E. M. (2008). Corruption inequality, and the rule of law: The bulging pocket makes the easy life. New York: Cambridge University Press. 24. Olken, B. A. (2007). Monitoring corruption: Evidence from a field experiment in Indonesia. Journal of Political Economy, 115(2), 200–249. 25. Rose-Ackerman, S. (2005). The challenge of poor governance and corruption. Especial 1 DIREITO GV L. Rev., 207. 26. Jain, A. K. (2001). Corruption: A review. Journal of Economic Surveys, 15(1), 71–121.

C. DiRienzo 27. World Bank (n.d.) http://www.worldbank.org/en/about/unit/integrity-vice-presidency/what-is-fraud-andcorruption Accessed January 5, 2018. 28. Aguilera, R. V., & Vadera, A. K. (2008). The dark side of authority: Antecedents, mechanisms, and outcomes of organizational corruption. Journal of Business Ethics, 77, 431–449. 29. Johnston, M. (1996). The search for definition: The vitality of politics and the issue of corruption. International Social Science Journal, 48(149), 321–335. 30. Nye, T. S. (1967). Corruption and political development: A cost-benefit analysis. The American Political Science Review, 61(2), 417–427. 31. United Nations Convention against Corruption. (n.d.) https://www.unodc.org/unodc/en/treaties/CAC/ Accessed January 5, 2018. 32. Tanzi, V. (1998). Corruption around the world: Causes, consequences, scope, and cures. Washington, D.C.: IMF Staff Papers, 45(4) International Monetary Fund. 33. Adepoju, A. (2005). Review of research and data on human trafficking in sub-Saharan Africa. International Migration, 43(1/2), 75–98. 34. Gajic-Veljanoski, O., & Stewart, D. E. (2007). Women trafficked into prostitution: Determinants, human rights and health needs. Transcultural Psychiatry, 44(3), 338–359. 35. Jac-Kucharski, A. (2012). The determinants of human trafficking: A US case study. International Migration, 50(6), 150–165. 36. American Civil Liberties Union. (n.d.) https://www.aclu.org/other/human-trafficking-modernenslavement-immigrant-women-united-states Accessed March 21, 2017. 37. Sigmon, J. N. (2008). Combating modern-day slavery: Issues in identifying and assisting victims of human trafficking worldwide.Victims and Offenders, 3(2–3), 245–257. 38. Bales, K., Trodd, Z., & Williamson, A. K. (2011). Modern Slavery: A Beginner's Guide. Oxford: Oneworld Publications. 39. United Nations Global Initiative to Fight Human Trafficking. (2008). Background paper-corruption and human trafficking: The grease that facilitates the crime. Vienna, Austria. 40. Bales, K. (2005). Understanding global slavery. Berkeley: University of California Press. 41. 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. 42. World Bank. (2016). http://www.worldbank.org/en/topic/governance/brief/anti-corruption Accessed March 20, 2017. 43. DiRienzo, C. E., & Das, J. (2015). Innovation and the role of corruption and diversity: A cross-country study. International Journal of Cross Cultural Management, 15(1), 51–72. 44. Holmes, L. (2009). Human trafficking & corruption: Triple victimization? Strategies against Human Trafficking: The Role of the Security Sector, National Defence Academy and Austrian Ministry of Defence and Sports. Vienna and Geneva, 83–114. 45. Uddin, M. B. (2014). Human trafficking in South Asia: Issues of corruption and human security. International Journal of Social Work and Human Services Practice, 2(1), 18–27. 46. Cho, S. Y. (2015). Modelling for determinants of human trafficking. Economics of Security Working Paper, No. 70. https://www.econstor.eu/bitstream/10419/119396/1/diw_econsec0070.pdf Accessed March 24, 2017. 47. Chattopadhyay, R., & Duflo, E. (2004). Women as policy makers: Evidence from a randomized policy experiment in India. Econometrica, 72(5), 1409–1443. 48. Bartilow, H. (2010). Gender representation and international compliance against human trafficking. Manuscript, the University of Kentucky. 49. Biesanz, J. C., Falk, C. F., & Savalei, V. (2010). Assessing mediational models: Testing and interval estimation for indirect effects. Multivariate Behavioral Research, 45(4), 661–701. 50. Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51(6), 1173–1182. 51. James, L. R., & Brett, J. M. (1984). Mediators, moderators and tests for mediation. Journal of Applied Psychology, 69, 307–321. 52. Judd, C. M., & Kenny, D. A. (1981). Process analysis: Estimating mediation in treatment evaluations. Evaluation Review, 5, 602–619. 53. Dollar, D., Fisman, R., & Gatti, R. (2001). Are women really the ‘fairer’ sex? Corruption and women in government. Journal of Economic Behavior & Organization, 46, 423–429. 54. Swamy, A., Knack, S., Lee, Y., & Azfar, O. (2001). Gender and corruption. Journal of Development Economics, 64(1), 25–65.

Compliance with anti-human trafficking policies: the mediating... 55. Esarey, J., & Chirillo, G. (2013). ‘Fairer sex’ or purity myth? Corruption, gender, and institutional context. Politics & Gender, 9(4), 361–389. 56. Sung, H.-E. (2003). Fairer sex or fairer system? Gender and corruption revisited. Social Forces, 82(2), 703–723. 57. Sung, H.-E. (2012). Women in government, public corruption, and the liberal democracy: A panel analysis. Crime, Law and Social Change, 58, 195–219. 58. Berg, E. (2001). How should corruption be measured? MSc economics extended essay. London: London School of Economics and Political Science. 59. Sampford, C., A. Shacklock, C. Connors, and F. Galtung, eds. (2006). Measuring corruption. Ashgate eBook, Grower House, Hampsire, England. 60. Jones, S., Dietsche, E., & Weinzierl, S. (2006). Corruption: Measuring prevalence and costs. Oxford: Oxford Policy Management. 61. Lancaster, T. D., & Montinola, G. R. (1997). Toward a methodology for the comparative study of political corruption. Crime, Law and Social Change, 27, 85–206. 62. Serra, D. (2006). Empirical determinants of corruption: A sensitivity analysis. Public Choice, 126, 225– 256. 63. Iqbal, Z., & Starr, H. (2008). Bad neighbors: Failed states and their consequences. Conflict Management and Peace Science, 25, 315–331. 64. Leicht, K. T., & Jenkins, J. C. (1998). Political resources and direct state intervention: The adoption of the public venture capital programs in the American states, 1974 – 1990. Social Forces, 76(40), 1323–1345. 65. Poddar, T., Goswami, M., Solé, J., and Icaza, V.E. (2006). Interest rate determination in Lebanon. IMF Working Paper, WP/06/94. Available http://www.mafhoum.com/press9/278E12.pdf Accessed January 8, 2018. 66. United Nations. (2012). Heinous, fast-growing crimes of human, drug trafficking will continue to ravage world’s economies without coordinated global action, Third Committee told. Retrieved from http://www. un.org/press/en/2012/gashc4039.doc.htm Accessed March 24, 2017.