The Socioeconomic Factors of Female Child Trafficking and Prostitution: An Empirical Study in the Capital City of Bangladesh

Abstract

Although Bangladesh is a constitutionally Islamic country, some brothels are regulated by state law. But these brothels are located in the suburbs, and there were legal brothels around the capital, which no longer exist. Thus, prostitution is observed in a variety of ways, including in residential hotels, resorts, homes, and open spaces. As prostitution and trafficking are inseparable, this paper adopts a quantitative approach to measure the socioeconomic variables associated with prostitution and trafficking in children in the capital city of Dhaka, Bangladesh. The data for this study were collected from 385 respondents, and the questionnaire format was open-ended. The proposed conceptual model is presented in a way that includes sociocultural and economic factors influencing prostitution. To examine the model, a three-level research design was applied. The sociodemographic data of the respondents were collected and analyzed in this study. This study finds that the significant economic factors are poverty and lack of employment opportunities. Moreover, sociocultural variables are closely associated with rape, harassment, divorce, insufficient support from household members, living in vulnerable conditions, social instability, lifestyle, and gender violence. These findings emphasize the need to implement existing anti-trafficking laws and raise awareness of children in Bangladesh to stop child trafficking for sex work.

1. Introduction

Prostitution or commercial sex in Bangladesh is completely covered by a legal-illegality fence. The constitution of Bangladesh is complicated, as Article 18(2) states that the state shall take effective measures to prevent prostitution, and Article 40 gives the right to every qualified citizen to take up any lawful profession. According to the Bangladesh Immoral Activities Control Act, 1933, known as the Prostitution Act, a woman above the age of 18 can voluntarily take up prostitution as a profession within the framework of the law. This complexity in the legal system led Bangladesh to the subsequent definition that commercial prostitution can be said to be partially legal (Huda 2019; Sultana 2015). However, in the social context of Bangladesh, commercial sex is stigmatized and operates under adverse social conditions (Huda 2019; Shewly et al. 2020; Amanullah and Huda 2012). Nevertheless, many Bangladeshi women and children engage in commercial sex voluntarily and involuntarily.

There are four types of sex workers commonly found in Bangladesh: home-based, hotel-based, park- and garden-based, and brothel-based (Islam and Nath 2012). About two-thirds of the professional sex workers in Bangladesh come from families of agricultural workers, small farmers, fishermen, etc., and the rest come from families of low-income rickshaw pullers or wheelbarrow drivers, sailors, etc. Currently, women from well-off families are also seen in this profession (Karim 2020). According to the United Nations Program on HIV and AIDS (UNAIDS 2023), more than 140,000 Bangladeshi females are involved in commercial prostitution as of February 2020; most of them enter commercial sex primarily through internal human trafficking (Biswas 2015).

Human trafficking is usually referred to as the “sexual trafficking” of women and children because it comprises their sexual abuse (Cimino 2017; Doezema 2002; Gerassi 2015). The United Nations (UN) Protocol to Prevent, Suppress and Punish Trafficking in Persons, Especially Women and Children for Sexual Exploitation defines trafficking as “any act of recruiting, transporting, transferring, harboring or obtaining. Persons, threatening or having control over a person or another person by force or by other forms of coercion, abduction, deception, fraud, abuse of power or position of weakness or giving or receiving money or favors to obtain consent, for exploitation” (United Nations Office on Drugs Crime 2006). The UN Trafficking Protocol involves two aspects. The first aspect is the forced or involuntary entry of women and children into commercial sex. The second aspect is the solicitation of children (under 18 years of age) to engage in sexual activity. As with the first context, the second dimension indicates whether children engaged in commercial sex are believed to be “trafficked” and whether they have agreed to engage in commercial sex (Cimino 2017; United Nations Office on Drugs Crime 2006; United Nations 2000; George and Sabarwal 2013; Sandy 2009; Servin et al. 2015).

In recent years, not only in Bangladesh but also in other Asian countries, women and child trafficking has been increasing alarmingly (Archavantikul 1998; Bennett 1999; Salma 1998). Unfortunately, little effort has been made to understand a significant solution to the local dynamics of such larger social, national, and international problems. This task is not easy, as it involves exploring a difficult field of regulatory perspectives, social systems, economic discrimination, and illegal exploits. Nevertheless, an exploration of the causes of trafficking and a significant explanation of its consequences are inevitable for any future intervention to address this humanitarian problem at the national, regional, and international levels. Despite the considerable number of papers, reports, and broadcast campaigns in various contexts on human trafficking alone (Sultana 2020; Huda 2019; Biswas 2015; Amin and Sheikh 2011), very rare methodical investigations have been carried out on female child prostitution and internal trafficking, especially in Bangladesh.

It is imperative that the understanding of institutions acting to prevent trafficking and support trafficked female children be copiously documented to define significant interventions to improve the effectiveness of current efforts. It is essential to adequately document the knowledge and experiences of agencies acting to resist trafficking and support trafficked female children to identify appropriate interventions to improve the usefulness of present endeavors. The basic objective of this study is to increase the awareness of female children in Bangladesh to implement the existing anti-trafficking laws and stop the trafficking of women and children for sex work. Therefore, this study was undertaken to provide a magnificent synopsis of the obtainable information on female child trafficking for prostitution in Bangladesh. Thus, this paper is expected to contribute to providing a framework for future studies to better understand female child trafficking issues and research gaps.

2. The Theoretical Background of This Study

We know that any research requires a theoretical framework at the outset, and this study is no exception. Female child trafficking pervades various institutions whose systematic operations involve many activities. Apart from describing the practices, processes, and pathways of trafficking, there is a considerable lack of firmness concerning theoretical frameworks for studying human trafficking (Bruckett and Parent 2002). The causes of trafficking vary over time and are interdependent and interconnected. (Truong 2001; Van Impe 2000). Female child traffickers may view human trafficking as a lucrative area for income, especially when it is seen that very few child traffickers are punished severely under a strong justice system for this type of crime (Ara and Khan 2022). Ineffective justice systems and societal responses to child trafficking reinforce the trafficking actions, increasing the exploitation of trafficked persons and providing opportunities for child traffickers to make financial gains from their crimes (Rafferty 2013). A crime is caused by many factors, so an integrated framework exploitation approach is needed to review and analyze the sequence of these events, which may explain why these criminals contribute to crimes like child trafficking (Lanier and Henry 2004). As we know, theory combination is a method of blending the ideal components of subsisting theories to better interpret the roots of transgressive behavior (Brown et al. 2007; Lanier and Henry 2004). Thus, this study combines ideas from demand, rational choice, economic, constructive criminology, and push–pull theories to describe female child trafficking for unwilled commercial sex.

Similar to other trades, rational choice theory is based on the premise that criminals engage in crime based on cost–benefit considerations. Criminal decision-making depends on self-determination and is deterministic in nature, which requires an examination of circumstances, opportunities, and conditions that may influence the effectual commission of a planned crime (Lanier and Henry 2004). Reasonable decision-making about the crime engages the type, method, and where and when it should be committed, and victims’ choices are determined by what to do (Brown et al. 2007). That is, criminals first find out the availability of potential victims, the area, the time they are most vulnerable, and the perfect way to get in easily. Additionally, they consider how to make their criminal activities easier to keep from state judicial authorities and other appropriate stakeholders. Nevertheless, some rationalist theorists argue that criminals make choices depending on their motives, skills, and perceptions, as well as their abilities, facilities, and capabilities, to enable them to cope with difficult situations (Lanier and Henry 2004).

There are three perspectives on the demand for commercial sex, namely the buyer of sex, the intermediary profiteer of sex, and the cultural attitudes of society towards this type of sex (Hughes 2004). This paper identifies and explains the growing social violence against these three factors as well as the cultural attitudes related to prostitution, especially with female children. From these factors, it is hypothesized that users of a certain category of commercial sex do not distinguish trafficked involuntary prostitutes from voluntary prostitutes but may be more concerned about receiving forced sexual services from trafficked female children. Some ideas exist within the victimology paradigm that explain why some children may be more vulnerable to victimization than others. Moreover, the victim–perpetrator interaction, in addition to the reasons for child victimization, is used to explain the nature of child trafficking victims for repeated victimization of the same incident and involuntary commercial sex in their social lifestyle. That is, the connections and interrelationships with traffickers, as well as their social, family, and personal lifestyles, and how often a child is trafficked into involuntary prostitution can be considered an important part of the process of trafficking in female children.

Economic theory could be employed to elucidate the actions, crimes, and behaviors that determine the costs and benefits gained from engaging in a particular activity. This theory essentially assumes that people make decisions about criminal activity in the same way as they make decisions about other normal activities (Witte and Witt 2002). It is believed that if the income generated from illegal work is more than that from any legitimate job, then people in society will turn to criminal activities. The principle of economic theory is that the spread of crime is organized only when the benefits of crime exceed all probabilities of considering its costs (Eagle and Betters 1998; Persson and Siven 2007). As pointed out by Pratt (2008) and Witte and Witt (2002), the main reason people engage in activities in society, whether legal or illegal, is the expectation of that action. Benefit exists and is motivated by the reality that expected profits from crime increase efforts to launder money compared to earnings from legal activities. Another element of the theory of economic crime is that the lower the compensation, the higher the human trafficking. Moreover, the likelihood of criminals being arrested and convicted and the expected punishment will describe the level of crime. This leads to another economic measure of trafficking which is provided by McCray (2006) where the author states that confidence is more significant than intensity. In short, once the rational decision is made, the vulnerability of the potential victims is confirmed and the child trafficking process begins again, taking into account the need for demand. Child traffickers could have studied the legal repercussions of child trafficking by then to ascertain the sanctions they would face if caught by criminal justice authorities.

The push and pull factors are a major factor that makes victims susceptible to falling prey to traffickers’ schemes. Push factors include economic poverty, dysfunctional families, and the tendency to flee social exclusion. These situations force the victim to believe the false claims made by the perpetrators which act as a pull factor. The driving factors behind this scenario are the false promise of a job, a better life, a working environment, and the security of a good income that lures victims into the trap of trafficking rackets. It should be noted that child trafficking and forced prostitution are not autonomous phenomena, but different actors and factors are closely related. The potential for these criminal activities creates a market that simultaneously fuels, promotes, and exacerbates the problem with a large number of actors.

To address the question of what are the strongest predictors of female child trafficking, “push and pull factors” can be used as relevant theories. Trafficking in female children in real terms is a response to a combination of push and pull factors. While push factors force victims to walk down the road of trafficking risk and create a trafficking-like environment, pull factors usually provide false promises and illusions that ultimately lead to exploitative conditions. The actual crime of human trafficking consists of several serious crimes. It is virtually impossible to commit the crime of trafficking in women and children in isolation from other forms of criminal activity. Any investigation into the criminal activities of female child traffickers should always focus on the related crimes that are likely to have been committed.

3. Literature Review

Since the mid-1980s, interest in different approaches to trafficking has grown (Doezema 2001; Kempadoo and Doezema 1998; Kempadoo 2005; Outshoorn 2004), with its origins in the late nineteenth and early twentieth centuries. This evolved into a movement that transcended international boundaries as feminists campaigned against “white slavery” as “sexual slavery” leading to some largely feminist international treaties (Day 2004; Outshoorn 2004; Kempadoo 2005; Sullivan 2003), adopted between 1903 and 1949. The inclusion of both prostitution and trafficking in the international agenda is the result of feminist pressure. This context has become the subject of lobbying by feminist discourse divided into two opposing and controversial perspectives, abolitionism and controls (Lozano and Milena 2011), which are represented in various transnational alliances against trafficking: the Coalition Against Trafficking in Women (CATW) and the Global Alliance Against Traffic in Women (GAATW). However, the intentional or unintentional prostitution of underage women and children continues to be studied from various perspectives of those who are still trafficked.

The recent literature points to trafficking as a primary cause of the entry of women and children into sex work (Gerassi 2015; Dahal et al. 2015). Recent studies have identified various factors such as poverty, social violence, social stigma, strong patriarchal society, job discrimination, lack of proper cooperation from household members and relatives, and lack of proper legislation that favored brokers and traffickers specifically in India (George and Sabarwal 2013; Devine et al. 2010; Swendeman et al. 2015), Iran (Karamouzian et al. 2016), the United Kingdom (Dodsworth 2015), and the United States of America (Gerassi 2015; Deshpande and Nawal 2013; Cobbina and Sharon 2011). These factors make women and children vulnerable to trafficking (Goldenberg et al. 2014). It has also been observed that traffickers target economically and socially vulnerable women and children who perform menial jobs to help their lives and family needs (Cimino 2017; Dodsworth 2015). Furthermore, pieces of evidence from other regions explore that human traffickers use psychological abuse, drug use, and physical abuse as tools to force women and children into prostitution, thus limiting their agency over their bodies, earnings, and safe associations with clients (Kaya and Erez 2018).

If we look at women’s engagement in voluntary commercial sex, we find a large amount of domestic and international literature. For instance, research on foreign women in Turkey (Kaya and Erez 2018) and immigrant women in Canada (Goldenberg et al. 2017) found that personal and interpersonal factors, especially financial issues, influence women’s voluntary involvement in commercial sex. The study also explored factors such as poverty, large economic disparities between women and men, very high unemployment rates, very limited employment opportunities, and low pay that motivated women’s desire for commercial sex as a source of income. However, there is insufficient scholarly literature on the involuntary inclusion of children in sex work in contexts of limited resources.

In particular, there is limited peer-reviewed research on the involuntary involvement of trafficked children in commercial sex in Bangladesh. Most of the studies are based on a literature review where the international trafficking of women and girls is given more importance than domestic trafficking (Biswas 2015; Amin and Sheikh 2011). Furthermore, a qualitative study by Huda et al. (2022) found that trafficking is the main reason why migrant women engage in sex work, as traffickers trap socially and economically marginalized women and sell them to madams, and because of social barriers, once they have engaged in sex work, they rarely return to mainstream society. Moreover, various reasons have been cited for women’s movement into forced and suo motu sex, including low income, lack of employment opportunities, and divorce (Sultana 2020). In the socioeconomic conditions of Bangladesh, females find themselves in some deprived situations due to several factors including patriarchal values, their economic dependability on men, and their disadvantaged situations in a society of gender discrimination (Huda et al. 2022). These determinants may be exploited in their participation in sexual activity (Huda 2019).

To our knowledge, no studies have explored how socioeconomic factors affect entry into sex work unintentionally, specifically from the perspective of child prostitution, except in brothels in Bangladesh. In this paper, the authors have chosen to use the term “female child prostitution”, which, while deliberated pejorative by some, is coherent with the more encyclopedic and recommended terminology of the Joint United Nations Programme on HIV/AIDS (UNAIDS 2015). This paper aims to examine how multiple factors, in particular socioeconomic and regulatory perspectives, affect Bangladeshi female children’s forced and involuntary involvement in commercial sex. This study examines female children’s entry into prostitution since it acts as a tapered end at which interventions could potentially be planned to help them cease their forced involvement in commercial sex and to help them quit (if they want to).

Conceptual Framework and Hypothesis Development

National and international organizations use a variety of definitions to describe trafficking. Analyzing these definitions, it can be seen that the main focus of trafficking is on the causes of trafficking and the age and gender of victims, as well as the various forms of coercion and violence associated with trafficking. For this study, a trafficked female child is defined as a child under the age of eighteen, who is intentionally or unintentionally transferred by an intermediary, with or without her consent, from one place to another within the country’s borders, forced or semi-forced or induced to engage in the commercial sexual exploitation of the child (Archavantikul 1998; Salma 1998). It is to be noted that under section 2 (k) of the Violence Against Women and Children Act 2000 in Bangladesh, a child is below sixteen years of age, but according to the UN Charter on the Rights of the Child, everyone below the age of eighteen years is counted as a child, so the age limit in this paper is below eighteen years old.

In Bangladesh, caused by growing landless households and limited investment in rural industrialization, job opportunities and skill development opportunities are limited, especially in the rural areas. Traditionally, Bangladeshi women have worked as unpaid housewives in a family, and loan/credit facilities have been controlled by them. Despite that, the demand for women and child labor in the urban informal economy, particularly in the garment industry and as domestic workers, has increased, and an increasing number of women and children are engaged in the labor force in urban areas (Amin et al. 1998). Thus, fewer employment opportunities for females in the countryside and increasing demand for labor in the formal and informal economic sectors in cities pull them to migrate from rural to urban areas. This migration trend originates vulnerable situations, particularly for female children, and provides an opportunity for human traffickers to exploit women and children.

The subordinate position of women in the patriarchal social system in Bangladesh is viewed from a feminist perspective (Shewly et al. 2020). Moreover, an unequal social status always exists between men and women in a patriarchal society, which feminists call gender class (Kandiyoti 1988). Indeed, men are the heads of the family because they are the breadwinners; women are dependent on men, and their roles are respected based on the criteria of reproductive ability by meeting the biological needs of men. This generates specific gender behaviors by attaching potential values and norms to everyone. On top of this, the idea of sexuality is built again, where women’s sexuality should be kept under the control of men. Moreover, this concept of women’s bodily control and protection of rights leads to a weak exploitative social system for women, where the slightest physical deviation, sexual deviance, or displacement socially pollutes them, leading to socially humiliating conditions.

The socialization systems in the household regulate the character of a female child as a full-time mother and wife in the patriarchal society (Liao and Cai 1995). Female children grow up with a dependent mindset in a patriarchal social environment, where their social contribution to the family is denied, as they are largely recognized by society as unpaid housewives. This situation is not confined to the rural social system but exists in both rural and urban areas. The girl child is prepared from her childhood so that she can establish herself well in the in-law’s house as a suitable housewife. Female children face the risk of sexual abuse even within the family; sometimes, the accommodation poses a threat to their safety.

The purpose of this study was to develop an integrated framework for describing the trafficking of female children for involuntary commercial sex. All variables were drawn from existing theory to create a magnificent policy framework. The main reasoning is that since the trafficking of female children involves a large group of actors from different backgrounds, there cannot be a sole interpretation for its cause. Given these realities, the responses to trafficking in involuntary prostitution should examine the significance of an integrated conceptual framework for the effective prevention and protection of female child victims of trafficking.

A proposed representation of the research model and hypotheses is shown in Figure 1 below.

Based on the above discussion, three hypotheses are presented for this study, which are as follows:

H1. 

There is a relationship between economic factors and female child trafficking and prostitution.

H2. 

There is an association between sociocultural variables and female child trafficking and prostitution.

H3. 

Regulatory drivers affect female child trafficking and prostitution.

In this study, economic factors are a combination of seven variables (primary needs, secondary needs, educational expenditure, living expenditure, family healthcare, poverty, and limited employment opportunities) identified based on the above discussion. Similarly, sociocultural variables were identified as a composite of fifteen variables (rape, harassment, exploitation, child marriage, divorce, limited support from family members, limited support from friends and relatives, feeling of disempowerment, desire to be autonomous, living in vulnerable conditions, social instability, lifestyle, gender violence, deceive with false promises, and falsely offering work). Finally, the regulatory drivers are calculated by including three important factors (no national identification, no punishment, and easily released from court).

4. Materials and Methods

The collection of primary data for this article had to overcome several challenges. Research-related challenges that may limit data quality and concerns about research impacts on participants, researchers, and others were identified. Challenges related to data quality include the age verification of study participants, procedures for obtaining informed consent from parents or legal guardians, and interviews. Concerns about the research implications were related to the fact that participation in research could further exploit female children, refraining from acting as a researcher, and providing advocacy. Because the process of collecting data from female children can be complex and fraught with ethical and practical challenges, we have presented our experiences sequentially.

4.1. Study Design

This study relied on a cross-sectional design with a quantitative approach. The survey questionnaire was finalized by conducting a pilot study with thirty respondents (Browne 1995). This pilot work was conducted to determine the final sampling method for this study. That is, this study was also performed to raise awareness of methodological issues that were not reported. Although we know that conducting a pilot study does not carry as much importance as the success of the original study, it greatly increases the likelihood of success. As such, it fulfills a wide range of important functions and provides valuable insights for other researchers. The process and results of the pilot study were discussed among the researchers themselves, leading to the final study. In this study, while conducting and analyzing the pilot trial, the researchers were able to identify the weaknesses that they encountered, which helped in the final design of this study and the initial improvement of the entire process.

4.2. Study Sample

It is very critical, although not impossible, to accurately ascertain the actual number of victims of female child sex trafficking and its various forms (Goodey 2008; Logan et al. 2009). Typically, the secrecy of such crimes makes estimating the true extent of the problem complicated because victims are often invisible to mainstream society, culture, and law enforcement. Moreover, there is currently no uniform system for collecting data on victims. In most cases, the nature of the crime is covert, and the victims may not self-identify as having been trafficked (Bales and Lize 2007; Goodey 2008; Hodge and Lietz 2007; Jones et al. 2007; Logan et al. 2009). Furthermore, other reasons for under-identification include that victims may fear the consequences of being identified as criminals and social victimization rather than being identified as victims of sexual exploitation or trafficking (Rafferty 2008), such as being arrested. Additionally, victims may experience feelings of shame and stigma, and their families may be aware of their victimization (Crawford and Kaufman 2008; Jones et al. 2007).

Dhaka is the capital of Bangladesh, and Dhaka Judge Court is the judicial court for about 1.5 million people of the entire district. The population and sampling framework for this study was determined by the logbook of women and children arrested for unlawful prostitution in this court. The quantitative survey took six months (15 May–15 September 2023). The recommendations of Krejcie and Morgan (1970) were followed in selecting the sample size for this study; those under eighteen years of age who appeared in the Dhaka Judge’s Court, Bangladesh, to be arrested for engaging in illegal sexual activity were identified. In the convenience sample selection process, 385 respondents were selected (Zhang et al. 2015).

4.3. The Study Site

The Dhaka Judge Court is a judicial court comprising fifty metropolitan police stations in Dhaka, the capital of Bangladesh. Every day from these police stations, people arrested for various crimes are produced in the court, and along with them, almost every day some women and children are arrested who claim in the police court that they were engaged in antisocial activities in different houses, parks, and residential hotels of the city. This antisocial act is essentially sex for money. Every day, approximately 400–500 women and children are brought to the court in different groups of 15–20 with the same rope. First, the officials of the concerned police stations present their documents to the court and keep the accused in a closed room where they have to stay for a long time. In the second stage, when all of them are presented to the court, the court gives three types of decisions after reviewing their documents and hearing them. Those suspected to be involved in the trafficking of women and children are taken into police custody for questioning at the behest of the police. Secondly, considering the seriousness of the crime, the court orders them to be sent to jail for further trial proceedings. Thirdly, even if they engage in sexual activity, they are freed by paying a financial penalty, instead of considering it as a social crime. It is to be noted that those below 18 years of age are handed over to their legal guardians or appropriate authorities in the rehabilitation center.

4.4. Data Collection and Analysis

For this study, after collecting information from the daily logbooks on those enrolled aged 18 and below with the permission of the court, their parents/legal guardians were identified through their appointed lawyers. For those who did not have parents, a representative of the rehabilitation center was identified as a legal guardian. Both representatives were informed in detail about the purpose of this study, the nature of data collection, the mode of analysis, and publication. A written informed consent form was also obtained from those interested in participating, and the time and place were determined. It is mentioned here that some were interested in providing information to the second author’s chamber, while others provided their home address, and the information was collected at that location. Those who were the representatives of the rehabilitation center gave their addresses for data collection. A written informed consent form was obtained from the competent authority of the rehabilitation center. Since all the participants of this study are Bangladeshi citizens, the consent forms were signed by those who have the legal right to sign according to the Bangladeshi regulatory perspectives; thus, 385 consent forms were collected for 385 respondents.

The survey questionnaire was developed in collaboration with the research team based on quantitative data collection and an extensive review of the existing literature directly related to this study (Amanullah and Huda 2012; Devine et al. 2010; Swendeman et al. 2015; Karamouzian et al. 2016; Deshpande and Nawal 2013; Zhang et al. 2015). As this study had purely a quantitative approach, considerable care was taken while developing the questionnaire. The questionnaire was open-ended, where questions were coded according to the appropriate level and scale, keeping consistency with each variable. All interviews were conducted in Bengali (participants’ mother tongue) under the overall supervision of the second author. After confirming the age of the interviewees, 4–5 interviews were conducted each day subject to the verbal consent of each participant on the condition of anonymity. The duration of the questionnaire interview was between 45 and 60 min.

The normality of distributions of latent variables with interval and ratio scales was tested by examining the shape of the distribution of the data collected, with a normal outcome defined by a bell shape (Neuman 2006). The normality of data distributions was tested by calculating the scores of kurtosis and skewness for the frequency of values for the explanatory variables. The P-P (Proportion–Proportion) probability plot is used as a normal distribution method that indicates that the variables are normally distributed about their means. Moreover, there were two major considerations: the data multicollinearity and autocorrelation, as Gujarati (2003) suggests that it is not necessary to carry out all the available assumption tests, as some were not relevant to this study. Consistency within all the items in the question matrix was ascertained by a reliability check using Cronbach’s Alpha. As long as the value of Cronbach’s Alpha is between 0.7 and 1.0, all of the statements in the question matrix can be used, while statements with a value below 0.7 are discarded. A reliability test provided the value of Cronbach’s Alpha (0.831) that confirmed the availability of consistency components in all the statements. The statistical software package IBM SPSS Statistics 26.0 (CAPCDR, Shariatpur, Post Code-8023, Bangladesh) was used to enable the statistical analysis.

4.5. Ethical Issues

Apart from in medical science, there is no institutional authority to provide ethical approval in Bangladesh for this kind of research nor is there any well-informed authority, so no institutional ethical approval was acquired. However, the second author of this study is ethically responsible as per the Bangladesh Legal Practitioners and Bar Council Order, 1972, as an honorable member of the Bangladesh Bar Council and Dhaka Bar Association in collecting data.

In studies in social sciences and humanities, data collection ensures a robust system by identifying measures to secure participants’ informed consent, guarantee their anonymity, achieved through the use of pseudonyms, anticipate risks, and avoid harm to all participants, including the researcher. From the beginning to the end of the fieldwork, four female research assistants arranged meetings with the participants whose presence helped to develop trust and openness. Several practical arrangements were made to ensure communication between all participants and the researcher (Sultana 2007). Meeting times, places, and spaces were arranged to avoid any participant being alone, or feeling alone, and participants were able to approach a female research assistant, in confidence, at any time, if they had concerns. Before consenting to participate, either in writing or via audio recording, participants were assured that they would be able to withdraw from this study, temporarily or permanently, at any time, for any reason or none, without repercussion. Recognizing the economic situation of prospective participants, meals were provided for their families, and drinks and snacks were provided during the meeting.

4.6. Sociodemographic Statistics

Table 1. Background of the participants (n = 385).

Sociodemographic Characteristics	n (%)
Age
12–14 years	61 (15.85)
15–16 years	123 (31.95)
17–18 years	201 (52.20)
Level of education
No formal education	21 (5.45)
1–5 years of education	79 (20.52)
6–10 years of education	107 (27.80)
11–12 years of education	178 (46.23)
Marital status
Never married	293 (76.10)
Married	46 (11.95)
Divorced	29 (7.53)
Widowed	17 (4.42)
Years of involvement in commercial sex
1–3 years	286 (74.28)
4–6 years	65 (16.89)
7–9 years	34 (8.83)
Monthly average total income (USD)
<100	130 (33.76)
101–300	188 (48.84)
301–500	22 (5.72)
501–700	19 (4.93)
701–900	17 (4.42)
901–1000	9 (2.33)

summarizes the sociodemographic backgrounds of the 385 participants in this study who participated in the survey questionnaire. The age range of the participants was between 12 and 18, and the average age was 15 years. Surprisingly, sixty-one (15%) participants were below 14 years old. Twenty-one had no formal education (5.45%); seventy-nine completed primary education (20.52%); one hundred and seven completed secondary education (27.80%); and one hundred and seventy-eight (46.23%) participants had higher secondary education. Two hundred and ninety-three were single, forty-six were married, twenty-nine were divorced, and seventeen participants were widowed. Female child participants in this study were engaged in commercial sex between the ages of one and nine. Participants’ monthly income ranged from BDT 5000 (USD 41) to BDT 100,000 (USD 833).

4.7. Empirical Model and Quantitative Results

The econometric model developed in Equation (1) was examined to determine the associations between female child prostitution and the independent variables. The same model has been used in several studies (Mannan and Farhana 2023; Mannan et al. 2020; Mannan 2017). It is a model that uses the “backward elimination” method of regression, which is a technique where the first step is to include all explanatory variables, and the final model is gradually eliminated by removing variables of value one by one, those that are highly insignificant. In this study, the backward elimination regression was conducted in three stages to determine a good-fit model that determines the utmost authoritative determinants. This conforms to the explication of data analysis by Slatten et al. (2011) and Juhdia et al. (2013).

However, before starting this backward elimination technique, statistically, it is necessary to follow particular elementary statistical contemplations of the collected data so that the results of this study are accepted beyond all doubt (Hocking 1976). To achieve this objective, the suggestions of Gujarati (2003) were followed: data normality was measured first, four types of multicollinearities were tested in the second step, and the level of autocorrelation was examined in the final step, while assumption tests were not significant to this study.

To construct a best-fit regression process, 25 prospective independent factors were determined for the operated model illustrated in Equation (1), as follows:

FDctP = β + β₁ PEnd + β₂ SEnd + β₃ EEex + β₄ LEex + β₅ EFhl + β₆ EPty + β₇ ELeo + β₈ SRpe +
β₉ SHnt + β₁₀ SEon + β₁₁ SCge + β₁₂ SDce + β₁₃ SLfm + β₁₄ LSfr + β₁₅ SFnt + β₁₆ DSus + β₁₇ SLvl +
β₁₈ SSin + β₁₉ LSle + β₂₀ SGvi + β₂₁ DSfp + β₂₂ FSow + β₂₃ LNni + β₂₄ NLpu + β₂₅ NErl + e₁

(1)

The model variable identification is given in Appendix A.

e₁ = error terms and π; the assumption is satisfied with the following:

(i) E(e₁) = 0; E(π₁) = 0 (zero mean);

(ii) E(e₁)² = ∂e²; E(π₁)² = ∂π² (constant variance);

(iii) e₁ and π₁; E(e_1j) = 0 and E(π_1j) = 0, where 1 ≠ j (there is no autocorrelation between errors).

The results of the first regression Equation (1) are presented in

Table 2. The results of the full model.

Determinants	Dependent Variable
	Female Child Trafficking and Prostitution (FDctP)					Collinearity Statistics
	Unstandardized Coefficient		Standardized	t-Value	p-Value	Tolerance	VIF
Primary needs	−0.488	0.139	−0.310	−3.341	0.002	0.247	4.031
Secondary needs	0.231	0.085	0.242	2.435	0.624	0.341	6.242
Educational expenditure	−0.810	0.546	−0.927	−3.278	0.757	0.478	5.397
Living expenditure	−0.810	0.329	−0.150	−2.424	0.015	0.457	2.184
Family healthcare	−3.866	0.916	−0.339	−3.487	0.657	0.456	4.681
Poverty	0.233	0.086	0.141	2.535	0.013	0.562	1.774
Limited employment opportunities	0.388	0.084	0.358	4.239	0.001	0.263	3.782
Rape	0.407	0.112	0.435	4.113	0.002	0.171	5.866
Harassment	2.011	0.426	0.355	4.632	0.001	0.321	3.107
Exploitation	0.008	0.079	0.021	0.205	0.837	0.688	6.126
Child marriage	0.034	0.224	0.023	0.168	0.852	0.395	2.522
Divorce	2.683	0.757	0.315	3.511	0.002	0.232	4.291
Limited support from family members	−0.289	0.083	−0.414	−3.222	0.002	0.126	7.862
Limited support from friends and relatives	0.257	0.468	0.972	0.268	0.618	0.497	2.438
Feeling of disempowerment	−0.019	0.973	−0.068	−2.049	0.728	0.896	3.788
Desire to be autonomous	2.683	0.757	0.355	4.632	0.002	0.232	4.291
Living in vulnerable conditions	0.233	0.086	0.141	2.535	0.001	0.562	1.145
Social instability	2.683	0.757	0.315	3.511	0.002	0.232	4.291
Lifestyle	2.011	0.426	0.355	4.632	0.001	0.321	3.107
Gender violence	0.233	0.086	0.141	2.535	0.011	0.562	1.774
Deceive with false promises	0.388	0.084	0.358	4.239	0.001	0.263	3.782
Falsely offering work	2.417	0.802	0.406	2.988	0.002	0.095	10.454
No national identification	0.884	0.669	0.088	0.861	0.818	0.757	2.877
No punishment	0.257	0.468	0.972	0.268	0.618	0.497	2.438
Easily released from court	0.327	0.129	0.355	4.632	0.001	0.321	3.107
Intercept							−45.231
R2							0.641
Adjusted R2							0.575
F-statistics							12.394
Sum squared residual							538.254
Durbin–Watson statistics (d)							1.782

.

For the first regression Equation (1), the results of the complete model are furnished in Table 2. In light of this equation, the experimental results show that some of the explanatory variables are statistically insignificant. The three values measured to express the accuracy of the equation are R² (0.641), F value (12.394), and a “p” value close to zero. Thus, these statistical outcomes suggest that all independent variables influence female child trafficking and involuntary prostitution.

Rules of thumb state that “Tolerance (TOL)” and “Variance-Inflating Factor (VIF)” values close to 1 are sufficient evidence that one independent variable is not collinear with another independent variable (Gujarati 2003). The values of TOL and VIF in

Table 3. The results of the second model.

Determinants	Dependent Variable
	Female Child Trafficking and Prostitution (FDctP)					Collinearity Statistics
	Unstandardized Coefficient		Standardized	t-Value	p-Value	Tolerance	VIF
Primary needs	0.165	0.128	0.058	03.461	0.677	0.588	1.698
Living expenditure	−0.441	0.131	−0.245	−4.260	0.001	0.357	2.793
Poverty	−0.412	0.280	−0.066	−1.466	0.003	0.591	1.691
Limited employment opportunities	0.504	0.054	0.486	9.112	0.002	0.427	2.336
Rape	0.391	0.081	0.405	4.876	0.003	0.175	5.666
Harassment	3.406	0.498	0.605	6.827	0.001	0.154	6.433
Divorce	2.594	0.483	0.320	5.361	0.000	0.341	2.924
Limited support from family members	−0.036	0.022	−0.061	−1.605	0.002	0.873	1.113
Desire to be autonomous	0.563	0.162	0.214	3.461	0.001	0.317	3.142
Living in vulnerable conditions	1.573	0.294	0.261	5.336	0.004	0.508	1.962
Social instability	0.217	0.101	0.084	2.147	0.002	0.617	1.616
Lifestyle	1.681	0.363	0.205	4.620	0.001	0.781	1.281
Gender violence	1.573	0.294	0.261	5.337	0.005	0.508	1.962
Deceive with false promises	0.155	0.138	0.057	03.562	0.697	0.578	1.898
Falsely offering work	1.681	0.363	0.205	4.622	0.002	0.781	1.281
Easily released from court	−0.024	0.082	−0.012	−0.301	0.764	0.611	1.634
Intercept							−31.283
R2							0.589
Adjusted R2							0.581
F-statistics							41.743
Sum squared residual							529.348
Durbin–Watson statistics (d)							2.112

explore that there is no multicollinearity remaining among the independent variables.

As mentioned earlier, variables are selected for regression in an order depending on their statistically non-significant “p-value” in the equation. For example, the regression results of the first Equation (1) in Table 2 show that R² = 0.641, and the adjusted R² = 0.575 is an acceptable value of d = 1.782. The removal method was initiated by rejecting the variable child marriage (SC_ge) which had the highest p-value (0.852) from the first equation. This elimination process continued until a best-fit model for the independent variables was established. The results of the complete backward removal method are given in Table 3. The result is the first best-fit model depicted in the following equation:

FDctP = β + β₁ PEnd + β₂ LEex + β₂₃ EPtv + β₄ ELeo + β₅ SRpe + β₆ SHnt + β₇ SDce + β₈ SLfm + β₉ DSus +
β₁₀ SLvl + β₁₁ SSin + β₁₂ LSle + β₁₃ SGvi + β₁₄ DSfp + β₁₅ FSow + β₁₆ NErl + e₁

(2)

The results derived by applying Equation (2) are revealed in Table 3. They indicate that the R² is a bit reduced (0.589) in proportion to the first equation (0.641) with 16 independent variables. We know that generally as the number of variables in a model increases, the value of R² increases and vice versa. The p-values of the three independent variables, primary demand (PE_nd: 0.677), deception by false promises (DS_fp: 0.697), and easy release from court (NE_rl: 0.764), were statistically insignificant. Consequently, a further backward removal method was conducted to obtain the best-fit model. This elimination process was completed until a best-fit model for the independent variable was achieved. The outcomes of the complete backward removal methods are provided in

Table 4. The results of the best-fit model.

Determinants	Dependent Variable
	Female Child Trafficking and Prostitution (FDctP)					Collinearity Statistics
	Unstandardized Coefficient		Standardized	t-Value	p-Value	Tolerance	VIF
Poverty	−0.472	0.101	−0.267	−4.717	0.001	0.381	2.631
Limited employment opportunities	−0.471	0.277	−0.075	−1.691	0.002	0.605	1.651
Rape	0.517	0.541	0.501	9.558	0.003	0.446	2.237
Harassment	0.401	0.081	0.413	4.992	0.001	0.176	5.633
Divorce	3.541	0.468	0.631	7.555	0.002	0.175	5.672
Limited support from family members	2.867	0.475	0.331	5.638	0.001	0.352	2.836
Living in vulnerable conditions	−0.038	0.022	−0.061	−1.682	0.002	0.887	1.125
Social instability	0.614	0.281	0.276	5.895	0.002	0.337	2.958
Lifestyle	1.667	0.282	0.233	3.891	0.001	0.853	1.461
Gender violence	0.417	0.441	0.401	8.558	0.001	0.346	2.137
Intercept							−30.542
R2							0.631
Adjusted R2							0.614
F-statistics							47.235
Sum squared residual							527.337
Durbin–Watson statistics (d)							2.031

.

The best-fit model result is provided in Equation (3):

FDctP = β + β₁ EPtv + β₂ ELeo + β₃ SRpe + β₄ SHnt + β₅ SDce + β₆ SLfm + β₇ SLvl +
β₈ SSin + β₉ LSle + β₁₀ SGvi + e₁

(3)

The best-fit model explored in Table 4 consisted of only ten independent variables with confidence levels of 1% to 5%. Both models, long and short regression, capture the degree of power of causality between the dependent and independent variables, namely, female child trafficking and prostitution, and the independent variables.

Analyzing Table 4 shows that the value of the adjusted R² is 0.614. And the significance of the overall results exists. Moreover, F-test results are also highly significant at a confidence level of 1% in both regression equations with all variables. Finally, the significant influential variables are poverty (EP_ty), limited employment opportunities (EL_eo), rape (SR_pe), harassment (SH_nt), exploitation divorce (SD_ce), limited support from family members (SL_fm), living in vulnerable conditions (SL_vl), social instability (SS_in), lifestyle (LS_le), and gender violence (SG_vi).

5. Discussion

In the first long regression model, nine variables found no statistically significant association, but the remaining sixteen variables were highly significant at the confidence level of 1% or 5%. At the second level (abbreviated regression), three variables were regressed which were statistically highly insignificant, namely, primary demand, cheated by false promises, and easy release from court. Thus, the significant variables are poverty, lack of employment opportunities, rape, harassment, divorce, insufficient support from household members, living in vulnerable conditions, social instability, lifestyle, and gender violence.

This study showed that initially no association was found with fourteen variables out of twenty-five variables consisting of seven economic, fifteen sociocultural, and three regulatory drivers. Although all these variables are related to female child trafficking and prostitution in some countries, the same form is observed in other countries. Among the significant variables were two economic (poverty and limited employment opportunities) and seven sociocultural (rape, harassment, divorce, limited support from family members, living in vulnerable conditions, social instability, lifestyle, and gender violence) variables for a total of nine but no regulatory variables. According to the push–pull theory, these nine variables are included in the push factors. Basically, in this study, the variables are mainly taken as the push factors, so this type of result is obtained. Therefore, based on this result, it can never be said that female child trafficking and prostitution are the result of only push factors and that no pull factors are effective in this case (Vanwesenbeeck 2013).

5.1. Economic Factors

Our study reveals covariance with two economic push variables that corroborate the first hypothesis (H₁). However, we could not find any similar studies in the context of Bangladesh. However, these findings are similar to those of some recent studies from different countries (Karamouzian et al. 2016; Goldenberg et al. 2014; Goldenberg et al. 2017; Ingabire et al. 2012). Although these studies find both push and pull variables present, our study did not take into account pull factors. Moreover, this type of bias may also be partially because these studies included participants who voluntarily chose sex work as employment. More specifically, this study found poverty to have a very strong relationship with child trafficking and prostitution. This result is consistent with other contemporary studies (Amin and Sheikh 2011; Aronowitz 2013; UNICEF 2005; IOM 2004). Moreover, the concept of unemployment has long become less and less significant in framing the critical directions of the interplay between young people and employment (Gunderson and Krashinsky 2014). Therefore, this study suggests that the state should take action to ensure that female children are not easily lured into trafficking and commercial prostitution as a practical reaction to accomplish their economic and family deficiencies and to support themselves.

5.2. Social Factors

The present study found that trafficked children engage in sexual activity due to several social push factors, which supports the first hypothesis (H₂). This may be because these children are often tricked into commercial sex and trafficked by trusting their relatives or do not understand the tricks and deceptions of the traffickers. Moreover, a possible reason is their insufficient awareness and lack of adequate knowledge about trafficking (Huda 2019). Furthermore, in the social system of Bangladesh, women are generally considered passive, submissive, and unpredictable (Huda et al. 2022). Culturally, women are only engaged in child rearing and food preparation at home and are completely dependent on men economically. Therefore, such a status of women in society, especially in rural areas, makes them vulnerable to social and cultural barriers such as family abuse and divorce, thereby leading to the belief of others that they cannot return to the sex profession after being cheated and trafficked (Huda 2019). Therefore, this study suggests that governments and other stakeholders should take the necessary steps to stop female child trafficking for commercial sex work. Moreover, it is expected to provide critical insight into the perspective of gender inequality in society.

5.3. Regulatory Factors

This study found no association with the legal variable, which means that the third hypothesis (H₃) is completely rejected. This view of trafficking and child sex work leads to a semi-legal status of commercial sex in Bangladesh (Huda 2019) which might increase the risk of child abuse and exploitation (Sabet and Ahmed 2012). However, Bangladesh is not yet a signatory to the Trafficking Protocol to Prevent, Suppress and Punish Trafficking in Women and Children (US Department of State 2017). Although the Human Traffic Deterrence and Suppression Act 2012 seeks to prevent child trafficking and seeks to prevent and suppress human trafficking and implement the protection and rights of victims of human trafficking, as well as to ensure safe migration, this enactment has resulted in a slight reduction in international trafficking but has not had much impact on internal trafficking (US Department of State 2017). Consequently, the government should emphasize the appropriate enforcement of existing trafficking rules and regulations by identifying trafficking syndicates to diminish the exploitation of children, improve their agency, and thus increase their socioeconomic capacity.

The main strength of this paper is that its core participants are the most vulnerable population in Bangladesh. The information on the involvement of female children trafficked for commercial sex that this article has generated may be helpful for future research in a similar context. Second, the present study comprehensively discusses the fundamental research areas that have explored the economic, regulatory, and sociocultural factors that shape children’s participation in prostitution in previous studies. Third, this study included different stakeholders in addition to children in trafficked sex work, providing different perspectives on forced engagement in commercial sex.

6. Conclusions

In this study, data from a primary survey were analyzed using a quantitative approach to understand what factors influence female child trafficking and prostitution. This study developed an econometric model with the best fit also found and applied in explaining the connection between female child trafficking and prostitution and factors. The analysis used fifteen sociocultural multipliers, seven economic variables, and three regulatory drivers, totaling twenty-five factors, and in-depth analysis. The results showed the initial association of 14 variables, and no association was found with 16 variables. After reanalysis by removing most of the uncorrelated variables, it was found that there were strong correlations with basically nine variables. Among them, there were two economic factors, namely, poverty and lack of employment opportunities. The significant sociocultural multipliers were rape, harassment, divorce, insufficient support from household members, living in vulnerable conditions, social instability, lifestyle, and gender violence.

This research also concludes that sociocultural factors have a greater influence on female child trafficking and prostitution. Overall, the regulatory factors were not found to be the major factors that influence trafficking and prostitution. Our study was limited to only trafficking (coercion) and excluded non-trafficking (economic). Apart from that, no data were collected from any registered brothels. Trafficking and female child prostitution were not analyzed separately in this study but may be included as a methodological direction for future analysis. Again, the sample size of this study is another limitation. By avoiding these limitations, research could be conducted with a larger sample that includes other stakeholders such as political representatives and human rights activists, comparing it with other commercial cities and suburbs in Bangladesh.

Our findings from this study have a few critical implications. First, it emphasizes the need to implement existing rules and regulations to diminish the trafficking of children for commercial sex. The appropriate enforcement of existing rules and regulations has the effect of dismantling human trafficking syndicates and protecting children from trafficking and prostitution. Additionally, the competent authorities should raise awareness about trafficking so that no children are trafficked for prostitution. Second, this study suggests safety net programs for children who are forced to engage in commercial sex due to their understanding of adverse social conditions. Finally, our findings emphasize that trafficking and sex work regulation can improve the survival strategies of trafficked and non-trafficked individuals.