Cross-Border Migration of Itinerant Immigrant Retailers in Ghana.

A few studies in West Africa describe the state of borders, particularly their poorly governed and porous nature. This study strove to add to the scant literature by assessing the cross-border migration patterns (modes of travel and routes of entry) of seven hundred and seventy-nine (779) West African itinerant immigrant retailers in the Accra Metropolitan Area of Ghana. The study was cross-sectional and adopted the concurrent nested mixed methods research approach. Interview schedule and interview guide were used to collect the data for analysis. The Statistical Package for the Social Sciences (SPSS) version 21 was used to analyse the quantitative data. The results show that most of the immigrants travelled to their destination (Ghana) by land, using buses and mostly through unapproved routes. Immigrants without travel documents and those who regarded border officials as extortionists were amongst those who mostly travelled through unapproved routes. Statistically, significant relationships were found between routes of entry and key demographic variables such as country of origin, gender, and marital status. It emerged that immigrants who were unmarried men and those from Togo, Burkina Faso, Niger, and Nigeria have the highest odds of migrating to the destination through unapproved routes. The study recommends that government through the Ghana Immigration Service should roll out effective border surveillance to ensure the orderly and regular migration of immigrants into the country. Also, any migration management policy which aims at stemming the current migration trend should be contextualised by taking into account the background characteristics of the immigrants involved.

Introduction

The world population is increasingly becoming more mobile due to demographic and economic disparities between countries, accessibility of modern telecommunication, media, and transport technologies (Population Reference Bureau [PRB], 2015; Vanyoro, 2019). These developments have exposed people in many parts of the world to life-changing opportunities elsewhere, motivating them to migrate to take advantage of such opportunities (Devillard et al., 2015). Today, the issue of transnational migration has become pivotal in national, regional, and global discourses and policy agenda. Asare (2012) argued that the increased attention on international migration stems from the fact that there has been an evolving international renaissance on the positive gains associated with migration. For example, many migrants are able to accumulate the needed financial, social, and human resources which enable them to overcome hardships and contribute to the advancement of their origin and destination areas (Ghana Statistical Service, 2014a, 2014b).

Even though the Economic Community of West African States’ (ECOWAS) protocol on free movement permits the free flow of goods, services, and people, there are country-specific requirements for border crossing in Africa. In Ghana, for instance, following the signing of the 2030 Sustainable Development Goals (SDGs) and also the promulgation of the Ghana Immigration Act (Act 573), the Ghana Immigration Service [GIS) has developed stringent immigration guidelines to guard against irregular migration practices (GIS, 2015; GSS, 2012). While all West Africans are free to migrate to any ECOWAS state for a ninety-day stay without visa, the cross-border migration patterns have not been described in the migration literature (Bakewell & De Haas, 2007; Bastide et al., 2017). In particular, issues involving the modes of travel and routes of entry amongst some West African immigrants who dwell in Ghana and operate as itinerant retailers are under-researched. Notwithstanding the scant literature which describes the state of Ghana’s borders as poorly governed and porous (Bruni et al., 2017), systematic empirical research concerning the cross-border migration configurations of the above immigrant group in the country is rare (Bakewell & De Haas, 2007). Itinerant immigrant retailers refer to a group of immigrants who mostly move by foot to sell foreign and locally manufactured wares, especially along the principal streets of various cities (Yendaw, Tanle, & Kumikyereme, 2019a). The main distinctive feature of their business is that they usually do not have permanent market structures where they sell their goods and their activities are generally unregulated.

In response to growing calls for data on cross-border movements and better use of migration data (GSS, 2012; International Organization for Migration [IOM], 2015), the present study assessed the nature of itinerant immigrant retailers’ cross-border movement into Ghana. The study was guided by these research questions: What is/are the most common mode(s) of transport used by itinerant immigrant traders to Ghana? What are the main entry routes used by itinerant immigrant retailers entering Ghana? And what factors influence itinerant immigrant retailers’ choices of entry routes into Ghana? The potential benefits of the results of this study are twofold: (1) it would contribute to improving migration data collection and management in Ghana, and (2) it would offer tailored policy recommendations for government, especially GIS and other stakeholders to enhance cross-border management, security, and immigrants’ safety.

Literature Review and Theoretical Framework

Literature Review

In the literature, ‘border’ is depicted as a dynamic process whereby humans and their activities are fixed territorially in space, by a range of actors but primarily by states (Durand, 2015). Following Durand (2015), the term ‘border’ as used in this paper is conceptualised as the locus of state territoriality with a territorial power (e.g., region, city) overseeing this spatial construction and using it as an identity marker to regulate the inflows and outflows of persons, goods, and services. Cross-border migration, therefore, signifies the transboundary flow of people from one country to another and this could be done through official/approved or unofficial/unapproved routes (Bruni et al., 2017). Whereas regular/legal migration involves the crossing of international borders using official migration passages designated by nation-states, irregular/illegal cross-border movement denotes the passage of international borders outside the regulatory norms of the country of origin, transit, or destination.

In Africa, the issue of irregular migration has become pervasive due to the courageous determination of people to overcome adversity and poverty (Bruni et al., 2017). In buttressing this view, Kobzar et al. (2015) maintained that though many immigrants generally strive to cross national borders using the available approved entry routes, irregular migration is accentuated by highly restrictive immigration policies. Kobzar et al. (2015) further stated that the driving factors of irregular migrations generally differ amongst different categories of migrants including family migrants who encounter difficulties accessing visas for their migration and economic migrants (e.g., immigrant retailers) who sometimes fail to fulfil the required employment qualifications for support. Related to the above, Düvell (2011) and Friebel et al. (2018) stated that because the status of immigrants as irregular migrants is usually assessed based on their level of compliance with domestic migration policies, the dynamics entailed in irregular border crossing and the challenges it presents for legal migrations vary amongst countries based on the obtainable routes of legal migration. De Haas (2009) and Kobzar et al. (2015), therefore, concluded that in situations where would-be migrants are not able to access approved migration routes, due to, for instance, high transaction costs, they tend to explore unauthorized migration routes using unsafe modes of transportation.

In the case of West Africa, though it is a region characterised by free movement of people following the ratification of the ECOWAS protocol, a large number of cross-border movements are undertaken illegally/irregularly (Bakewell & De Haas, 2007; IOM, 2015). Nationals of ECOWAS member countries, particularly amongst young unmarried men with lower levels of education, think that interregional movements are permissible, and hence do not fully appreciate the illegality of their migration (Adepoju, 2001; Bruni et al., 2017). Many immigrants in West Africa leave their countries of origin without observing migration protocols, by failing to obtain national passports and other travel documents before embarking on their journeys through unofficial entry routes (Adepoju, 2001; IOM, 2015). Similarly, GSS (2012) and IOM (2015) bemoaned that some immigrants, as a result of immigration regulations, enter their host countries through unrecognised entry points or without appropriate entry documentation, particularly entry visas or ECOWAS identity cards. The above observations resonate with the position of Bastide et al. (2017) and Friebel et al. (2018) who postulate that restrictive migration policies tend to heighten the volumes of unauthorized border crossings, where perilous means of transportation, including trekking by foot, are common.

According to Arthur (1991), antismuggling measures often increase immigrants’ vulnerability to migration laws on migration routes, as illegal entry into a country often entails some risk-taking. Some immigration officials usually monetize their positions by abusing their authority to allow immigrants or potential immigrants to obtain the relevant entry documentations or simply allowing them to proceed along migration journeys illegally (Bruni et al., 2017; IOM, 2015). Immigrants often bribe officials for crossing of borders or checkpoints, or in obtaining fake documents (Zohry, 2011). Moreover, Stenou (2004) and Bruni et al. (2017) argued that despite the ECOWAS treaty on free movements, immigrants are still harassed by law enforcement agents along the routes and the borders. Consequently, many immigrants prefer to cross over boundaries at unmanned posts despite the risks associated with such unofficial migratory conduct. Studies by Anarfi (1982), Zohry (2011), and Bastide et al. (2017) revealed that many immigrants in such irregular situations face the danger of being caught by occasional border patrol officials or being extorted by scouts who take unsuspecting travellers across for a fee.

Stenou (2004) and Bruni et al. (2017) added that physical fatigue usually characterises the migration process of most immigrants within the subregion. According to Stenou (2004), migration in the region could be very tedious. That is, from trans-Saharan trade times through the colonial era to the present, it has been done mainly by land. For example, Painter (1992) has observed that the vast majority of immigrants from Niger and Mali travel to La Côte d’Ivoire in buses. The distances usually covered by many of these immigrants, according to Stenou (2004), are incredibly long. For example, the round trip from Niger to La Côte d’Ivoire is about 3500 km and journeys over such distances take days and traverse several international boundaries.

Theoretical Framework

Although arrays of multidisciplinary theories have been espoused to explain migration, the reality is that most of the existing theories (if not all) were developed at the time when cross-border movements were relatively less restrictive (Friebel et al., 2018; Zohry, 2011). Consequently, most migration theories from Ravenstein’s laws of migration did not anticipate the current situation where migration has become increasingly securitized across nation-states (Jaskulowski, 2018). Because of the difficulty in finding theoretical frameworks for theorizing the cross-border migration behaviour of immigrants, I adapted the macro, meso, and micro level analytic framework for understanding illegal cross-border migration (Zohry, 2011) for the study (Fig. 1). This framework was adapted because most of its basic underpinnings are applicable for studying the general nature of cross-border movements.

Fig. 1

Framework for understanding the cross-border migration pattern of immigrants. Source: Adapted from Zohry (2011)

The framework posits that any analysis regarding the cross-border movement of people must take into account macro-level factors (political, economic, and legal frameworks) and meso-level influences (migration networks) between countries of origin and destination on the one hand and personal characteristics such as sex, age, education, and country of origin of the immigrants involved, on the other hand (Fig. 1). The above proposition clearly underpins the fundamental assumptions of Bronfenbrenner’s ecological model which hypothesised that the inner or personal resources of people and the wider environmental influences (consisting of internal and external systems) interact in a complex way to influence people agency including taking risky migration decisions such as illegal cross-border migration (Babatunde-Sowolea et al., 2016).

In this adapted framework, the political context in the country of origin relates to political stability and democracy. It posits that politically fragile states are likely to have ineffective border control mechanism (Zohry, 2011). So, immigrants from unstable countries are expected to find it easy to cross borders irregularly using unapproved routes and unsafe means of transportation. Additionally, the framework describes democracy as something that provides hope for potential immigrants to remain in their countries of origin. Good democratic governance comes along with absolute respect for human rights and dignity by the state and the system gravitates towards fulfilling the needs of young people in the society such that the propensity for irregular migration declines (Zohry, 2011). The framework describes the political situation in the country of destination to include the state’s disposition or attitudes regarding migration and the forces for or against migration. In the case of Ghana for example, the Immigration Act, 2000 (Act 573) enjoins all foreigners to migrate through the official border posts using valid travel identification. Thus, immigrants who lack these entry requirements may resort to illegal cross-border migration.

The economic framework in the country of origin and destination, according to Zohry (2011), comprises factors related to labour market regulations, unemployment rate, welfare systems, and structure of the economy. These factors could negatively or positively influence the tendency for regular or irregular border crossings and even the migration decisions of potential immigrants. The economic performance of a country and its ability to create productive work opportunities, for instance, can curtail the possibility of people migrating using hazardous means of transportation through unapproved routes. Studies by Bade (2004) and Sebastiano et al. (2005) indicate that immigrants who undertake irregular migrations through unofficial routes are usually absorbed in the informal economy of the country of destination especially in retail trade. This observation may as well explain why a country like Ghana, with large informal sector and better economic opportunities than several West African countries, continues to entice many immigrants from the Sahelian region of Benin, Burkina Faso, Mali, and Niger.

The legal framework in the country of origin and destination involves factors related to international, regional, and bilateral treaties, as well as national-level legal frameworks that regulate migration. For instance, in West Africa, the ECOWAS treaty on free movement has been instituted to guarantee unimpeded flow of people, goods, and services. Hence, per the provisions of this protocol, all member states are enjoined to stop demanding visas and residence permits from ECOWAS nationals (ECOWAS Commission, 2014). With the above treaty in operation, the prevalence of irregular cross-border migration would be expected to reduce and or end in the subregion.

At the meso level, the analysis relates to migration networks and the linkages between areas of origin and the national Diasporas in the countries of destination (Zohry, 2011). Haug (2008) stated that migration networks and ethnic affinities play critical roles in stimulating migration streams between countries of origin and destinations. Strong network connections and ethnic antecedents between illegal migrants and nonmigrants could positively or negatively influence the cross-border migration behaviours of nonmigrants. For instance, Awumbila et al. (2017) reveal that migrants do not migrate blindly to destinations in search of jobs, but are usually supported to arrange their journeys and solve problems in the migration process. For example, nonmigrants with strong network connections to migrants who have succeeded in undertaking their journeys could be assisted to embark on irregular migration with ease.

According to Zohry (2011), the microindividual-level framework deals with the individual characteristics of migrants and potential migrants. Factors in the microindividual-level analysis include sociodemographic characteristics of people in both countries of origin and destination (Zohry, 2011). Factors associated with the sociodemographic context such as age, gender, education, employment status, marital status, household characteristics, and migration intentions could influence the cross-border migration behaviour of immigrants, including unauthorized cross-border migration (Zohry, 2011). The present study builds on the above observation that the background characteristics of the immigrants wield a significant effect on their choice of entry routes to Ghana (see Fig. 1).

Materials and Methods

Study Context and Design

The data used for the analysis was extracted from a much wider research project on Migration Patterns and Livelihood Activities of West African Immigrant Retail Traders conducted by the researcher in the Accra Metropolitan Area (AMA). The choice of AMA (Fig. 2) as the study setting was informed by the fact that it is the most industrialized urban setting in Ghana where most commercial activities are executed (GSS, 2012, 2014a, 2014b). The city of Accra, therefore, serves as a key destination for many West African migrant retailers in Ghana (GSS, 2012). It has been estimated that over 21% of West African immigrants dwell in the Greater Accra Region (the highest in Ghana) and AMA has the highest proportion of these immigrants (GSS, 2013). The availability of requisite infrastructure and job opportunities continues to entice people from all parts of Ghana and beyond, including West African itinerant immigrant retailers, for various business activities (GSS, 2012; Yendaw, Tanle, & Kumikyereme, 2019a).

Fig. 2

Map of Accra Metropolitan Area showing the study communities. Source: GIS Unit of the Department of Geography and Regional Planning (2017)

Considering the built-up nature of the study area coupled with the fact that the immigrants settle in clusters (GSS, 2012; Yendaw, Tanle, & Kumikyereme, 2019a), nine communities noted as the key dwellings of West African immigrant retailers were purposively selected as the study areas. These were Abossey Okai, Abeka, Ablekuma, Agbogbloshie, New Fadama, Sukura, Lartebiokorshie, Nima, and Mamobi (see Fig. 2).

I identified these communities during a reconnaissance survey aided by immigrant key informants in the study area. Even though earlier research on this immigrant population abounds in the AMA, the focus has been on their livelihood status (Yendaw, Tanle, & Kumikyereme, 2019a), migration motivations (Yendaw, Borbor, et al., 2019b), and issues on the policy and security implications of their retail trade (Yendaw, 2019). Consequently, information regarding their cross-border migration configurations such as mode of travel and routes of entry are rare.

This study is cross-sectional and adopted the concurrent nested (embedded) mixed methods research approach to unravel the cross-border migration patterns of the immigrant retailers. Creswell (2012) has emphasised that researchers using mixed methods are required to demonstrate the various stages of mixing data in their research. Accordingly, the design of this study includes one phase of data collection in which the quantitative approach was given priority, while the qualitative approach was embedded, serving an ancillary role. Indeed, the quantitative and qualitative data were combined and the qualitative results were used to elaborate and confirm the quantitative findings.

The qualitative data was collected to complement the quantitative data due to the nature of the research questions which required the immigrants to give a thorough account of their cross-border migration experiences and so a mono research approach (i.e., either quantitative or qualitative) could not have sufficiently address the research problem. Additionally, a number of studies (Bruni et al., 2017; GSS, 2012) have underscored the inherent complexities associated with measuring the cross-border migration patterns of immigrants in West Africa mainly because of the surreptitious nature of many border crossings. Consequently, Bruni et al. (2017) have recommended the use of innovative research strategies to obtain an in-depth understanding of the phenomenon. Indeed, the amalgamation of the two research approaches (quantitative and qualitative) was timely as it aided the researcher to unearth certain key nuances which a single research approach could not have been able to do.

Sampling and Data Collection

The target population included West African immigrants who were engaged in itinerant retail trade in the AMA for a minimum period of 1 year and were not dual nationals, naturalized foreign citizens, or nationals by marriage. Given the general difficulties in obtaining representative samples of mobile populations like itinerant immigrant retailers (Vigneswaran, 2007), the snowballing technique was used which produced a sample frame of 842 immigrant retail traders from the nine study communities (see Fig. 2). The initial intention for undertaking the snowball sampling was to produce a sufficient sampling frame for a randomized sample selection due to the lack of accurate statistics on this group of immigrants in Ghana (GSS, 2012). However, being mindful of the fact that the study used the mixed methods research approach and the argument in the central limit theorem that a large sample size is more likely to generate a normal distribution in a data set than a smaller sample, the entire sample frame of 842 was enrolled in the study. As regards the qualitative dimension of the study, nine leaders of the immigrant retailers (one key informant, representing the 9 study sites) who possessed key knowledge of the cross-border migration of the immigrant retailers were purposively chosen for key informant interview.

Structured and semistructured questionnaires (specifically interview schedule and interview guide) were used to collect the data from the field. The interview schedule enabled the interview questions which were written in English to be interpreted to the immigrant retailers who were not proficient in English (GSS, 2012). The instruments were sectioned into three. Part 1 consisted of questions on their sociodemographic characteristics (country of origin, gender, age, educational level, and marital status). Part 2 tapped into their main modes of travel (by foot, by bus/car/vehicle, by motorcycle, by air, and by water) and routes of entry (approved and unapproved). Part 3 assessed the key factors which influenced their choices of migration routes (e.g., lack of travel documents and bribery/extortion by border officials). The interview guide consisted of open-ended questions and prescripted probes. Prior to the actual field data collection, the instruments were pretested and reworked in the Kumasi Metropolis (the second-highest destination of West African immigrants) which improved the face and content validity of the instrument.

All ethical questions that guide the conduct of social sciences research were adhered to strictly. With respect to informed consent, respondents for both the survey and the interview guide were informed about the rationale of the study and the possible implications of their participation in the study. Before the interview, a consent form was read and interpreted to them, stating their right to withdraw or avoid answering questions they deemed infringed on their privacy. As regards anonymity, all personal identifiers that could link the data to the immigrants such as name, house number, or addresses were avoided. Besides, the research protocol was reviewed and approved by the University of Cape Coast Institutional Review Board (with protocol number UCCIRB/CHLS/2016/23) before the fieldwork started.

The instruments were administered by the researcher with the assistance of three trained final year masters’ students from the University of Ghana who could communicate in the indigenous languages (Asante Twi and Ga), French and Hausa which are the main languages spoken by the immigrants. All the surveys and key informant interviews were conducted face-to-face and most of them were carried out at the immigrants’ places of abode and a few at their workplaces. With prior permission of the immigrant key informants, a voice recorder, complemented by note-taking of the deliberations, was used to record the interviews. The average duration for each key informant interview lasted between 30 and 45 min. As regards the demographic attributes of the key informants, all the respondents were men and were aged between 28 and 45 years. Four out of them had no formal education, one was proficient in Arabic, two attained basic education and the remaining two (one each) acquired basic and secondary level education. They were all Muslims, and apart from two participants who were single, the rest were all married which contradicts GSS (2012)) and Yendaw, Tanle, and Kumikyereme’s (2019a) findings that most itinerant immigrant retailers are generally single uneducated Muslim men.

Data Analysis

Data from the interview schedules were validated by reading through the responses to ensure they were correctly filled and complete. Out of the 842 questionnaires audited, about 779 were found useful, giving a response rate of 92.5%. The data from these questionnaires were processed and analysed using Statistical Package for the Social Sciences (SPSS) version 21. Descriptive statistics (frequency distributions, percentages, cross-tabulations) and an inferential analytical technique (binary logistic regression model) were used to analyse and report the findings. Binary logistic regression was used to confirm whether key explanatory variables such as gender, age, marital status, level of education, and country of origin of the immigrants had any influence on their main routes of entry (e.g., the dependent variable). Routes of entry denoted the various passages or channels used by the immigrants to get into their destination (Ghana) and were treated as a dichotomous variable (approved/official and unapproved/unofficial routes).

Data from the key informant interviews were also validated by playing and listening to the contents of the audio recordings several times to confirm that they were thorough and completed. The data were then processed and analysed manually by means of the thematic analysis technique. The thematic analysis procedure was employed to identify, analyse, and report patterns that arose from the interviews. The thematic analysis was carried out following the four steps Yendaw (2019) used: data preparation and close reading of text to get familiar with the raw data, development of categories from the raw data into a model or framework (coding of data), searching for themes and recognizing relationships (e.g., axial coding), and refining of themes through coding consistency checks such as independent parallel coding. To guarantee the credibility and validity of the findings, the respondent validation technique was used by contacting three of the participants to authenticate the interview transcripts and summaries. The subsequent sections present the results, discussion, conclusions and implications, and study limitations.

Results

Sociodemographic Characteristics

Table 1 shows that most of the immigrant retailers came from Niger (42.2%) and Mali (22.6%). About 89.1% of them were men. A little over six out of ten (62.0%) were in the age group of 20–29. About 62.3% of the immigrants were unmarried (single). Over a sixth (66.7%) of them were uneducated. Almost all of them (97.0%) were Muslims. Over one-third (43.0%) of them had household sizes of between one to four members. They were generally into the sale of leather belts, footwear, and leather bags (18.0%). A higher percentage (65.3%) of them travelled into the country between 2007 and 2011. Concerning their length of stay, the results reveal that most of them (63.1%) had resided in the study area for 1–4 years, and about 81.3% of them had been involved in their retail businesses for 1–4 years (Table 1).

Table 1

Sociodemographic characteristics of the immigrant retailers

Sociodemographics	N	Percentage
Country of origin
Benin	39	5.0
Burkina Faso	32	4.1
Mali	176	22.6
Niger	329	42.2
Nigeria	146	18.7
Togo	49	6.3
Other	8	1.1
Gender
Men	694	89.1
Women	85	10.9
Age (completed years)
<20	89	11.4
20–29	479	61.5
30–39	164	20.0
40–49	55	7.1
Marital status
Single	462	59.3
Married	294	37.7
Widowed	13	1.7
Divorced/separated	10	1.3
Educational attainment
No formal education	520	66.7
Basic education	188	31.1
Secondary/tertiary	16	2.2
Religious affiliation
Christianity	23	3.0
Islam	756	97.0
Household size
1–4	335	42.5
5–8	250	32.6
>8	194	24.9
Wares traded
Electronics/accessories	125	12.9
Herbal medicines	165	17.0
Food items	72	7.3
Garments/fabrics	73	7.5
Clothing	119	12.2
Belts/foot wear/bags	173	17.8
Ornaments	88	9.1
DVDs/LCDs	157	16.2
Period of entry into Ghana
1997–2001	13	1.7
2002–2006	56	7.2
2007–2011	509	65.3
2012–date	201	25.8
Length of stay in Ghana
1–4	492	63.1
5–9	261	24.3
>=10	26	3.3
Duration in retail trade
1–4	633	81.3
5–8	129	16.6
>=9	17	2.1

Cross-Border Migration Patterns

This section presents results on migrant retailers’ modes of travel and routes of entry. The rationale was to ascertain the common modes of transport/travel used into the country, and understand whether they used the approved routes of entry. Table 2 revealed that about sixth (58.2%) of the immigrants migrated into the country through unapproved routes. As regards their modes of travel, the analysis indicated that all of them travelled into the country by land often with vehicles, trucks, and/or buses (72.0%) followed by those who trekked by foot (28.0%). This result was confirmed in the interviews. During deliberations with leaders of the immigrants, they unanimously admitted that most West African immigrant retailers travel to Ghana by land, using vehicles but generally through unapproved routes. In a confirmatory statement, a 45-year-old participant who resides at Agbogbloshie remarked that

Table 2

Modes of travel and routes of entry by sociodemographics

	Mode of travel (N=779)		Route of entry (N=779)
	By bus/vehicle	By foot	Unapproved routes	Approved routes
Variable	(%)	(%)	(%)	(%)
Travel mode and route	71.9	28.1	58.2	41.8
Country of origin
Benin	87.2	12.8	71.8	28.2
B. Faso	70.0	30.0	84.4	15.6
Mali	94.5	5.5	43.2	56.8
Niger	93.7	6.3	53.2	46.8
Nigeria	100.0	0.0	56.8	43.2
Togo	63.5	36.5	75.3	24.7
Other	92.1	7.9	32.5	67.5
Gender
Men	81.6	18.4	56.5	43.5
Women	98.0	2.0	24.7	75.3
Age (in years)
˂20	89.6	10.4	54.5	45.5
20–29	81.8	18.2	63.1	36.9
30–39	93.9	6.1	32.5	67.5
40–49	100.0	0.0	20.0	80.0
Marital status
Married	100.0	0.0	34.0	66.0
Single	80.7	19.3	66.5	33.5
Widowed	90.0	10.0	10.0	90.0
Divorced	100.0	0.0	38.5	61.5
Educational level
No edu.	79.5	20.5	63.1	36.9
Basic edu.	87.6	12.4	43.8	56.3
Higher edu.	93.9	6.1	32.5	67.5

B. Faso, Burkina Faso; Edu., education; Higher (secondary and tertiary edu.)

The majority of us you see in this business travelled into this country (Ghana) by land using vehicles. Some of them too travel by foot especially those who cannot finance their trips. Talking of routes of entry, I believe you are aware that many of the immigrants have their way of getting in here (Ghana). Most of them for some reasons usually undertake their journeys using alternative routes other than the approved entry points. (45-year-old participant from Niger)

The study also assessed the influence of sociodemographic characteristics on modes of travel and routes of entry into the country. Sociodemographic characteristics such as gender, age, level of education, marital status, and country of origin were cross-tabulated with modes of travel and routes of entry (Table 2). The results showed that while all Nigerians (100.0%) travelled to Ghana by bus, their Togolese counterparts (63.5%) were the least group that travelled by bus. On immigrants who travelled into the country by foot, those from Togo (36.5%) and Burkina Faso (30.0%) were amongst the highest. Concerning their routes of entry, Burkinabes (84.4%) and Togolese (75.3%) often used unapproved routes to the destination. Nationals of Cote d’Ivoire and Senegal (67.5%) followed by those from Mali (56.8%) generally used approved routes (see Table 2).

The qualitative data obtained from interviews with key informants amongst the immigrants confirmed the findings from the quantitative data. During the in-depth interviews, all participants contended that immigrants from Togo and Burkina Faso are noted for often travelling through unapproved routes to Ghana by foot. In their view, proximity of the two countries to the destination and their ethnic connections with the destination are the main driving factors for the high incidences of irregular cross-border migration of their nationals. In elaborating on this issue, two of the participants made this exposé:

Illegal border crossing of immigrants from Togo and Burkina Faso is largely due to the proximity of the two countries to Ghana coupled with their ethnic affinities with the republic of Ghana. Take the case of the ‘Ewe’ ethnic group in Togo and the Mossi’ in Burkina Faso, the same ethnic groups are also in Ghana. These ethnic connections are such that you have some Togolese or Burkinabes having their relations in Ghana. These ethnic links render Ghana’s borders very porous with several unapproved routes. (38-year-old participant from Togo, Abossey Okai; 39-year-old participant from Burkina Faso, Nima)

In terms of gender, while 98.0% and 75.3% of women immigrants mostly travelled to the country by bus using approved routes, 81.6% and 43.5% of their men counterparts did so via the same means of travel and routes of entry. Table 2 further reveals that immigrants’ modes of travel and routes of entry differed across age. While 100.0% of immigrants who were older (40–49 years) undertook their journeys by bus, about 81.8% of those who were younger (20–29 years) travelled using the same mode of transport. Likewise, whereas 63.1% of immigrants who were less than 20 years migrated through unapproved routes, the data shows that about 80.0% of those who were older (40–49 years) travelled through approved routes. In respect to marital status, the results indicate that while 100.0% of those who were married and divorced travelled into the country by vehicle, about 80.7% of those who were single travelled using the same mode of transport.

In effect, single immigrants (19.3%) usually travelled into the country by foot. Regarding marital status and routes of entry, whereas 66.0% of immigrants who were married mostly migrated into Ghana using approved routes, 66.5% of their counterparts who were single largely travelled into the country through unapproved entry points. Again, 93.9% of immigrants who attained higher education (secondary/tertiary) mostly used buses to Ghana when compared with their counterparts of other educational categories. While about 20.5% of those without formal education migrated into the country by foot, slightly lesser number of persons with higher education (6.1%) did the same. Additionally, while over a sixth (67.5%) of those with higher education migrated to the destination through approved routes, 63.1% of their counterparts without formal education migrated using unapproved entry points (Table 2).

Further interactions with informants in an interview show that young unmarried men generally use unapproved routes. A 28-year old Nigerian who lives in Abeka, Accra revealed that

Immigrants who engage in this illegal border crossing pattern are usually young men who are not married. You know young unmarried men are generally daring and can do anything no matter the risks. Women hardly participate in this kind of risky migration perhaps due to the vulnerabilities associated with illegal border crossings.

Considering that most of the immigrant retailers migrated into the country through unapproved entry points, a binary logistic regression model was used to estimate the level of influence of their sociodemographic characteristics on their choice of unapproved routes (see Table 3). Prior to estimating the logit model, unapproved routes were denoted as zero signifying decreasing odds in favour of unapproved routes while approved routes were captured as one signifying increasing odds in favour of approved routes. The binary logistic regression analysis shows that demographic characteristics demonstrated to be a good predictor of their main routes of entry as shown by the Omnibus Tests of Model Coefficients (model=163.798; df=14; p<0.050) and the Hosmer and Lemeshow Test (χ2=10.159; df=8; p>0.050). According to Pallant (2005), for a model to be regarded as a good predictor, the alpha value of the Hosmer and Lemeshow Test has to be greater than 0.05, and in the case of the output of this data, the alpha value is 0.381, indicating strong reliability of the model.

Table 3

Binary logistic regression estimates of choice of unapproved routes

	Unapproved entry routes
Sociodemographics	p value	Odds ratio
Country of origin
Benin	0.104	1.514
Burkina Faso	0.000***	0.082
Mali	0.062	1.467
Niger	0.00***	0.141
Nigeria	0.039*	0.450
Togo	0.002**	0.210
Other (RC)
Gender
Men	0.002	0.361
Women (RC)
Age (in years)
Less than 20	0.087	0.519
20–29	0.560	0.146
30–39	0.217	0.208
40–49 (RC)
Educational attainment
No education	0.142	0.447
Basic education	0.462	1.507
Secondary/tertiary (RC)
Marital status
Married	0.284	17.770
Single	0.046	2.352
Widowed	0.689	1.391
Divorced (RC)

Omnibus Tests of Model Coefficients (model=163.798; df=14; p<0.050)

Hosmer and Lemeshow test (χ2=10.159; df=8; p>0.050*)

Percentage of variance explained =33.3%

The model explained about 33.3% of the immigrants’ routes of entry. Despite the significance of the model in predicting the immigrants’ routes of entry, it could not predict the significance of certain variables. Three out of the five predictor variables were found to be significant in the model. These include country of origin, gender, and marital status (Table 3).

The analysis showed that country of origin was a significant predictor of the choice of entry routes. In particular, respondents from Togo, Niger, Burkina Faso, and Nigeria were more likely to use unapproved entry routes to Ghana when compared to the reference category (see Table 3). However, while the results revealed that respondents from Togo (0.210) and Burkina Faso (0.082) were more likely to pass through unapproved entry points, their counterparts from Mali (1.467) and Benin (1.514) were more likely to use approved entry points to Ghana. Furthermore, gender significantly impacts the choice of route of entry. Table 3 shows that men had a 36.1% likelihood of using unapproved entry points relative to their women. Age and educational background did not significantly influence the choice of entry routes. Yet, respondents between the ages of 20 and 29 years had more odds of using unapproved routes than those in other age categories. Likewise, those respondents with no formal education were more likely to use unapproved routes when compared with respondents with secondary and tertiary education. Also, marital status was a significant predictor of choice of entry routes. The individuals in the single category, compared to those divorced, were more likely to use unapproved routes. Those who were single had 2.352 decreasing odds in favour of the use of approved routes to Ghana.

The study finally examined factors which influence the immigrants’ choices of entry routes (Table 4). This was necessary because such important information could provide useful policy directions for state and nonstate actors to address the issue of irregular migration, a phenomenon which has over the years been on national and international policy agenda. As summarised in Table 4, seven key factors influence the immigrants’ choice of routes of entry into the country. The factors, which are no money to finance the trip, extortion by border officials, lack of travel documents, safety/convenience, attitudes of some border officials, unaware of any alternative entry routes, and possession of travel documents, influenced the immigrants’ routes of entry differently.

Table 4

Factors which influenced choices of entry routes by routes of entry

Reasons for entry routes	Routes of entry
	Unapproved (%)	Approved (%)
Has no money to finance trip	95.8	4.2
Extortion by border officials	100.0	0.0
Have no travel documents	100.0	0.0
Safety and convenience	25.6	74.4
Attitudes of border officials	66.7	33.3
Unaware of any alternative route	31.0	69.0
Have my travel documents	1.6	98.4

For instance, all immigrants who stated that the lack of travel documents and frequent extortions by some border officials influenced their choice of travel route migrated to the country through unapproved routes. This was followed by those who mentioned that they chose their entry routes because they had no money to finance their trips (95.8%). On the other hand, the results showed that immigrants who had their travel documents (98.4%) and those who regarded official border crossings as safer and convenient (74.4%) often travel to Ghana through approved entry routes (Table 4). Moreover, the use of approved entry routes was equally high amongst immigrants who were oblivious of any alternative entry routes (69.0%).

Results from the qualitative interviews provided further evidence in support of the findings of the quantitative data. Some key informants amongst the immigrants reasoned in the interviews that they migrated through unapproved routes because of their lack of travel documents. According to these respondents, border officials prey on their lack of travel documents to extort them financially, creating opportunities for them to travel through unapproved routes. This contributes to the rise in irregular migration in West Africa. In relation to this issue, one of the participants made these observations:

Many West African immigrants enter Ghana illegally mainly because they usually lack the required travel documents. I think some border officials also contribute to this problem because some of them harass immigrants for petty cash before being allowed to cross the border. The worst part is that immigrants who are unable to pay such bribes are sometimes detained at the border for days. Some of these experiences normally compel many immigrants to explore alternative entry routes. Due to these attitudes, many people from my country (Burkina Faso) who lack travel documents sometimes undertake their journeys through ‘Sankaase’ (a small community in between Ghana and Burkina Faso) into Ghana just to avoid being harassed for money or detained. (39-year-old participant from Burkina Faso, Nima)

Discussion

This study is an assessment of the modes of travel and routes of entry of West African immigrants who migrate to Ghana for work in the informal sector as itinerant retailers. The results of the study show that most of the immigrants travel to Ghana by land, often by vehicles. However, evidence from the study showed that immigrant women, those older (40–49 years), and those married used vehicles the most into the destination. It was again found that while immigrant retailers from Nigeria and Mali were amongst those who frequently travelled to Ghana by vehicles/buses, their Togolese and Burkinabe counterparts frequently migrated to the country by foot. This affirms previous studies on West Africa intraregional migration patterns which suggested that many immigrants in the region travel to other neighbouring states mostly by land, using buses (Bakewell & De Haas, 2007; Bruni et al., 2017; Stenou, 2004).

The above findings further consolidate what Painter (1992) had stated that the vast majority of immigrants from Niger and Mali travel in buses and small trucks covering several distances across the Sahel desert. Again, the evidence that most of the immigrants travelled using vehicles could be explained by what GSS (2012)) and Yendaw, Tanle, and Kumikyereme (2019a) observed that many West African immigrant retailers generally have lower socio-economic status and thus may lack the requisite financial wherewithal to patronise air transportation. It was equally noticed that some of the immigrants from Togo and Burkina Faso travelled to the destination by foot owing to their proximity and ethnic affinities with the destination country causing porous borders and availability of numerous illegal or unapproved entry points as evinced in the qualitative narratives. The above exposé finds credence in Adepoju’s (2004) position that some West Africans especially from countries with shared borders usually perceive their migration as an extension of their homeland territories (Adepoju, 2004).

Findings from the study further revealed that most of the immigrants migrated through unapproved routes and this was particularly higher amongst those who had no valid travel documents and those who perceived border officials as being extortionists. The above evidence contradicts what Bruni et al. (2017) have noted that most migrations from West Africa are regular and that the majority of them migrate with valid passports and other travel documentation. Nevertheless, the current revelations support the assertion made by Adepoju (2001) and IOM (2015)) that most border crossings in West Africa are undertaken illegally through unmanned border posts despite the dangers associated with it. These findings further confirmed what Afrifa (2004), Bakewell and de Haas (2007), and GSS (2012)) have reported that many ECOWAS nationals generally sidestep the immigration laws of Ghana with impunity. Meanwhile, as per Ghana’s immigration policy, these forms of migration constitute a gross violation of the statutes of the Ghana immigration Act (Act 2000) which requires foreigners to acquire a valid documentation/identification at any of Ghana’s missions outside. The above findings further authenticate what De Haas (2009), Zohry (2011), and Kobzar et al. (2015) indicated that immigrants who are unable to find legal migration routes with ease (due to harassments, extortions, lack of valid travel documents, etc.) or are unable to bear the transaction costs of border crossings generally engage in unauthorized border crossings. This phenomenon is further articulated in the conceptual framework for this study (macro–meso–micro model for irregular migration) where payment of bribes is the norm in negotiating the crossing of borders or checkpoints, or in obtaining fake documents (Zohry, 2011).

Results from the binary logistic regression test demonstrated a statistically significant relationship between the immigrants’ routes of entry and key microindividual-level variables such as country of origin, gender, and marital status. Country of origin was a significant predictor of the immigrants’ choice of entry routes. In particular, immigrant retailers from Burkina Faso, Niger, Togo, and Nigeria were more likely to migrate to the country through unapproved entry points. This finding could be associated with what Yeboah (1986) and GSS (2012)) found amongst West African migrants that certain trans-frontier-ethnic groups in the region consider borders as both abstract and inefficient and hence regard surreptitious border crossings as normal. Moreover, aside from proximity, the historical connections Nigeria, Burkina Faso, and Togo shared with Ghana with respect to their migration and ethnic antecedents might have contributed to the current irregular border crossing of such immigrant retailers. This migration pattern implies that state actors’ attempt to nip irregular migration in the bud and ensure migrants’ safety in the region remains unfruitful since most of the immigrants surveyed generally migrated through unmanned border posts.

Also, immigrant men had higher odds of using unapproved routes as compared to women. This finding supports the general observation in the migration literature and the qualitative narratives in this study where women are generally risk averse as compared to men in the general migration process (GSS, 2012). Although age and educational status had no significant effects on their choices of entry routes, it was noticed that immigrant retailers who were young (≤20–39 years) and those who had no formal education had more odds of migrating through unapproved or unauthorised routes. As regards marital status, the analysis demonstrated that those who were single were more likely to use unapproved routes. That is, immigrants who were unmarried had the highest odds of migrating through unapproved entry channels as compared to those who were married. These findings are said to be in sync with the literature on migration studies which suggests that young unmarried adult men with no or lower level of education are more likely to use unsafe means of transportation and clandestinely through unauthorized routes in migrating (Adepoju, 2004; GSS, 2012). The overall consequences of these findings are that aside from the potential loss of revenue to the state, such irregular cross-border movements could pose serious security and or health threats to the nation particularly in an era characterised by increasing terrorism and disease outbreaks (e.g., the current coronavirus pandemic).

Conclusions and Implications

Based on the findings, the study concludes that most West African itinerant immigrant retailers in the AMA travelled to Ghana irregularly, and this was particularly high amongst those who had no travel documentations such as passports or identifications and those who viewed border officials as blackmailers. The current cross-border migration pattern accentuates what has been debated in the literature that the type and nature of migration policies imposed by nation-states shape the regularity or irregularity of border crossings including choice of entry routes and modes of travel. Significant relationships were established between key sociodemographic variables and choice of entry routes. Generally, unmarried immigrant men from Burkina Faso, Niger, Togo, and Nigeria had the highest odds of migrating into the country through unauthorised routes. The finding aligns with the basic tenets of the macro-, meso-, and micro-level analytic framework for understanding illegal cross-border migration (Fig. 1) that the individual characteristics of immigrants have the proclivity to influence their cross-border migration behaviour. One key implication of this finding is that any future academic research that seeks to further understand the cross-border migration pattern of this immigrant group must take into consideration issues of their background characteristics, especially gender, age, marital status, and country of origin. I recommend detailed research on their cross-border migration dynamics nationwide to get a broader picture of their cross-border migration behaviour.

Ensuing from the above conclusions, the study further recommends that any migration management programme which aims at stemming irregular migration into the country should target unmarried immigrant men and those coming from neighbouring countries of Burkina Faso, Niger, Togo, and Nigeria. That said, key migration institutions responsible for managing migration in Ghana, particularly GIS, should, as a matter of priority, map out concrete strategies, including the creation of a reliable database on all immigrant retailers, roll out intensive border surveillance, and team up with various diplomatic consulates and immigrant associations to ensure orderly and safe migration by addressing the underlying causes of illegal migration to Ghana. This is imperative because if the current cross-border migration pattern is not addressed, Ghana is exposed to the risk of various communicable diseases (e.g., Ebola and Coronavirus) and immigration-related crimes including terrorism. To this end, anticorruption measures are equally needed to decisively deal with border officials who allegedly extort monies and other resources from unsuspecting immigrants at the various country borders.