Flipping the page: exploring the connection between Ghanaian migrants’ remittances and their living conditions in the UK

ABSTRACT

Migrants’ remittance to close relatives and acquaintances back home has been widely studied. However, little is known about the impact remittances have on the living conditions of migrants who send these remittances. This paper hypothesises that migrants often deprive themselves of better living conditions in their host countries in order to regularly remit home. We test this hypothesis using survey data from 129 Ghanaian migrants living in the United Kingdom. We develop a Living Condition Index based on migrant participants’ self-assessment of how remittance has impacted their living conditions in the UK and specify a hierarchical multiple regression model to test this hypothesis. Controlling for migrants’ demographic, migration, employment, and remittance information, we find that there is a 19.8% decrease in migrants’ living conditions when they remit more than £150 per month, thereby confirming our hypothesis that remittances adversely impact migrants’ living conditions in their host countries. The paper reflects on the implications of the findings in light of migrants’ living conditions under the current global health pandemic and offers practical suggestions that could help lessen this financial burden on migrants while supporting sustainable businesses back home.

KEYWORDS:

• Remittances
• living conditions
• Ghanaian migrants
• the UK

1. Introduction

Moving abroad and away from immediate family members, social circles and familiar institutional settings poses considerable challenges to migrants. Some migrants deplete their life savings to embark on trips abroad. Others risk their lives through dangerous routes to seek greener pastures. Migration from the global South to the North in the pursuit of better employment and education opportunities, quality of life, and social welfare has been a preferred pathway for different population groups, especially skilled workers and students (c.f King and Sondhi 2018). For West Africans and Ghanaians in particular, migration to more affluent (typically Western) countries has been a long-established tradition, with the United States, United Kingdom and Germany being three key destinations (Peil 1995). Despite the struggles that migrants may encounter during their journey and in their new life abroad, they invariably strive to send money home. Remittances form a crucial resource for the remaining family, friends and local economy of the home country (see Carling 2014; Diko and Tipple 1992; Mazzucato 2008; Yang 2011; Yeates and Owusu-Sekyere 2019). In 2018, annual remittance flows to low- and middle-income countries increased to US$ 529 million, a 9.6% increase over the 2017 figure (World Bank 2018). In the face of this increase in remittances, it appears logical to hypothesise that the pressure of remittance-sending on the households of senders increases too, and adds strain to their living conditions and opportunities.

The structural and household effects of remittances sent by the immigrants to their home countries are substantial and have been widely researched. There are two dominant analytical paradigms and datasets that scholars interested in the subject often subscribe to. The first group of scholars demonstrate strong leaning towards research on how remittance inflows impact developing economies at the macro-level (Amuedo-dorantes and Pozo 2004; Ayana Aga and Martinez Peria 2014; Chami et al. 2008; Ratha and Mohapatra 2007). Among this group, the themes often explored include the impacts of remittance on origin countries with regards to investment (Adams, Cuecuecha, and Page 2008), exchange rate volatility (Amuedo-dorantes and Pozo 2004), and financial development (Gupta, Pattillo, and Wagh 2009). Quite often, existing cross-sectional and panel data are used as evidence in these studies (Amuedo-dorantes and Pozo 2004; Gupta, Pattillo, and Wagh 2009). On the other hand, are scholars and development experts interested in how remittance inflows impact households in developing countries. These studies often draw on empirical data to explore the contributions remittance inflows make towards several aspects of household wellbeing, including but not limited to: household infrastructure financing and adaptation to climate change impacts (Musah-Surugu et al. 2018), housing investment in origin countries (Kuuire et al. 2016; Obeng-Odoom 2010), increasing investment in education, school enrolment and educational attainment (Gyimah-Brempong and Asiedu 2015), and improvement in child health and reduced incidence of child mortality.

While these two strands of research have significantly illuminated the dynamism of migrant remittance and its importance for national economies and household wellbeing in origin countries, little is known about the austere life and sometimes harsh realities of the migrants who send these remittances (Sayed Mahadi, Hussin, and Khoso 2017; Shooshtari et al. 2014). In other words, we do not know what it costs a migrant to regularly remit home, or the trade-offs they have to negotiate to constantly keep sending monies home.

Against this backdrop, the research pursues two goals: first, it aims to explore the relationship between the amount Ghanaian migrants regularly remit and their living conditions in the UK, and second, to identify other predictors that significantly impact Ghanaian migrants’ living conditions in the UK. Data were gathered through a survey of 129 Ghanaian migrants in the UK who regularly remit money to their home country. The paper argues that remittance-sending denies migrants the opportunity to use the money in ways that improve their living conditions, resulting in a negative and statistically significant relationship between the amount of money migrants remit and their living conditions abroad. By exploring this relationship, the study brings into focus the precarity and crucial trade-offs that most migrants living abroad have to endure or negotiate in order to sustain their current remittance behaviour. Also, by bringing this perspective to the fore, we seek to open an avenue to further engage with some of the existing theoretical arguments on migrants’ motivations to remit, which tends to dwarf the important trade-off migrants make whenever they remit.

The rest of the paper is organised as follows. Section two reviews the literature on the determinants of migrants’ motivations for sending remittances. This is followed by a review of related literature on migrants’ living conditions abroad, specifically highlighting the situation of Ghanaian migrants. In Section three, we outline the study context, research methods and data collection. The research findings, including the model specification and results, are presented in Section four. Section five discusses the results from the model, its implications and the conclusion.

2. Literature review

2.1. Remittance behaviour beyond altruism, risk-sharing and self-interest

There are several reasons why migrants send remittances to their home countries. The literature so far points to three main motives. They include altruism, risk-sharing and investment (Lucas and Stark 1985; Stark and Lucas 1988; Bollard et al. 2011; de Haas 2007; Mughal 2013). While it has not been empirically established whether all of these three motives come into play when a migrant is sending remittances home, the evidence suggests that a combination of country-specific factors, the extent of migrants’ ties back home, one’s stage in life, and the prevailing conditions back home are often responsible for the difference in the importance of each motive (Batista and Umblijs 2016).

Scholars who lean towards the altruistic motive (Agrawal and Horowitz 2002; Niimi, Pham, and Reilly 2008) tend to argue that, where there is a strong social tie between migrants and their households back home, altruism often influences their remittance behaviours. This relationship is mostly modelled using the utility (e.g. the sense of responsibility) migrants derive from the wellbeing of relatives back home (Becker 1974). An empirical study from Ghana shows that Ghanaian international migrants in Amsterdam invest in their families by building houses for their parents back home. This brings a great sense of accomplishment to the migrants in their social standing both back home and among their countrymen abroad (Kabki, Mazzucato, and Appiah 2004). The model from the empirical study of Agrawal and Horowitz (2002) on Guyanese migrants’ remittance motivations suggests that the per-migrant remittance to multiple-migrant households declines significantly if altruism is the motivation. Conversely, in single-migrant households, however, they report that per migrant remittance is positively significant if altruism is the motivation.

Proponents of the risk-sharing motive of remittance (Stark and Levhari 1982) contend that families and households consider their migrant members as a risk-diversifying strategy against income loss since migrant members are constantly expected to remit to the family. Using insurance as a motive for migrants’ remittance behaviour, Niimi, Pham, and Reilly (2008) explain that the more risk-averse migrants are in their destination countries, the more often they remit back home. This finding is also consistent with earlier findings by Lucas and Stark (1985) that Botswana migrants significantly remit to household members in their home country for insurance purposes such as supporting family following the failure of crops and loss of livestock.

Self-interested motivations to remit may originate from the aspiration to inherit, acquire and maintain assets in the home country (investment), or future return intentions (Carling 2008). The latter might include preparing the base for the physical capital, strengthening ties with relatives and friends and upgrading social status (Lucas and Stark 1985). This notwithstanding, some scholars have established that using altruism, risk-sharing and investment motives separately to explain migrants’ remittance behaviour does not present a complete picture of migrants’ remittance motivations (Lucas and Stark 1985; Gosh 2006; Niimi, Pham, and Reilly 2008). In Vietnam for example, Niimi, Pham, and Reilly (2008) use the 2004 Vietnam Migration Survey Data to examine the key determinants of internal migrants’ remittance and find that altruism alone does not provide a sufficient explanation of Vietnamese migrants’ behaviours.

In light of this, Lucas and Stark (1985) have proposed a more integrative model of analysing migrants’ motives and remittance behaviour such as tempered altruism and enlightened self-interest where remittance can simultaneously be considered a return to the household investment in migration as part of a household risk-sharing strategy and as a source of investment capital into areas like education, as they note that ‘in the end, one cannot prove whether the true motive is one of caring or more selfishly wishing to enhance prestige by being perceived as caring’ (904). Following this argument, Carling (2008) indicates that the explanation of remittance motives rests on the profound contextual differences, such as the migration dynamics (temporary, permanent, circular), the nature of families and households (e.g. traditional patriarchal structure), and the normative settings (norms and values creating substantial pressure to remit).

2.2. Remittances and migrants living conditions

Capturing migrants’ living conditions abroad is often not a straightforward task. One is presented with different aspects of life, including but not limited to the physical, economic, social, psychological, emotional, mental or some combination of different aspects of life. Scholars have tended to draw insights from the housing circumstance of migrants (Kuuire et al. 2016; Owusu 1998), the nature and conditions of their employment (Obeng-Odoom 2010), their consumption and investment patterns (Kuuire et al. 2016; Martin 2016) and sometimes, their involvement in communal or collegial activities in their host countries (Andoh, Berrones-Fleming, and Dornberger 2019).

For example, reflecting on Ghanaian returnee migrants’ living conditions while they were in Germany, Martin (2005) illuminates the ordeal suffered by different generations of Ghanaian migrants who travelled to Germany for varied pursuits. She observes that while the immediate post-independence Ghanaian scholars, sponsored by the first President, Dr Kwame Nkrumah, to study in Germany enjoyed quite favourable living conditions, subsequent generations of migrants were not so fortunate due to radical socio-political and economic changes in both the home and destination countries. She indicates that these Ghanaian migrants’ quest to continue living in Germany against immigration rules juxtaposed with their pride in maintaining social prestige back home, which in most cases is measured by how much money these migrants remit for developmental and investment projects, compelling them to take up jobs and lead lives often incommensurate with their level of education. She notes for example that:

the migrants are perceived as socially successful, modern and wealthy in the Ghanaian context, and at the same time as backward, poor and marginalised in Germany. The gaining of status in one context means the loss of status in the other (212).

Investigating the economic impacts of remittances sent by Netherlands-based Ghanaian migrants, Kabki, Mazzucato, and Appiah (2004) observed that it is hard for remittance recipients in the Ashanti Region to believe that their migrant relative abroad cannot financially help them when they make demands, although these migrants often live a ‘deprived life’ according to Western standards. This unbelief often makes remittance recipient unsympathetic to their migrant relatives, and sometimes leads to straining transnational relationships. In some cases, to register their displeasure, the unbelieving migrant recipient would withdraw services run on behalf of migrant relatives, including supervising a building project, business venture or checking up spouse and children left behind. To avoid such undesirable outcomes and maintain the transnational relationship, migrants may strive to send remittance even when they live deprived life abroad, which could have detrimental effects on their wellbeing and living conditions.

In Australia, Obeng-Odoom’s (2010) study of housing-related remittance of Ghanaian migrants found that although the average Ghanaian earned about US$454 a week, in order to build or buy a house worth about US$100,000, 66% had to keep two jobs. Also, 48% of the interviewees remitted a third of their monthly income while 26% had to invest half of their monthly income to build a home. Those with more stable jobs and financially supported by their spouses committed a quarter of their monthly income towards this housing project back home.

Relatedly, a study of homeownership among Ghanaian migrants in Canada (Owusu 1998) showed that although migrants highly valued the idea of homeownership, they could not afford to buy or build a house. Owusu (1998) further noted that the homeownership rate of these Ghanaian migrants was 11%, compared to 56% of the Canadian-born population and 62% of other immigrants. Owusu’s research discussed factors contributing to such outcomes and the one that stood out was the financial support to relatives and family in Ghana: 98% of his respondents remitted money (Owusu 1998).

Describing the precarious living conditions of Ghanaian migrants in London, Diko and Tipple (1992) observed that, although before they migrated to Britain, Ghanaian migrants lived in high-status company houses or had rooms in a family house, in Britain, they often become tenants of either council flats or sub-tenants in houses rented to Ghanaians. They add that most of them live in the lower end of the London rental property market so as to minimise cost and maximise their savings (most of which they send back home).

Regarding employment, wages and living conditions of Indian emigrants in the United Arab Emirates (UAE) Zachariah, Prakash, and Rajan (2002) explain that the gradual decline in the demand for manpower for construction works in the UAE because of an economic reorientation to sectors like tourism and banking has presented enormous challenges for migrants without professional, technical and top managerial experience. The situation not only has a negative effect on the mental wellbeing of these migrants, but is also riskier for their continued survival. Hence, a significant percentage of Indian migrants working in sectors like construction are forced to work for more than the average 8-hour day.

Reflecting on the entrepreneurial contributions of Ghanaian migrants in four German cities – Hamburg, Berlin, Hannover and Frankfurt – Andoh, Berrones-Fleming, and Dornberger’s (2019) survey of 54 respondents revealed that, ‘paying for their employees to acquire training to improve their skills and knowledge’ and ‘offering to help build community facilities’ received the least scores in terms of their social contributions, each achieving a mean score of 1.63 and 1.39 respectively out of a total score of 5. Regarding their economic contributions in Germany, they found that ‘investing in properties in Germany’, ‘making financial investment in shares, bonds, stocks, fixed deposits’, ‘number of business employees in all other branches’, and the number of business branches all achieved a mean score below 1.5.

The insights uncovered from the literature thus far do not seem to present desirable living conditions for Ghanaian migrants living abroad. The extent to which monies remitted back home contribute towards their seemingly perverse living conditions is the empirical quest pursued in the subsequent sections of the paper. Following the insights uncovered from the literature, we hypothesise that an analysis of the relationship between migrant living conditions abroad and the amount of money remitted back home will reveal a negative and statistically significant relationship, controlling for other variables. Details regarding the model specification are elaborated in Section 4.3 of the paper. The article now turns to the study context and elaborates on the methodology.

3. Study context and methodology

3.1. Study context

The United Kingdom (UK) was chosen as the case study for this research because Ghanaians have long migrated to the UK for studies, work or both (Wolf and Mulder 2018). Recent data from the Office of National Statistics indicates that there are 114,000 Ghanaians resident in the UK, of which 110,000, reside in England, 2,000 in Scotland, and 1,000 in Wales (ONS 2020). London alone accommodates 71,000 Ghanaians with many of the rest clustered in other metropolitan areas like Manchester, Birmingham, Nottingham, Bristol, Liverpool. These metropolitan areas attract Ghanaian migrants for two reasons: first is the existing (ethnic) social networks that facilitate access to accommodation and jobs, and the second is that these areas are typically multicultural cities with large immigrant populations thereby facilitating easier integration.

3.2. Research design

An online survey was adopted as the research design primarily because although the Ghanaian population living in the UK is known, more granular information about them, including their contact information and residential address is difficult to come by, partly because they constitute personally identifiable information which must be protected. Thus, by using an online survey, we were able to gain access to Ghanaians in different parts of the UK without having direct access to them. Also, because the research was not funded, the online survey was found to be both a cost and time-saving approach to collect empirical data (Lefever, Dal, and Matthíasdóttir 2007; Wright 2005).

3.3. Sampling

Considering that online surveys rely on voluntarism and participants cannot be ascertained a priori (Wright 2005), a combination of purposive and snowballing sampling approach was adopted to identify potential respondents. Among those purposefully sampled were individuals from recognised diaspora groups, leaders of churches with Ghanaian membership, the officials from the Ghana High Commission in London, Ghanaian associations in London, Glasgow and Cardiff and other people within the researchers’ social networks. This purposive sampling approach was similarly adopted in Obeng-Odoom’s (2010) remittance research of Ghanaian migrants in Sydney, Australia. The snowball sampling, on the other hand, involved asking the purposefully sampled respondents to introduce us to other Ghanaian diaspora communities they know about in other parts of the UK. This way, respondents’ selection bias by the researcher can be minimised while reaching out to people outside the researchers’ social networks (Bethlehem 2010). That said, it is worth mentioning that, by design online survey designs tend to suffer from under-representation of certain groups, notably the aged and those without access to internet connectivity, and our survey is no exception (Wright 2005). The purposive sampling entailed the researchers identifying initial contacts among Ghanaian migrants across the UK from their social networks and sending them a link to the survey with a request to complete and share the URL to the survey with other Ghanaians within their social networks. Our primary contacts comprised. While the online survey has some limitations such as targeting people with access to and active use of the internet, it provided the opportunity to reach out to a wider audience beyond the networks of the primary contacts (Wiersma 2013).

3.3. Research instrumentation

To collect primary data, a close-ended survey questionnaire was created using Qualtrics. We found a close-ended questionnaire useful for this exercise because it allows for quantitative analysis of pre-coded survey responses (Lefever, Dal, and Matthíasdóttir 2007).

The questionnaire was organised into four main parts. The first part captured information about respondents’ demographic and socio-economic circumstances. This comprised information related to gender, age bracket, educational attainment and marital status. The second part captured respondents’ migration information. This included migrants’ home region in Ghana, their UK destination and length of stay in the UK, residency status and housing tenure. The third part asked questions relating to the employment situation, namely, employment skills, employment tenure and monthly income bracket. Next, we gathered data on migrants’ remittance behaviour. This comprised information about the frequency of remittance, the amount remitted, the beneficiaries of the remittance and the reasons for remitting. The final part of the survey asked respondents to score six statements formulated to capture aspects of their living conditions following the insights gathered from the literature. The six statements comprised the following:

1. Sending money home negatively impacts my ability to maintain financial stability

2. Sending money home negatively impacts my ability to go overseas for holidays

3. Sending money home negatively impacts my ability to pursue further education and/or training in the UK

4. Sending money home negatively impacts my choice of housing tenure in the UK

5. Sending money home negatively impacts my ability to make economic investments in the UK

6. Sending money home negatively impacts my ability to make charitable donations in the UK

These statements were measured on a Likert Scale of 1–5, showing varying levels of agreement, where a score of 1 signalled strong disagreement while a 5 depicted strong agreement.

3.4. Survey dissemination and data collection

We followed both our purposive and snowballing sampling approaches and a range of social media platforms to disseminate our survey questionnaire. After using Qualtrics to design the survey questionnaire, a unique URL that redirected people to the survey was generated. This URL was then sent to Ghanaians in our social networks. This entailed directly messaging or emailing the survey links to people in our network asking them to complete the survey. Messaging platforms such as WhatsApp, Telegram, Facebook, Twitter, LinkedIn. These primary contacts were also implored to forward the link to their contacts and member of Ghanaian diaspora groups in the UK they are aware of. Additionally, the survey link was to the Ghanaian High Commission entreating them to help circulate it to Ghanaians in the UK. Finally, we Finally, the survey link was posted on social media profile pages of the first and third authors, for wider visibility. The survey remained open for three months – from December 2016 to February 2017, during which time a total of 206 participants completed the questionnaire. After cleaning the data, 129 responses were valid and used in the analysis. The data cleaning exercise entailed removing responses from participants who: merely opened the survey link but did not answer any question but was counted by the Qualtrics algorithm (n = 15), completed only portions of the survey, such that their demographic, migration, remittance and perceptions about the impact of remittance on their living conditions could not be matched (n = 62). This final sample used in the analysis proved to be adequate for the intended inferential statistical analysis adopted and elaborated below (Norman 2010).

3.5. Data analysis

The cleaned data were analysed using both descriptive and inferential statistics after profiling the respondents. The descriptive statistics entailed generating measures of central tendencies, such as mean, standard deviation, minimum and maximum scores on the six statements formulated to capture migrants’ living conditions. Regarding our inferential statistics, we first created a ‘Living Conditions Index’ (hereafter LCI)’ from the respondents’ self-reported scores of their living conditions after validating the internal consistency of the six statements. Using the scores from this index, we specified a Hierarchical Multiple Regression model where we regressed respondents’ demographic, migration, employment and remittance information on the LCI. Hierarchical models are useful for evaluating the contributions that each predictor variable included in the model makes above and beyond other variables previously entered into the equation (Martin 2007). The entry of the variables into the model is often based on existing theory or a conceptual framework or hypothesis posited by the researcher as in the case of this paper (Tabachnick and Fidell 2013). The model primarily aimed to explore the relationship between migrants’ living conditions and the amount of money remitted when controlling for other variables. In keeping with the study’s hypothesis, the amount of money remitted was taken as the main independent variable of interest in the base model. The next section presents the findings from the survey.

4. Findings

The first part of the findings presents a profile of the respondents while the second part summarises the descriptive statistics on the statements capturing migrants’ living conditions.

4.1. Demographic profile of respondents

In terms of the gender of the respondents, there were nearly as many males (51.9%) as females (48.1%). The age profile of the respondents is predominantly youthful with nearly half (49.6%) of them aged between 25 and 34 years and another 29.5% within the 35–44 age bracket, suggesting that it is the youth who frequently migrate. Also, the majority of the respondents (60.5%) were married and more than half (57.4%) had between 1 and 3 dependants. Consistent with the ‘brain drain’ thesis and migration of the highly skilled (Bredtmann, Martínez Flores, and Otten 2019), 36.4% and 49.5% of the respondents held tertiary and post-tertiary level education respectively. These figures are substantially higher than the proportion of graduates in both Ghana (5%) (Government of Ghana 2019, 19) and the UK (40%) (ONS 2017). It also reinforces Peil’s (1995, 348) argument that Ghana has become ‘a major exporter of educated people’. This finding, however, differs from the situation in Germany where Andoh, Berrones-Fleming, and Dornberger (2019) report that 13% and 22.2% of Ghanaian migrants respectively hold a diploma/HND and a bachelors/postgraduate qualification (Table 1).

Table 1. Demographic characteristics of respondents.

Demographic characteristics	Number (N)	Percentage (%)
Gender
Male (1)	62	48.1
Female (2)	67	51.9
Age bracket
15–24 (1)	8	6.2
25–34 (2)	64	49.6
35–44 (3)	38	29.5
+ 45 (4)	19	14.7
Marital status
Single (1)	38	29.5
Married (2)	78	60.5
Separated (3)	3	2.3
Divorced (4)	7	5.4
Widowed (5)	3	2.3
Educational attainment
Basic education (1)	4	3.1
Junior High School (2)	1	0.8
Senior High/Technical/Vocation School (3)	13	10.1
Tertiary level (Bachelors & Higher Diploma) (4)	47	36.4
Post-tertiary (Masters & PhD) (5)	64	49.5

Source: Authors’ online survey results (2017).

Regarding respondents’ migration information, a relative majority migrated from the regions with the largest cities in Ghana, namely, Ashanti (27.1%), Greater Accra (21.7%) and Western (13.2%). This data corresponds to official statistics in Ghana on regions receiving the highest proportion of remittance sent by migrants living abroad (see Table 2). Also, in the UK, the regional distribution of the respondents is fairly consistent with statistics on Ghanaians in the UK with a disproportionate majority (50.4%) living in London (in all 33 boroughs) – the most expensive and economically prosperous city in the UK (Atkinson 2020).

Table 2. Respondents’ migration information.

Migration Information	Number (N)	Percentage (%)
Region of out-migration in Ghana
Greater Accra (1)	28	21.7 (80.3)^1
Ashanti (2)	35	27.1 (61.9)^1
Western (3)	17	13.2 (91.9)^1
Northern (4)	1	0.8 (21.4)^1
Central (5)	15	11.6 (75.1)^1
Volta (6)	6	4.7 (44.2)^1
Eastern (7)	13	10.1 (69.7)^1
Brong Ahafo (8)	7	5.4 (65.8)^1
Upper West (9)	4	3.1 (29.6)^1
Upper East (10)	3	2.3 (18.3)^1
Current destination in the UK	65	50.4 (62.3)^2
London (1)	3	2.3 (2.5)^2
North West (2)	4	3.1 (0.8)^2
North East (3)	5	3.9 (2.5)^2
Yorkshire & Humber (4)	9	7.0 (-)^2
East Anglia (5)	12	9.3 (6.1)^2
West Midlands (6)	8	6.2 (0.8)^2
South West (7)	11	8.5 (8.8)^2
South East (8)	5	3.9 (1.7)^2
Scotland (9)	2	1.6 (-)^2
Northern Ireland (10)
Residence status	35	27.1
Indefinite leave to remain (1)	41	31.8
Limited leave to remain (2)	28	21.7
British passport holder (3)	25	19.4
Holder of British and Ghanaian passport (4)
Housing tenure	99	76.7
Renting (Including PRS and Social housing) (1)	30	23.3
Homeownership (2)

Note: Figures in parenthesis in Column 1 are original codes for the categorical data. Figures in ()¹represent the percentage of remittances received from migrants living abroad across different regions in Ghana (Source: GLSS 7). Figures in () ²represent the percentage of Ghanaian migrants across different regions in the UK (N=114,000).

Source: (ONS 2020) and Authors’ online survey results.

Regarding the length of stay in the UK, a little over a fifth (22.5%) and nearly half (47.3%) have lived in the UK for between 5 and 10 years and over 10 years respectively. Also, 68.2% of respondents have some settled residence status in the UK. An overwhelming majority (76.7%) of the respondents lived in rented accommodation while a little over a fifth (23.3%) owned their dwelling. This confirms Diko and Tipple’s (1992) earlier assertion that the majority of Ghanaians in London are renters.

In terms of employment, about 8 in every 10 of the respondents possessed a high or intermediate skill level in their current employment. For example, the majority (47.3%) had professional skills while 16.3% and 25.6% possessed highly technical and intermediate skills. Because most respondents are graduates (see Table 1), this relatively high proportion of professionals and other skilled workforce relative to unskilled workers reinforces the argument that migration from developing countries to more prosperous Western countries tend to deprive the countries of origin of the highly skilled workforce who could drive local economic development and prosperity (Docquier and Marfouk 2006) (Table 3).

Table 3. Respondents’ employment information.

Employment Information	Number (N)	Percentage (%)
Skills required for current employment	61	47.3
Professional (1)	21	16.3
Highly technical skills (2)	33	25.6
Intermediate skills (3)	14	10.9
No skills (4)
Employment tenure
Part-time (1)	86	66.7
Full-time (2)	43	33.3
Monthly income (In £)
100 – 500 (1)	12	9.3
501 – 1000 (2)	15	11.6
1001 – 2,500 (3)	63	48.8
2,501+ (4)	39	30.2

Note: Numbers in the parenthesis are the original codes used for the survey questions.

Source: Authors’ online survey results (2017).

Also, two-thirds (66.7%) of respondents work on a part-time basis while the remaining one-third (33.3%) work full-time. It is not immediately clear why there are twice as many part-time workers as there are full-time workers. A plausible reason may be labour market difficulties migrants experience in host countries, where they rarely find jobs that meet their academic training or professional qualification and hence are forced to either find as many jobs that help them get by or downgrade their job preference (see Nowicka 2012). Income-wise, nearly half (48.8%) earn between £1,001 and £2,500 and a little less than a third (30.2%) earn above £2,500. Compared to an average monthly income of Ghc 898.65 (Government of Ghana 2015) (£157) back home, these earnings might seem much higher. However, considering that majority of the respondents live in London, an ‘Alpha City’ (Atkinson 2020), and as renters, they might not be so much better off when rent and other bills are deducted from their incomes.

Concerning their remittance practices, the majority of respondents (45%) send money home monthly. This is followed by those who remit ‘as and when a need arises’ (34.9%) and those who remit ‘as and when they have money’ (10.9%). Also, in terms of the money sent home, a little more than a quarter (26.4%) regularly send between £150 and £200, followed by those who remit above £200 (24%). Regarding the money sent home, it appears parents back home are the biggest beneficiaries either as exclusive recipients (26.4%) or in combination with siblings (19.4%) and others (28.7%). This finding is consistent with other studies on Ghana which suggests that parents tend to be the main recipient of overseas remittance (Kabki, Mazzucato, and Appiah 2004) (Table 4).

Table 4. Respondents’ remittance information.

Remittance information	Number (N)	Percentage (%)
Frequency of remittance
Weekly (1)	2	1.7
Monthly (2)	50	43.5
Semi-annually (3)	4	3.5
Yearly (4)	5	4.3
As and when I have money (5)	13	11.3
As and when a need arises (6)	39	33.9
Amount regularly remitted (In £)
10 – 50 (1)	7	5.4
50.1 – 100 (2)	25	19.4
100.1 – 150 (3)	22	17.1
150.1 – 200 (4)	34	26.4
200.1+ (5)	31	24
None at all (6)	10	7.8
Beneficiary of remittance
Only parents (1)	34	26.4
Only spouse and children (2)	3	2.3
Only siblings (3)	4	3.1
Both parents and siblings (4)	25	19.4
Only extended relatives (5)	9	7
Only friends and non-relatives (6)	6	4.7
Others including parents (7)	37	28.7
Others including spouse and children (8)	5	3.9
Others including siblings (9)	5	3.9
Others including friends and non-relatives (10)	1	0.8
Reasons for remitting
Upkeep of parents (0, 1)	41: 88	31.8: 68.2
Education and training of siblings (0,1)	74: 55	57.4: 42.6
Nuclear family business support (0,1)	91: 38	70.5: 29.5
Private business and investment (0,1)	75: 54	58.1: 41.9
Retirement in Ghana (0,1)	101: 28	78.3: 21.7

Source: Authors’ online survey results (2017).

Whilst reasons for remitting include education and training of siblings, family needs, business support, and others, it is ‘the upkeep of parents’ that majority of the respondents (68.2%) selected as the prime reason for remitting. This reinforces the argument about the inadequate welfare support for the elderly in Ghana (Kabki, Mazzucato, and Appiah 2004; Musah-Surugu et al. 2018).

4.2. Descriptive statistics on migrants’ living conditions

This section presents the descriptive statistics on the six statements capturing migrants’ living conditions. Overall, respondents did not show either weak or strong agreement with each of the six statements as they tended to moderately agree with each of the six statements with a mean score marginally above 2.5. However, the negative impact of remittance on respondents’ ability to go overseas for holidays seems to resonate more with respondents than the other reasons, recording a mean score of 3.02 (see Table 5). The Cronbach’s Alpha – a statistic reported when developing scales ‘intented to measure attitudes or affective construct’ often a Likert Scale (Taber 2018, 1273) was subsequently computed to gauge the internal consistency of these six statements purporting to measure migrants living conditions (Field 2018). It yielded a score of 0.854, indicating a strong internal consistency and meaning that the six statements are measuring the same construct (Taber 2018).

Table 5. Respondents’ score on constructs capturing their living conditions.

Constructs of living conditions	N	Min	Max	Mean	SD
Remittance negatively impacts my ability to maintain financial stability	129	1	5	2.88	1.279
Remittance negatively impacts my ability to go overseas for holidays	129	1	5	3.02	1.346
Remittance negatively impacts my ability to pursue further education and/or training in the UK	129	1	5	2.72	1.212
Remittance negatively impacts my choice of housing tenure in the UK	129	1	5	2.85	1.281
Remittance negatively impacts my ability to make economic investments in the UK	129	1	5	2.56	1.249
Remittance negatively impacts my ability to make charitable donations in the UK	129	1	5	2.63	1.275
Scale reliability – Cronbach’s Alpha (a) = 0.854, Mean = 16.66, Variance = 33.76, SD = 5.810, N of items = 6

Note: Respondents were asked to score each statement on a 1–5 where 1 indicated weak agreement with the statement and 5 indicated strong agreement. SD denotes Standard deviation.

Source: Authors’ online survey results (2017).

Following the scale reliability test, we proceeded to create an index aimed at generating a composite score of migrants’ living conditions. We are mindful that creating indices from different variables often tends to homogenise variables, some of which may not be intuitively related. However, we have drawn on the insights from the literature regarding migrants’ living conditions in creating this index. The Living Condition Index is an aggregate score of each respondent’s score on all of the six statements, expressed as a percentage of the aggregate maximum score one could achieve for all six statements. This is illustrated in the formulae below and the descriptive statistics on the LCI is reported in Table 6: $\mathrm{Living}\mathrm{Condition}\mathrm{Index}(\mathrm{LCI})={\displaystyle \frac{\mathrm{Aggregate}\mathrm{score}\mathrm{on}\mathrm{all}\mathrm{six}\mathrm{constructs}}{\mathrm{Aggregate}\mathrm{Maximum}\mathrm{score}\mathrm{on}\mathrm{all}\mathrm{six}\mathrm{statements}}\times 100\text{\%}}$

Table 6. Descriptive statistics on the Living Condition Index (LCI).

Variable	N	Minimum	Maximum	Mean	Standard deviation
Aggregate score	129	6	30	16.6589	5.81014
Living Condition Index	129	20	100	55.5297	19.36714

Source: Authors' online survey results (2017).

This approach to creating indices is intuitive and has recently been applied in both academic (Chakrovorti and Chatvurvedi 2019; Ehwi 2020) and industry research (ITU 2019).

The article now turns to the research hypothesis and model specification.

4.3. Model specification

The hypothesis guiding the study is that there is a statistically significant and negative correlation between the amount of money migrants remit and their living conditions in their host country, controlling for other variables. Thus, we want to test both the nature and strength of the relationship between migrants’ LCI in the UK and the amount of money they remit. Against this backdrop, we specified a Hierarchical Multiple Regression model where the base model comprises migrants’ LCI as the dependent variable and the amount of money regularly remitted as the main explanatory variable of interest. Also, following literature on the determinants of remittance (see, for example, Carling 2008), we introduced other control variables sequentially following the approach by Ayuso et al. (2011) to examine the changes in the explanatory power of the model (Adjusted R²) and the level of significance of the main independent variable in each model (see the Appendix for categorisation of variables). The hierarchical models are summarised as follows:

Model 1: LCI = ß ₀ + ß₁ Amnt + e;

Model 2: LCI = ß ₀ + ß₁ Amnt + ß₂ Demg + e;

Model 3: LCI = ß ₀ + ß₁ Amnt + ß₂ Demg + ß₃ Migr + e;

Model 4: LCI = ß ₀ + ß₁ Amnt + ß₂ Demg + ß₃ Migr + ß₄ Empl + e;

Model 5: LCI = ß ₀ + ß₁ Amnt + ß₂ Demg + ß₃ Migr + ß₄ Empl + ß₅ Remt + e;

Where ‘Amnt’, ‘Demg’, ‘Migr’, ‘Empl’, and ‘Remt’ respectively denote amount remitted, demographic, migration, employment and remittance information. The breakdown of these variables and how they were coded are summarised in the Appendix. Also, ß ₀ denotes the correlation of the constant when no amount has been remitted, while ß_(1-5) are standardised correlation coefficients of the independent variables and e denotes the error term.

4.4. Model results

In the base model – Model 1 (see Table 7) – we show the relationship between the amount of money remitted and migrants’ LCI without any control variable. The explanatory power of this base model (Adjusted R²) is 13.6% of the variance explained in the LCI, however, the model is not statistically significant and hence, the 12.7% decrease in migrants’ living conditions any time the remitted amount increases above £150 can be ignored. In Model 2, we control for respondents’ demographic information, namely age bracket, gender, marital status and gender. The explanatory power of this model increases to 23.7% and the model fit becomes statistically significant.

Table 7. Hierarchical Regression of Migrants’ Living Condition Index and Amount remitted controlling for other variables.

Variables	Model 1	Model 2	Model 3	Model 4	Model 5
	ß (SE)	ß (SE)	ß (SE)	ß (SE)	ß (SE)
Constant	(2.963)***	(3.937)***	(4.485)***	(6.969)***	(7.026)
Amt (Ref = ≤ £ 150)	−0.127 (2.254)	−0.138 (2.167)	−0.166 (2.513)***	−0.177 (2.270)***	−0.198 (2.347)***
Demographic variables Gend (Ref=Female)		−0.256 (3.275)***	−0.269 (3.3266)***	−0.268 (3.422)***	−0.198 (3.690)***
Age_bra (Ref = ≤ 34 years)		0.030 (3.343)	0.038 (3.417)	0.053 (3.515)	0.088 (3.721)
Mar_ stat (Ref=Non-coupled HH)		−0.140 (3.360)	−0.116 (3.371)	−0.084 (3.550)	−0.092 (3.751)
Edu_ att (Ref= ≤ Tertiary)		−0.213 (3.285)*	−0.229 (3.332)*	−0.234 (3.451)*	−0.241 (3.567)*
Migration variables Hm_Reg (Ref= Gt. Accra & Ashanti)			−0.145 (3.292)	−0.150 (3.408)	−0.086 (3.638)
Reg_UK (Ref= London)			−0.028 (3.374)	−0.038 (3.467)	−0.028 (3.545)
Res_sta (Ref = Unlimited leave)			0.132 (3.710)	0.093 (3.967)	0.045 (4.139)
Employment variables  Emp_sk (Ref=Professionals)				0.053 (3.837)	0.006 (4.088)
Emp_ten (Ref = Part-time)				0.026 (3.829)	0.033 (3.879)
Inc_bra ≤£1000 (Ref=Not selected) £1001–£2,500 (Ref=Not selected) >£2,500 (Ref = Not selected)				−0.075 (4.982) −0.038 (4.081) −0.033 (4.073)	−0.156 (5.848) −0.040 (4.179) −0.036 (4.068)
Remittance variables  Remt_Freq  (Ref=Structured)					−0.150 (3.733)
Ben_Rmt (Ref=family exclusive)					−0.304 (3.543)
Parents’ upkeep (Ref=Not selected) Education & training of siblings (Ref=Not selected) Family business (Ref=Not selected) Private business (Ref=Not selected) Retirement in Ghana (Ref=Not selected)					−0.086 (3.619) 0.141 (3.553) 0.132 (3.937) −0.078 (3.590) 0.014 (4.266)
Model Fit	N=129 Adj. R^2 = 0.139 Sig. = 0.150	N=129 Adj. R^2 = 0.237 Sig. = 0.003	N=129 Adj. R^2 = 0.268 Sig. = 0.0040	N=129 Adj. R^2 = 0.280 Sig. = 0.020	N=129 Adj. R^2 = 0.330 Sig. = 0.045

Note: *p<10%, **p<5% ***p<1%.

Source: Authors’ survey results (2017).

Also, the negative correlation between the amount of money remitted and migrants’ LCI increases to 13.8% but is still not statistically significant. Meanwhile, males compared to females experience a 25.6% decrease in their LCI. This inverse relationship is statistically significant and could be linked to more males being breadwinners in their households and hence expected to send more money back home than their female counterparts. Also, being educated at the post-tertiary level relative to possessing a tertiary level of education reduces the LCI by 22.9%. This statistically significant inverse relationship could be interpreted to mean that higher educational attainment is associated with being prosperous, which could potentially lead to an increase in financial demands from relatives back home (Shooshtari et al. 2014).

In Model 3, we control for respondents’ migration information, namely, their home region in Ghana, region in the UK and residency in the UK, and this increases the model’s explanatory power to 26.8%. Here, there is still a statistically significant negative correlation between the amount of money migrants remit and their LCI, with an increase in the amount remitted above £150 resulting in a 16.6% decrease in a migrant’s LCI. Also, educational attainment still has the same statistically significant but negative effect on migrants’ LCI. However, the home region in Ghana, the region in UK and migrants’ residency do not have a statistically significant relationship on migrant’s LCI despite having −14.5%, −2.8% and 13.2% correlation co-efficient respectively.

In Model 4, we control for migrants’ employment information, namely, employment skills, employment tenure and income bracket and the explanatory power of the model increases to 28%. An increase in the amount remitted beyond £150 still results in a 17.7% statistically significant decrease in migrants’ living conditions. Also, gender and educational attainment have the same relationship with the LCI as witnessed in model 3.

In Model 5, we include other remittance information, namely, frequency of remittance, beneficiaries of remittance and reasons for remittance, and the overall explanatory power of the model increases to 33%. In terms of the correlations between the amount remitted and migrants’ LCI, we still observe a nearly 20% decrease in the latter following an increase in the amount sent beyond £150. This inverse correlation is still statistically significant. Both gender and educational attainment still have a statistically significant negative impact on migrants’ LCI, with being a male and having a post-tertiary education resulting in a 19.8% and 24.1% decrease in migrants’ LCI respectively.

Besides the remitted amount, age and gender, other variables also correlated negatively with migrants’ LCI which are worth highlighting, despite them not being statistically significant. First, being a coupled household leads to a 14% decrease in LCI. This is possibly because having a coupled household could result in two sources of income, which could lead to an increase in demand for remittance back home in Ghana. Second, having migrated from outside the Greater Accra and Ashanti regions leads to a 14.5% decrease in LCI. This is possibly because there are fewer economic opportunities for those left behind in these areas, which can likely lead to an increase in demand for economic support. Third, living outside London leads to a 2.8% decrease in LCI. This is perhaps because of the relatively fewer economic opportunities and limited social capital from ethnic communities in these areas, potentially making meeting the cost of living hard in these areas, and thereby making any demand for remittance more acutely felt. Fourth, earning monthly income over £1000 increases the decline in LCI from 3.3% to 7.5%. This is perhaps because earning more often signals economic prosperity from the perspective of relatives back home, leading to growing demands for remittance. Sixth, remitting in an unstructured manner leads to a 15% decrease in LCI, possibly because this approach leaves little room for good financial planning. Furthermore, sending money for parents’ upkeep and personal business respectively lead to an 8.6% and 7.8% decrease in LCI, possibly due to the regularity of these transfers.

On the other hand, several variables correlated positively with migrants’ LCI. First, having a limited leave to remain in the UK leads to a 4.5% increase in LCI. A possible explanation is that migrants are aware that the obligation to remit would potentially cease when they return home. Second, not being a professional worker leads to a 5.3% increase in LCI, possibly due to family and societal expectations that professionals should remit more money. Thirdly, LCI correlates positively (14.1%). with sending money for the education and training of siblings This may be linked to the assumption that when siblings are better educated and trained, they would stop requiring remittance from abroad, as they would be in a better position to look after both themselves and their parents, thereby relieving migrants abroad of this burden. Finally, there was a positive correlation (1.4%) between LCI and planning to retire in Ghana. This is possibly due to the perception that monies being sent are ultimately being saved up for use when migrants retire in Ghana.

6. Discussion and conclusion

This article tested the hypothesis that sending remittances back home has a negative and statistically significant impact on migrants’ living conditions in their host countries. Using survey data collected from 129 Ghanaian migrants living in the UK, which was used to create a Living Condition Index (LCI) from six statements related to migrants’ living conditions abroad, we specified a hierarchical multiple regression model where migrants’ LCI was regressed on the amount of money they regularly remit, and controlled for their demographic, migration, employment and remittance situation. The result suggests that there is nearly a 20% decrease in migrants’ LCI when they remit more than £150, thereby confirming the hypothesis formulated. This finding is important for several reasons.

First, this study has contributed to the existing literature on Ghanaian migrants’ living conditions abroad, many of which are either qualitative or draw mainly on descriptive statistics that provide no insights into both the nature and magnitude of the relationship between the amount of money Ghanaian migrants remit and their living conditions (Diko and Tipple 1992; Kabki, Mazzucato, and Appiah 2004; Obeng-Odoom 2010; Owusu 1998). This study has filled this gap by providing the much-needed quantitative insights into both the nature and magnitude of the relationship between remittance sent home and migrants living conditions.

Also, while most of these studies point out that Ghanaian migrants in Western countries often do not enjoy the best living conditions that they possibly could have afforded, the trade-offs they endure to keep remitting are often dwarfed by the arguments regarding their motivations to remit. Indeed, the existing mainstream theoretical arguments underpinning migrants’ motives for remitting, including (tempered) altruism and (enlightened) self-interest (Lucas and Stark 1985), tend to portray migrants who frequently remit to their relatives back home as cheerful, socially conscious and good-hearted people who have no qualms with remitting, and who do not suffer the consequences of their remittance behaviour. However, while the reasons to send money home are well-founded, the opportunity cost of consistently sending money, and the impact on migrants’ lives as a result of this remittance behaviour, demands attention. This is because any money a migrant sends home means a lost opportunity to use that money for their consumption or investment in their host country. This opportunity cost, adverse effects and the implications for migrants, including emotional pressures (cf. Amoyaw and Abada 2016), however, remain overshadowed in the literature by the motives underpinning migrants’ remittance.

The importance of social norms and cultural practices may have also played a key role in this important relationship not receiving the warranted level of academic attention. For example, studies show that there is a significant social recognition that households with migrants abroad who regularly remit enjoy in Ghanaian society (Kabki, Mazzucato, and Appiah 2004) as well as among the migrants’ ethnic groups abroad. Thus there is a social stigma and scorn from communities in both their home country and abroad for failing to remit or complaining about remittance, as this has become the overriding logic for having a relative abroad (Amoyaw and Abada 2016; Shooshtari et al. 2014). This social stigma and scorn, as sociological institutionalists (Dacin 1997; Harty 2005) suggest, is powerful enough to make migrants rationalise this practice, knowing very well that they are leading a precarious life in their host countries.

Furthermore, motivations to remit even under precarious conditions are, to a great extent, determined by social and emotional connotations of the remittance-sending behaviour (Amoyaw and Abada 2016). Through monetary remittance, migrants being away from home country can maintain, manage and reproduce social relationship with the non-migrant relatives and friends remaining in the home country. In line with findings from transnationalism literature (Mazzucato 2008, Carling 2014; Kuuire et al. 2016), such form of connection to the place of origin represents a transnational social practice that allows migrants to live ‘dual lives’ and maintain identity across borders, without the need to dissociate themselves from the home country due to the physical distance. In this sense, remittance-sending not only makes a financial contribution to the wellbeing of the family and friends or have investment purposes, but it becomes the key transnational social practice for migrants to demonstrate their affection and emotional connection to the home country, and, hence, is sustained despite the possible financial difficulties and aggravated living conditions.

However, in light of the negative correlation found between the amount of remittance sent home and migrants’ living conditions identified in this research, the sustainability of this practice and its potential long-term effects should be questioned. As mentioned previously, remittances to lower and middle-income countries for over a decade have surpassed overseas development assistance, and are even predicted to overtake foreign direct investment in the next couple of years after 2020. This is despite the current 20% decrease in the global volume of remittances (from US$ 544 in 2019 to US$ 455) owing to the COVID-19 pandemic (World Bank 2020). What this projection means in light of the identified relationship between the amount Ghanaian migrants remit and their living conditions, is that the current precarity migrants endure is likely to persist and potentially become further exacerbated. This is partly because the economic devastation of this pandemic has considerably interrupted livelihood and economic opportunities in developing countries, including Ghana. For example, Ghana’s estimated GDP growth in 2020 had been downgraded from a target of 6.8% to 2.6% and workers in the tourism industry were predicted to be disproportionately affected by the pandemic (Deloitte 2020).

Moreover, according to the Ghana Statistical Service, 77.4% of households in Ghana experienced a decrease in income since 16 March 2020 following the introduction of COVID-19 restrictions, with another 77.4% reporting an increase in food prices (Government of Ghana 2020). As the countercyclical hypothesis of remittance suggests, migrants tend to remit more money back home during crises in both their host and home countries (Frankel 2009; World Bank 2020). Thus, in light of the economic hardships that many households are currently facing as a result of the COVID-19 pandemic, Ghanaian migrants abroad, including those in the UK, will be expected to remit more in order to provide sufficient economic succour and a social safety net to their relatives in Ghana. However, a recent study by Public Health England (2020) has found that Black, Asian and Minority Ethnic (BAME) groups in England, of which Ghanaian migrants are part, tend to have poor socio-economic conditions and hence experience poor health outcomes. The report further pointed out that overcrowding is more common among BAME groups than their white counterparts with 30% Bangladeshi households, 16% black African households and 18% Pakistani households suffering from overcrowding compared to 2% white households. Although these reports attribute the health and housing precarity experienced by BAME groups in the UK to structural factors like inequalities and racism (Public Health England 2020), Diko and Tipple (1992) have earlier pointed out that, for Ghanaian migrants, their quest to save more of their income and remit for construction projects back home in Ghana predispose them to become tenants in the lowest end of the housing market.

It remains uncertain whether the living conditions of Ghanaian migrants in the UK will improve while they keep remitting substantial sums of money back home in the wake of the novel coronavirus. However, it is also true that over the past few decades, remittances have proven to be a reliable source of international direct transfer to both households in Ghana (Gyimah-Brempong and Asiedu 2015; Musah-Surugu et al. 2018) and the Ghanaian macroeconomy (Adams, Cuecuecha, and Page 2008; Adenutsi and Ahortor 2008). Hence, it would be impractical to suggest that migrants stopped or cut down on the amount remitted to improve their living conditions in the UK. Rather, some alternatives could be explored to help relatives in Ghana to become more self-sustaining and less dependent on remittances from overseas. Such an alternative can include helping them to set up local businesses or investing in established local businesses that have the potential to generate regular streams of income to support their livelihood. Migrants can also share the knowledge, discipline and workplace skills they have developed from their working years abroad to support these businesses to succeed. Often, just sending money home has proven not to be enough if not accompanied by some form of oversight and support.

Also, since the cost of money transfer also adds to the cost of remittance, we will echo Yeates and Owusu-Sekyere’s (2019) suggestion that the UK government could help lessen the financial impact of remittances on migrants’ precarious living conditions by viewing ‘familial social payments as an allowable expense on sender’s tax returns or by extending credits to overseas care provision’ (151).

Despite the important insights into the relationship between Ghanaian migrants’ remittances and their living conditions in the UK, there are some limitations of our study which future studies can address to deepen our understanding of this aspect of the remittance literature. First, the six variables used in constructing the LCI do not constitute an exhaustive list of factors that determine migrants’ living conditions and hence future studies could expand the list to capture different aspects of migrants’ living conditions, such as mental health, working conditions, household composition, and others. Second, unlike other studies (Shooshtari et al. 2014), the LCI was conceptualised as a static variable and hence did not account for inter-temporal dynamics of migrants’ living conditions from the time they arrived in the UK, or found their first job, to when they settled and obtained indefinite leave to remain. Hence, future studies can examine the strength of the relationship at different critical junctures regarding migrants’ stays abroad. Finally, a comparative approach to examine the relationship between remittance and migrants’ living conditions among other African migrant groups in the UK and potentially in other Western countries could offer new insights.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Appendix. Variables included in the model, variable information and variable coding.

Table

Variables	Variable information	Old and new code	Variable type	N & (%)
Dependent variable LCI	Migrants’ Living Condition Index measured in percentage	N/A	Discrete	100
Independent/Control variable
Amnt	Regular amount respondents remit to home country	≤ £150 (1-3,6) = 0, >£150 (4-5) = 1	Dummy	64:65 (49.6:50.4)
Demography	Demographic characteristics of respondents
Gen Age_bra Mar_ stat Edu_ att	Respondents’ gender Respondents’ age bracket Respondents’ marital status Respondents’ educational attainment	Female (1) = 0, Male (2) = 1 ≤34 years (1-2) = 0, >34 years (3-4) = 1 Non-coupled HH (1,3-5)= 0, Coupled HH (2) = 1 ≤ Tertiary (1-4) = 0 > Tertiary (5) = 1	Dummy Dummy Dummy Dummy	67:62 (51.9:48.1) 72:57 (55.8:44.2) 51:78 (39.5:60.5) 65:64 (50.4:49.6)
Migr	Respondents’ migration information
Hm_ Reg Reg_UK Res_ sta Emp_ sk Emp_ ten Inc_brk	Respondents’ home region in Ghana Region where respondents’ currently live in the UK Respondents’ residence status in the UK Respondents’ employment skills Respondents’ employment tenure in the UK Respondents’ income bracket	Gt. Accra & Ashanti = (1-2) 0, Other regions (3-10) = 1 London (1) = 0, Other regions (2-10) = 1 Unlimited leave to remain (1,3-4) = 0, Limited leave to remain (2) = 1 Professional (1) = 0, Others (2-4)= 1 Part-time (1) = 0, Full-time (2) = 1 Other income brackets (3-4) = 0, ≤£1,000 (1-2) = 1 Other income brackets= 0 (1,2,4), £1,001-2,500 (3) = 1 Other income brackets (1-3) = 0, >£2,500 (4)= 1	Dummy Dummy Dummy Dummy Dummy Dummy Dummy Dummy	63:66 (48.8:51.2) 65:64 (50.4:49.6) 88:41 (68.2:31.8) 61:68 (47.3:52:7) 86:43 (66.7:33.3) 102:27 (79.1:20.9) 66:63 (51.2:48.8) 90:39 (69.8:30.2)
Remittance	Respondents’ remittance information
Rmt_Freq Ben_Rmt Rzn_Rmt	Respondents frequency of remittance Beneficiaries of remittance sent home Reason for remitting	Structured (1-4) = 0, Unstructured (5-6) = 1 Exclusively for family members (1-6) = 0 For family members and others (7-10) = 1 Other reasons = 0, Parental upkeep =1 Other reasons = 0, Education and training of siblings = 1 Other reasons = 0, Family business support = 1 Other reasons = 0, Personal business = 1 Other reasons = 0, Retirement in Ghana =1	Dummy Dummy Dummy Dummy Dummy Dummy Dummy	70:59 (54.3:45.7) 75:54 (58.1:41.9) 41:88 (31.8:68.2) 74:55 (57.4:42.6) 91:38 (70.5:29.5) 75:54 (58.1:41.9) 101:28 (78.3:21.7)

Note: Figures in parenthesis in column 3 are the original codes (see Tables 1–5).

Source: Authors’ survey data (2017).