Elsevier

Economic Modelling

Volume 79, June 2019, Pages 141-153
Economic Modelling

Remittances and household expenditure behaviour: Evidence from Senegal

https://doi.org/10.1016/j.econmod.2018.10.007Get rights and content

Highlights

  • We analyse the impact of internal and external remittances on several household budget shares in Senegal.

  • When considering average impact of remittances on expenditure behaviour, we find an overall productive use of remittances.

  • The impact of remittances on household expenditure behaviour disappears when marginal spending behaviour is considered.

  • Contrary to some existing evidence, we show that remittances are treated just like any other source of income.

Abstract

We use different econometric techniques, from propensity score matching to multinomial treatment methods, to assess the impact of internal and external remittances on several household budget shares in Senegal. When only considering the average impact of remittances on the household expenditure behaviour, we find an overall productive use of remittances. However, the impact of remittances disappears when the marginal spending behaviour is considered, i.e., households do not show a different consumption pattern with respect to their remittance status. The marginal spending behaviour therefore suggests that, in the decision on how to allocate expenditure, remittances are treated just as any other source of income.

Introduction

The main objective of this paper is to contribute and extend the debate on whether remittances impact economic development and, in particular, on how remittances are spent or used by the recipient households. The literature presents three views on how remittances are used, which depends on how they are perceived by the recipient household. The first view, which is part of the permanent income hypothesis, is that remittances are transitory income and therefore are spent, at the margin, in more ‘productive’ activities like human and physical capital. If this is the case, then remittances should have a long-term impact on growth and development of the receiving country. The second view is that remittances are compensatory income and therefore spent more on consumption rather than investment goods. While this could result in generating domestic production perhaps, it can also lead to an indirect effect on inflation in a number of developing countries (Narayan et al., 2011). The final view regards remittances as just any other source of income and therefore no difference in the expenditure behaviour emerges from the households' remittance status.

We conduct the analysis using migration and remittance data from a much-neglected region in migration research, Africa. More precisely, we use the data from Senegal, which has recently become one of the leading out-migration (both internal and international) countries in sub-Saharan Africa. The survey data, collected in 2009, was part of the African Migration Project, led by the World Bank. The data allows us to identify households with at least one current migrant (‘migrant household’ hereafter) and households receiving remittances (‘recipient households’ hereafter). We assume that each household has to allocate its expenditure on several commodities and we want to understand whether receiving remittances has any impact on the household decision. We are able to identify four types of goods: food, durable goods, education, and other type of items such as expenditure on funerals, engagements and weddings.1

The remittance analysis is based on three types of households: non-recipients; internal recipients (remittances received from within Senegal) and external recipients (remittances received from international destinations). The reason for considering the origin of remittances is not only because internal and external migrants might have different motivations for remitting to their families but those families who receive external remittances might perceive, treat and use them in a different way than those receiving internal remittances (Azizi, 2018a).2 In fact, several empirical studies have found that internal and external remittances affect differently the consumption behaviour of households in terms of consumed and investment goods. For instance, Adams (1996) finds that internal remittances have an equalizing impact on income distribution while external remittances have the opposite effect (see also Clément, 2011; Adams and Cuecuecha, 2010b; Adams et al., 2008b; Castaldo and Reilly, 2007).

Migrant and recipient households are not randomly selected – characteristics associated with a particular household rather than their status of being a remittance recipient can potentially have an impact on their expenditure behaviour. As we could not find a suitable instrument in the data to correct the bias, we use propensity score matching (PSM) to evaluate the impact of receiving a “treatment”, i.e., receiving remittances, on household expenditure behaviour, at the average level.3 The propensity score matching shows that internal transfers do not have a strong impact on household expenditure decision whereas external transfers negatively affect the expenditure on food while the effect on education expenditure is positive. It therefore shows that external remittances are used in more productive activities like investment in human capital rather than on consumption.

The PSM results provide us with a benchmark against which it is possible to evaluate the Working-Leser model, which relates budget shares linearly to the logarithm of total household expenditure. For the Working-Leser model, we carry out the estimates using the Multinomial Treatment Regression (MTR). We employ this method because of the potential bias coming from the selection of unobservables. It confirms that external remittances have a negative effect on food expenditure and a strong positive effect on education expenditure, at the average level. Moreover, receiving internal remittances has a positive effect on both food and education which does not emerge from the PSM analysis.

Additionally, the Working-Leser model allows us to compute marginal budget shares and elasticities for the different types of goods. When we explore household consumption decision looking at the marginal behaviour, we do not find any significant difference in how households allocate their expenditure. We also find that different types of recipient households perceive expenditure items in quite similar ways, i.e. in terms of necessity, normal or luxury goods. The demand elasticities and marginal consumption results show that remittances do not change the household expenditure behaviour, i.e., remittances are treated just like any other income.

The rest of the paper is structured as follows. Section 2 contains a brief outlook on Senegal and its emergence as an important emigration country while Section 3 presents the relevant literature on the relation between remittances and household consumption patterns. Section 4 describes the dataset used in this study and Section 5 presents the propensity score matching techniques and the Working-Leser model. Section 6 discusses and compares the empirical findings and the last section concludes the paper.

Section snippets

Senegal: a brief background

Sub-Saharan Africa (SSA) is becoming an important emigration region. The rate of migration from Africa has evolved dramatically in early 2000s, with the growth rate in net migration was over 275% between 2000 and 2005 (Naudé, 2010). The stock of emigrants reported in 2013 for the SSA region was 23.2 million or 2.5 per cent of population (World Bank, 2016). The relatively high rate of outmigration is due to the interplay of different factors: political and economic instability, violent

Literature review

The household is the first unit which takes decision on the use of remittances and therefore, in essence, it determines the role remittances play in the development process of the receiving country. Remittances are received under imperfect information, uncertainty and with different regularity (Seshan, 2012; Chami et al., 2005) and therefore how they are perceived by the households is not straightforward. Based on the previous empirical studies, the impact of remittances on household

Data

We investigate household expenditure behaviour using data from a recent Migration and Remittance Household Survey in Senegal.7

Methodology

The Engel curve approach is generally adopted to analyse the impact of remittances on household expenditure patterns. The main challenge of this approach is to address the concern linked to the endogeneity of remittances. The usual way to deal with the endogeneity is to use the instrumental variable approach. The literature provide us with a large number of potential instruments to address the endogeneity of remittances. Some of those instruments are related to economic conditions in the

Estimates from PSM

Table 4 contains the summary statistics of the explanatory variables used in the empirical analysis, including information about the household members who are currently abroad. Differences exist in the household size as well as the composition of the two types of households: recipient vs non-recipient. The size of the household is much larger for those household receiving remittances compared to the one who do not. Moreover, the presence of children and elderlies is higher for the recipient

Conclusions

Migrant's transfers can potentially play an important role in developing countries and it is important to understand how recipient households perceive and use them. The question on what remittances represent for the households is still a topic of debate. The way remittances are spent – on consumption or investment goods – is strictly determined by the context of the analysis. We contribute to the existing debate by investigating the impact of remittances on household expenditure behaviour in

References (45)

  • ANSD (Agence Nationale de la Statistique et de la Démographie) du Sénégal

    Troisième Recensement de la Population et de l'Habita (RGPH-III)

    (2002)
  • S. Azizi

    Altruism: primary motivation of remittances

    Appl. Econ. Lett.

    (2017)
  • S. Azizi

    Why Do Migrants Remit? the World Economy

    (2018)
  • S. Azizi

    The impact of workers' remittances on human capital and labor supply in developing countries

    Econ. Modell.

    (2018)
  • S. Bertoli

    Networks, sorting and self-selection of Ecuadorian migrants

    Ann. Econ. Stat.

    (2010)
  • S. Bertoli et al.

    Migration, remittances and poverty in Ecuador

    J. Dev. Stud.

    (2014)
  • R. Brown et al.

    Comparative effects of migrants' remittances on composition of recipient household income in two small, island economies

    Appl. Econ.

    (2011)
  • M. Caliendo et al.

    Some practical guidance for the implementation of propensity score matching

    J. Econ. Surv.

    (2008)
  • A. Castaldo et al.

    Do migrant remittances affect the consumption patterns of Albanian households

    S. E. Eur. J. Econ.

    (2007)
  • G. Cerulli

    Econometric Evaluation of Socio-economic Programs: Theory and Applications

    (2015)
  • R. Chami et al.
    (2005)
  • F. Cisse

    Senegal

  • Cited by (25)

    • Adaptation to climate change: Air-conditioning and the role of remittances

      2023, Journal of Environmental Economics and Management
    • Do remittances reshape household expenditures? Evidence from Nepal

      2022, World Development
      Citation Excerpt :

      Despite a rise in both internal and international remittance flows in recent years, the ongoing COVID-19 pandemic poses a threat to migration, employment opportunities, and associated remittance flows, increasing the need to better understand how households utilize their remittance income. Broadly, remittances can influence household spending in sizeable ways, making it an important source of income, both at the micro and macro levels (Randazzo & Piracha, 2019; Rapoport & Docquier, 2006). In this article, we investigate the impact of remittances on household expenditure patterns in Nepal.

    • An analysis of food demand in a fragile and insecure country: Somalia as a case study

      2021, Food Policy
      Citation Excerpt :

      The final piece of our analysis examines how our elasticity estimates are impacted once we take account of the likelihood of a household receiving some form of remittance income from outside the country. The reason for examining this issue is that remittances are an important source of income in developing countries and regions such as Sub-Saharan Africa (SSA) (Randazzo and Piracha, 2019). The importance of remittances to household food security in Somalia is noted by Majid et al. (2018) who report that this source of income, estimated to be $1.4 billion in 2016, enables households to buy more food and more diverse types of food.

    View all citing articles on Scopus
    View full text