Relationship between Remittance and Household Expenditure in Benin: Evidence from ARDL Model

Abstract

This study examines the short- and long-run relationship between remittance and Benin household expenditure. We analysed 45 years of time series data, from 1974 to 2019, from the World Bank Open data repository using the autoregressive distributed lag (ARDL) to cointegration model. The findings suggest no significant long-run relationship between remittance receipts and household expenditure. However, the short-run relationship is positive and significant indicating that household remittance increases their expenditure solely in the short run. Development aid and trade openness used as control variables are negatively associated with household expenditure. These do not significantly increase household expenditure in the short and long-run. We recommend that policymakers in developing countries like Benin adopt instruments that encourage remittance inflows and promote household efficient use to meet short- and long-run expenditure needs.

Plain language summary

Previous studies did not provide conclusive information on the long-run importance of remittance for households in the context of developing countries. We contribute to the literature by evaluating the short and long-run relationship between remittances and household expenditure taking evidence from Benin. We retrieved data on remittance, household expenditure, development aid, trade openness, and inflation from the World Bank open data repository and analysed them using an econometric model. The long-run results show no significant association between household remittance and their expenditure. In the short-run, household remittances increase their expenditure. We recommend that policymakers in developing countries like Benin adopt instruments that encourage remittance inflows and promote household efficient use to meet short- and long-run expenditure needs.

Keywords

remittances household expenditure ARDL model development aid Benin

Introduction

The increasing volume of global remittance shows its importance as one of the largest sources of foreign capital inflows into developing countries. According to the World Bank (2021a), remittances represent about 739 billion US dollars in 2021 and account for about 0.8% of the global GDP. About 77% of global remittances are received by the low-and-middle-income countries (World Bank, 2021b). The remittances are nearly three times the net official development aid received by these countries. In Benin, a lower-middle-income country, remittance remains one of the most important foreign capital inflows. Between 2011 and 2021, remittances increased by 20% (equivalent to 59 million US dollars), accounting for 1.3% of the Benin GDP in 2021, and standing as the second-largest foreign capital inflow after development aid (World Bank, 2021c). It is sent by 14% of the Benin population representing migrants (INStaD, 2024) and accounts for 2% of household income (Kinkpe, 2024).

Remittance refers to the money and goods transmitted to households by migrant workers working outside their origin communities (Adams, 2011). According to the World Bank (2021b), remittances are part of the compensation paid to migrant employees and are sent home as personal transfers. These consist of all transfers in cash or kind received by households from non-resident households (World Bank, 2021b). Remittances result from migrants’ decision to send money to their home of origin which is explained by the theories of pure altruism, self-interest, and tempered altruism or enlightened self-interest (Lucas & Stark, 1985). The theory of pure altruism posits that migrants remit to support their families due to emotional ties, mostly observed among migrants from developing countries (Azizi, 2019). The self-interest theory suggests that migrants remit money to embark on investment projects or repay a loan (Lucas & Stark, 1985). According to tempered altruism or enlightened self-interest, remittances are part of an intertemporal, mutually beneficial contractual arrangement (investment or risk sharing) between migrants and their homes (Lucas & Stark, 1985). Against the motivation to remit, remittances can take various natures in household income affecting how households spend it. Existing literature suggests three potential expenditure channels through which households allocate remittances. First, households disburse the remittance receipts as compensatory income for the consumption of food and durable goods (Ajefu & Ogebe, 2021; Akobeng, 2022; Mishra et al., 2022; Mondal & Khanam, 2018; Nguyen et al., 2017; Thapa & Acharya, 2017). Second, households spend the remittance receipts as transitory income to invest in productive activities, health, and education (Askarov & Doucouliagos, 2020; Dash, 2023; Mohammed & Karagöl, 2023; Nanziri & Mwale, 2023). Thus, remittances generate long-run income streams for households through financial and human development. Third, remittances are like any other income source and do not necessarily affect household expenditure (Makina, 2024). Based on existing migration data, Benin households employ mostly the first two expenditure channels through which they allocate 46% of remittance receipts to food, followed by health (21%), education (12%), ceremonies (7%), and other various expenditures accounting for 14% (INStaD, 2024).

Despite the growing remittances in Benin and its importance in receiving households, no empirical evidence exists supporting its positive effect on household consumption expenditure in Benin. We study Benin analyzing remittance data from 1974 to 2019 to address this gap. The long-run implications of remittances on household consumption expenditure are also very limited in the literature, specifically in the sub-Saharan African context. The few existing long-run studies including Musakwa and Odhiambo (2019) in Botswana, Aslam and Sivarjasingham (2020) in India, Lamsal (2024) in Nepal, and Makina (2024) in Lesotho also led to inconsistent and contradictory findings that can hardly be extended to other developing countries like Benin. While the studies of Aslam and Sivarjasingham (2020), Lamsal (2024), and Makina (2024) found a positive long-run relationship, Musakwa and Odhiambo (2019) found no long-run relationship between remittances and household consumption expenditure. By employing the autoregressive distributed lag model (ARDL), which simultaneously estimates short- and long-run effects, we provide new evidence contributing to the literature on the long-run relationship between remittances and household expenditure.

Following this introduction, a summary of past empirical studies on the relationship between remittances and household expenditures in low- and middle-income countries is provided. Data, variables, and the econometric model are presented. After discussing the results, some concluding remarks and potential policy implications are provided.

Literature Review

Existing empirical studies on the relationship between remittances and household expenditure can be categorised into three groups. The first body of literature addresses the question of the actual effect of remittances on total household expenditure. Using various methods and with different country focus, these studies suggest that, on average, household remittances increase their total spending. Démurger and Wang (2016), Randazzo and Piracha (2019), and Nanziri and Mwale (2023) analysed household survey data using the propensity score matching (PSM) technique in China, Senegal, and Zambia respectively, and found that remittance receipts have a positive impact on household expenditure. Besides, Yamada et al. (2022) and Akobeng (2022) adopted an instrumental variable (IV) estimation approach which also led to the outcome that remittances have an increased household expenditure effect in Tajikistan and Ghana respectively. In Tajikistan, Yamada et al. (2022) pinpoint the heterogenous effect of remittances among household heads given their sex, age, and education level. According to the authors, the remittance effect on household expenditure is more pronounced among male, older, and less-educated head-leading households. Having assessed the remittances and household consumption expenditure nexus among farming households in Ghana, Akobeng (2022) found that remittance plays an income-compensating role as households use it to compensate for agricultural income loss. We have the studies of Adeseye (2021) and Chintamani and Kulkarni (2023) which relied on primary data at a county level in Nigeria and India respectively. Employing the ordinary least square (OLS) and logit/probit methods, the studies found a significant positive association between remittances and household consumption expenditure. Beyond increasing household expenditure, remittance is found to have a multiplier effect of reducing poverty (Awan et al., 2017; Muhammad Abubakar et al., 2016; Nanziri & Mwale, 2023; Wagle & Devkota, 2018; Yoshino et al., 2017) and inequality (Basanta & Malvika, 2016; Ellyne & Noxolo, 2017). Moreover, remittance can significantly reduce household consumption volatility in the short and long-run (Mondal & Khanam, 2018).

The second body of literature delved into each household consumption expenditure component. These studies demonstrate that remittances affect each component differently. Using the propensity score matching (PSM) method, studies such as Mahapatro et al. (2017) in India, Awan et al. (2017) in Pakistan, Thapa and Acharya (2017) in Nepal, and Dey and Ahmed (2024) in Bangladesh, conclude that remittance-receiving households increase their expenditure on food consumption, and invest in health, education, and housing. Using primary data and regression models, Dhakal and Oli (2020) also show that remittance positively affects household food consumption and education expenditure. Ajefu and Ogebe (2021) and Mishra et al. (2022) employed an instrumental variable (IV) technique and found a result consistent with the latter. Besides, they provide additional consideration to the effect of remittances on household expenditure. Mishra et al. (2022) show that remittance is negatively associated with household expenditure on alcohol and tobacco in Vietnam. Considering five sub-Sahara African countries such as Burkina Faso, Kenya, Nigeria, Senegal, and Uganda, Ajefu and Ogebe (2021) found the remittance effect on food, health, and education expenditure consistent across different household expenditure quantiles. Kakhkharov and Ahunov (2022) assess the remittance effect on household wedding expenditures in Uzbekistan and unveil that remittance-receiving households are less likely to consume conspicuously. While most studies show a positive remittance effect on household expenditure, a few other studies found mixed results. Adams and Cuecuecha (2010) and Maara et al. (2019) found that remittance-receiving households spend less on food and more on education and durable goods in Guatemala and Kenya respectively. Nguyen et al. (2017) and Raihan et al. (2022) found a significant relationship between remittances and household spending on durable goods but not education in Vietnam and Bangladesh respectively. Furthermore, Wang et al. (2021) suggest that remittance only has a small increasing effect on food and health expenditures and no effect on other household consumption expenditures in Kyrgyzstan.

The third body of literature investigated the short and long-run effects of remittances on household expenditure. We identified four long-run studies conducted in Botswana (Musakwa and Odhiambo, 2019), India (Aslam and Sivarjasingham, 2020) Nepal (Lamsal, 2024), and Lesotho (Makina, 2024). Musakwa and Odhiambo (2019) examined the impact of remittances on poverty reduction. The study employed time-series data from 1980 to 2017 and considered two poverty proxies including household consumption expenditure. Using the ARDL model, the results show that remittances have no association with household expenditure in the short and long-run. Aslam and Sivarjasingham (2020) employed the ARDL model on time-series data from 1975 to 2018, and found that remittances have a positive short and long-run relationship with household consumption expenditure. Lamsal (2024) also employs the ARDL model and draws a similar conclusion. Even with the Johansen cointegration technique and the Engle-Granger residual approach, Makina (2024) found a positive long-run relationship between remittances and household consumption expenditure. Contrary to the latter, Makina (2024) found a negative short-run relationship between remittances and household consumption expenditure.

In light of all the above, we observe that the long-run assessments of the relationship between remittances and household expenditure report inconsistent and sometimes contradictory results. For instance, while some studies find a positive long-run association between remittances and household expenditure, others report no significant relationship. This suggests that the effects of remittances on household expenditure may be context-specific and potentially influenced by methodological differences. This inconsistency underscores a critical research gap, particularly for developing countries striving to achieve the Sustainable Development Goals (SDGs), where remittances could play an important role in enhancing household consumption, contributing to economic growth, and alleviating poverty. Therefore, further long-run assessments are essential to generate valuable insights to support development policymaking in Benin and other similar countries.

Methodology

Data and Variables

We analysed 45 years of time-series data (1974–2019) from the World Bank Indicator repository. We considered variables such as household final consumption expenditure, personal remittances, net official development aid, inflation, and trade openness.

Household final consumption expenditure, expressed in US dollars, constitutes the market value of all goods and services, including durable products such as cars, washing machines, and home computers purchased by households. It includes payments and fees to governments to obtain permits and licenses. It can be used as an official poverty measure. Besides, it explains economic differentials in education, unemployment, and healthcare (Srivastava & Mohanty, 2010). It is a better measure of achieved living standards (Stoyanova & Tonkin, 2018). Final consumption expenditure is preferred as a measure of economic well-being in low-income countries, as it reflects the long-run economic status of households (Friedman, 1957). It is measured more accurately than household income (Brewer et al., 2017). Its data collection process involves adjustments for household size, composition, and price level (Srivastava & Mohanty, 2010).

Personal remittances are expressed as a percentage of GDP and used in the model to capture international remittances flows. The importance of choosing remittances as a percentage of GDP is to account for the country’s sizes (Musakwa & Odhiambo, 2019).

Official development aid is also expressed as a percentage of GDP. It consists of disbursements of loans made on concessional terms (net of repayments of principal) and grants by official agencies. It is included in the model because of its economic welfare promotion and poverty reduction implications (OECD, 2018). Following Ellyne and Noxolo (2017), the study determines whether all foreign capital inflows generally increase household expenditure.

Inflation shows the rate of price change in the economy and is measured by the annual growth rate of the GDP implicit deflator. According to Rashid Talukdar (2012), high inflation lowers the value of cash holdings, real income, and purchasing power. Inflation is included in the model as a control variable because it can determine household expenditure.

Trade openness is the sum of exports and imports of goods and services measured as a share of gross domestic product. Trade openness benefits local producers as they can easily access foreign inputs for use in their production processes (Pradhan & Mahesh, 2014). It also benefits consumers because of the availability of increased variety and cheaper products, which raises national income and triggers poverty reduction (Pradhan & Mahesh, 2014). Trade openness is included in the model to explore its association with household expenditure, similar to Berg and Krueger (2003) regarding its correlation with economic growth.

Econometric Model

The study employed the autoregressive distributed lag (ARDL) to cointegration model developed by Pesaran and Shin (1999). In the model, household final consumption expenditure represents the dependent variable while personal remittances constitute the main explanatory variable. The other variables are used as control variables.

The choice of the model lies in its advantages. The model is free from residual correlation as all variables of the model are assumed to be endogenous (Nkoro & Kelvin, 2016). The ARDL approach is reliable for small samples and simultaneously estimates short and long-run effects (Pesaran et al., 2001). The model provides unbiased estimates irrespective of the endogeneity of some regressors (Pesaran & Shin, 2012). Given this characteristic, addressing endogeneity is usually not a concern within the ARDL framework.

According to Pesaran and Shin (1999), the general ARDL model is expressed as:

y_{t} = α_{0} + α_{1} t + \sum_{i = 1}^{p} \emptyset_{i} y_{t - i} + β^{'} X_{t} + \sum_{i = 0}^{q - 1} {β_{i}}^{*^{'}} Δ x_{t - i} + μ_{t}

(1)

Δ X_{t} = P_{1} Δ X_{t - 1} + P_{2} Δ X_{t - 2} + \dots + P_{s} Δ X_{t - s} + ε_{t}

(2)

where:

- $X_{t}$ is the ó-dimensional variables that are not cointegrated among themselves.

- $μ_{t}$ and $ε_{t}$ are serially uncorrelated disturnces with zero means and constant variance-covariances.

-and $P_{i}$ are ó x ó coefficient matrices such that the vector autoregressive process in $Δ X_{t}$ is stable.

We logarithmically transformed all the variables to address potential heteroscedasticity. Thus, the model is expressed as:

\begin{matrix} Δ {logHCE}_{mt} = α_{0} + \sum_{i = 1}^{n} α_{1 i} Δ {logHCE}_{mt - i} \\ + \sum_{i = 0}^{n} α_{2 i} Δ logRe m_{t - i} + \sum_{i = 1}^{n} α_{3 i} Δ logOD A_{t - i} \\ + \sum_{i = 1}^{n} α_{4 i} Δ logT O_{t - i} \\ + \sum_{i = 1}^{n} α_{5 i} Δ logIn f_{t - i} + ϑ_{1} logHC E_{t - 1} \\ + ϑ_{2} logRe m_{t - 1} + ϑ_{3} logOD A_{t - 1} \\ + ϑ_{4} logT O_{t - 1} + ϑ_{5} logIn f_{t - 1} + μ_{t} \end{matrix}

(3)

where:

- $α_{0}$ is a constant.

- $α_{1 i} - α_{5 i}$ are coefficients that explain the short-run relationship between household final consumption expenditure and the explanatory variables.

- $ϑ_{1} - ϑ_{5}$ are coefficients that explain the long-run relationship between household final consumption expenditure and the explanatory variables.

- $μ_{t}$ is the error term

- $\log$ is the logarithm of base 10

- $HCE$ , $Rem$ , $ODA,$ $TO$ , and $Inf$ are household final consumption expenditure, personal remittances, net official development aid, trade openness, and inflation respectively.

According to Nkoro and Kelvin (2016), if there exists a long-run relationship between variables, both the short and long-run information can be captured by parametrizing the ARDL model into the Error Correction Model (ECM). This is expressed as:

\begin{matrix} Δ logHC E_{mt} = α_{0} + \sum_{i = 1}^{n} α_{1 i} Δ logHC E_{mt - i} \\ + \sum_{i = 0}^{n} α_{2 i} Δ logRe m_{t - i} \\ + \sum_{i = 1}^{n} α_{3 i} Δ logOD A_{t - i} \\ + \sum_{i = 1}^{n} α_{4 i} Δ logT O_{t - i} + \sum_{i = 1}^{n} α_{5 i} Δ logIn f_{t - i} \\ + γ_{1 i} EC M_{t - 1} + μ_{t} \end{matrix}

(4)

where:

- $α_{0}$ is a constant.

- $α_{1 i} - α_{5 i}$ are coefficients that explain the short-run relationship between household final consumption expenditure and the explanatory variables.

- $μ_{t}$ is the white noise error term and, $ECM$ is the error correction model term.

-The definition of other variables remains the same as in Equation 3.

We replaced household final consumption expenditure with GDP data for a robustness check. GDP is the first indicator considered when evaluating a country’s economic growth and the well-being of its inhabitants.

Results and Discussion

Unit Root Test

The ARDL model to cointegration is the preferable technique when dealing with variables that are integrated of a different order, I (0), I (1), or a combination of both. It crashes when variables are integrated of order I (2) or higher (Nkoro & Kelvin, 2016). For this reason, the existence of unit root was checked at a 5% level of significance, using two different testing strategies which include the Augmented Dickey-Fuller (ADF) test (Dickey & Fuller, 1981) and Phillips and Perron (PP) test (Phillips & Perron, 1988). Under the ADF and PP tests, variables are said to be stationary, that is, have no unit root (integrated), when the absolute value of the t-statistic is greater than the absolute critical value, and vice-versa. An application of this shows that only the variable inflation is stationary at a level that is, integrated in the order I (0) (see Table 1).

Table 1.

Unit Root Test with Intercept and Linear Trend.

Variable	Level	ADF		PP		Status
Variable	Level	T-Test	CV	T-Test	CV	Status
Household expenditure	I (0)	−2.3428	−3.45	−2.3048	−3.54	Non-stationary
Remittances	I (0)	−2.7058	−3.45	−4.2204	−3.54	Non-stationary
Development aid	I (0)	−2.2230	−3.45	−2.1615	−3.54	Non-stationary
Trade openness	I (0)	−2.2266	−3.45	−2.6645	−3.54	Non-stationary
Inflation	I (0)	−4.2768	−3.45	−4.6297	−3.54	Stationary

Note. 5% level of significance. ADF Test Cv = −3.45; PP Test Cv = −3.45.

The variables such as household expenditure, remittances, development aid, and trade openness are found to be stationary at the first difference I (1), but not at level, that is, not integrated of order I (0) (see Table 2). The mixed stationary state of the variables confirms the appropriateness of employing the ARDL model for cointegration.

Table 2.

Unit Root Test with Intercept and No Trend.

Variable	First Difference	ADF		PP		Status
Variable	First Difference	T-Test	CV	T-Test	CV	Status
Household expenditure	I (1)	−7.339	−2.91	−7.307	−2.89	Stationary
Remittances	I (1)	−6.292	−2.91	−6.584	−2.89	Stationary
Development aid	I (1)	−7.686	−2.91	−8.343	−2.89	Stationary
Trade openness	I (1)	−6.656	−2.91	−8.343	−2.89	Stationary

Note. 5% level of significance. ADF Test Cv = −2.91; PP Test Cv = −2.89.

Cointegration Test

One of the most important steps in the ARDL to cointegration analysis involves testing the existence of a meaningful long-run relationship between the dependent variable and the set of regressors. This is done using a cointegration test (see Table 3). Cointegration is an econometric concept that mimics the existence of a long-run equilibrium among underlying economic time series that converges over time (Nkoro & Kelvin, 2016). In practice, when the bound F-statistic (or W-statistic) value is lower than the lower critical bound, there exists no long-run relationship among variables, that is, variables are not cointegrated. When the value is greater than the upper critical bound, there exists a long-run relationship among variables, that is, variables are cointegrated. In case the value falls between the lower and the upper critical bounds, the result is inconclusive.

Table 3.

Cointegration Test.

Test	Value	5% – I (0)	5% – I (1)	10% – I (0)	10% - I (1)	Status
F-statistic	2.3840	3.2775	5.1303	3.2637	4.384	Not Cointegrated
W-statistic	11.92	19.3875	25.6517	16.3184	21.9468	Not Cointegrated

Note. Function: F (Household expenditure | Remittances, Development aid, Trade openness, Inflation). I (0) = Lower critical bound; I (1) = Upper critical bound.

In this study, the results show that both at 10% and 5% levels of significance, the F-statistic (and/or W-statistic) value is less than the lower critical bound. This implies that variables are not cointegrated. Furthermore, it implies that there is no significant long-run relationship between the dependent variable (household consumption expenditure) and the regressors (remittances, development aid, trade openness, and inflation). Thus, the results suggest no long-run relationship between remittances and household expenditure in Benin, consistent with the findings of Musakwa and Odhiambo (2019) in Botswana.

Relationship Between Remittances and Household Expenditure

We only discussed the short-run model results (see Table 4) as suggested by the cointegration test. The analysis was conducted using the Schwarz Bayesian Criterion (SBC) for the optimal lag length of the model. This SBC was chosen because it is a consistent model selection criterion and performs better in most of the experiments. From the results, we observed that all the regressors are significantly associated with household expenditure except inflation.

Table 4.

ARDL Model Results.

ARDL (1,1,0,0,0) selected based on Schwarz Bayesian Criterion
Variable	Coefficient	Std-Error	T-Ratio	Prob.
Household expenditure (−1)	0.60	0.098	6.12	0.0000***
Remittances	−0.089	0.086	−1.03	0.310
Remittances (−1)	0.16	0.073	2.28	0.028**
Development aid	−0.26	0.081	−3.21	0.003***
Trade openness	−0.45	0.163	−2.74	0.009***
Inflation	0.006	0.019	0.34	0.735
Constant	3.94	0.916	4.3	0.000
Trend	0.01	0.003	3.55	0.001
R-Squared	0.98
R-Bar-Squared	0.98
S.E. of Regression	0.054
F-Statistic	333.41
Prob.	0.000
Residual sum of squares	0.11
Equation Log-likelihood	71.47
DW-statistic	1.77
Schwarz Bayesian Criterion	56.24
Akaike Info. Criterion	63.47

Note.** and *** denote statistical significance at 5% and 1% levels respectively.

The first lag coefficient of remittances is positive and significant at 5%. This means that a unit change in remittances is associated with a 0.16 change in household expenditure on average, ceteris paribus, at a 5% statistical significance level. It can be inferred that remittances received by households are positively associated with their expenditure. Remittances come as a compensatory income spent by households to purchase more goods and services. Chintamani and Kulkarni (2023) made a similar inference about remittance-receiving households in India. The authors analysed primary data collected from remittance-receiving and non-receiving households using the probit model and found that remittances increase household income. The result is also consistent with the findings of Yamada et al. (2022) who employed an instrumental variable approach on nationwide household data in Tajikistan and confirmed a positive association between remittances and household total expenditure. Contrary to Aslam and Sivarjasingham (2020), Lamsal (2024), and Makina (2024), our result is only significant in the short-run. The difference between our result and the latter could be explained by household behaviour variation across countries. Most likely, households in Benin use remittances to meet short-term expenditure needs and do not invest in any productive activities that could generate additional income in the long-run.

The coefficient of development aid is found to be negative and significant at 1%. This implies that a unit change in development aid is associated with a −0.26 change in household expenditure. This means when household expenditure increases, development aid will tend to decrease and vice-versa. Countries demand development aid to promote economic growth and reduce poverty (OECD, 2018). Achieving this aim implies that household welfare has improved, and likely translated into an increased consumption expenditure. In such an economic situation, countries will be less demanding of development aid as shown by the result. In other words, development aid will no longer be necessary when increased consumption expenditure is observed among households. The result is supported by Asongu (2014) suggesting that high-growth African countries would require less development aid given its fewer positive effects in these countries.

The coefficient of trade openness is also negatively related to household expenditure at a 1% statistical significance level. This implies that a unit change in trade openness is associated with a −0.45 change in household expenditure on average, ceteris paribus, and vice-versa. Contrary to what could have been predicted, if increased, trade openness would not be beneficial to households. This result is supported by Dowrick and Golley (2004) testing whether trade benefits vary over time and across countries. Findings from the study showed that export specialization in primary commodities harms growth. Benin intervenes in the international market, mostly through the export of agricultural commodities and the import of manufactured commodities (International Trade Centre, 2019). Therefore, we argue that the country could experience a negative effect of trade on growth suggesting a negative spillover effect on household final consumption.

Diagnostic Tests

The diagnostic tests performed on the ARDL model include serial correlation, functional form, normality, and heteroscedasticity test, at a 5% level of significance (see Table 5). The results show that the null hypothesis of no serial correlation in the time series cannot be rejected. Furthermore, the functional form of the model fits the data at a 5% significance level. Similarly, the null hypothesis of normality and heteroscedasticity assumptions cannot be rejected at a 5% level of significance.

Table 5.

Diagnostic Tests.

Test Statistics	LM Version	F version
(a) Serial Correlation	0.52 (0.472)	0.419 (0.521)
(b) Functional Form	1.35 (0.245)	1.11 (0.298)
(c) Normality	1.779 (0.411)	Not applicable
(d) Heteroscedasticity	1.635 (0.201)	1.621 (0.210)

Note. (a) Lagrange multiplier test of residual serial correlation (b) Ramsey’s RESET test using the square of the fitted values. (c) Test of skewness and kurtosis of residuals (d) The regression of squared residuals on squared fitted values.

We plotted the cumulative sum of recursive residuals (see Figure 1) and the cumulative sum of squares of recursive residuals (see Figure 2) to check for the model stability. The results show that the model lies within the critical bounds at a 5% significance level. This implies that the model is stable, and the results from the model are robust.

Figure 1.

Plot of the cumulative sum of recursive residuals (CUSUM).

Figure 2.

Plot of the cumulative sum of squares of recursive residuals (CUSUMSQ).

The results remain consistent after we replaced household final consumption expenditure with GDP (see Annex).

Conclusion

Remittance is one of the largest sources of foreign capital inflows in developing countries. In Benin, remittance has increased in the past few years and accounted for a non-negligible percentage of the GDP as the second-largest foreign capital inflow. Despite the growing inflow of remittance in Benin, no empirical evidence is available to support its positive effect on household consumption expenditure in Benin. Moreover, the long-run implication of remittances on household consumption expenditure is not sufficiently explored in the literature. The few existing long-run studies suggest findings that are misleading and inconclusive. We contributed to the existing literature by examining the short-and-long-run relationship between remittances and household consumption expenditure employing the autoregressive distributed lag (ARDL) model. The findings suggest no significant long-run relationship between remittances and household expenditure in Benin. In the short-run, remittances received by households are positively associated with their expenditure. This implies that households increase their overall expenditure with an additional remittance received. We conclude from the results that Benin households use remittances mainly to meet short-term expenditure needs.

We used other control variables such as inflation, development aid, and trade openness that could also affect household expenditure. We found that inflation does not play any significant role in influencing household expenditure. Development aid and trade openness are negatively associated with household expenditure. Households seem to not benefit when the country continues relying heavily on development aid and trade involving the exchange of raw agricultural commodities for manufactured commodities.

Based on the findings, we recommend policymakers in developing countries like Benin design and implement policies that encourage and promote efficient use of remittance by receiving households. Setting up a favourable business environment and encouraging remittance use in productive activities could lead to long-run benefits among households. Future studies could delve into the short and long-run effects of remittances when used by households to consume and invest in productive activities.

Footnotes

Appendix

Annex.

ARDL Model Results with GDP as a Dependent Variable.

ARDL (1,1,0,0,0) selected based on Schwarz Bayesian Criterion
Variable		Coefficient		Std-Error	T-Ratio	Prob.
GDP (−1)		0.59		0.1075	5.46	0.000
Remittances		−0.12		0.0815	−1.49	0.146
Remittances (−1)		0.17		0.0682	2.60	0.013
Development aid		−0.20		0.0743	−2.77	0.009
Trade Openness		−0.36		0.1497	−2.45	0.019
Inflation		0.006		0.0181	0.37	0.713
Constant		4.01		0.9789	4.09	0.000
Trend		0.01		0.0036	3.47	0.001
R-Squared		0.99
R-Bar-Squared		0.99
S.E. of Regression		0.051
F-Statistic		440.45
Prob.		0.000
Residual sum of squares		0.095473
Equation Log-likelihood		74.65
DW-statistic		1.81
Schwarz Bayesian Criterion		59.42
Akaike Info. Criterion		66.65
Test	Value	5% – I (0)	5% – I (1)	10% – I (0)	10% – I (1)	Status
F-statistic	2.6417	3.8775	5.1303	3.2637	4.3894	Not Cointegrated
W-statistic	13.2083	19.3875	25.6517	16.3184	21.9468	Not Cointegrated
Note. Function: F (GDP\| Remittances, Development aid, Trade openness, Inflation)
Test Statistics		LM version				F version
(a) Serial Correlation		0.34 (0.558)				0.28 (0.602)
(b) Functional Form		2.09 (0.148)				1.75 (0.194)
(c) Normality		1.30 (0.522)				Not applicable
(d) Heteroscedasticity		0.80 (0.371)				0.77 (0.382)

Acknowledgements

We are deeply grateful to Prof. Hans De Steur for his contribution at the early stage of this work. We want to acknowledge especially that “the article processing charge was funded by the Open Access Publication Fund of Humboldt-Universität zu Berlin.

ORCID iDs

Martial Houessou

Wannes Slosse

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

Datasets generated and/or analyzed during the current study are available in the World Bank Open Data Repository accessible via

References

Adams

R. H.

(2011). Evaluating the economic impact of international remittances on developing countries using household surveys: A literature review. Journal of Development Studies, 47(6), 809–828. https://doi.org/10.1080/00220388.2011.563299

Adams

R. H.

Cuecuecha

(2010). Remittances, household expenditure and investment in guatemala. World Development, 38(11), 1626–1641. https://doi.org/10.1016/j.worlddev.2010.03.003

Adeseye

(2021). Migrant remittance and household expenditure pattern in Nigeria. Open Journal of Political Science, 11(1), 73–98. https://doi.org/10.4236/ojps.2021.111006

Ajefu

J. B.

Ogebe

J. O.

(2021). The effects of international remittances on expenditure patterns of the left-behind households in Sub-Saharan Africa. Review of Development Economics, 25(1), 405–429. https://doi.org/10.1111/rode.12721

Akobeng

(2022). Migrant remittances and consumption expenditure under rain-fed agricultural income: Micro-level evidence from Ghana. Oxford Development Studies, 50(4), 352–371. https://doi.org/10.1080/13600818.2022.2077924

Askarov

Doucouliagos

(2020). A meta-analysis of the effects of remittances on household education expenditure. World Development, 129, 1–11. https://doi.org/10.1016/j.worlddev.2019.104860

Aslam

A. L. M.

Sivarjasingham

(2020). Remittances and Consumption Expenditure in India: An ARDL Investigation. Theresa Journal of Humanities and Social Sciences, 6(2), 26–50.

Asongu

(2014). Development thresholds of foreign aid effectiveness in Africa. International Journal of Social Economics, 41(11), 1131–1155. https://doi.org/10.1108/IJSE-01-2013-0014

Awan

R. U.

Waqar

Rahim

Sher

(2017). Impact of remittances on expenditures and poverty at household level in Pakistan. Pakistan Economic and Social Review, 55(2), 415–434. https://doi.org/10.2307/26616720

10.

Azizi

(2019). Why do migrants remit? World Economy, 42(2), 429–452. https://doi.org/10.1111/twec.12681

11.

Basanta

P. K.

Malvika

(2016). Impact of remittances on poverty: An analysis of data from a set of developing countries. Economics Bulletin, 36(1), 108–117.

12.

Berg

Krueger

(2003). Trade, Growth, and Poverty: A Selective Survey (WP/03/30). IMF Working Paper.

13.

Brewer

Etheridge

O’Dea

(2017). Why are households that report the lowest incomes so well-off? Economic Journal, 127(605), F24–F49. https://doi.org/10.1111/ecoj.12334

14.

Chintamani

B. G.

Kulkarni

(2023). Determinants and effects of international remittances: Evidence from Ratnagiri district of rural Maharashtra. Indian Journal of Labour Economics, 66(2), 617–637. https://doi.org/10.1007/s41027-022-00426-1

15.

Dash

R. K.

(2023). Do remittances crowd-in or crowd-out domestic investment? An empirical analysis of 24 low-income countries. Journal of the Knowledge Economy, 14(2), 1177–1193. https://doi.org/10.1007/s13132-022-00948-5

16.

Démurger

Wang

(2016). Remittances and expenditure patterns of the left behinds in rural China. China Economic Review, 37, 177–190. https://doi.org/10.1016/j.chieco.2015.12.002

17.

Dey

S. R.

Ahmed

(2024). International Remittances and Household Expenditure Patterns in Bangladesh. The Indian Economic Journal, 72(4), 689–710. https://doi.org/10.1177/00194662241238937

18.

Dhakal

Oli

S. K.

(2020). The impact of remittance on consumption and investment: A case of province five of Nepal. Quest Journal of Management and Social Sciences, 2(1), 27–40. https://doi.org/10.3126/qjmss.v2i1.29018

19.

Dickey

D. A.

Fuller

W. A.

(1981). Likelihood ratio statistics for autoregressive time series with a unit root. Econometrica, 49(4), 1057. https://doi.org/10.2307/1912517

20.

Dowrick

Golley

(2004). Trade openness and growth: Who benefits? Oxford Review of Economic Policy, 20(1), 38–56. https://doi.org/10.1093/oxrep/grh003

21.

Ellyne

Noxolo

(2017). The impact of remittances on poverty in Africa: A cross-country empirical analysis. In 14th African Finance Journal Conference (pp, 17–18). https://www.researchgate.net/publication/316655304

22.

Friedman

(1957). A Theory of the consumption function (Vol. 79, No, 1911). Princeton University Press. http://www.nber.org/chapters/c4403

23.

INStaD. (2024). Deuxième Édition de l’Enquête sur la Migration au Bénin (EMB-2). https://instad.bj/images/docs/insae-statistiques/enquetes-recensements/EMB/EMB-2_Rapport%20d’analyse.pdf

24.

International Trade Centre. (2019). Diversifying trade in Africa: New strategy approaches for the African continental free trade area. Geneva: Internationa Trade Centre.

25.

Kakhkharov

Ahunov

(2022). Do migrant remittances affect household spending? Focus on wedding expenditures. Empirical Economics, 63(2), 979–1028. https://doi.org/10.1007/s00181-021-02163-8

26.

Kinkpe

A. T.

(2024). Economy-wide effects of agricultural value-chain enhancement in agriculture-based economies [Humboldt-Universität zu Berlin]. https://edoc.hu-berlin.de/items/3ca1ac91-7ae6-4cfc-9b02-3f3357106e37

27.

Lamsal

M. R.

(2024). The impact of workers’ remittances on household consumption expenditure in Nepal: A time series analysis. Nepalese Journal of Management Science and Research, 7(1), 1–19. https://doi.org/10.53056/njmsr-2024.7.1.001

28.

Lucas

R. E. B.

Stark

(1985). Motivations to remitx: Evidence from Botswana. Journal of Political Economy, 93(5), 901–918. https://www-jstor-org-443.web.bisu.edu.cn/stable/1833062

29.

Maara

J. N.

Manda

D. K.

Kiiru

J. M.

(2019). Remittances and household expenditure allocation behavior in Kenya. African Journal of Economic Review, 7, 85–108.

30.

Mahapatro

Bailey

James

K. S.

Hutter

(2017). Remittances and household expenditure patterns in India and selected states. Migration and Development, 6(1), 83–101. https://doi.org/10.1080/21632324.2015.1044316

31.

Makina

(2024). The long-run relationship between remittances and household consumption: Evidence from Lesotho. Cogent Economics and Finance, 12(1), 1–13. https://doi.org/10.1080/23322039.2024.230709

32.

Mishra

Kondratjeva

Shively

G. E.

(2022). Do remittances reshape household expenditures? Evidence from Nepal. World Development, 157, 1–10. https://doi.org/10.1016/j.worlddev.2022.105926

33.

Mohammed

Karagöl

E. T.

(2023). Remittances, institutional quality and investment in Sub-Saharan Africa. Journal of Business and Socio-Economic Development, 3(4), 322–338. https://doi.org/10.1108/jbsed-07-2022-0077

34.

Mondal

R. K.

Khanam

(2018). The impacts of international migrants’ remittances on household consumption volatility in developing countries. Economic Analysis and Policy, 59, 171–187. https://doi.org/10.1016/j.eap.2018.07.001

35.

Muhammad Abubakar

Rizwan

Tariq

Muhammad Abubakar Siddique

Shehzadi

Rizwan Manzoor

Tariq Majeed

. (2016). Do international migration and remittances reduce poverty in developing countries? Science International, 28(2), 1519–1524.

36.

Musakwa

M. T.

Odhiambo

N. M.

(2019). The impact of remittance inflows on poverty in Botswana: An ARDL approach. Journal of Economic Structures, 8(1), 1–13. https://doi.org/10.1186/s40008-019-0175-x

37.

Nanziri

L. E.

Mwale

M. L.

(2023). Remittances, crowd-in effect, and household welfare. Scientific African, 19, 1–11. https://doi.org/10.1016/j.sciaf.2022.e01521

38.

Nguyen

D. L.

Grote

Nguyen

T. T.

(2017). Migration and rural household expenditures: A case study from Vietnam. Economic Analysis and Policy, 56, 163–175. https://doi.org/10.1016/j.eap.2017.09.001

39.

Nkoro

Kelvin

(2016). Autoregressive Distributed Lag (ARDL) cointegration technique: Application and interpretation. Journal of Statistical and Econometric Methods, 5(3), 63–91. https://doi.org/10.1002/jae.616

40.

OECD. (2018). Development Co-operation Report 2018: Joining Forces to Leave No One Behind, Development Co-Operation Report. Paris: OECD. https://www.oecd.org/en/publications/development-co-operation-report-2018_dcr-2018-en.html

41.

Pesaran

M. H.

Shin

(1999). Econometrics and economic theory in the 20th century. In Strom

(Ed.), Econometrics and economic theory in the 20th century (pp. 371–413). Cambridge University Press. https://doi.org/10.1017/ccol521633230

42.

Pesaran

M. H.

Shin

(2012). An autoregressive distributed-lag modelling approach to cointegration analysis. In S. Strøm (Ed.), Econometrics and economic theory in the 20th century (pp. 371–413). Cambridge University Press. https://doi.org/10.1017/ccol521633230.011

43.

Pesaran

M. H.

Shin

Smith

R. J.

(2001). Bounds testing approaches to the analysis of level relationships. Journal of Applied Econometrics, 16(3), 289–326. https://doi.org/10.1002/jae.616

44.

Phillips

P. C. B.

Perron

(1988). Testing for a unit root in time series regression A. Biometrika, 75(2), 335–346.

45.

Pradhan

B. K.

Mahesh

(2014). Impact of trade openess on poverty: A panel data analysis of a set of developing countries. Economics Bulletin, 34(4), 2208–2219.

46.

Raihan

Uddin

Ahmmed

(2022). Impact of foreign remittances on the household spending behaviour in Bangladesh. Migration and Development, 11(3), 1104–1126. https://doi.org/10.1080/21632324.2020.1870835

47.

Randazzo

Piracha

(2019). Remittances and household expenditure behaviour: Evidence from Senegal *. Economic Modelling, 79, 141–153. https://doi.org/10.1016/j.econ mod.2018.10.007

48.

Rashid Talukdar

. (2012). The effect of inflation on poverty in developing countries: A panel data analysis. Texas Tech University.

49.

Srivastava

Mohanty

S. K.

(2010). Economic proxies, household consumption and health estimates. Economic and Political Weekly, 45(16), 55–63.

50.

Stoyanova

Tonkin

(2018). An expenditure-based approach to poverty in the UK an expenditure-based approach to poverty in the UK (Working Parper 11). https://unece.org/fileadmin/DAM/stats/documents/ece/ces/ge.15/2018/mtg1/D._UK.pdf

51.

Thapa

Acharya

(2017). Remittances and household expenditure in Nepal: Evidence from cross-section data. Economies, 5(2), 1–17. https://doi.org/10.3390/economies5020016

52.

Wagle

U. R.

Devkota

(2018). The impact of foreign remittances on poverty in Nepal: A panel study of household survey data, 1996–2011. World Development, 110, 38–50. https://doi.org/10.1016/j.worlddev.2018.05.019

53.

Wang

Hagedorn

Chi

(2021). Remittances and household spending strategies: Evidence from the Life in Kyrgyzstan Study, 2011–2013. Journal of Ethnic and Migration Studies, 47(13), 3015–3036. https://doi.org/10.1080/1369183X.2019.1683442

54.

World Bank. (2021a). Net official development assistance and official aid received (current US$)—Benin. https://data.worldbank.org/indicator/DT.ODA.ALLD.CD?locations=BJ

55.

World Bank. (2021b). Personal Remittances received (current US$). https://data.worldbank.org/indicator/BX.TRF.PWKR.CD.DT

56.

World Bank. (2021c). Personal remittances, received (current US$)—Low & middle income. https://data.worldbank.org/indicator/BX.TRF.PWKR.CD.DT?locations=XO

57.

Yamada

Shimizutani

Murakami

(2022). Remittances, household welfare, and the COVID-19 pandemic in Tajikistan. Asian Development Review, 39(2), 147–174. https://doi.org/10.1142/S0116110522500159

58.

Yoshino

Taghizadeh-Hesary

Otsuka

(2017). International remittances and poverty reduction: Evidence from Asian developing countries (759; ADBO Working Paper). https://www.adb.org/publications/international-remittances-and-poverty-reduction