How does the flow of remittances affect the trade balance of the Middle East and North Africa?

ABSTRACT

The Middle Eastern and North African (MENA) economies have one of the highest degrees of dependency on received remittances worldwide. In this study, we have examined the role of remittances in the trade balance of 11 labour abundant MENA countries. Our panel regression analysis showed that the inflow of remittances has fostered the trade deficit. We also found that the final effect of remittances depends on the level of domestic capital formation. The results are robust after controlling for other drivers of trade deficit such as income, inflation, exchange rate and institutions as well as country and year fixed effects.

KEYWORDS:

• Remittances
• Middle East and North Africa
• trade balance
• panel fixed effects

JEL CLASSIFICATION:

• C23
• F22
• F14

Acknowledgements

Sherif M. Hassan acknowledges the financial support of Yousef Jameel Academic Program. We thank helpful comments of the Editor (Judith Clifton) and two anonymous referees.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental data for this article can be accessed here.

Notes

1. One possible reason for the trade deficit is the existence of a current account deficit due to the shortfall in domestic national savings relative to investment, known as ‘twin deficit hypothesis’ (Congressional Budget Office [CBO], 2000; Elwel 2007). However, we eliminate this possibility, as LA-MENA countries have been experiencing an increasing general trend in their current account balances in the last three decades (World Economic Outlook 2018). A trade deficit may also point to a highly productive and a growing economy, as some manufactured imports can be used as intermediates for the production of future exports. Nevertheless, the sample of countries in this study and the MENA region in general has the lowest global share of manufactured exports (World Bank 2018).

2. Appendix A is available online at https://works.bepress.com/mr_farzanegan/34/download/.

3. In literature, we also encounter studies, which have used regional trade models such as gravity model in order to investigate the intensity of trade among countries. According to this model, the main drivers of intensity of trade between two countries are economic size of importing and exporting countries as well as their geographical distance. Tinbergen (1962) was the first who used gravity to explain trade flows. The main conclusion of gravity models is (1) bigger countries trade more, and (2) more distant countries trade less. In our study, we are also benefiting from theoretical gravity model by including variables such as income per capita and other local macroeconomic indicators but also controlling for country and year fixed effects in addition to factors which were not in the focus of gravity such as channel of remittances in the MENA region. For a review of application of gravity models in international trade and their limitations see Shepherd (2013).

4. The assumption of no tariffs and remittances as exogenous source permanent income may help simplify the concept. We revisit these assumptions in empirical part and especially consider the case that remittances are not exogenous but can get affected by trade balance. We will use two-stage least squares to address the possibility of treating remittances as an endogenous variable.

5. Bruegel provides a large database for real effective exchange rates. Retrieved from http://bruegel.org/2012/03/real-effective-exchange-rates-for-178-countries-a-new-database/.

6. We have re-estimated the panel fixed effects model in Table 1 using lagged values of remittances and other independent variables. The results of both models at levels and lagged are similar, the only difference is that the coefficients of remittances in models with the non-oil trade balance as dependent variable, are slightly smaller in the lagged model relative to the model in level (these results are available upon request).

7. Using the lags of variables as instruments in 2SLS method has been used in literature such as Biswas, Farzanegan, and Thum (2012), Imai et al. (2014) and Catrinescu et al. (2009) among others.