Harnessing remittances for the poor: the role of institutions

ABSTRACT

This paper investigates the direct effect of institutions on poverty and explores whether the remittances and poverty link can be strengthened by institutions in Sub-Saharan Africa. The results suggest that a country characterized by sound financial development and good and stable government with systems to control corruption and attract investment will provide the enabling environment to reduce the rate, depth and severity of poverty. Such quality institutional attributes reinforce the effectiveness of international remittances in reducing poverty.

KEYWORDS:

• Institutions
• remittances
• poverty

JEL CLASSIFICATION:

• J6
• N4
• O1
• O5

Acknowledgements

I am grateful to my advisors, Dr. Barbara Roberts and Dr. Jesse Matheson, for their support. I am also thankful to the Department of Economics at the University of Leicester in the UK, and Business Studies Department of the Lancaster University (Ghana campus). My sincere gratitude extends to the anonymous referees and the Editor-in-Chief of the Journal of Economic Policy Reform, Professor Judith Clifton, for their constructive comments. All remaining errors are mine.

Notes

1. Note that Liberia, Djibouti, Equatorial Guinea, Sao Tome and Principe, Somalia, Eritrea and Zimbabwe were excluded due to data limitation.

2. Some of the macroeconomic studies on remittances and development that use remittances to GDP ratio are Acemoglu and Robinson (2010), Aggarwal, Demirgüç-Kunt, and Pera (2011), Anyanwu and Erhijakpor (2010), Giuliano and Ruiz-Arranz (2009) and Gupta, Pattillo, and Wagh (2009).

3. According to Page and Plaza (2006); IFAD (2009), informal remittances flow through post offices, saving cooperatives and microfinance institutions, and mobile money services which are not included in the International Monetary Fund estimates could be greater than or equal to the official estimates for Sub-Saharan Africa.

4. See for example Karemera, Oguledo, and Davis (2000), Vogler and Rotte (2000) and Hatton and Williamson (2002).

5. The p-values of the Hansen J-statistics of the estimated models are above 0.1.