Help or hindrance? U.S. aid on growth

ABSTRACT

We distinguish between U.S. aid and non-U.S. aid and study the effects on growth in recipient countries. Our analysis exploits time variation in aid due to changes in the supply of U.S. aid and cross-sectional variation in a country’s tendency to receive any U.S. aid. We find that U.S. aid has a positive effect on growth. In particular, U.S. economic aid has a positive effect on growth while U.S. military and food aid have no effect on growth. There is also no evidence of U.S. aid crowding out aid from other countries. The effect of U.S. aid on growth is smaller for countries that are well endowed with natural resources, less ethnically polarized, and more aligned with the U.S.

KEYWORDS:

• U.S. aid
• growth

JEL CLASSIFICATION:

• F35
• O10
• O40

Acknowledgments

We are grateful to Daniel Trefler and seminar participants at University of Oklahoma for helpful comments and discussions. We also appreciate the generosity of Nathan Nunn and Nancy Qian in making available the datasets used in their paper.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ The existing evidence is mixed. Clemens, Radelet, and Bhavnani (2012), Doucouliagos (2019) and Drher and Langlotz (2017) provide a survey of key papers on the impact of aid on growth. Some key papers include: Clemens, Radelet, and Bhavnani (2012), Alesina and Weder (2002), Bauer (1971), Burnside and Dollar (2000), Collier and Dollar (2002), Dalgaard, Hansen, and Tarp (2004), Hansen and Tarp (2001), Roodman (2004), Svensson (2003), Werker, Ahmed, and Cohen (2009), and Dreher et al. (2017).

² Region classification is based on World Bank classification and consists of the following groups: South Asia, East Asia and Pacific, Europe and Central Asia, Latin America and Caribbean, Middle East and North Africa, and sub-Saharan Africa.

³ In the aid literature, aid is scaled to GDP. This is because the real cost of providing goods and services may increase with GDP.

⁴ The instruments used in Rajan and Subramanian (2008) cannot be used to study the effect of U.S. aid on growth because most of their instruments are time-invariant characteristics of donor and aid-recipient countries except interactions with aid-recipient’s population over time. In this paper, the time-invariant characteristics of the U.S. and aid-recipient countries are absorbed by the country fixed effects. The aid-recipient countries’ population over time, on the other hand, could be endogenous and directly related to GDP per capita growth.

⁵ The number of observations is reduced for this robustness check because we do not have information on some countries’ neighbouring countries.

⁶ We instrument non-U.S. aid using the previous year’s total non-U.S. to the rest of the world $\times$ a country’s propensity to receive non-U.S. aid.