Does one good turn deserve another? Evidence from China’s trade and aid policy

ABSTRACT

In this article, we explore the effects of China’s foreign aid on its exports. To do this, we use a sample of 165 countries during the period 2000–2014 and employ a gravity model. We find that the return on Chinese exports for every dollar spent on foreign aid is from $0.156 to $0.4, at the aggregate level. The aid provided in past periods continues to promote China’s exports. We also show, while taking into account aid heterogeneity, that China experiences a higher return in terms of exports when providing development aid intended for infrastructure, to the recipients. Additionally, we find that China’s international aid helps the country to trade more with similar income-level economies. Hence, it can, to a certain extent, foster South-South trade relations.

KEYWORDS:

• Emerging donors
• aid-trade nexus
• gravity model
• China

Acknowledgment

We thank the Editor and three anonymous referees for constructive remarks and suggestions that helped us improve the article. Special thanks to Inmaculada Martinez-Zarzoso, Dimitrios Asteriou, Jarko Fidrmuc, Hendrik Kruse, Pedro Cerqueira, Matthieu Bussière, Scott Hegerty, Kiril Tochkov, Mary-Françoise Renard, Barry Naughton and Lina Song as well as the participants at the “China in the Global Economy” international conference (Orléans 2023) for their helpful comments and remarks on a previous version of the article. Camelia Turcu acknowledges the support obtained through the APR CriseReactGlobal 21063 research grant. Yunzhi Zhang acknowledges the support received through Project GD22XYJ03 financed by the Guangdong Social Science Fund.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/14631377.2024.2349391.

Notes

1. Authors’ calculation based on data taken from World Bank – World Development Indicators.

2. Data is from the Global Chinese Official Finance Dataset.

3. By contrast, conditionality means that aid is provided under specific conditions which have to be met by the recipient country (e.g. good governance and democracy) (Liu et al., 2018).

4. This is because most recipients are based in Africa (Dreher et al., 2017).

5. We also explored, in an extension of this study, whether Chinese imports are influenced by China’s foreign aid. Our results did not indicate a consistent and significant impact of China’s foreign aid on Chinese imports. Hence, we do not insist on this dimension. Results are available upon request.

6. Singapore, Hong Kong China, or Macau China do share the same language as China. However, as they represent an extremely small share of China’s exports, this dummy is omitted. Moreover, as China does not share the same currency with any other trading partner, this variable is ignored in our model as well.

7. In our empirical analysis, due to the structure of our panel, we introduce both country and time fixed effects.

8. However, as we have already explained, due to the structure of our panel we cannot jointly use country and time fixed effects.

9. The country list is shown in Table 7 in Appendix.

10. Correlation Matrix of all the variables is shown in Table 8 in Appendix.

11. The static terms are calculated as follows: ${\beta }_{\mathit{Aid}}=\frac{\mathrm{\partial X}}{\partial \mathit{Ai}{\mathit{d}}_{\mathit{cjt}}}\times \frac{\mathit{Ai}{\mathit{d}}_{\mathit{cjt}}}{\mathit{X}}$, thus, $\frac{\mathrm{\partial }\mathit{X}}{\partial \mathit{Ai}{\mathit{d}}_{\mathit{cjt}}}={\beta }_{\mathit{Aid}}\times \frac{\mathit{Av}\left(\mathit{X}\right)}{\mathit{Av}\left(\mathit{Ai}{\mathit{d}}_{\mathit{cjt}}\right)}=0.014\times \frac{6.54\mathit{e}+09}{2.29\mathit{e}+08}=\mathrm{0.4.}$

12. Table A10 in Appendix presents in detail the results obtained while including MR terms.

13. We compute the OLS with MR terms without fixed effects. All the variables of interest keep the same sign except gravity variables. As underlined by the literature (i.e Lavallee and Lochard, 2019), gravity variables might change signs when MR terms are computed.

14. We take the lags from five (t − 5) to one (t − 1) year: only the results of t − 1 and t − 2 are statistically significant.

15. Table A11 in Appendix presents in detail the results obtained while including MR terms.

16. The results obtained using t − 2 are presented in Table A18 in the appendix.

17. For more information on the aid sectors, see Figure A3 in Appendix A. These details come from the Global Chinese Official Finance Dataset. The sectors are classified similarly to the Chinese Credit Reporting System.

18. The data availability is extremely low for Oceania, hence we can not analyse econometrically the effects for this area.

19. Countries classification per income level is taken from the World Bank.

20. In the dataset, there are certain flows of aid oriented towards high-income countries. They can take the form of grants, loans, and scholarships to students. The aid provided by China to high-income countries includes also the opening and operation costs of the Confucius Institutes

21. Given the notable increase in aid in 2009 and 2010, as shown in Figure A1, we also examine the two distinct subsamples: 2000–2008 and 2009–2014. We obtain similar results that are available upon request.

22. Most of the zero aid flows concern the developed countries, for example, Australia, Austria, Belgium, Canada, and Switzerland.

Additional information

Funding

The work was supported by the Conseil Régional du Centre-Val de Loire [APR IA CriseReactGlobal - 2021 00149486]; Conseil Régional du Centre-Val de Loire [APR Mutmond]; Guangdong Social Science Fund [GD22XYJ03].