Impacts of Foreign Aid to Melanesia

Abstract

This paper investigates aid effectiveness in Melanesia, a region consisting of Fiji, Papua New Guinea, the Solomon Islands, Vanuatu and New Caledonia. These countries are of great interest since they have not performed well despite being rich in resources and receiving large amounts of foreign aid. The paper examines the impact of foreign aid on agricultural growth and overall economic growth in Melanesia. The impact on agricultural growth is important since the majority of people in Melanesia live in rural areas, reliant on agriculture for their livelihoods. Using the econometric analysis of data for the period 1980 to 2001, results provide no evidence that foreign aid has impacted on the agricultural sector. However, the paper does find evidence that foreign aid has impacted favourably on economic growth. A number of explanations and policy recommendations are provided.

Keywords:

• Foreign aid
• agriculture
• economic growth
• rural sector
• Melanesia

JEL-Classifications:

• F35
• O11
• O13
• O40

Acknowledgements

This paper is partly based on fieldwork conducted in Melanesian countries during March and April 2004. The author is grateful for funding from the Australian Research Council, World Vision of Australia and the Australian Agency for International Development (AusAID). The views expressed in this paper are those of the author and not necessarily those of the Commonwealth of Australia. The Commonwealth of Australia accepts no responsibility for any loss, damage, or injury resulting from reliance on any of the information or views contained in this paper. The author is also grateful to AusAID for financial support, to James Gilling, Tim Fry, Sinclair Davidson, Nikunj Soni, Ron Duncan and two anonymous referees for helpful comments and advice and to feedback obtained from participants of an RMIT University seminar, 8 October 2004 and an AusAID seminar, Canberra, 29 November 2004.

Notes

^a 2000.

^b 1999.

1. External remittances, or money sent back from overseas, are large for Fiji but are not significant for other Melanesian countries. However, internal remittances are believed to be important for all Melanesian recipients. Unfortunately accurate data on such remittances do not exist.

2. A relevant exception is Norton et al. (1992) who examine the impact of foreign assistance on agricultural growth. Using data for 98 developing countries covering the period 1970 to 1985, the study concludes that foreign aid has improved agricultural productivity in Asia and Sub-Saharan Africa but not in the Middle East or in Latin America.

3. Roodman (2004) applies a battery of diagnostic tests to the specifications of a number of these studies, testing the strength of each. The results from this extensive testing lends most support to the Dalgaard et al. (2004) study, which finds that, on average, aid works but not in countries located in the tropics. Weakest support is found for the Burnside and Dollar finding that aid effectiveness is contingent upon the policy environment. Moreover, Easterly (2003) shows that the Burnside & Dollar result is not robust to changes in other plausible definitions of aid, policies and growth and Easterly et al. (2003) find that Burnside–Dollar is not robust when they add additional years and countries to their sample. The debate over the importance of policy for aid effectiveness continues. There are also questions regarding what good polices are, how they should be defined and when policies become ‘good’.

4. Short-impact aid relates to aid flows that can be expected to increase GDP per capita within approximately four years. This time period was chosen since most previous studies using cross-country data use observations averaged over a four or five year period. Clemens et al. (2004) argue that such aid includes budget support and project aid for infrastructure or to support transportation, communications, energy, banking, agriculture and industry. Long-impact aid relates to aid flows that might be expected to increase GDP per capita but which is unlikely to do so within four years of its disbursement. It is argued that such aid includes technical cooperation, social sector investments in health, education, population control and water. Clemens et al. (2004) also categorise some flows as ‘humanitarian’, which relate to emergency assistance and food aid. Humanitarian aid is not expected to impact on growth.

5. Feeny (2005) also provides an alternative explanation; that structural adjustment might be more effective at spurring growth when supported by foreign aid.

6. To obtain the data for the dependent variables it was necessary to use a variety of sources. These sources include the Asian Development Bank's (ADB) Key Indicators of Developing Asian and Pacific Countries, the World Bank's World Development Indicators, the National Statistics Office of Vanuatu, the New Caledonia Institute of Statistics and Economic Studies (ITSEE), the Central Bank of the Solomon Islands, and the United Nations National Accounts Main Aggregates database.

7. The budget balance includes current and capital revenue and grants received, less total expenditure and lending minus repayments from the World Bank (2004). It is recognised that a budget balance variable without the inclusion of foreign grants is preferred since foreign aid is also included as an explanatory variable in the regression models.

8. Real (Effective) Exchange Rate is defined as the nominal effective exchange rate (a measure of the value of a currency against a weighted average of several foreign currencies) divided by a price deflator or index of costs. Data come from the World Bank's World Development Indicators for Fiji, Papua New Guinea and the Solomon Islands. Authors own calculations for New Caledonia and Vanuatu.

9. The initial level of GDP per capita is often included as an explanatory variable in empirical models of growth to capture convergence. However, given the similarity of the five countries under consideration a convergence term would not be meaningful in the current study. The number of hectares of arable land per capita and the number of tractors per hectare were also initially included as explanatory variables. These data are available from the UN's Food and Agricultural Organisation (FAO). However, these variables are only estimated every five years and exhibit very little variation over time. Since the coefficients attached to these variables were not found to be statistically significant across a number of specifications, they were subsequently excluded from the analysis.

10. It is also recognised that including both the aid variable and the real exchange rate might also be problematic given a potential relationship between the two. However, results change little when either of the variables are excluded from the regression equations.

11. Using the investment data that are available, results from further regression analysis finds fairly strong evidence that foreign aid is an important determinant of investment in Melanesia. Note that this finding in itself is very encouraging. Investment is usually found to be a robust determinant of economic growth in empirical studies thereby indicating that aid is likely to have a positive impact to economic growth in Melanesia. However, it should also be noted that investment is widely cited as being unproductive in most of the Melanesian recipients under consideration.

12. Results from these tests are available from the author on request.

13. Tavares (2003) and Collier et al. (2004) have used the total aid budget of donors when instrumenting for foreign aid in studies examining the impact of aid on corruption and capital flight. As Tavares (2003) argues, these budgets are good instruments since, when OECD countries increase their total aid budgets, countries that are culturally and geographically closer to donors experience an exogenous increase in aid.

14. Adopting an instrumental variable approach without the presence of endogeneity leads to inefficient standard errors and misleading results.

15. For example, indices of political and civil rights were obtained for Fiji, Papua New Guinea, the Solomon Islands and Vanuatu from Freedom House. Data are not available for New Caledonia. However, coefficients on these variables were never statistically significant when applied to the four countries for which data are available, due to relatively little variation across both time and country.

16. Results from these F-tests for the joint statistical significance of the individual country effects are available from the author on request.

17. STABEX is a stabilisation scheme offered by the EU to African Caribbean and Pacific (ACP) States. The scheme began in the late 1970s and entails financial transfers to the governments of ACP countries to compensate them for large falls in their export earnings. The basic principle is that STABEX transfers replace the amounts that would have been paid to producers if normal market conditions had prevailed. Donors recognised the difficulties encountered by agricultural sectors of these countries due to falls in their export earnings, whether these were due to falling world prices, natural disasters or a combination of the two. Fluctuations in export earnings disrupt investment planning, public finances and lead to deteriorations in the balance of payments situation. Recently STABEX has been replaced by FLEX.

18. For the sake of comparison, exactly the same explanatory variables were used in the model to investigate the impact of aid on economic growth and in the model explaining the impact of aid on the rural sector. However, if total population growth is used instead of rural population growth, results remain unchanged. Increases in population growth are associated with increases in economic growth.

19. This result is encouraging given that technical assistance has been such a major component of the recent Enhanced Cooperation Program (ECP) to Papua New Guinea and the Regional Assistance Mission to the Solomon Islands (RAMSI).