Abstract
Danish bilateral development assistance is aimed at reducing poverty in the partner countries. Even so, bilateral assistance may have secondary, or knock-on, effects, which are beneficial for Denmark. An important secondary effect is the prospect of increased export from Denmark to the partner countries. This Evaluation Study presents an econometric analysis of Danish exports to 144 countries over the period from 1981 to 2010. The analysis is based on the gravity model of bilateral trade; a structural model developed over decades and now the central model in analyses of bilateral trade flows and trade policies. The main result of the study is that Danish bilateral aid has a positive and statistically significant impact on Danish exports to the recipient countries.
Bilateral development assistance may affect exports through several channels. Three of the main channels are direct aid tying; increasing recipient income where higher income leads to higher imports, and decreased trade costs, say due to improved information about cultural and administrative customs and practices. Thus, as for preferential trade arrangements, bilateral aid has two potential economic effects; trade creation working through income and benefiting all countries, and trade diversion by increasing donor exports at the expense of other exporters. Both effects have been taken into account in recent econometric studies.
The Evaluation Study briefly summarizes the results of nine other econometric studies of bilateral aid and exports. Five of the studies estimate the impact of aid on bilateral trade for groups of donors, while the remaining four studies estimate the impact for a single donor county. The studies of groups of donors all include Denmark in the sample. The unanimity in the results regarding the links between
bilateral assistance and donor exports is noteworthy. All nine studies find that bilateral aid has had a statistically significantly positive impact on exports to the recipient countries. The impact is not statistically significant for all periods in all models, as is also the case for Denmark, but the overall result is surprisingly well founded in the data spanning the past 30-50 years.
The parameter of interest in the econometric models estimates the percentage increase in exports following a percentage increase in aid to the specific country (the export/aid elasticity), or following a percentage point increase in aid relative to disposable GNI in the specific country (the export/aid semielasticity). This indicates that the estimated return in terms of dollars of increased exports per dollar of additional aid varies over time and across countries. The overall estimated average export return is about 30 cent per aid dollar when it is estimated using model formulations and methods that are directly comparable to the existing studies. The overall average masks a difference in returns of 5 cent per aid dollar for the Danish Priority Countries and 1.5 dollar per aid dollar for the group of countries in the Neighbourhood Programme. These results reflect that the impact of additional aid on exports will be higher if export is already high compared to the aid to a given country (aid/export ratio). Using an alternative specification where the percentage increase in exports following a one percentage point increase in the aid-to-GNI ratio is estimated, gives somewhat lower results – 8 cent increase in export pr. 1 dollar increase in aid to priority countries and 24 cent with regards to the Neighbourhood Programme countries.
The study also suggets that the impact of aid on exports was much higher in the 1980s compared to in particular the most recent decade. This may be related to factors such as untying of aid in the same period.
The econometric analysis has two important limitations. First of all, the model can only give information about marginal changes in aid. As a decision to give bilateral aid to a new country is not a marginal decision, the impact should not be evaluated based on the results presented in the study. Likewise, the impact on exports following a decision to stop all aid to a country cannot be deduced
from the regression results. Thus, the results of this Evaluation Study are not useful as tools for country selection in the aid allocation decision. Second, although the export return is positive it is difficult to give a precise estimate of the size, both for individual countries and for country groups. The authors of this Evaluation Study clearly prefer a model formulation that indicates low overall average export returns (8 cent per dollar) and low returns for aid to the Neighbourhood Programme countries (24 cent per dollar) but the other model formulation cannot be excluded. Thus, model uncertainty leads to quite
substantial uncertainty in the country specific export return.
Bilateral development assistance may affect exports through several channels. Three of the main channels are direct aid tying; increasing recipient income where higher income leads to higher imports, and decreased trade costs, say due to improved information about cultural and administrative customs and practices. Thus, as for preferential trade arrangements, bilateral aid has two potential economic effects; trade creation working through income and benefiting all countries, and trade diversion by increasing donor exports at the expense of other exporters. Both effects have been taken into account in recent econometric studies.
The Evaluation Study briefly summarizes the results of nine other econometric studies of bilateral aid and exports. Five of the studies estimate the impact of aid on bilateral trade for groups of donors, while the remaining four studies estimate the impact for a single donor county. The studies of groups of donors all include Denmark in the sample. The unanimity in the results regarding the links between
bilateral assistance and donor exports is noteworthy. All nine studies find that bilateral aid has had a statistically significantly positive impact on exports to the recipient countries. The impact is not statistically significant for all periods in all models, as is also the case for Denmark, but the overall result is surprisingly well founded in the data spanning the past 30-50 years.
The parameter of interest in the econometric models estimates the percentage increase in exports following a percentage increase in aid to the specific country (the export/aid elasticity), or following a percentage point increase in aid relative to disposable GNI in the specific country (the export/aid semielasticity). This indicates that the estimated return in terms of dollars of increased exports per dollar of additional aid varies over time and across countries. The overall estimated average export return is about 30 cent per aid dollar when it is estimated using model formulations and methods that are directly comparable to the existing studies. The overall average masks a difference in returns of 5 cent per aid dollar for the Danish Priority Countries and 1.5 dollar per aid dollar for the group of countries in the Neighbourhood Programme. These results reflect that the impact of additional aid on exports will be higher if export is already high compared to the aid to a given country (aid/export ratio). Using an alternative specification where the percentage increase in exports following a one percentage point increase in the aid-to-GNI ratio is estimated, gives somewhat lower results – 8 cent increase in export pr. 1 dollar increase in aid to priority countries and 24 cent with regards to the Neighbourhood Programme countries.
The study also suggets that the impact of aid on exports was much higher in the 1980s compared to in particular the most recent decade. This may be related to factors such as untying of aid in the same period.
The econometric analysis has two important limitations. First of all, the model can only give information about marginal changes in aid. As a decision to give bilateral aid to a new country is not a marginal decision, the impact should not be evaluated based on the results presented in the study. Likewise, the impact on exports following a decision to stop all aid to a country cannot be deduced
from the regression results. Thus, the results of this Evaluation Study are not useful as tools for country selection in the aid allocation decision. Second, although the export return is positive it is difficult to give a precise estimate of the size, both for individual countries and for country groups. The authors of this Evaluation Study clearly prefer a model formulation that indicates low overall average export returns (8 cent per dollar) and low returns for aid to the Neighbourhood Programme countries (24 cent per dollar) but the other model formulation cannot be excluded. Thus, model uncertainty leads to quite
substantial uncertainty in the country specific export return.
| Translated title of the contribution | Dansk Eksport og Dansk Bilateral Bistand |
|---|---|
| Original language | English |
| Place of Publication | Copenhagen |
|---|---|
| Publisher | Ministry of Foreign Affairs of Denmark. Danida |
| Volume | 2014 |
| Edition | 2 |
| Number of pages | 46 |
| ISBN (Print) | 978-87-7087-841-8 |
| ISBN (Electronic) | 978-87-7087-840-1 |
| Commissioning body | Danida, Ministry of Foreign Affairs of Denmark |
| Publication status | Published - Jun 2014 |
| Series | Evaluation Study |
|---|---|
| Number | 2 |
| Volume | 2014 |