Anna Whittaker and Dr. Dan Nielson, Political Science
Although bilateral and multilateral aid donors have given away hundreds of billions of dollars in foreign aid, it is still unclear whether providing loans and launching major development projects in poor countries can really cause economic growth, helping these countries to provide for their own needs. Technical assistance is a unique type of foreign aid: instead of giving low income countries money to spend on development projects, donors provide their own experts to consult with governments, plan future development projects, or supervise the completion of current projects. Instead of providing physical resources for projects, donors give of their knowledge and organizational expertise to build up struggling governments or assist with technical projects when recipient countries lack training and skills.
On the surface, technical assistance seems like an attractive alternative to traditional foreign aid. Instead of just giving countries money or spending money on projects that might not cause real economic growth, donors can share knowledge and expertise in order to help developing countries create their own development solutions. However, some argue that in practice, technical assistance does just the opposite–undermining local capacity and leadership. In this research project, I attempted to test whether or not technical assistance can improve the quality of governance in the developing world by decreasing corruption.
Some scholars, including Collier , agree that technical assistance can help to turn a country around for the better in the right situation. In some cases leaders are open to liberal ideas and willing to make reforms. However, these leaders lack the technical skills and training to implement reforms that would make government operations more transparent and democratic. In situations like this, technical assistance could be given to create more transparent government practices, or to train officials. Once officials receive this training, they are in a better position to make economic and political reforms to help the economy to grow.
However, other scholars argue that technical assistance can actually harm a country’s prospects for effectively governing itself and rising out of poverty. According to this argument, donors are rarely able to make lasting changes in government by sending in foreign experts and resources. Leaders might not be open to change, or corruption might be so entrenched that technical assistance will not sustain reforms. If the government is very unstable, training current leaders does not lead to lasting improvements, since power changes hands so frequently. In other cases, foreign aid workers simply come in, oversee projects themselves, and do not effectively train local leaders to take over after they leave. Aid agencies often take the most talented people away from jobs in the government in order to help oversee aid projects. Thus, local capacity can actually be undermined if the most educated and talented people start working for foreign aid agencies instead of working directly through the national government to make reforms. Instead of becoming motivated to solve their own problems, governments begin to turn to aid agencies to help them with organization and leadership.
In my paper, I developed measures to test whether or not technical assistance aid can effectively decrease corruption in developing countries. I wanted to see which of these competing theories about technical assistance actually holds up in the real world—can technical assistance provide governments with the skills and tools they need to begin to operate more transparently, or does it tend to undermine local leadership and prolong poor governance and corruption?
To test these theories, I gathered data on technical assistance aid commitments from bilateral donors, collected by the OECD. The most comprehensive measure I could find for corruption was the ICRG rating system, which used cross-country survey data to rank each country’s level of corruption from 0 (high corruption) to 6 (low corruption).
I set up a complex regression analysis to try to isolate the causal effect of technical assistance on corruption. There are many factors besides technical assistance which could potentially cause a country’s level of corruption to change over time, such as a country’s level of economic development and its colonial heritage. I included nine main control variables from an influential work on the causes of corruption in order to be able to isolate the effect of technical assistance. Only after I found how much these other factors were influencing corruption could I really tell how big of an impact technical assistance has on its own.
I also had to be careful of reverse-causation: even if I found that countries which receive the most technical assistance have the lowest amount of corruption, this does not imply that technical assistance actually lowered corruption. In fact, it could be exactly the opposite. It is possible that donors choose to give technical assistance to developing countries with the least amount of corruption because they feel their money is better spent. In order to make sure that my results did not simply reflect this type of endogeneity, I developed an instrumental variable. I predicted that donor countries with the highest number of consultants (the largest service sectors) would be pressured by their citizens to give the most technical assistance aid. Countries receiving the most aid from these bilateral donors would get higher levels of technical assistance aid than other countries, regardless of their level of corruption. By using this instrument to indirectly estimate the amount of technical assistance received, I could test whether or not technical assistance actually caused differences in corruption across countries and over time.
I ran two different regressions, a two-stage least squares regression with controls for corruption, as well as a fixed effects two-stage least squares model (which controlled for all fixed variables within each country that would affect its level of corruption). In both cases, I found that the effect of technical assistance aid on corruption was very small and statistically insignificant. There are many improvements that could be made to this initial model in order to more fully test the impact of technical assistance aid. However, these initial results do suggest that technical assistance has been ineffective at controlling corruption and improving governance in the developing world. Donors should look for ways to either improve technical assistance programs, or to spend their money more effectively on other development projects.
References
- Collier, Paul. 2007. The Bottom Billion: Why the Poorest Countries are Failing and What Can Be Done About It. Oxford University Press.
- Knack, Stephen. 2001. Aid dependence and the quality of governance: Cross-country empirical tests. Southern Economic Journal 68 (Oct): 310-329.
- Treisman, Daniel. 2000. The causes of corruption: A cross-national study. Journal of Public Economics 76 (3): 399-457.