9 September 2006 (Sat), GRIPS Campus in Tokyo 14:00-17:00

"Measuring Utilization Efficiency of Foreign Aid in Asian Countries:
Data Envelopment Analysis (DEA) Approach"
by Dr. Seungkee Baek and Prof. Kaoru Tone (GRIPS)

Presented by
Dr. Seungkee Baek
(GRIPS Visiting Research Fellow)


In this workshop, VDF-Tokyo welcomed Dr. Seungkee Baek, a Visiting Fellow of GRIPS, to present his co-authored paper with Prof. Kaoru Tone of GRIPS. This research focuses on measuring utilization efficiency of foreign aid by Data Envelopment Analysis (DEA) approach, and then analyzing the results with some Asian countries during 1999-2001.

Dr. Baek began his presentation by talking about their motivation of doing the research. It was shown that foreign aid in Korea in 1960s and 1970s and currently in other Asian developing countries played a critical role as a source of seed money for economic growth and development. According to him, however, it still remained a question that economic positions of the receiving countries were really improved by such foreign aid flows, and therefore their research would focus on the role of foreign aid and relevant utilization policy by using DEA as a tool to evaluate. In this research, foreign aid was considered with two sources, i.e. loan and grant. The initial assumption of this research was that utilization policies for foreign aid contributed to economic performance of the receiving countries, though there might be causality between them. The question of the research, therefore, was how much such policies could contribute to economic performance.

Going further with theoretical background, Dr. Baek gave some definitions of such important indicators and variables of the research as efficiency, weighted efficiency, input, and output in DEA approach. In this research, the authors defined input as the amount of foreign aid, and output as economic performance, and thereby efficiency of decision-making unit (DMU) would be identified by the ratio between weighted output and weighted input. In addition to measuring efficiency, Dr. Baek also elucidated inefficiency problem by using concept of slacks, which were defined as shortfall in output or excess in input. With such definition, DMU could be considered as efficient if there would be no output shortfall or input excess. Dr. Baek explained more about this definition by using slack-based measurement model, which was proposed by Prof. Tone in 2001, and he also pointed out the authors’ contribution through introducing additional information of resource mix to decompose the slacks. Some simple numerical examples were provided to illustrate this theoretical framework.
[For more information about the model, please read the presentation slides attached to this summary].

To show how to measure the impacts of foreign aid utilization policies on economic performance, Dr. Baek used available data of 13 South and Southeast Asian countries during 1999-2001 as the receiving countries of Japan. Loan and grant were used separately as input of the model, while economic performance was measured by values of exports and gross domestic product (GDP). Moreover, additional information of development projects such as social infrastructure, economic infrastructure, and production facility, or on sector such as material or service was also provided in the model to get more detailed information on efficiency. The estimated results indicated that slacks and efficiency scores were much different for each country before and after applying such addition information. The results were also discussed with some policy implications for each country in utilizing their foreign aid sources.

Mr. Vu Tuan Khai (Yokohama National University-YNU & VDF-Tokyo) started the discussion section by a comment that multiplier effects might not be taken into DEA model as both efficiency and weighted efficiency must not be greater than 1. According to him, such effects were crucial in the model with macroeconomic variables. Dr. Baek said that in DEA model, input and output were assigned with different weights, and it assumed that weighted output could not be greater than weighted input.

Being concerned with units that were used in the DEA models, Prof. Kenichi Ohno said that scale problems could happen and make different estimates if variables were measured or changed with different units. Though Dr. Baek responded that DEA could measure both monetary and non-monetary units, Prof. Kenichi Ohno asked him to make more detailed explanation of such possibility.

In addition to the above comment, Prof. Kenichi Ohno (GRIPS & VDF) also said that choosing exports and GDP as outputs for the models might not be appropriate to measure foreign aid utilization efficiency. According to him, donor countries would care about other macro indicators such as poverty reduction, health care system, and human rights rather than exports. Sharing this opinion, Prof. Izumi Ohno (GRIPS) added that, besides exports and GDP, there would be many other things that could be considered in the model, and she wondered why the authors used only such aggregate indicators to measure efficiency. In his response, Dr. Baek said that the most important reason that they must use macro data was that data was not available at lower levels. Given that response, Prof. Izumi Ohno suggested that the authors focus on specific sectors in the sample countries, or just one country because, according to her, different countries would have different governance capacity for foreign aid projects and country-specific characteristics, and thus they might produce different results with the same amount of loan and grant.

Going further with assumptions of the model, Prof. Izumi asked Dr. Baek how to treat loan and grant in the model because they could provide different incentives for receiving countries to use for their economies. Dr. Baek agreed with Prof. Izumi about her comments, but he said that loan and grant were treated similarly in the model, and they might be studied more with particular aspects.

Also being interested in the models of the research, Mr. Giang Thanh Long (GRIPS & VDF-Tokyo) said that DEA model did not require any functional form, and it created the best-practice DMU among all, and then compared other DMUs with the best one in order to get relative efficiency levels. Therefore, according to him, more data could provide more precise estimates of efficiency, and he wondered about the estimates in the research with only three years. Dr. Baek agreed with Mr. Long, and said that limitations of data would surely prevent the estimated results of the research, and he wished to get more recent data for the research.

Given limitation of DEA approach, Prof. Kenichi Ohno supplemented that modeling of slacks might not also be appropriate with initial conditions of different countries such as economic growth and policy strategies. He also said that in the receiving countries, private sector development might be more important than government for donors to consider their loan and grant. DEA could explain in narrow scope, but could not interpret precisely with broader scope. Prof. Izumi Ohno shared some information about different countries with various strategies for attracting and using foreign aid to support this comment. Moreover, Prof. Kenichi Ohno wondered how to interpret the estimates of slacks with 100 percent because various situations could happen such as lagged effects of loan and grant, and more importantly, measuring impacts of foreign aid on economic performance might be difficult, even impossible, as we could not know their causality, and there was no counterfactual factor for such study. Mr. Nguyen Duc Thanh (GRIPS & VDF-Tokyo) agreed with Prof. Ohno, and he additionally said that economic system was crucial factor because various constraints of specific systems might result in different social and economic performances. Responding to these comments and questions, Dr. Baek again said that the authors really wanted to have further research with numerous indicators as suggested, but it was impossible due to data availability.

In addition to comments on technical issues of the model, Mr. Pham Truong Hoang (YNU & VDF-Tokyo) asked a question about how to know the ways that Japan decided to give loan and grant to the studied countries, and whether the estimated results were still relevant for current policies of Japan and receiving countries as the data was during 1999-2001. In his answer, Dr. Baek said that the main objective of the research was to introduce DEA approach as a tool to evaluate policy, and the estimated results would illustrate this approach. However, he admitted that many factors needed to be considered in order to have more appropriate interpretation of foreign aid policies.

At the end of the presentation, Dr. Baek thanked all participants, and he hoped to have more exchange information on the research issues.           

We had another 45 minutes for exchange information section. Current studies of the participants, information of research in both VDF-Hanoi and VDF-Tokyo were informed among us. 

Paper (Not available at this moment) | Slides (PDF61KB)                  

(By Giang Thanh Long)

  WS Schedule | VDF-Tokyo HOME