Download Budget support: Conditional Results PDF

TitleBudget support: Conditional Results
LanguageEnglish
File Size2.4 MB
Total Pages246
Document Text Contents
Page 1

IOB Evaluation | no. 369 | Budget support: Conditional Results | IOB Evaluation | no. 369 | Budget support: Conditional Results | IOB Evaluation | no. 369 | Budget support: IOB Evaluation | no. 369 | Budget support: Conditional Results | IOB Evaluation | no. 369 | Budget support: Conditional Results | IOB Evaluation | no. 369 | Budget

IOB Evaluation

Review of an instrument (2000-2011)

Budget support:
Conditional Results

Page 2

Review of an instrument
(2000-2011)

Budget support:
Conditional Results

IOB Evaluation

Page 123

Direct impact of budget support funds

| 122 |

countries have noticed that the transaction costs during implementation have decreased
because the standard government procedures can be used (IDD and Associates 2006: S5).
Donors observe that they could not have increased aid to the same extent without budget
support. This was true, for example, in Rwanda (IDD and Associates 2006) and for DfID
in many countries (National Audit Office 2008). The Ghana evaluation concluded that the
transaction costs for budget support were lower than for ‘other modalities’ (Lawson et al.
2007). The World Bank only needed half of the transaction costs per dollar of aid for the
PRSC than it needed for projects in Uganda (Miovic 2004, quoted in IDD and Associates:
52). In a 2005 study on PRSCs, the World Bank concludes that for each dollar of preparation
costs, US$142 could be spent on budget support, as opposed to US$32 for other modalities
(in NAO 2008). An estimate of four types of transaction costs in Nicaragua does suggest
that they are usually higher than for an average project, but per dollar of delivered aid they
were certainly lower between 2005 and 2009, both for donors and recipients (Dijkstra and
Grigsby 2010). For the World Bank, participation in joint donor systems for budget support
means that the transaction costs will increase compared to a situation in which the World
Bank acts on its own (IEG 2010). But for the recipient country the opposite is probably true,
at least if there are more donors providing budget support.

Bigsten et al. (2011) use data collected by Sida for the year 2010 about various kinds
of administration costs: a) developing policy documents, b) implementing policy
for development cooperation, c) collaboration and consulting, both national and
international, and d) management, guidance and support. These costs are kept track of
separately for various aid modalities. Table 4.19 has been compiled based on these data.

Table 4.19: Transaction costs per aid modality in Sweden

Administration costs per
Swedish krona spent

Administration costs in
relation to project aid (%)

Programme­based aid 3.8 33

Budget support 2.2 19

Technical assistance 5.2 46

Project aid 11.3 100

Total 6.5 57

Source: Bigsten et al. 2011, p. 79

The average administration cost per Swedish krona spent was 6.5%, but the costs for project
aid were much higher than for programme-based aid modalities (column 1). The second
column shows that the transaction costs for budget support were only 19% of those for
project aid. These data are only available for Sweden, but there is no reason to assume that
this would be much different with other donors.52

52 Easterly and Pfutze (2008) have made an estimate of the administration costs in relation to the total aid
budget per donor. For bilateral donors they arrive at an average of 7%, and the average for multilateral
donors is 12%. 6.5% for Sweden is close to the average of 7%.

Page 124

Budget support: Conditional Results

| 123 |

4.6.2 Transaction costs in six countries

Whether transaction costs decreased after the introduction of budget support has not
been analysed for all six countries. The conclusion for Ghana is that transaction costs for
the government have decreased, but not for donors. The work of donors has changed, not
decreased. But it seems that this conclusion does not take the increasing volume of aid into
account. The transaction costs for every euro spent have presumably decreased for donors
as well. It has been noted that transaction costs are still high in Tanzania, partly due to the
emergence of new donors there, such as China, as well as the vertical funds. The conclusion
for Nicaragua and Vietnam is that transaction costs for budget support are lower than for
project aid, both for the governments and donors.

4.7 Conclusions

Over the long term, programme aid had not increased since 2000. Between 1986 and 1993,
during the glory days of structural adjustment programmes, programme aid comprised
about 10% of total (bilateral) aid, while between 2000 and 2010 it was between 3% and
4%. This percentage was much higher in some countries, however. In our six case study
countries, the share of programme aid in total aid varied between 10% (Vietnam) and
42% (Tanzania).

The predictability of budget support varied strongly from country to country and from
period to period, but it was not always better than with project aid. A great deal of budget
support, moreover, was disbursed at the end of the year, which often resulted in expensive
domestic borrowing. In African countries, including our four African case studies,
predictability did improve over time. A study by Celasun and Walliser (2006) shows that
the greater the differences are between commitments and actual disbursements, the fewer
budget support funds are used for additional expenditures: in such cases, it is much more
likely that domestic debt increases or reduces. When there are fewer actual disbursements,
government investment decreases as well, but the opposite is not true.

In at least three of the six case study countries, part of the budget support was used to
increase domestic reserves or pay off domestic debt (Ghana, Mali and Nicaragua). It appears
that almost all of Nicaragua’s budget support was used for this, partly because all funds were
disbursed in the last quarter. In four other countries, government expenditure increased
as a result of budget support. In Vietnam, budget support was too small compared to the
government budget to determine whether it had an impact.

Our own cross-section analyses of budget support’s impact on the internal balance show
that every additional 1 percentage point of budget support (in % of GDP) leads to an increase
in government expenditure of about 0.6 percentage points of GDP. There was no statistically
significant impact on tax revenue. We can conclude with a fairly high degree of certainty
that budget support did not cause tax revenue to decrease. This conclusion corresponds
with other literature in this area. We can also conclude that the lion’s share of budget

Page 245

| 244 |

Published by:

Ministry of Foreign Affairs
Policy and Operations Evaluation Department (IOB)
P.O. box 20061 | 2500 EB The Hague | The Netherlands
www.rijksoverheid.nl/BZ-evaluaties

© Ministry of Foreign Affairs | September 2012 | ISBN: 978-90-5328-426-1

Authors:
Geske Dijkstra, Antonie de Kemp, Denise Bergkamp

Translation:
Mark Speer (Contactivity)

Photography:
Cover: Budget support: Pu�ing all one’s eggs in one basket or walking on eggshells?
(Hollandse Hoogte)
Chapter 1: Pollling station Tanzania (2010) (Hollandse Hoogte)
Chapter 2: Health in Mali (2012) (Hollandse Hoogte)
Chapter 3: Basic Education in Zambia (2011) (Antonie de Kemp)
Chapter 4: Vietnam (2012) (Hollandse Hoogte)
Chapter 5: Elections in Ghana (2012) (Hollandse Hoogte)
Chapter 6: Market in Nicaragua (2010) (Hollandse Hoogte)
Chapter 7: Mozambique (2009) (Hollandse Hoogte)

Page 246

IOB Evaluation | no. 369 | Budget support: Conditional Results | IOB Evaluation | no. 369 | Budget support: Conditional Results | IOB Evaluation | no. 369 | Budget support: IOB Evaluation | no. 369 | Budget support: Conditional Results | IOB Evaluation | no. 369 | Budget support: Conditional Results | IOB Evaluation | no. 369 | Budget

Published by:

Ministry of Foreign Affairs
Policy and Operations Evaluation Department (IOB)
P.O. box 20061 | 2500 EB The Hague | The Netherlands
www.rijksoverheid.nl/BZ-evaluaties
© Ministry of Foreign Affairs | September 2012

12BUZ612266 | E



In the past decade, donors, including the
Netherlands, have increasingly supported the
budgets of governments in developing countries
directly. It was felt that this general budget
support would make aid more efficient and
effective, and generate more sustainable results.
It would also give donors the opportunity to
encourage reform through the accompanying
policy dialogue.

IOB has concluded that the instrument
contributed to economic growth and the
extension of public services. Public finance
management and democratic control improved
as well. Budget support is not suitable for
pursuing major reform, however, unless the
recipient government takes ownership of it.

Similer Documents