Constituency Development Funds Scoping Paper

Constituency Development Funds:
Scoping Paper
18 January 2010
Contact details:
Alison Hickey Tshangana
PDG
PO Box 53123, Kenilworth, 7745
70 Rosmead Avenue, Kenilworth, Cape Town
Tel: (021) 671 1402 Fax: (021) 671 1409
Cell: 083 280 2759
Email: [email protected]
Executive Summary
Constituency Development Fund (CDF) schemes are decentralisation initiatives which

send funds from the central government to each constituency for expenditure on
development projects intended to address particular local needs. A key feature of CDF
schemes is that Members of Parliament (MP) typically exert a tremendous degree of
control over how funds are spent. This scoping paper provides an overview of CDFs:
when, where and how they have emerged worldwide; identifies the key features of CDFs
which impact on their performance; outlines the arguments and evidence available for
and against CDFs, and; investigates the opportunities and possible future research from
an advocacy perspective.
Despite the variety of forms and approaches in different countries, the following essential
elements are identified which help to distinguish CDFs from other decentralization
initiatives or community-based development programmes. First, funds are raised by
national government and disbursed at local level. Second, funds are allocated per
constituency and MPs have some degree of control over the spending. Lastly, funds are
intended for development projects which reflect localized needs and preferences.
Other critical features of CDFs which vary from country to country will have a direct
bearing on the performance of the programme, including: quantum of funding available
per constituency; method for allocating funds between constituencies (degree of
targeting); and the existing legislative framework.
From a transparency and accountability perspective, the separation of powers is most

fundamental criticism of CDFs. The paper considers other issues which have emerged in
the implementation of CDF schemes worldwide. These issues include:
 Impact on MP-constituent relations;
 Public participation
 Selection of non-priority projects due to MP influence
 Development expenditure in the absence of coherent, long-term planning
 Intergovernmental coordination to avoid project duplication
 Targeting of allocations to poor areas
 Corruption and mismanagement of funds
 Need for greater professionalisation and improved project management
 Additionality and sustainability
 Administrative costs and impact on local authorities
 Transparency and public access to information
Civil society groups have formulated a number of proposed reforms to address these
implementation issues including proposals to improve the representivity and capacity of
the CDC Committee and to improve and strengthen citizen participation and
accountability.
In summary, the paper argues that emerging examples of good practices and the mere
variety of CDF schemes worldwide would suggest that it is possible to build safeguards
into the system to ensure transparency, promote public participation and substantially
-i-
curb corruption. From a civil society perspective concerned with poverty alleviation,
transparency and accountability, the critical features of a more effective CDF would be:
1. The existence of a legislative and regulatory framework which:

 Promotes and protects public participation processes in project
prioritization and identification
 Provides for a body—apart from Parliament—which is empowered to
manage and conduct oversight
 Establishes funding flows and disbursement procedures which remain in
control of government officials, as opposed to MPs
 Ensures representivity of CDF committee at local level
 Requires CDF project and budget information to be publicly available
2. Allocation formula which prioritizes poorer constituencies
-ii-
Contents
1. Defining and understanding CDFs ............................................................... 1

1.1. Introduction ..................................................................................................1
1.2. Methodology and sources ............................................................................ 2
1.3. Typology of CDFs ........................................................................................3
2. Advantages and disadvantages in practice ................................................. 5

2.1. Impact on MP-constituent relations .............................................................. 5
2.2. Public participation .......................................................................................6
2.3. Selection of non-priority projects due to MP influence .................................. 8
2.4. Development expenditure in the absence of coherent, long-term planning... 9
2.5. Intergovernmental coordination to avoid project duplication ....................... 10
2.6. Targeting of allocations to poor areas ........................................................ 10
2.7. Corruption and mismanagement of funds................................................... 11
2.8. Need for greater professionalisation and improved project management ... 12
2.9. Additionality and sustainability ................................................................... 13
2.10. Administrative costs and impact on local authorities .................................. 14
2.11. Transparency and public access to information ......................................... 15
3. Proposed reforms and alternatives to CDFs.............................................. 17

3.1. Views of donor organisations ..................................................................... 17
3.2. Reforms proposed by civil society organisations ........................................ 18
3.3. Directions for future research and advocacy .............................................. 19
References ............................................................................................................21
Appendix A. Summary of country experience of CDFs .......................................... 24

Appendix B. India .................................................................................................. 27
Appendix C. Kenya ................................................................................................ 30
Appendix D. Uganda ............................................................................................. 37
Appendix E. Tanzania ........................................................................................... 40
Appendix F. Philippines ......................................................................................... 44
Appendix G. Pakistan ............................................................................................ 45
1. Defining and understanding CDFs
1.1. Introduction
Constituency Development Fund (CDF) schemes are decentralisation initiatives which
send funds from the central government to each constituency for expenditure on
development projects intended to address particular local needs. A key feature of CDF
schemes is that Members of Parliament (MP) typically exert a tremendous degree of
control over how funds are spent. The practice was first adopted in India, but gained
prominence when Kenya established a CDF in 2003. Based on the perceived success of
the Kenya model and various political and historical drivers, the trend has spread to
other African countries and across the world in recent years. The following countries
have adopted some form of a Constituency Development Fund:
 Southern Sudan  Jamaica  Zambia

 Philippines  Solomon Islands  Uganda
 Honduras  Tanzania  Ghana
 Nepal  Malawi  Malaysia
 Pakistan  Namibia  India
The major argument in favour of CDFs is that they skirt bureaucratic hassles which
weaken the efficiency and effectiveness of the usual government development
programmes. Theoretically, by sending funds directly to constituency level and enabling
communities to identify their own local development priorities, funds are spent faster,
and spent on the right things. Those supporting CDFs believe it is a vital and innovative
means to achieve tangible development outcomes at a grassroots level.
However from the citizen perspective, there are three main concerns with these
arguments. In the view of most civil society groups concerned with accountability and
governance issues, the first and foremost issue is that CDF schemes violate the
separation of powers principle on which democratic constitutional governments are built.
CDFs put MPs in the position of determining and/or implementing government
programmes, instead of focusing on their legislative and oversight functions. In the most
glaring example, Parliament approves the establishment of the CDF and funds are then
transferred directly to the bank accounts of MPs, who are responsible for choosing
projects and approving expenditure. The body responsible for managing the funds—
typically the national Ministry—reports back to Parliament, whose members are the
implementers and, often beneficiaries, of the fund. Without adequate checks and
balances built into this system to ensure independent oversight, the opportunity is wide
open for corruption, fraud, wastage and use of funds for personal or political gain by
MPs. 1
1
The best articulation of the legal argument against the constitutionality of the Kenyan CDF is captured by
Ongoya and Lumallas. In their 2005 appraisal of the CDF Act, Ongoya and Lumallas asserted that the
legislation is “inconsistent with the constitution and incapable of implementation.” Their analysis was
based on three main arguments. First, the CDF has no specific development agenda meaning that its role in
development is not clearly defined compared to the LGA or national ministry. Second, the CDF system
lacks proper governance and accountability systems because the members of the National Assembly make
the law, implement the law, and monitor their own expenditure. Third, the Act blurs the role of the MPs
and violates the constitutional principle of separation of powers. The strength of this argument will be
tested when MUHURI lodges a court case opposing the CDF Act in 2009.
-1-
Constituency Development Funds – Scoping Paper
Second, a key argument in favour of CDFs—that they enable communities themselves

to choose how funds are spent—is eroded if in reality the public are not aware of or
involved in CDF processes. Experience in several countries has shown that excessive
powers of the MP are often accompanied by very poor public participation in project
prioritisation and inadequate access to information, which undermines communities’
ability to hold authorities accountable for how funds are used. Unless legislation and
institutions are designed to curb MP’s influence and protect genuine public participation,
then the benefits of delivering funds directly to the grassroots to be prioritised by the
community will not be realized.
Even if a legislative framework is put in place to limit MP participation, ensure adequate

checks and balances, and enable genuine participation, there is a third issue which
relates to the relevance and need for the programme. The push towards decentralization
for development in some African countries has resulted in a number of decentralization
schemes being introduced in recent years. The result is a layering and duplication of
programmes as multiple schemes are set up to send funds to local level. Often these
initiatives have not reaped the anticipated development benefits due to problems with
bureaucracy, sufficiency of funding, and capacity. The concern is that the introduction of
the CDF represents another potentially problematic decentralization scheme which will
further compound the administrative burden of local authorities and result in the same
problems encountered by current initiatives. Government would do better to focus on
improving the efficiency, effectiveness and quantum of funding under existing
decentralization schemes.
Apart from the more theoretical debate around the constitutionality and relevance of
CDFs, practical issues have arisen around implementation. There is little comparative
data on design and practical impacts of these programmes despite a proliferation of CDF
schemes on multiple continents. Such information is needed by civil society groups who
wish to engage in debates in their own country around the possible establishment,
proposed expansion, or suggested reforms of CDFs.
The purpose of this scoping paper is therefore to:

 Provide an overview of CDFs: when, where and how they have emerged
worldwide;
 Identify the key features of CDFs which impact on their performance;
 Outline the arguments and evidence available for and against CDFs, and;
 Investigate the opportunities and possible future research from an advocacy
perspective.
1.2. Methodology and sources
This paper is primarily a desk study, supplemented by phone interviews with a select
number of civil society groups engaged in the CDF debate. The study is therefore
essentially limited to secondary information (apart from phone interviews) including:
online news service articles, blog discussions chiefly by civil society members, position
papers by civil society groups, and a limited number of academic articles.
The primary obstacle with this research is the lack of information. There is asyet little
academic research on CDFs and only a handful of studies/surveys on their impact or
-2-
public awareness related to CDFs. Kenya surpasses other countries in terms of making
documentation—including allocation information—available by government online.
However public access to information is one of the key critiques of CDFs in many
countries; primary documentation must therefore be obtained through country visits
which were beyond the scope of this report. The scarcity of sources for this paper is
discussed in Section 2 as one of the issues related to CDFs.
1.3. Typology of CDFs
Appendices B-F contain descriptions of the CDF schemes in a select number of

countries where CDF programmes are more entrenched and greater information was
available, while Appendix A contains a summary chart of the main characteristics of
selected CDF schemes. 2 Despite the variety of forms and approaches in different
countries, commonalities can be identified which help to distinguish CDFs from other
decentralization initiatives or community-based development programmes. The following
are the essential elements of a CDF:
1. Funds are raised by national government and disbursed at local level
2. Funds are allocated per constituency and MPs have some degree of control
over the spending
3. Funds are intended for development projects which reflect localized needs and
preferences.
Other critical features of CDFs which vary from country to country will have a direct
bearing on the performance of programme. They are:
 Quantum of funding available per constituency
 Method for allocating funds between constituencies (degree of targeting)
 Legislative framework
 Governance: institutions and systems used to select, implement and monitor
projects
 Checks on corruption
 Public access to information
 Community participation in project prioritization
From a transparency and accountability perspective, the separation of powers is the

most fundamental criticism of CDFs. As shown in the Tanzanian example (Appendix E),
some civil society organizations have elected to take a position which opposes the
establishment of a CDF in totality on the basis that it violates this core constitutional
principle. However where this battle has been lost and CDFs established despite
opposition, the country experiences gives useful examples of institutional and procedural
safeguards which can be built into CDF schemes to better protect the separation of
powers between the executive and legislative branches and to check undue influence of
MPs in project selection, committee selection, and CDF operations.
2
Information on the origin, design, processes and impact of the CDF schemes in these countries is drawn
from government documentation as well as research reports and position papers by civil society
organizations in-country. The focus of these brief country reports is on how and why the scheme was
initiated; the legislative framework governing its operations, if any; the governing institutions and financial
and administrative procedures; and finally the response by civil society groups.
-3-
Table 1 isolates the critical design features which act as the key determinants of the
degree of control afforded to the MP. Examples of various approaches are given which
allow a greater or lesser degree of control by the MP.
CDF Design Elements Impacting the Degree of MP Control
Authority to appoint CDC Signatories on

Funding flow
Committee cheques
High degree of  MP appoints all members  Funds  MP is signatory
MP influence of CDC Committee transferred
 MP serves as chair of CDC directly to MPs
personal
 MP serves on Committee account
as patron instead of chair
 MP is not signatory;
 Legislation requires only 2 signatories
representation from certain who are civil servants
groups
 Potential members of  Legislation

Committee nominated by requires that
public separate
commercial  3 or more
Low degree of  Public votes on members account set up signatories, including
MP influence of Committee for CDC district representation
CDF schemes also vary widely in the amount of funds allocated to each constituency or
MP. In some countries, the amounts are quite small, hardly enough to deliver on
substantial development projects. Other schemes were originally distributing non-
substantial amounts, but the allocations have been ratcheted up quite quickly to sizable
kitties (e.g. Philippines).
Table 2 attempts to display how some of the CDF schemes compare in terms of
quantum of funding per constituency. The table can only be interpreted as a rough
picture, as the amounts are best understood in the context of the national economy and
budget, poverty levels, and electoral systems. The amount should also be considered in
relation to the overall amount of government funding that goes to local government.
Table 1 Amounts allocated per MP in USD
Amount
GDP Total Exchange Amount allocated
Local allocated
(billion population rate to per MP
currency per MP
USD) (millions) USD (local currency)
(USD)
3
Philippines 166.91 90.35 46.8384 200,000,000 Peso $ 4,270,001
Kenya 34.51 38.53 75.5226 60,000,000 Shilling $ 794,464
Malaysia 194.93 26.99 3.4605 2,000,000 Ringgit $ 577,951
Jamaica 15.07 2.69 87.65 40,000,000 Dollar $ 456,361
India 1217.49 1139.96 47.5296 20,000,000 Rupee $ 420,790
3
24 Senators receive P200 million each; 238 House members R70 million each.
-4-
Sudan 58.44 41.35 2.3094 733,333 Pound $ 317,543

4
Pakistan 168.28 166.04 83.21 5,000,000 Rupee $ 60,089
Malawi 4.27 14.28 140.5 3,000,000 Kwacha $ 21,352
5
Tanzania 20.49 42.48 1308 18,000,000 Shilling $ 13,761
Uganda 14.53 31.66 1928 10,000,000 Shilling $ 5,187
Source: Exchange rates from http://www.xe.com on 5 October 2009.
However the figures do help to put some of the CDF debate over the potential
consequences of abuse and corruption into perspective. Although there are grave
concerns with the Ugandan scheme which delivers funds directly to the MP’s account,
the funds are quite small amounts. In contrast, the amount allocated to each
constituency in Kenya has increased rapidly since the CDF’s inception and now amounts
to approximately 790 000 USD per MP. Relative to the national economy the amounts
allocated per constituency in the Philippines, Kenya and Jamaica are substantial,
making the installation of proper management and oversight of the CDF all the more
important.
2. Advantages and disadvantages in practice
The growing body of experience on the ground with CDFs provides evidence of the
practical consequences emerging in the implementation of CDFs. Using examples and
citing evidence available from various countries, in this section we examine the critical
problems arising with CDF schemes from a citizen and civil society perspective. 6
2.1. Impact on MP-constituent relations
In some countries the stated rationale for creating the CDF has included the need to
provide funds for MPs who are burdened with constituent requests for resources for
personal hardship or community development needs. The tendency of MP-constituent
relations to be driven by MP contributions has been documented in Kenya and Uganda.
 The National Anti-Corruption Campaign Steering Committee (NACCSC) report

from Kenya gives evidence of high levels of voter dependence on monetary hand-
outs from politicians. The report found that in virtually all instances money was
used to influence voting.
 In Uganda, the Africa Leadership Institute (AFLI) compiled a score card assessing
the performance of all the MPs of the 7th parliament (elected in 2001) , and
findings were then disseminated to constituents in four constituencies. The results
indicate that voters were basing their view of MP’s performance on community
projects—“material things that the member was able to bring to the
constituency…Clearly the legislative role of the MP was not well recognized or
4
R20 million for each National Assembly member; each Provincial Assembly member receives 5 million
rupees.
5
Approximation; amount is not known publicly. See Appendix E.
6
Where possible, I cite some of the evidence on these questions. The bulk of the available research
conducted at constituency level is from Kenya (NACCSC, 2008) (CDFAP, 2008) (CCGD, 2009) although
there is also research findings from Uganda (UDN, 2008) (AFLI, 2007). Methodologies of the studies vary,
using both quantitative and qualitative data, varying sample sizes, surveys of citizens, MP interviews, other
stakeholder interviews and focus groups.
-5-
given the prominence it deserves.” The UDN report captured views from MPs
which reinforce this notion: a number of MPs felt that it was important that the MP
control the CDF funds because then constituents could recognize the contribution
of MPs in developing their area.
The criticism of the CDF is that, by providing a kitty to MPs for this type of expenditure,
the CDF reinforces MP-constituent relations which are based on commodities or benefits
in return for voter support. The evidence from constituent surveys in Uganda and Kenya
certainly suggest that, in the minds of both constituents and MPs, the CDF is
contributing to the view that MPs are agents of development in their constituencies and
are thus expected to contribute to community projects and deliver material outcomes to
their area.
A study in India by Keefer and Khemani also found that political context was a key
determinant of effort by legislators on behalf of constituents. Their main finding was that
legislator effort is significantly lower in constituencies where voters are more attached to
political parties and in constituencies that are reserved for members of socially
disadvantaged groups or are seen as candidate strongholds. “Legislators substantially
reduce effort to provide public works to their constituents when their constituency is a
party stronghold.”
If delivery can garner votes, non-delivery may cost them. The CDF Accountability Project
undertook some investigations and learnt of various incidences where voter anger over
MP non-performance or poor performance in spending CDF funds likely cost the
politician the re-election. The report cites statistics from the Electoral Commission of
Kenya (ECK) which indicate that 60% of the legislators from the 9th parliament did not
return after the 2007 elections.
The ECK statistics in collaboration with those from the CDF National Board and the
United Nations Development Programme (UNDP) Secretariats demonstrate that a
majority of the MPs who fell at the last general election across the country had
accumulated billions of CDF cash lying un-used in bank accounts with scores of
th
incomplete projects. Most of those who made it back to the 10 parliament had lesser
7
amounts of un-used CDF cash in bank accounts.
The ECK study does not prove the direct causation between CDF performance and
voter behaviour. However anecdotal evidence suggests that voters in some
constituencies notice non-performance and are holding their MP accountable. The CDF
Accountability Project concludes: “The resounding lesson to the 10th parliament seems
evident, manage your CDF prudently or else face the consequences in 2012.”
2.2. Public participation
One of main arguments in favor of CDFs is that the institution enables greater
involvement by citizens in identifying development priorities. 8 The key measure of public
participation are the representivity of the CDC committee and the inclusivity of the
processes used for project identification and selection.
7
“CDF – A Double Edged Sword.” www.cdfproject.org
8
Policy Forum. “Policy Forum Position Paper on the Constituency Development Fund.”
-6-
The Kenyan CDF cites low/non-involvement of local communities in project identification

and selection as one of the key challenges of the CDF. This is evidenced by data from
the NACCSC report that showed low levels of public participation: nearly 60% of
Kenyans are not given the opportunity to be involved in project selection or prioritization.
Approximately 25% of respondents were involved in CDF projects in some manner
(project identification or prioritization, project management, project monitoring)
(NACCSC, 2008).
The degree and manner of public participation in project selection/prioritization appears

to vary between constituencies. However the MP and CDC committee are key drivers of
project selection. Figure 1 from the NACCSC reveals the methods used in project
identification.
Figure 1 Methods Used in Project Identification/Prioritisation (Kenya)
30 Community CDF
consensus Committee
25% 25%
25
MP's suggestion
21%
20
Suggestion by
Don't know MP's close
%
15 13% associates
12% Extract from
District
10 Development
Plan
5%
5
0
Method of project identification/prioritisation
Source: National Anti-Corruption Campaign Steering Committee (June 2008). “The

Constituency Development Fund: An Examination of Legal, Structural, Management
and Corruption Issues in Kenya.”
With regard to representation on the CDF committee, the Kenyan system does make
provisions for community representation on the committee, with representatives
appointed by the MP. 9 Two places are reserved for persons representing religious
organizations, and one place each for persons representing the youth and active NGOs
in the community. Furthermore, the legislation reserves two slots on the committee for
women. 10
In Uganda, the lack of criteria or proper guidelines for appointing committee members
opened the opportunity for MPs to manipulate the composition of the committee in their
favour (AFLIA, 2008). The UDN report recommends, “Members of the committee should
not be politicians but people with integrity in the communities.” (UDN, 2009).
In Tanzania, limited efforts have been made in the new legislation to ensure NGO and
community involvement and consultation. The legislation limits the CDC Committee to
9
Legislation mandates that the committee, appointed by the MP, shall be chaired by the MP unless he/she
opts out and shall have between 12 and 15 members.
10
The legislation does not specifically indicate who appoints these representatives, but does state that the
MP will constitute and convene the CDF committee.
-7-
six members, one of which must be a person nominated by the Committee from among
the active NGOs in the area. (The other members are ward and district officials and
councilors.) Section 10.4 of the CDCF Act also stipulates that each ward shall come up
with a list of priority projects to be submitted to the CDC committee.
The situation in India is quite different. The explicit objective of the Members of
Parliament Local Area Development Scheme (MPLADS) is to enable MPs to
recommend development projects; there is no community committee establishment at
grassroots level. According to the Ministry guidelines, the projects recommended by the
MPs should be works which “meet the locally felt community infrastructure and
development needs” but there are no requirements placed on the MP to solicit
community input or set up structures which enable representivity and involvement by
constituents in project selection.
In summary, while Tanzania and Kenya have legislated a degree of community

representation on the CDF committees, on the whole, public participation in CDF
processes at grassroots level appears to be falling far short of civil society expectation.
MPs and their appointees on the CDF committees tend to dominant project prioritisation
and selection decisions and edge out public participation processes.
2.3. Selection of non-priority projects due to MP influence
Despite policy intentions of public participation and prioritization of projects which speak
to local community needs, the research shows that many constituents feel resources are
not flowing to the intended beneficiaries, primarily as the result of improper influence by
the MP. The NACCSC reported instances where the CDF office was located in the MP’s
home or rented from the MP’s building (2008). 11 This led to a situation where the
majority of respondents felt the MP favored people of his or her ethnic
group/clan/community or allocates more resources to the region where the MP was
from. Figure 2 shows the data from the NACCSC report on constituent perceptions on
how CDF resources were allocated.
11
Apart from legal and ethical issues, the personal attachment of the CDF to the MP personally also creates
practical transitional issues when the MP is not re-elected and the institution must then be disengaged from
the person. The CDF Accountability Project has highlighted instances where there was the difficulties
installing the new MP and sustaining CDF projects because the CFC office had been based at the former
MP’s premises. “A Storm Brews in CDF Kitties.” www.cdfproject.org
-8-
Figure 2 Perceptions by Constituents of CDF Resource Distribution (Kenya)
More to MP's
majority ethnic
group
15% More to MP's
Rich benefit clan
more 8%
19%
Poor benefit
more
1%
More go to MP's
Fairly distributed close associates
25% or cronies
32%
Source: National Anti-Corruption Campaign Steering Committee (June 2008). “The

Constituency Development Fund: An Examination of Legal, Structural, Management and
Corruption Issues in Kenya.”
As a result of low levels of public participation and tremendous power accorded to the
MP in CDF management, the project identification and prioritization processes are not
resulting in outcomes which meet the CDF’s policy goals of pro-poor targeting,
redistribution and grassroots development. The NACCSC report in Kenya found the
uneven distribution of projects was a problem: projects were not targeting the right
beneficiaries and projects were not reaching all community members. 78% of
respondents reporting funding of non-priority projects, such as white elephants which
remain un-utilized and projects which do not benefits the most poor and needy
(NACCSC, 2008).
2.4. Development expenditure in the absence of coherent, long-term

planning
On one hand, the purpose of the CDF is to address those specific local development
needs which may have been neglected by centrally run development programmes or to
let local communities move ahead with much-needed projects without having to wait on
central government. However, the concern has also been raised that public resources
are not maximized if they are expended on thousands of small, diverse projects which
fail to move the country together towards common goals. When the total envelope
available to the CDF is relatively small, this is less of a concern, but where political
pressures are expanding the slice of the national budget which is spent on CDF projects,
the danger increases that a sizable amount of public funds will be spent on projects
which ultimately do not have a long-term impact.
Critics in Kenya have argued that funds are most often and most easily being spent on
short-term projects which benefit a small number of residents; the larger struggle is to
-9-
identify relevant and viable projects to fund which have long term value. 12 The view
expressed by one Kenyan MP was that the CDF must be tied to national goals:
Is it appropriate to leave CDF to 210 MPs—do they have a vision in the first place. We
should lock CDF to particular projects for the next five years so that the whole country
grows together. So for instance we can all build dispensaries, or all fund water projects.
We need to tie CDF to the national goals of the country for a particular number of years
13
so that we are all growing at the same pace.
In order to maximize the impact of resource utilization, each constituency should

prioritize their projects in a coherent strategy, which is linked to national goals and has a
medium term time frame of 5 or more years. 14
In Jamaica, a reform was introduced to address this same issue. Each MP is required to
submit a 5-year development plan for their constituency, developed within the context of
national priorities. 15 While this reform promotes long-term planning, it increases the
similarity between the CDF projects and government’s regular development programme
and begs the question why a separate decentralized development programme is
required.
2.5. Intergovernmental coordination to avoid project duplication
A common concern raised in Kenya, Uganda and Pakistan among others, is that CDF
processes do not adequately protect against the duplication of development projects.
MPs eager to garner political support may initiate ‘new’ projects under their name which
are essentially duplications of their predecessor’s work. In Kenya the District Project
Committees receive the list of projects from their constituencies and are responsible for
ensuring there is no duplication. However the Kenyan CDF itself has acknowledged that
the failure to follow planning norms has led to projects being implemented close
together.
Apart from duplication within the CDF scheme, there is also the danger of duplication of
projects funded by other decentralization schemes (i.e. across national programmes)
such as the Local Authority Transfer Fund (LATF). 16 A coordinating mechanism at local
level is required to monitor and streamline the multiple sources of funding flowing to
grassroots level. In some cases, the CDF might be used to contribute to the budget of an
existing project for which there are insufficient funds available. The Joint Memo by SPAN
in Kenya recognised this need when it included broader recommendations to set up
institutions to coordinate decentralisation schemes at local level.
2.6. Targeting of allocations to poor areas
In some countries an equal amount of funds is allocated to each constituency. Other

countries—including Kenya and Tanzania—have included an equity and redistribution
12
Juma, Victor. “Bill seeks to curb wastage of CDF cash.” Posted 24 July 2009.
13
Report on Kikuyusforchange Discussion Forum on Constituency Development Funds, held 18 July 2009.
14
15
“Contractor General to Monitor Spending Under Constituency Development Fund.” Jamaica Information
Service. 13 June 2009.
16
- 10 -
objective in the CDF legislation and have thus provided for an allocation process which
is biased towards poorer areas. In Kenya, 75% of the Fund is equally divided among 210
constituencies while the balance is allocated to the poorest constituencies. 17 In his 2008
paper on the Kenyan CDF, Bagaka investigates whether CDF allocations per
constituency are based on district population characteristics i.e. size and poverty. The
evidence shows that densely populated districts with high poverty indices do indeed
receive more funds than less populated districts with fewer people living in poverty. He
concludes that the allocation formula used in the Kenyan model succeeds in promoting
equity and allocative efficiency.
Tanzania adopted a similar approach to Kenya but went further to include geographic
size and population size (in addition to poverty indices) as factors in the allocation
formula. In contrast, the CDF schemes in India, Zambia, Malawi, Uganda and Southern
Sudan allocate funds equally or on population basis. 18
Although India allocates an equal amount per MP, the development policy objectives are
promoted via Ministry guidelines which require that MPs set aside a certain amount for
vulnerable groups with particular development needs within their constituency. MPs
must recommend works costing at least 15% of funds for areas inhabited by Scheduled
Caste population and 7.5% for areas inhabited by Scheduled Tribe Population—this
translates into Rs. 30 lakh (63 000 USD) and Rs. 15 lakh (32 000 USD) respectively. 19
If funds are not prioritized towards poorer constituencies, it provides further evidence to
suggest that the main objective of the CDF scheme is to bolster the MPs’ stature. The
effectiveness of the formula used to target poorer areas will depend upon the factors and
data included. As yet, there is no study which compares changes in development
indicators per constituency to relative share of CDF funds, in order to assess impact of
targeting formulas.
2.7. Corruption and mismanagement of funds
Issues of corruption are a common thread linking all CDF schemes, backed by research
in many areas. Beyond anecdotal evidence, there is research from Kenya that clearly
documents the corruption present in CDF operation in many constituencies. Public
surveys have also recorded constituent perceptions of corruption on the part of CDF
management.
The issue is also acknowledged by the Government of Kenya which identified corruption
as one of the key problems to be investigated by the Task Force. 20 The various forms of
corruption reported in Kenya include (NACCSC, 2008) (Gikonyo, 2008):
 Funding of non-priority projects which benefit a particular few, or are ‘quick-wins’
as opposed to more long-term development projects which are difficult to
implement.
 Favouring of particular geographic areas of MP support in selecting projects
17
Oxford Analytica, “Africa: Wide CDF adoption belies limited efficacy.” 2 April 2009.
18
19
A lakh equals 100 000 rupees.
20 At the launch of the Task Force in May 2009, the Planning Minister admitted that at least 20% of funds
go to waste. Kagira, Anthony. “New Team to Review CDF Laws.” Capital News, 22 June 2009.
- 11 -
 Collusion in the awarding of tenders and committee officials/MPs acting as

suppliers
 Bribery in order to secure contracts
 Double-funding of projects
 Starting new projects instead of following through on the implementation of
existing ones, in order for an MP to tie their name to particular project and point to
their impact
 Tendering and procurement procedures are unclear and tenders are un-
advertised. Single sourcing and irregular expenditure may result.
 Poor or little contract management, leading to contractors being paid for
incomplete work or sub-standard work.
The NACCSC report gives data on the frequency of different forms of corruption and
found that nepotism and sub-standard delivery by contractors were predominant:
 Nepotism: 64%
 Shoddy implementation of projects: 60%
 Awarding of tenders irregularly: 54%
 Payment of bribes: 39%
Corruption is a result of other weaknesses in CDF management—such as lack of

reporting, weak contract management, poor oversight—because it creates an enabling
environment for abuse. The corruption is also enabled by the failure of strong systems
for redress in the CDF—due to weak institutions and the lack of political will. The CDF
Accountability Project reports that despite numerous complaints, only three cases of
CDF corruption have been prosecuted in Kenya over past six years. 21 At the same time,
the 2007 Amendments to the CDF Act stripped the National Management Board of the
powers it needed to go after corruption.
Recommendations for reform to address corruption in the Kenyan CDF centre around
the need to establish a proper process for utilizing and managing the funds. It is clear
that without a proper legislative framework, avenues for redress, and strong systems for
accountability, decentralization schemes including the CDF are more likely to bring
corruption down to the local level than achieve greater gains in development.
2.8. Need for greater professionalisation and improved project

management
The low capacity and skill level of members of the CDC Committee and Project
Committees has emerged as another key issue in Kenya and elsewhere. Where
nepotism impacts on MP appointments to the CDC, the experience and skills base of the
Committee membership may suffer.
This gap can partially be addressed by sourcing external support. According to the
Kenyan CDF, there is minimum or non-existent consultation between CDCs and
professionals in the community and low utilization of the existing technical capacity in the
district because the Act did not spell out the role of technical officers in the programme
implementation. It has also been suggested by government and civil society
21
“Failure of Redress in CDF.” www.cdfproject.org CDF Insight January 2009.
- 12 -
stakeholders that one remedy for this problem is for professionals with relevant skills to
volunteer their time to the local CDC Committee.
In Kenya recent efforts have been made to professionalize CDFs and improve their
capacity. As part of 2007 reforms, Fund or account managers were recruited and posted
to each constituency by the Board, to improve management and compliance. Their role
was to advise on the viability of proposed projects and to act as the link between the
Board and the CDFC. However some fund managers have reported of intimidation and
hostility from the MPs and Committees. The CDF Accountability Project reports that
many have come under intimidation or have met serious resistance from MPs and CDC
Committees when they raise questions around record-keeping and expenditures. 22
In Jamaica, recent efforts have been made to improve the capacity at local level. Project
managers were deployed to the 6 designated regions with oversight responsibility for
approximately 10 constituencies each. 23 The CDF-PMU assigns Project Managers to
each region, and then project officers assigned to each parish within regions to work with
MPs in project implementation. 24 To be successful, such efforts must include provisions
to support and empower professional deployees.
2.9. Additionality and sustainability
Proponents of the CDF argue that the system supplies additional discretionary funds at
the local level and makes those resources available to communities for development
projects of their own choosing. However its not always clear whether the CDF
allocations are additional injections or are diversions from other budget priorities. If the
funds are ‘new’, is their source donor funding or revenue from new or higher taxes?
In Kenya, where legislation mandates 2.5% of ordinary government revenue is allocated

to the CDF, Parliament has boosted that percentage beyond legislative requirements to
7.5%. It would be necessary to undertake a tax analysis to check if increases in taxes
collected are associated with increased allocations to the CDF, and whether the CDF
allocation is increasing at higher rate than the overall budget (implying that it is crowding
out other budget priorities).
A separate but related question is whether the CDF funding flow is supplementing or
edging out other local government revenue sources. A budget analysis at constituency
level would be necessary to determine if the CDF has resulted in a net increase in
development funds available at local level. One of the concerns raised by Kimenyi
(2005) in his analysis of the political economy of the CDF is that the scheme will weaken
fiscal effort by local governments instead of complementing local revenue-raising efforts.
However in practice, this appears to be less of a concern on the part of civil society
groups than the fear that CDFs will displace other development funding streams from
national government, instead of increase the net funds available at local level.
22
“CDF Managers’ Trials and Tribulations” www.cdfproject.org
23
“Contractor General to Monitor Spending Under Constituency Development Fund.” Jamaica Information
Service. 13 June 2009.
24
Government of Jamaica Office of the Prime Minister. “Constituency Development Fund.” 2008.
Downloaded at http://www.opm.gov.jm/library/ministry_papers/constituency_development_fund_cdf on
16 August 2009.
- 13 -
The danger of the CDF creating fiscal illusion among residents at a local level is also
explored by Bagaka in his 2008 paper on the Kenyan CDF. Similar to Kimenyi, he labels
the CDF as a delegated form of fiscal decentralization, whose main attribute is that it
allows local people to make expenditure decisions based on their own development
priorities, within the boundaries of directions from central government. Bagaka was
chiefly interested in the impact of the CDF on central government’s recurrent and capital
health budget—how do decentralization schemes impact the size of the public sector?
His analysis suggests that because regular maintenance of capital projects and salaries
of new employees tasked with running the new facilities fall under operating costs, the
CDF projects have the effect of growing the operating budget. The scheme essentially
has created “a mismatch between projects’ benefits at the constituency level and the
‘true’ operating cost of such projects.” In essence, the CDF results in the export of the
tax burden outside the constituency and the growth in government spending overall.
To remedy this mismatch, Bagaka recommends that the Act be amended to that a
portion of a constituency’s annual CDF allocation is set aside to cover the operating and
maintenance costs associated with CDF projects so that the full costs of CDF projects
are internalized at the local level.
Our understanding of the full impact of CDF schemes on the national budget is clouded
by another issue. The PF in Tanzania has noted that often only the direct costs of CDFs
are considered i.e. project costs (capital) and administrative costs. The Kenyan CDF Act
requires the line departments to be involved during project development in order to
ensure that budget estimates are ‘realistic’; in other words, to ensure that the full long-
term costs of the project are tallied, including running costs. These on-going operational
costs (salaries, maintenance etc.) must be sourced from the regular budget of line
departments. Thus theoretically every CDF project allocation must carry with it an
associated allocation in the main budget for running costs. However in practice, the
budget process does not provide for the intergovernmental coordination which would
ensure that the capital and operating costs of all CDF projects are planned for over the
long term.
In some cases, the CDF scheme becomes the platform for a tussle between levels or
departments over budget responsibilities. The CDF management may feel its
responsibility ends with capital financing and other government bodies must share the
total burden. For example, India is experiencing the same debate over the responsibility
for operational costs as Kenya. The National Ministry in India takes the position that
operations and maintenance costs are not to be funded via the MPLADS. The view of
the Ministry is that these running costs are the responsibility of the owner/user and it is
the task of the District Authority to address this issue before approving the project
(MPLADS, 2006).
The net result is that poor intergovernmental coordination undermines proper budgeting
and planning, and the underfunded projects become unsustainable. The Kenyan CDF
has acknowledged that the lack of collaboration between line departments on staff
requirements especially has led to low utilization of completed facilities especially Health
institutions and cattle dips.
2.10. Administrative costs and impact on local authorities
- 14 -
From an institutional perspective, what impact do the CDFs have on the power of local
authorities in development? The danger is that the role of local government in
development projects will be undermined, marginalized, or duplicated.
Critics of the CDF in Kenya have argued that the scheme sets up a parallel
administrative structure which is expensive, unnecessary and burdensome on the local
authority. In India, the funds flow through existing administrative structures which avoids
the set-up of a new structure. However the MPLADS puts significant additional
monitoring and reviewing responsibilities on the States which already suffer from staff
shortages. The Ministry has prohibited the use of MPLADS funds, including the
contingency set-aside, to finance additional administrative capacity for the State Nodal
departments. The view expressed by the Ministry is that the central government via the
MPLADS is already financing development projects that otherwise would be financed by
the State; thus “it would, therefore, be appropriate that overhead/establishment
expenses for the State Nodal Department is borne by the respective State/UT
Government.” (MPLADS, 2006). Restrictions on CDF expenditure which limit its use to
capital costs mean that operational cost burden (which must come from another funding
source, often local) may not be properly coordinated and/or planned for, or may be
insufficient to ensure the functioning of the facility.
Apart from the administrative and financial burden on local authorities, there is also
evidence that putting funds in the direct control of the CDC committees and MPs sets up
conflict between them and the local authorities. For example in Kenya some MPs have
argued that government officials obstruct utilization of funds and MPs should thus have
more direct control. 25
2.11. Transparency and public access to information
CDF data, including budget and expenditure figures, progress reports and
tendering/procurement documentation, is often very difficult for citizens to access. CDC
Committees, national and local government officials and MPs often hinder access, delay
publication of data, or simply refuse to release documentation. There is documented
evidence of this in Kenya. The CDF Accountability Project conducted field visits in
Nairobi’s constituencies in 2008 and found that constituents are basically unable to
access CDF project information from their local CDF office or from the national CDF
Board. During field visits, CDFCs were ‘extremely reticent’ to provide any project
information: “Officials at the constituency level in all 8 constituencies refused to provide
CDF records without the express permission of either the Board or the MP. This ‘official
secrets mentality’ facilitates a culture of corruption and impunity that must change if
widespread local economic development is to occur.” 26
Similarly, the NACCSC study in Kenya found that 88% of respondents felt that
transparency in the CDF management was poor or lacking entirely.
The same problem was encountered by MUHURI and other civil society groups in
Tanzania attempting to gain CDF project information from CDCF offices in the
constituencies.
25
Wanambisi, Laban. “Kenya MPs want more say in CDF use.” Capital News, 17 June 2009.
26
The CDF Accountability Project. “Exercising the Right to Know: The Constituency Development Fund
(CDF) in Nairobi Constituencies.” CDF Case File Report No 1, September 2008.
- 15 -
Countries with longer-running CDF programmes, such as India and Kenya, have made
progress publishing information online. The official CDF website of the Government of
Kenya is the best example of publicly available information (www.cdf.go.ke) however
there are substantial deficiencies. The website ostensibly provides information on CDC
committee members, project allocations, project status (included amount spent to date),
disbursements and reallocations per project by constituency, for 2003/04 to 2008/09.
However it is incomplete for some constituencies for some years, outdated and/or
vague.
Civil society groups say the struggle to access CDF progress and expenditure reports,
as well as information on tendering and procurement, is not particularly unique to the
CDF but applies as well to other decentralization schemes which government
bureaucrats are reticent to share information.
Efforts to access information are substantially impeded by the non-existence of Freedom

of Information legislation. In Kenya, FOI legislation has been under debate since 2005
and a Bill was actually introduced in Parliament in May 2007. However legislation has
not yet been passed. 27 In Tanzania, Article 18 of the Constitution of Tanzania gives
citizens a broad general right to information and freedom of expression, there is no
specific Act to detail and protection citizen access to information. The Newspaper Act of
1976 actually restricts public information by granting the President powers to prohibit any
publication to be imported or printed if it jeopardizes national interest.
As described in Appendix B on India, utilization rates of CDF funds in India sharply

improved following the release of the Auditor’s report and the related media coverage.
MP disbursement improved markedly once the media focused on the parliamentarians’
poor spending rate to date. Keefer and Khemani also make that the point that it was the
combination of independent credible data from the CAG and the dissemination of that
information by the media which succeeded in making an impact on government
performance.
Public access to information and transparency of the MPLADS continued to improve and
India now offers some best practices for CDF transparency. The 2005 Guidelines from
the Ministry included specific provision to protect the right of citizens to access
information on “any aspect of the MPLAD Scheme and the works
recommended/sanctioned/executed under it” as per the Right to Information Act, 2005. A
web-based system has also set up which the District Collectors use to enter the monthly
progress data which is required for the release of the next MPLADS installment. Reports
are also available to the public online, showing expenditure per State per
MP/constituency, as well as progress reports on work done in priority sectors (e.g.
drinking water, education, electricity). Although Kenya has similar information available
online, India differs from Kenya in that a Freedom of Information Act is in place, and
secondly, that Act is specifically cited in government regulations to emphasize the
applicability of the Act to CDF data.
27
Access to information was further hindered by amendments to the 1998 Communications Bill which were
approved by the President in January 2009. Section 88 of the amended act gives the Minister responsible
for Internal Security the power “to take temporary possession of any telecommunication apparatus or any
radio communication station or apparatus within Kenya,” upon “the declaration of any public emergency or
in the interest of public safety and tranquility.”
- 16 -
3. Proposed reforms and alternatives to CDFs
3.1. Views of donor organisations
The long gestation period for the Tanzanian CDCF legislation is an interesting example
of donor influence. In Tanzania a number of domestic and international NGOs have
been involved in the Policy Forum coalition opposed to the Act, including Action Aid
which currently serves as chair. 28 The Forum includes other international partners such
as: Norwegian People’s Aid; Oxfam-Great Britain; Save the Children; Accord; Catholic
Relief Services; Family Health International; and the Danish Association for International
Cooperation. Action Aid is of the view that the CDF compromises the independence of
Parliament and thus undermines their ability to provide oversight of the executive. There
is already a democratically-elected institution at local level—local government—and thus
there is no need for the parallel system which the CDF introduces.
One of the reasons the CDCF Act was not passed earlier in 2007 was that donors
refused to allow the Tanzanian government to utilize General Budget Support funds for
the CDF. Tanzania is one of the largest recipient’s countries of foreign aid in Sub-Sahara
Africa with approximately 35% of government spending dependent on foreign aid (Fiscal
Year 2008/2009). 29 Government took a decision to fund the CDCF from local sources
before the Act went through in 2009. 30 Thus civil society, teaming up with donor
organizations, were a factor in the delayed establishment of the CDCF but were not
successful in blocking its passage.
On the whole it appears that, with regard to CDFs, donor organizations have more
concerns than confidence. How those views manifest in public action varies. Some
donors have expressed explicit opposition to the establishment of CDFs. It is the view of
the German Embassy in Tanzania that the country does not need a CDF. Their
opposition is firstly based on the separation-of-powers issue. From a political
perspective, it is the view of the Embassy that the CDF favours incumbents, while from a
practical perspective, the CDF is unnecessary duplication given that donors already
support a local development fund intended to foster grassroots development. The
German Embassy has brought forward these arguments to the Tanzanian government
on many occasions, in unequivocal terms. 31 However the Embassy has not provided
grant support to organizations involved in advocacy on the matter.
In contrast, other donor organizations have not directly taken a position on CDFs but
instead relied upon grantees who undertake civic and public participation initiatives
around the issue. The Ford Foundation, Office for Eastern Africa, views the primary
advantage of the CDF is that it can serve as a critical entry-point to mobilize people at
grassroots level for change, accountability and control of resources earmarked for their
28
Phone interview with Rose Mushi, ActionAid Tanzania, Dar es Salaam, Tanzania, 2 November 2009.
29
According the Development Partners Group Tanzania (http://www.tzdpg.or.tz ), aid management in
Tanzania is guided by the Joint Assistance Strategy (JAST) jointly developed by the government and
development partners. There are three modalities used to provide assistance to Tanzania: General Budget
support (GBS), Basket Funds (BFs) and direct project funds, with the GBS being the most preferred mode
since it is consistent with the government’s legal framework and processes.
30
Email from Irenie Kiria, Tanzania Youth Action Volunteers. 29 September 2009.
31
Personal email correspondence from Dr. Guido Herz, German Embassy, Tanzania, 4 November 2009.
- 17 -
livelihoods. 32 However with regard to the Kenyan situation, their concern is that the CDF
may become a slush fund for patronage and direct bribes to Members of Parliament to
allow them to perpetuate corruption at the grassroots level and promote impunity. The
Foundation has supported NGOs and CBOs involved in advocacy around CDFs.
3.2. Reforms proposed by civil society organisations
Civil society groups have formulated a number of proposed reforms to address the
implementation issues described above. Several organizations have worked to highlight
best practices and documented the various interpretations and implementation schemes
adopted in different constituencies.
A number of the suggestions for reform have zoomed in on improving the representivity
and capacity of the CDC Committee. In Kenya, CCGD cites a best practice where CDC
Committee members are democratically elected by the community (CCGD, 2009). 33
Kikuyus For Change has argued that the CDF is a constituency fund and not a political
fund; therefore instead of a winner-take-all approach to CDC Committee composition
(where the MP selects all CDC Committee members), representation on the committee
could reflect share of vote in the local election. 34
It is also suggested that representation on the committees should be opened up to other

groups, including the private sector, CBOs, NGOs etc. Best practices are also
highlighted in Kenya where deliberate measures were taken to proactively improve
gender equity and include the participation of marginalized and vulnerable groups on the
Committee (CCDG, 2009).
In some areas, specific actions were taken to boost the capacity of the CDC Committee,
including incorporating professionals on the Committee and organizing capacity-building
programmes for both CDFC and Project Committee members (CCDG, 2009).
Other good practices provide examples of measures to improve and strengthen citizen
participation and accountability, such as: the public disclosure of the financial status of
the CDF projects; opening up the project identification and selection process to better
participation by the community; and making tendering procedures transparent were also
identified as best practices. In some instances MPs themselves took steps to connect to
constituents and increase accountability, through progress reports by CDF committees
and monthly meetings with citizens; open procurement meetings; and processes for
public nomination of CDF committee members. 35
Another set of proposed reforms centre on strengthening the capacity and resources of
MPs to carry out their oversight function. The rationale behind these recommendations is
that CDFs were instigated by a misguided understanding of the nature of MP-constituent
relations which casts the MP in the role of resource provider. Civic education efforts are
32
Personal email correspondence with Willy Mutunga, Ford Foundation, Office for Eastern Africa,
Nairobi, Kenya, 27 October 2009.
33
The 2009 report from the Collaborative Centre for Gender and Development in Kenya documents a
number of best practices and provides case studies of specific CDF projects, based on research carried out
in 10 constituencies.
34
Kikuyusforchange hold online conversations on CDF. 24 June 2009.
35
Gikonyo, Wanjiru. “Why CDF is a bucket full of holes.” The East African, 29 June 2009.
- 18 -
needed to change the mind-set of the citizens so that they fully understand the oversight
and law-making function which is the core of the MP’s function. The role of the MP is to
contribute to the parliamentary process to shape laws which promote the national good,
not to aid the welfare of individuals by intervening in personal hardship cases.
Finally it is noted that one of the reasons CDFs evolved was that opposition MPs had
little to no impact on budget-making and thus, after campaign promises, are unable to
return to constituencies and point to any concrete projects which they have effected.
CDFs thus partly arose from public frustration with the main budget process which does
not allow minority input or local responsiveness at grassroots level. This weakness
points to need to alter budget process so there are more real opportunities for input into
the budget process by legislators so that the budget can better reflect the priorities of
their constituencies.36
3.3. Directions for future research and advocacy
There is a clear need for sound research to document the impact and consequences of
CDF schemes in various countries. As can be seen from the dominance of research
from Kenya cited in this paper, the majority of constituent perception surveys and studies
of CDF processes and institutions have provided a clear understanding of the issues in
Kenya but have not enabled us to understand how different design features used in
other countries may impact on programme outcomes. This research will benefit civil
society groups undertaking advocacy campaigns for CDF reform.
Three key areas must be investigated. First, analysis of budget and expenditure
information is needed, from a public finance perspective, to understand how the
efficiency of CDF spending relates to other decentralisation schemes (utilisation rates
per constituency for each decentralisation fund). Historical budget analysis can indicate:
 From an expenditure perspective, whether the installation of CDF
schemes has resulted in a net increase in funds spent on grassroots
development; and
 From a revenue perspective, how the growth in allocations for CDF
schemes are financed and what impact they have on the budget.

Second, the relatively quick escalation of allocations per constituency (in Kenya and the
Philippines especially) raises questions around the political and social drivers of CDF
initiatives. A stakeholder analysis may be able to give insight into why CDFs take off in
some countries and are slow to be established in others.
Third, as yet we have no research telling us whether CDF schemes actually make a
difference on the ground in terms of development indicators. Despite the concerns
raised in this paper, it is important to note that survey and interview data reflect positive
views of the CDF from citizens. In Kenya, the CDF is relatively popular and well-known
by citizens, in comparison to other local level funding sources/programmes. The level of
community satisfaction with the Fund was found to be approximately 50% of
respondents (NACCSC, 2009). Although many respondents feel that the CDF can be
better implemented, on the whole it has had a positive development impact on
36
Phone interview with Vivek Ramkumar, 24 September 2009.
- 19 -
communities, in concrete tangible ways: increasing health facilities, building classrooms

and schools, providing bursaries, improving and building roads, enabling access to
drinking water, and constructing social and community facilities in villages (CCGD,
2009). For many, the CDF has become the most popular and visible vehicle for
delivering social projects at grassroots level (Mshana, 2009).
The AFLI report expressed similar sentiments from Ugandans. Views varied on whether
the funds had been well-utilized and levels of awareness of the programme ranged from
47% to 14% (AFLI, 2007). However the study found that generally people felt the fund
was a good approach to poverty alleviation and, even where there were concerns of
mismanagement of funds, people felt there was a need for the programme.
If arguments opposing the establishment of a CDF on the basis of violation of separation

of powers are disregarded and legislation is approved, civil society groups may be
forced to take the legal route and push the courts to declare the legislation
unconstitutional. The success of this approach will depend on the strength of civil
society, the court system and the rule of law.
However the emerging examples of best practices and the mere variety of CDF
schemes worldwide would suggest that it is possible to build safeguards into the system
to ensure transparency, promote public participation and substantially curb corruption;
although the violation of the separation of powers would remain a fundamental issue
This review indicates that, from a civil society perspective concerned with poverty
alleviation, transparency and accountability, the critical features of a CDF are:
1. The existence of a legislative and regulatory framework which:

 Promotes and protects public participation processes in project
prioritization and identification
 Provides for a body—apart from Parliament—which is empowered to
manage and conduct oversight
 Establishes funding flows and disbursement procedures which remain in
control of government officials, as opposed to MPs
 Ensures representivity of CDF committee at local level
 Requires CDF project and budget information to be publicly available
2. Allocation formula which prioritizes poorer constituencies
According to Gikonyo, “A CDF project can be said to be successful if it enjoys public

involvement and support, is transparently managed, and answers the development
needs of the electorate.” The evidence reviewed in this paper would suggest that these
ingredients in the design of a CDF scheme would contribute to such success.
- 20 -
References
Africa Leadership Institute. “Development or Politics of Patronage?: A study report of the

Constituency Parliamentary Debates on the Constituency Development Fund and the
Parliamentary Score Card as piloted in Nyabushozi, Makindye, Aruu and Bukedea
Constituencies in Uganda.” May 2007.
Bagaka, Obuya. “Fiscal Decentralization in Kenya: The Constituency Development Fund

and the Growth of Government.” 10 October 2008. Munich Personal RePEc Archive
MPRA Paper No. 11813, posted 29 November 2008.
CDF Accountability Project. “Exercising the Right to Know: The Constituency

Development Fund (CDF) in Nairobi Constituencies.” CDF Case File Report No 1,
September 2008.
Chiwambo, Yakobe and Wandwi, Abdul. “MPs to legalize Constituency Development

Fund soon.” Posted 16 July 2009.
Chweya, Dr. Ludeki. “Constituency Development Fund: A Critique.” The African

Executive. 2009.
Comptroller and Auditor General of India (1999). Audit Report of the Comptroller and
Auditor-General (CAG) on MPLADS 1997-2000.
Collaborative Centre for Gender and Development (2009). “Best Practices in

Constituency Development Fund (CDF).” Nairobi, Kenya.
Francis, K., Nekesa, P., and Ndungu, B. (2009). “Best Practices in Constituency
Development Fund (CDF)”. Collaborative Centre for Gender and Development: Nairobi,
Kenya.
Gazette of the United Republic of Tanzania. The Constituencies Development Catalyst

Fund Act, 2009. Special Bill Supplement No. 2, 30 June 2009.
Gikonyo, Wanjiru (2008). “The CDF Social Audit Guide.” Open Society Initiative for East
Africa: Nairobi, Kenya.
Government of India Ministry of Statistics and Programme Implementation. (2005).

“Guidelines on MPLADS.”
Government of Jamaica Office of the Prime Minister. “Constituency Development Fund.”

2008.
Government of Kenya. Constituencies Development Fund (Amendment) Act, 2007.
- 21 -
Keefer, P. and Khemani, S. When Do Legislators Pass on “Pork”?: The Determinants of

Legislator Utilization of a Constituency Development Fund in India.” Policy Research
Working Paper, the World Bank, Development Research Group. May 2009.
Khan, Shadiullah. (2006). “Local Government and Participatory Rural Development: The
Case study of District Government in North Western Pakistan.” Thesis submitted for
Phd. Department of Public Administration/Gomal University, Dera Ismail Khan.
Kimenyi, Mwangi S. “Efficiency and Efficacy of Kenya’s Constituency Development
Fund: Theory and Evidence.” University of Connecticut Department of Economics
Working Papers. 2005.
Magubira, Patty. “Tanzania: CDF Bill Tabled Amidst Objection.” Posted 24 July 2009.
Mapesa, B. and Kibua, T. (2006). “An Assessment of the Management and Utilisation of
the Constituency Development Fund in Kenya.” Institute of Policy Analysis and
Research, Discussion Paper No. 076/2005. Nairobi, Kenya.
Member of Parliament Local Area Development Scheme (MPLADS) (2006). “A

compendium on instructions/clarifications issued on MPLADS guidelines.”
Mshana, Vera. “Constituency Development Fund in Tanzania: the Right Solution?”

Policy Forum. July 2009.
National Anti-Corruption Campaign Steering Committee (June 2008). “The Constituency

Development Fund: An Examination of Legal, Structural, Management and Corruption
Issues in Kenya.”
Ongoya, Z.E. and Lumallas, Eunice. “A Critical Appraisal of the Constituency

Development Fund Act.” November 2005.
Policy Forum. “Policy Forum Position Paper on the Constituency Development Fund.”
July 2008.
Ramkumar, V. and Kidambi, S. “Twataka Pesa Zetu” (We Want Our Money): A Public
Budget Hearing in Kenya.” Muslims for Human Rights (MUHURI).
Social and Public Accountability Network (SPAN) (August 2009). “Constituency
Development Fund: A Joint Memorandum.”
Uganda Debt Network (2007a). “Briefing Paper on the Constituency Development Fund
(CDF) in Uganda.” May 2007.
Uganda Debt Network (2007b). “Final Report on Constituency Development Fund.”
Telephone interviews
Wanjiru Gikonyo, The CDF Accountability Project, Nairobi, Kenya. 2 October 2009.
- 22 -
Ali Asghar Khan, Executive Director, Omar Asghar Khan Development Foundation,
Islamabad, Pakistan. 25 September 2009.
Semkae Kilonzo, Coordinator, Policy Forum, Dar es Salaam, Tanzania. 29 September

2009.
Irenei Kiria, Executive Director, Youth Action Volunteers (YAV), Dar es Salaam,
Tanzania. 29 September 2009.
Rose Mushi, ActionAid Tanzania, Dar es Salaam, Tanzania. 2 November 2009.
Vivek Ramkumar, Manager of Open Budget Initiative, International Budget Partnership

(IBP), Washington DC. 24 September 2009.
- 23 -
Appendix A. Summary of country experience of CDFs
Year started and Legislative/regulatory Institutional Key issues emerging

Budget and allocation
original intent framework framework in press/literature
India Members of − Every MLA receives 2 Funds paid into separate MPs recommend Low spending rates in early
Parliament Local crore rupees each year account for each projects to District period increased to over
Area Development (420 790 USD) constituency Authority (District 90% currently
Scheme instituted − Small amount originally Collector) responsible for Significant authority to
1993 and then sizably scaled up implementation District Collector; degree of
For development: since introduction No CDF committee influence of MP over DC
durable community − State legislators also State Nodal Department varies
assets based on have own funds; differs coordinates with Ministry
locally felt needs from state to state
Kenya 2003 2009/10 KShs 60 million Clear legislation in place. National CDF Civil society research
Address regional to each of 210 CDF Act (2003); Management Board – gives evidence public
imbalance from constituencies (795 464 Amendments to Act semi autonomous awareness is very low;
harambees; USD) (2007); currently CDF Committee for cannot access info
redistribution. 75% equally distributed, Taskforce undertaking each constituency Corruption and mis-
Make funds remainder allocated review of CDF appointed by MP management of funds
available at according to poverty Project Management Utilization rates fairly high;
grassroots level to levels Committee for each CDF most well-
fight poverty Act mandates 2.5% of project – responsible for known/’successful’ of
ordinary govt revenue for implementation multitude of
CDF; increased to 7.5% District Project decentralization schemes
Committees responsible
for coordination
Tanzania July 2009 Total envelope and Constituencies CDCC convened and Strong opposition by civil
Implement amounts per constituency Development Catalyst chaired by MP; District society activists arguing
development not known publicly Fund (CDCF) modeled Planning Officer is CDF is unconstitutional
projects and reduce Approximated at 15-20 on Kenyan legislation secretary because violates
burden on MPs million TZS (11000-15000 separation of powers
from constituent USD) Concern with introduction
requests 25% divided equally; 45% of CDF prior to 2010
in relation to population; general election; potential
20% poverty margin; 10% misuse by MPs for
- 24 -

geographical size campaigns
Uganda September 2005 Current disbursement: 10 − 2005 Parliamentary − MP establishes Low levels public
Intended to relieve million shillings per Committee created committee and chairs awareness
MPs from constituency (5 187 USD) guidelines which not − District Project Corruption and mis-
constituent approved in plenary or Committee monitors management of funds
pressure for followed in all − Money released to
financial support for constituencies. individual MP who is
development − No comprehensive responsible for
projects legislation in place. accounting
Sudan − Total of 141 million Issue of who has signing
pounds in 2007 power for expenditure
− Each of 171 MPs get
733 333 pounds (317 543
USD)
Malaysia − Each BN ADUN in
Selangor is allocated RM2
million (577 951 USD)
annually per constituency
− Increased from RM500
000 originally
Philippines Priority Senators P200 million −
Development each (4.27 million USD);
Assistance Fund 238 House members P70
million (1.5 million USD)
each
Jamaica February 2008 − Every MP allocated J$40 − Project applications − Recent efforts made to −Reporting 98.5%
To empower MPs million (456 361 USD) in submitted to CDF-PMU improve accountability: expenditure for 2008
to respond to needs 2008, same in 2009 then referred to specially project managers −In 2009, will establish
articulated by appointed Parliamentary deployed to the 6 Constituency Project
constituents Committee which designated regions with Oversight Committees
approves oversight responsibility −Concerns over separation
for approximately 10 of powers
constituencies −Debate in Parliament on
− Also project officers possible cutting of CDF to
working with MPs to reduce budget expenditure
- 25 -

implement
Malawi To enable K3 million annually (21
communities to act 352 USD)
on small projects
without waiting for
central government
Pakistan 1985, evolved in R20 million (240356 USD)
various forms since for each National
then Assembly member; each
Provincial Assembly
member receives 5 million
rupees (60 089 USD)
- 26 -
Appendix B. India
Origin and Design of the MPLADS
India has two CDF-style schemes: the Members of Parliament Local Area Development
Scheme (MPLADS) at the national level and the Member of Legislative Assembly Local
Area Development Fund (MLA-LAD) for the Legislative Assembly of each of India’s 28
States.
The MPLADS scheme was instituted in India in 1993 under the dominant national party,
the Congress Party. Under the MPLADS, an equal amount is allocated annually to each
single-member parliamentary constituency; the funds are to be used for “works of
developmental nature with emphasis on the creation of durable community assets based
on the locally felt needs.” 37
The Ministry of Statistics and Programme Implementation has overall responsibility for
managing the funds. Each State government must designate a Nodal Department which
is responsible for coordination with the Ministry and effective supervision (including
physical inspection) of the work on site. The MPs recommend projects which are
sanctioned by the District Authority who is directly responsible for implementation. The
District Authority identifies the agency to be used to execute the project; this may be a
local government, government agency or NGO. The District Authority is also responsible
for enforcing the provisions of the guidelines with regard to admissible expenditure.
In FY 1999-2000 the amount allocated for each MP was doubled to Rs 2 crore—

approximately 420 790 USD. 38 The Ninth Report of the Lok Sabha committee on
MPLADS suggested increasing the allocation to Rs 5 crore per MP in view of cost
escalations. If funds are unspent in a given year, they are not returned to central
government but are pushed forward into the next financial year with no penalty.
MPLADS performance
The MPLADS was introduced at a time when the Congress Party was vulnerable at state
level; “The timing and manner of programme initiation..suggests that MPLADS may have
been conceived as a vehicle for the dominant national party to channel funds to its MPs
in the growing number of states controlled by the opposition.” (Keefer and Khemani,
2009). Keefer and Khemani suggest that in this context, there was no political incentive
for the national party to publicize the programme and inform voters of these resources
available for constituency service. As a result, the programme was largely unknown and
disbursement levels were quite low for the first five years. From 1993 to 1999, MPLADS
disbursements in the average and median districts were only 36 percent of available
funds (Keefer and Khemani, 2009).
37
Funds may not be spent for maintenance, repairs, grants and loans, acquisition of land, assets for
personal reward, recurrent expenditure, places of worship or religious groups. Ministry of Statistics and
Programme Implementation. “2005 Guidelines on MPLADS.”
38
1 USD = 47.5296 INR (http://www.xe.com/ 5 October 2009). 1 crore equals 10 million Indian rupees.
- 27 -
The context changed in 1999 when the Comptroller and Auditor General (CAG) released
an audit of the MPLADS programme which covered a select few states. 39 The report
drew attention to the fact that central government transfers MPLADS funds directly to the
District Authority thus the usual checks and balances for regular government
expenditure were not applicable. However the Ministry had failed to develop specific
accounting controls for the MPLADS funds to make up for this, resulting in poor financial
management (CAG, 1999).
Numerous problems were identified by the CAG Audit, including:

 Low utilization of funds due to: MPs not recommending projects and DCs delaying
in sanctioning work.
 Failure of the Ministry to obtain utilization certificates to show appropriate use of
the funds, and yet the disbursements of further installments had continued.
 Mis-reporting of financial progress of work by the District Collectors (DCs).
 Great time delays in implementation, by the DCs in approving projects and by the
implementing agencies.
 Use of funds for inadmissible purposes, including commercial and private
organizations and repair and maintenance works, purchase of stores, works on
private land etc.
Table 3 shows cumulative expenditure as a share of cumulative funds released from

1997 to 2000, as reported in the CAG audit. As of March 2000, 35.8% of cumulative
funds released had still not been spent.
Table 2 MPLADS utilization rates, 1997-2000 (cumulative)

Cumulative Cumulative
Percentage
Date up to release of funds expenditure
utilisation
(Rs in crore) (Rs in crore)
31 March 1997 2349.8 1285.5 54.71%
31 March 1998 2837.8 1549 54.58%
31 March 1999 3627.3 2315.4 63.83%
31 March 2000 5017.8 3221.21 64.20%
Source: Ministry of Statistics and Programme Implementation, as quoted in 1999
CAG Audit Report on MPLADS 1997-2000.
The CAG report also found that the physical progress on works had actually declined
between the 1st audit period (1993-97) and the 2nd audit period (1997-2000). A smaller
portion of works were completed in the latter period.
In summary, the overall finding was decidedly negative, arguing that the Ministry had
clearly failed to monitor the MPLADS and that the implementation of the scheme has
actually become worse since 1997.
When the CAG report publicized the extent of un-utilized funds, pressure was put on
government to revise guidelines to improve performance and prevent the
mismanagement of funds. In 1999 national elections brought in a new cohort of MPs.
Faced with media scrutiny of their MPLADS performance, MPs stepped up their
39
Comptroller and Auditor General of India (1999). “Audit Report of the Comptroller and Auditor-General
(CAG) on MPLADS 1997-2000.”
- 28 -
spending: between 1999 and 2003, the median MP office disbursed 85% of accumulated
funds (Keefer and Khemani, 2009).
However the study by Keefer and Khemani shows that even after the jump in utilization
rates in 1999, some MPs were still showing low expenditure rates and there were wide
variations between constituencies. 30 % of district still had spent less than 75% of
accumulated allocations by 2004.
Utilization rates have improved since then. As of 25 September 2009, a total of 20713
Rs. crore in funds are available. 40 Of this, 18522 Rs. crore or 91.35% has been
utilized. 41
Role of MPs
Notably there has been no act of Parliament passed to govern the MPLADS and MLA-
LAD. 42 Instead guidelines issued by the government set the terms for the
implementation of the scheme. Under those guidelines, most recently updated in 2005,
MPs ‘recommend’ projects to be funded. According to the Ministry: “Hon’ble MPs have
the full authority to select the works of their choice provided these are eligible under the
Guidelines.” 43 However the India scheme also gives significant authority to the District
Authority who sanctions the projects and has technical, financial and administrative
authority. Funds are not deposited in the MP’s account but are paid into a separate
account to be opened for each MP and maintained by the DC (civil servant). (The
Ministry Guidelines specifically forbid the funds being deposited in the State/UT
Government Treasury accounts). Depending on the strength of the local MP and party
politics, the MP’s influence over the District Collector will vary but in official terms the
funds remain in state government with oversight executed by the civil service and not
politicians. 44
Notably MPLADS Guidelines state that the work and site which the MP selects for a
project cannot by changed by the District Authority except with the concurrence of the
MP. However the CAG report gave evidence that in a large number of instances, DCs
were not approving projects which the MP had put forward: DCs did not sanction 24% of
the projects recommended by MPs. Furthermore DCs incurred expenditure on Rs3.97
crore worth of projects which were never recommended by MPs.
40
Figures are cumulative, since the inception of the programme in 1993. The amount available quoted here
includes interest earned.
41
Ministry of Statistics and Programme Implementation. “Statewise Status of Fund Release and
Expenditure (as on 25/09/2009). Available online.
42
Email from Sowmya Kidambi, Mazdoor Kisan Shakti Sangathan (MKSS), Rajasthan, India.
43
Member of Parliament Local Area Development Scheme (MPLADS) (2006). “A compendium on
instructions/clarifications issued on MPLADS guidelines.”
44
Phone interview with Vivek Ramkumar, 24 September 2009.
- 29 -
Appendix C. Kenya
Origin of the CDF
The CDF idea took hold in Kenya 10 years after its establishment in India. Hon. Muriuki
Karue (former MP for Ol Kalou) is credited with initiating the CDF in Kenya in 2003.
According to him, the original intention of the fund was to ensure that money was made
available right at the grassroots level development. The existing budgetary system was
not succeeding in getting sufficient funds to local level for expenditure on priority projects
defined by the communities themselves. 45 Opposition MPs had also expressed the view
that the distribution of development funds throughout regions in Kenya was biased
against the opposition areas. By providing funds directly to each constituency for fighting
poverty, the proposed CDF would assist to iron out regional imbalances due to
patronage (Bagaka, 2008). Proponents also sought to frame the initiative as a means of
ensuring equitable distribution of development funds for equitable growth (Gikonyo,
2008). Thus both equity and efficiency arguments were used to support the
establishment of the Fund.
There was also a historical impetus for the emergence of the CDF in Kenya. The long-
standing tradition of community events, termed ‘harambees,’ held to generate support
and build communities was an established feature of Kenyan life. Under the rule of
President arap Moi, some political leaders began to abuse the system for electoral
support and other MPs felt building pressure to support local harambee drives with
significant personal financial contributions. The CDF system was therefore designed to
replace the harambee system and reduce corruption by institutionalizing MPs control of
funds. 46
Finally, the CDF was located within a variety of decentralized funds which are currently
operational in Kenya, including the Secondary School Education Bursary, Local
Authority Transfer Fund, Roads Maintenance Levy Fund, Rural Electrification Program
Levy Fund and Free Primary Education, most of them established in the 90’s (Gikonyo,
2008). The CDF followed on these programmes as part of the national drive to shift the
implementation of development and service delivery to the local level in order to boost
ownership, responsiveness, and participation. 47
Total envelope
One of Karue’s main goals was to create a statutory obligation on national government
to transfer a certain amount of the budget to local level for development projects.
Although the original hope was for a 5% set-aside, the CDF Act of 2003 stipulated that
2.5% of total government revenues were to be disbursed to constituencies. 48
45
See http://www.kikuyusforchange.com/?p=194#respond
46
47
48
- 30 -
Figure 3 Kenya: CDF allocations 2003-08 as percentage of ordinary government revenue
12 4
3.5
Total annual allocations (Ksh billions)
% of ordinary government revenue

10
3
8 Total annual
2.5 allocation
(Ksh billion)
6 2
1.5 % of ordinary
4 government
1 revenue
2
0.5
0 0
4
8
/0
/0
/0
/0
/0
03
04
05
06
07
20
20
20
20
20
Source: Gikonyo, Wanjiru (2008). “The CDF Social Audit Guide." Open Society
Initiative for East Africa: Nairobi, Kenya. Pg. 6.
Figure 3 shows the total annual allocation to the CDF from 2003/04 to 2007/08, in Ksh
billions and as a percentage of ordinary government revenue. In 2006/07, the total
allocation increased to 3.5% of ordinary government revenue and then in 2007/08 the
allocation dropped to 2.7% (Gikonyo, 2008). The total kitty has increased ten-fold since
the Fund’s inception: from 1.26 Ksh billion in 2003/04 to 10.1 Ksh billion in 2008/09.
Since its launch in 2004, total of Sh44 billion has been devolved through CDF framework
(approximately 580 million USD). 49 A motion seeking to increase the allocation to 7.5%
of revenue was recently passed in Parliament. 50
Allocation of funds
The total amount of CDF funds allocated to each constituency depends upon poverty
levels in the area. 75% of the CDF is equally distributed among Kenya’s 210
constituencies. The remaining 25% is allocated to constituencies according to the
National Poverty Index multiplied by the constituencies poverty index. 51
In 2009/10 each constituency was allocated approximately KShs 60 million (794 464
USD). 52 There is a proposal to increase that amount to Ksh183 million in the next
financial year. 53
The CDF Act (2003) and CDF Amendment Act (2007) place the following limits on how
the funds may be utilized:
 Each constituency must keep 5% aside for emergency reserves.

 The maximum each constituency can spend on each area is as follows:
49
Kagiri, Anthony. “New team to review CDF laws.” Capital News, 22 June 2009.
50
Kenya Institute for Public Policy Research and Analysis (KIPPRA). See
http://www.kippra.org/Constituency.asp
51
Constituency Development Fund website http://www.cdf.go.ke/
52
1 USD = 75.5226 KES (http://www.xe.com/ 5 October 2009)
53
Gikonyo, Wanjiru. “Why CDF is a bucket full of holes.” The East African, 29 June 2009.
- 31 -
o Administrative costs: 3%
o Operations and maintenance (vehicles, equipment and machinery): 3%
o Sports Activities: 2%
o Monitoring and evaluation: 2%
o Environmental activities: 2%
 Total number of projects in a year must be a minimum of five and a maximum of
25.
 Funds cannot be used to support political or religious activities and may not
include personal awards.
 Funding must cover a complete project, or defined phase or element of a project.
 Expenditure must be on capital costs of development projects; recurrent costs may
not be included.
Governance and structures
The main governing bodies for the CDF are as follows:
 The Constituency Fund Committee is the select committee of parliament

responsible for determining the allocation and distribution of the CDF.
 The Board of Management of the CDF is a semi-autonomous body of 18 members,

composed of top government officials and representatives from civil society groups
nominated by the Minister of Planning. The Board reports to the CFC, administers
the fund and approves all payments.
 Below the Board, the District Projects Committees are comprised of all MPs in
each district as well as other local government heads and other officials. The MPs
table their list of projects to the DPC which is responsible for ensuring there is no
duplication across the district. The 2007 Amendments to the Act removed
procurement responsibility from the DPC and limited their role to primarily
coordination.
 In each constituency, the relevant MP constitutes and convenes a Constituency

Development Committee (CDC) with a maximum of 16 members. The CDC,
chaired by the MP, is responsible for deliberating on and prioritizing project
proposals.
 Finally the Projects Committee set up for each project monitors the actual
implementation of the project with the assistance of the relevant government
department.
Process for approval and disbursements
Funds are sent from the Ministry of Planning to the CDF Board which then disburses
them directly to constituency committees, into commercial bank accounts set up for each
CDC. 54 Project proposals are submitted to the MP, then forwarded to Clerk of the
National Assembly. The approved list is reviewed by the national CFC, which presents
final recommendations to the Finance Minister.
54
One of the amendments to the Act in 2007 shifted the CDF from the Ministry of Finance to the Ministry
of Planning.
- 32 -
Similar to India, unspent funds may be rolled over to the next budget year (Section 10 of
the CDF Act). IN order to improve monitoring of rollovers and unspent funds, the CDF
Amendment Act (2007) inserted two new schedules into the reporting system(Section 9).
Each year constituencies must use the Fourth Schedule to submit details on unspent
funds which they intend to re-allocate. Constituencies use the Fifth Schedule to submit
their request to re-allocate those funds to a new or ongoing eligible project. These re-
allocations must be approved by the Board (Section 9).
One of the potential dangers of permitting rollovers of unspent funds is the possibility
that MPs may influence project approvals and implementation in order to delay
expenditure until immediately prior to an election, in order to take credit with voters for
the delivery of tangible projects in the area. The Act allows rollovers of funds which ‘are
not utilized for whatever reasons’ and the Amendments are not clear on the criteria to be
used by the Board in approving the re-allocation of unspent funds. Detailed information
on the percent of funds which are rolled over, per constituency and per project, should
be provided via the website in order to improve accountability and identify any abuse of
the rollover mechanism.
Checks and balances
Section 25 of the CDF Act (2003) stipulates that there must be three signatories to the
CDC fund: one signature from a nominee of the DPC and one signature from the CDC.
The MP may not be a signatory on the account.
The CDF Implementation Guide issued by the National Management Committee is more
specific than the legislation with regards to the required signatures:
 Two nominees of the DPC, one of whom must be the District Accountant; 55 and
 The treasurer and secretary of the CDC (both appointed by the MP).
The authority of the District Development Officer (under the Ministry of Planning and
National Development) is also required for any payment to be made from the account.
Although theoretically the district official could veto approval of projects via his authority
as signatory, in practice the power to determine projects lies with the MP and his
appointees. Where the district administration is weak and district officers are unable to
stand up to the MP, the Act does not sufficiently empower the officers to enforce
regulations. The tremendous influence of the MP is thus a primary criticism of the
Kenyan system, as it allows for essentially no independent oversight. 56
Section 34 of the Act is explicit in requiring that all funds received shall be audited and
reported upon by the Controller and Auditor-General. However Gikonyo reports that a
nationwide audit of the CDF has never occurred to date. Considering that each of 210
constituencies has 5-25 projects, the likelihood that the Controller and AG have the
necessary capacity to audit thousands of projects is not high. It is recommended that an
annual national audit of the CDF must be resourced and institutionalized, and audit
findings made public in a timely manner.
55
The District Accountant is the district officer under the Ministry of Finance and is responsible for
maintaining the CDF books and ensuring compliance with procurement regulations, in addition to being
one of the required signatories on CDF cheques.
56
Phone interviews with Wanjiru Gikonyo (2 October 2009) and Vivek Ramkumar (24 September 2009).
- 33 -
With regard to procurement and tender procedures, the DPC, CDC and PCs are all
public entities under the CDF Act and therefore are subject to government procurement
regulations. All CDF procurement must be done by tender committees set up for this
purpose by the CDC, DPC and PC, However in practice most procurement is done by
the CDC without setting up a formal tender committee (Gikonyo, 2008). MPs are not
allowed to sit on the tender committees.
Sustainability and inter-governmental coordination
CDF projects are implemented by the relevant government department, which is also
has an instrumental role to play in ensuring the sustainability of projects. CDF funds may
not be used for recurrent expenditure, but this means that such costs must be planned
for and built into the budget of the relevant line department. If coordination isn’t achieved
to organize sufficient public funding for operations, staffing and maintenance, CDF
capital expenditure can result in empty or under-utilized facilities (Gikonyo, 2008).
Section 24 of the CDF Act places the responsibility on the CDC to liaise with the relevant
government department to make sure that cost estimates are realistic i.e. that recurrent
costs are included in the budget and appropriate funding sources found. In June 2009
Planning Minister Wycliffe Oparanya stated: “According to the law, the CDF committees
ought to be submitting their plans to our Ministry by March of every year after which we
are to forward them to Treasury so that the maintenance of the projects is factored in the
allocations to various ministries. This is however not happening and we want to see how
best to address it.” 57
Clarifying the roles and responsibilities of the CDC and the line departments may be one
means to address this. Some MPs have drafted a formal MOU with the heads of relevant
government departments to spell out responsibilities and contributions and set up
procedures to ensure accountability. These agreements may include minimum amounts
for remuneration of government employees and norms and standards for quality of
service (Gikonyo, 2008).
Proposed reforms
A number of Kenyan civil society groups have become involved in research and
advocacy efforts to enhance the transparency, accountability and performance of the
CDF, including the Kenya Institute for Public Policy Research and Analysis (KIPPRA),
the CDF Accountability Project, Institute of Policy Analysis and Research (IPAR),
Collaborative Centre for Gender and Development (CCGD), and Muslims for Human
Rights (MUHURI).
In August 2007, MUHURI worked with community members of Changamwe constituency

in Kenya’s Cost Province to undertake a social audit whereby constituents investigated
CDF expenditure in their community and held authorities to account for their
performance. Government documents on CDF expenditure were accessed, analyzed
and distilled into easily-understandable assessments which were shared with community
members. Residents assisted to identify problems and make on-site visits to verify
delivery of these public projects. An open public hearing was then held at which
57
Kagiri, Anthony. “New team to review CDF laws.” Capital News, 22 June 2009.
- 34 -
residents presented their findings and citizens questioned the public officials responsible
for delivery of the CDF projects and directly interrogated the area MP. 58
The social audit process proved a powerful tool, and subsequently the Open Society
Institute for East Africa (OSIEA) put together a guide for communities in 2007 which
walks them through the process of undertaking a social audit of CDF projects.
Developed by Wanjiru Gikonyo, the guide can be used as a training manual to help
communities understand how the CDF works in Kenya, how they can participate
effectively in the different stages of the project cycle, and the basic steps of a social
audit:
a. Organization of the social audit and gathering records from the government
b. Training community activists and analyzing information
c. Educating and mobilizing the public, and raising awareness of the social audit
public meeting
d. Inspecting project sites to verify information
e. Public hearing
f. Follow up
Apart from providing a practical instrument for constituents to monitor CDF spending, the
Guide also offers an analysis of the main problems with the CDF in Kenya, examples of
best practices, and suggestions for improvement. The recommendations include:
 the installation of routine audits;
 improved public participation in project selection and better access to information;
 the development of clear bursary guidelines;
 community selection of project committee members;
 strengthening of the CDF Board so that it is empowered to enforce compliance of
CDF rules and regulations;
 elimination of the duplication of roles of MPs, by making the CDFC locally elected;
 limiting the MPs’ involvement to participation in an ex-officio capacity; and
 the establishment of a monitoring and evaluation framework.
In 2008 the National Anti-Corruption Campaign Steering Committee released a study on

the CDF which they had commissioned to look into possible solutions to the reported
corruption in the CDF. Their analysis of the problems with the CDF and the required
remedies echoes the views of civil society groups (Table 4).
Table 3 Findings of the Kenyan National Anti-Corruption Campaign Steering Committee

(NACCSC), June 2008
Key problems Recommendations
 Flaws in the CDF legislation which create  Further amend CDF Ac (2007) to clarify
loopholes and permit MPs excessive roles and powers of NMB
leeway in the management of the fund  Set up mechanisms for transition of CDF
 Structural weaknesses and unclear role of committees
NMB  Increase national allocation of CDF from
 Small size of CDB 2.5% to 5%
 Low community awareness and  Carry out civic education and increase
58
Ramkumar, V. and Kidambi, S. “Twataka Pesa Zetu” (We Want Our Money): A Public Budget Hearing
in Kenya.” Muslims for Human Rights (MUHURI). See also documentary on the event, available at
www.internationalbudget.org
- 35 -
participation public vigilance and scrutiny

 Tribalism, clanism, nepotism and equity in  Undertake and institutionalize audit of CDF
the selection of CDF committees and the monies
awarding of tenders  Increase community participation and make
 Low community participation in CDF membership on CDF committees rotational
management and project prioritisation  Improve on project prioritization, reach and
 Failure to undertake feasibility studies or impact
environmental impact assessments  Streamline tendering and accountability, and
 Low project prioritization, reach and enforce compliance
impact, including duplication of projects  Sensitization and advocacy by NACCSC
and insufficient monitoring and planning
 Lack of professionalism and gender bias in
membership of CDF committees
 Lack of clear tendering and procurement
procedures; tenders often unadvertised
Source: National Anti-Corruption Campaign Steering Committee (June 2008). “The Constituency
Development Fund: An Examination of Legal, Structural, Management and Corruption Issues in Kenya.”
Five years on from enactment of the CDF, the Kenyan government announced they
would undertake a review to see how programme can be improved. The CDF Task
Force was set up in June 2009 by the Minister for Planning, National Development and
Vision 2030, Mr Wycliffe Oparanya, and is headed by Muriuki Karue. 59
The 16-member task force has been holding extensive public hearings, meeting with
stake holders and distributing questionnaires in order to solicit wide input. The
anticipated output is a review of the CDF legislation, structures and performance, and
suggestions on ways to improve accountability of the Fund and eliminate
mismanagement. The Task Force’s report is due in October 2009. 60
In response to the CDF Review Task Force, a group of civil society organizations came
together in 2009 as the Social and Public Accountability Network (SPAN) and formulated
a Joint Memorandum for the Task Force. The memo is an “appeal to official
stakeholders to take a sober view of local development in Kenya and institute reforms
that will transform not only our country but also establish professional standards of local
development to be emulated regionally and across the globe.” (SPAN, 2009).
Apart from recommending a clarification of the roles of the key institutions of the CDF,
the memo goes further to address necessary reforms for the decentralization sector
more broadly. For example, its recommended that the Board be reconstituted as a
Commission of Constituency Development and the CFC be transformed into a
Committee addressing decentralization. The memo suggests other mechanisms for
better collaboration in the decentralization sector, improved accountability, a
comprehensive citizen engagement framework and avenues for redress. It remains to be
seen whether the report of the Task Force will sufficiently reflect the main concerns of
civil society groups and the public’s input.
59
60
Wanambisi, Laban. “CDF task force report due in October.” Capital News, 26 August 2009.
- 36 -
Appendix D. Uganda
Although the Kenya CDF was established before Uganda’s, the notion of allocating
funds to MPs for development purposes in their constituencies reportedly stretches back
a number of years in Uganda (AFLI, 2007). The idea gained ground during the 7th
Parliament when the President held discussions with MPs and pledged to ease the
burden MPs experienced from pressure put to them by constituents asking for financial
support for development projects in their areas. The Ugandan government sent a
mission to Kenya to study how the Kenya CDF worked. Subsequently the 2005 State of
the Nation Address included an announcement by President Yoweri Museveni that MPs
would be given funds for development in their constituencies as part of the proposed
CDF. During a plenary session 9 September 2005, Parliament adopted the proposal to
allocate 2.95 Billion shillings for the CDF (1.5 million USD) (UDN, 2007). 61
A Parliamentary Committee was then set up in October 2005 and tasked with the job of
developing interim guidelines and procedures for the establishment of the CDF for
2005/06. The interim guidelines were discussed by the Committee but never approved
by the plenary. However despite no legislative framework being put in place,
disbursement of the funds went ahead and each constituency was given Ushs10m (5187
USD). The Clerk to Parliament released the money to the individual MP’s personal
accounts in November 2005 which was near to the conclusion of the 7th parliamentary
term and approaching elections in early 2006 (UDN, 2007). The focus on the elections
was a reason cited for the Parliament not to give adequate attention to putting legislation
and policy in place to govern the CDF (AFLI, 2007).
The governance structure proposed in the interim guidelines calls for every MP to
establish a CDF Committee of 5 people composed of him/ herself as the Chairperson, a
Secretary, a Treasurer and two other members. The District Project Committee to be
established in each district would receive and consider project proposals from the CDF
Committees in its district. The DPC would play a monitoring role, headed by the District
Community Development Officer (UDN, 2007).
Each MP would be responsible for accounting on the expenditure of the CDF funds to
the Accounting Officer (Clerk to Parliament). According to a briefing paper by the
Uganda Debt Network, the interim guidelines restrict the use of the funds to activities
that directly increase household incomes and productivity; interventions that can trigger
rapid rural transformation and economic development; and agro-processing and
marketing of produce in the respective constituencies. The stated intention was that the
money would not be used on development of infrastructure projects already underway
via local government initiatives or central government programmes. Furthermore the
funds could not be used for political and/or religious activities. The regulations
concerning how the funds can be used thus echo those established in Kenya.
Up to the present, no comprehensive legislation has been put in place to govern the
CDF and reportedly the interim guidelines developed by the Parliamentary Committee
have not been followed in most cases (UDN, 2007) (AFLIA, 2007). The current
61
1 USD = 1 928 UGX (http://www.xe.com/ 5 October 2009).
- 37 -
disbursement is 10m Ugandan shillings per constituency each year, paid directly into
MPs’ accounts (ALFIA, 2007).
Concerned that there was no law to govern the CDF which could enforce accountability
and effective and appropriate utilization of the funds, the Uganda Debt Network
conducted a study in November-December 2006 (UDN, 2007) which included a desk
review, field visits and MP interviews. Their main findings were that the interim
guidelines were not only inadequate and largely disregarded by the MPs. UDN
advocated for a clear legislative and regulatory framework to be put in place to govern
the CDF and eliminate mismanagement.
According to the UDN study, field visits to the different districts showed that over 87% of
respondents did not have knowledge of the CDF. In interviews with MPs it was shown
that the over 70% of MPs could not provide detail on which projects were funded and
were also not aware of the guidelines to be followed in spending the money.
The Africa Leadership Institute (AFLI) in Kampala also conducted a CDF study at about
the same time and their findings echoed those of the UDN. According to the AFLI report,
constituents were largely unaware of the fund and its purpose and processes were not
transparent. The report argues that although the CDF assisted to strengthen the link
between constituents and MP, “At the grassroots, the voters have no knowledge or say
on how the funds should be utilized. At all depends on the initiative of the area MP.”
Furthermore, the AFLI found evidence of mismanagement of funds and concluded that
this was primarily due to: lack of relevant law and policy; MPs unlimited powers; and no
independent signatories on the account.
Table 5 summarises the recommendations advocated by UDN and AFLI, who have been
the most vocal civil society organizations in the CDF debate in Uganda.
Table 4 Civil Society CDF Recommendations in Uganda (2007)

Recommendations from the Uganda Debt Recommendations from the Africa Leadership
Network on the CDF: Institute for improving the CDF:
 The beneficiaries of the CDF in the  There should be a five year and annual CDF plan
constituencies should be involved in the and budget. This should be backed up by a
selection and planning of the projects, so that comprehensive policy and law to guide the use of
they can participate in project implementation, CDF in the country.
monitoring and evaluation. The chosen projects
should be submitted and explained by the MPs  A CDF committee be established at the
to the Local Governments’ planning constituency, sub-county, district and national
committees, to ensure there is no duplication of levels and must include key stakeholders and
the projects funded by the government. constituency representatives. The CDF
committee at constituency level should be
 The CDF money should not be banked on the headed by the area MP and be responsible for
MPs’ personal accounts or mixed with their selecting projects to be funded.
other emoluments, but be banked on a
separate account of Local Governments where  CDF funds should go to an account controlled by
the Chief Administrative Officers should be part the CDF Committee at constituency level and not
of the CDF management. Whenever CDF the personal account of an MP.
money is disbursed, it should be publicized to
create citizens’ awareness and participation in  There should be civic education to all
the utilization and accountability of the fund. stakeholders about CDF to improve
- 38 -
accountability.
 The Clerk to Parliament should work closely
with the Chief Administrative Officers to  CDF should be increased from 10 million to 50
ascertain the existence of a credible million annually to cause more impact. Bigger
Constituency Committee to oversee the constituencies should receive more funding.
management of the CDF.
 All projects and funds should be displayed for
 The CDF accountability and auditing community to review and build consensus on the
procedures should be a function of the Clerk to beneficiaries of projects.
Parliament and the Auditor General,
respectively.
 The CDF should not simply be paid towards

the end of any Presidential/ Parliamentary term
and or impending elections to avoid a risk of
exploiting the fund for personal political gain.
- 39 -
Appendix E. Tanzania
Origins of the CDCF
Similar to Kenya, Tanzania has been in ongoing process of decentralization since Policy
Paper on Local Government Reform I 1998 and the resultant LG Reform Program
(LGRP) were set in motion (Mshana, 2009). Tanzania’s approach to local government
reform is to decentralize through devolution (D by D), thus alleviating poverty by
improved service delivery brought about by increased political, administrative and fiscal
autonomy at the local level. Intergovernmental transfer systems which sent funds from
central government to local level were set up to support these development objectives,
including the Tanzania Social Action Fund (established in 2000), followed by the Local
Government Capital Development Grant System in 2004. Key weaknesses in these
systems have emerged, including the need for better role clarification of levels of
government and increased fiscal and administrative control by local government
authorities.
Against this backdrop, the implementation of a CDF in Tanzania has been under debate
for a number of years. Similar to Uganda, a government study tour to Kenya was
conducted in July 2008 to glean lessons from the Kenyan experience. The possible
establishment of a CDF in Tanzania was considered in Parliamentary Committee and
then in President Jakaya Kikweta proposed the establishment of the CDF in August
2008, in order to assist MPs in implementing development projects and to reduce the
‘daily nuisances’ that MPs face in their constituencies. 62 One of the stated intentions was
to relieve MPs embarrassment at constituent requests which they could not
accommodate. On 3 July 2009 the Constituencies Development Catalyst Fund (CDCF)
Act was gazetted in Parliament and in August was signed by the President. 63
Mechanics of the Act
Tanzania took the Kenyan practice of allocating 25% of the fund according to the poverty
index and went further to include factors of geographical size and population in their
allocation formula. According to Section 3 of the Act, 25% of the total amount will be
divided equally between all constituencies. The remaining 75% will be split between the
constituencies as follows: 64
 45% in relation to the constituency’s population
 20% in relation to the poverty margin, and;
 10% in relation to the geographical size of the constituency.
62
President Kikwete’s Address to Parliament on 21 August 2008, as quoted in Mshana, Vera (2009),
“Constitutuency Development Fund in Tanzania: The Right Solution?”
63
According to Kilonzo, the insertion of the word ‘catalyst’ in the title of the fund is an acknowledgement
that the quantum of funding is insufficient to bring about genuine development and instead must be
approached as a top-up to existing revenue streams for development. The concern that the funds allocated
to each constituency were insufficient was one of the criticisms posed by the Policy Forum in the early
drafts of the Act.
64
Earlier drafts of the Act allocated the funds equally to all constituencies. Concern that the funds were not
biased towards poorer areas was one of the criticisms raised by the Policy Forum.
- 40 -
It is unclear how the poverty margin is calculated and what data is used for the formula.
Furthermore, the size of the total envelope allocated to the CDCF is not publicly
known. 65 The CDCF Act was passed retrospectively, effective for the FY 2009/10
beginning 1 July, yet civil society groups report that the budget guidelines and ministerial
speech did not indicate the amount allocated to the CDCF for 2009/10. 66
With respect to rollover funds, the Tanzanian CDCF legislation takes the same approach
(and largely identical language) as the Kenya CDF Act of 2003. However the Tanzanian
legislation does not set out a procedure for re-allocating unspent funds.
Similar to Kenya, the Constituency Development Catalyst Committee (CDCC) for each
constituency is convened and chaired by the MP. However it is limited to 6 members,
including the District Planning Officer who serves as Secretary.
Each ward generates a list of priority development projects which is submitted to the
CDCC of the constituency for consideration. The CDCC considers and approves
projects. All projects approved for financing by the CDCC are to be then implemented by
the relevant Council, overseen by the Council Planning Officer.
The required signatories on the account differ from the Kenyan model. Only two
signatories are required in Tanzania: one must be either the Council Director or Planning
Officer, and the second signature must be from the Council Treasurer (appointed by the
Council Director) or the Council Accountant. The signatories must be public servants.
Role of civil society
From the first discussion of a Tanzanian CDF in 2004, civil society groups opposed the
idea on the grounds that it would violate the separation of powers. The Policy Forum a
network of 96 civil society organizations in Tanzania, headed a coalition which was
joined by other groups opposed to the proposed CDCF, including international
organizations and the teachers and students union. 67 When the Bill was introduced in
Parliament in July 2009, there was heated debate with most MPs appearing to favour
the CDCF and civil society organizations, including the PF and the Tanzania Association
of Non Governmental Organizations, opposing the entire Bill outright. 68
The civil society coalition argued that the CDF is unconstitutional as MPs would be
taking on the implementation role of government instead of restricting themselves to
oversight and law-making. 69 The PF also argued that the funds were not actually
additional resources for development as assumed because they would be taken from the
65
In 2007/08 Parliament approved 7.2 billion TZS (1.5 million USD) for a CDF but legislation establishing
the Fund was never approved and it is unclear what happened to the funds. Given that Tanzania has over
300 constituencies, it is estimated that the amount per constituency could be in the region of 15-20 million
TZS (11000-15000 USD). (1 USD = 1308 TZS http://www.xe.com/ 5 October 2009).
66
Phone interview with Irenei Kiria, Tanzania Youth Action Volunteers. 29 September 2009.
67
Phone interview with Semkae Kilonzo, Policy Forum. 29 September 2009.
68
“Activists oppose Constituency Development Fund.” 16 August 2009. Daily News online edition.
Downloaded from http://www.dailynews.co.tz/home/?n=3159 16 August 2009.
69
Chiwambo, Yakobe and Wandwi, Abdul. “MPs to legalize Constituency Development Fund soon.”
Posted 16 July 2009. Downloaded from http://csrtanzania.blogsport.com/2009/07/mps-to-legalise-
constituency.html
- 41 -
general budget. 70 Concern was also raised with the timing of the Bill, and the possible
release of funds prior to the 2010 general election when MPs may be tempted to misuse
the funds for campaigning purposes. On the whole, the submission of the Policy Forum
at the public hearing in July was brushed aside, and the Act was approved 2 days later.
In a paper by Vera Mshana, published on the Policy Forum website, it is argued that the
real problems are a) systemic issues within the existing development financing
arrangements for local government; and b) the need to strengthen the Office of
Parliament to “reorient the relationship between MPs and their constituents to its
democratic rather than (apparent) financial basis.” (Mshana, 2009). The CDF is the
wrong solution to these problems for a number of reasons.
First, as demonstrated by the experience in Uganda and Kenya, the CDF system
compromises the integrity and legitimacy of parliamentarians and will negatively impact
on MP-constituent relations. MPs feel obliged to provide financial contributions to
constituents in order to get re-elected:
However, given that CDF is sourced from domestic revenue, it is a misuse of public funds
if they will be used to meet these individual assistance claims or small development
projects that MPs are asked to personally finance (such as provide business capital or
pay school fees). More importantly, creating a ‘development fund’ for this purpose does
not resolve but actually perpetuates the underlying (and problematic) financial basis of
constituent-member relations. Rather, MPs must not concede to the demands of their
constituents to provide personal financial assistance, as that is not the nature of their
representative role in a democratic society. (Mshana, 2009)
Second, even if the MPs were not involved in the CDF, the vehicle would still bring little
added value to the mission of development in Tanzania. It is argued that the CDF is an
expensive, redundant, parallel funding mechanism which is “unlikely to bring about
substantial developmental gains to citizens that cannot be realized through the LGCDG
system and TASAF, if these systems are integrated and strengthened.” The focus of
government should be on reforms to the local government system which enable the
actual realization of the objectives of the decentralization by devolution.
Instead of introducing a new function for legislators which is misplaced and potentially
undermining of the legislators’ legitimate role, the PF argued that measures should be
taken to strengthen the oversight function of MPs within the existing system, within
existing legislation (see Table 6). This would include resources to strengthen MPs’
offices and their capacity to follow up on issues and conduct research and analysis.
Furthermore, public awareness campaigns were needed to educate voters that MPs
should play a representative role and not a philanthropic one. 71
Table 5 Alternatives to the proposed CDF, suggested by Policy Forum in Tanzania (2009)
 Ensure that all MPs (including nominated or special seats MPs) have offices, preferably
in their constituencies for elected MPs and in either Dar es Salaam or Dodoma for
special seats.
70
Chiwambo, Yakobe and Wandwi, Abdul. “MPs to legalize Constituency Development Fund soon.”
Posted 16 July 2009. Downloaded from http://csrtanzania.blogsport.com/2009/07/mps-to-legalise-
constituency.html
71
- 42 -
 Ensure that all MPs offices have basic essential furniture and equipment (including IT
equipment where possible).
 Empower every MP with the ability to engage staff who are able to conduct research,
follow up issues and assist them in performing their oversight function.
 Set aside a budget for special research projects or commissioned studies to strengthen
their performance in Parliament.
 Establish accountability mechanisms to ensure that resources allocated in subsections a
to d are effectively used for their intended purpose.
Source: Policy Forum. “Policy Forum Position Paper on Constituency Development Fund.” July 2008.
From a civil society perspective, there are a number of specific concerns with the Act.
First, unlike the Kenyan model with its National Management Board which manages and
oversees the fund, the Tanzanian scheme has no national management or oversight
body. The CDCC of each constituency reports on their disbursements to the Minister
directly with no intervening body to monitor performance and take corrective action if
there is abuse of the funds. Overall management of the scheme lies with the Minister
and there is no separate oversight body for the CDCF, apart from the regular
Comptroller and Auditor-General responsible all government departments.
Second, the Act is quite vague on the content of projects to be funded, citing only that
they must be development projects which are community-based. There is no clarity or
guidance on the criteria to be used by CDCCs in selecting projects.
Given that the CDCF Act has now been signed into law by the President despite stated
opposition by civil society organizations in Tanzania, the Policy Forum now plans to shift
their strategy from lobbying to litigation. They intend to launch a court case, led by the
Legal and Human Rights Centre, to overturn the Act on the grounds that it is
unconstitutional. There is precedent in Tanzania of court cases successfully launched by
civil society groups, and the PF is optimistic of the strategy. 72
72
- 43 -
Appendix F. Philippines
In the Philippines, a CDF-type fund has been in existence in various forms for a number
of years. The Priority Development Assistance Fund (PDAF) is a kitty of discretionary
money allocated to congressmen, intended for infrastructure projects; at one stage the
fund was called the ‘countrywide development fund’ CDF. Proponents of the funds argue
that it’s a developmental tool and is particularly important for addressing the
developmental needs of rural constituencies that are often neglected in national
programmes. Opponents argue the funds are a mechanism by the executive to buy
votes of Congress and an opportunity for lawmakers to collect bribes and kickbacks from
contractors.
The amount of discretionary funds available to each legislator has risen rapidly over the
years. In 1990 the amount per representative was approximately P12.5 million. In 2004,
the amount allocated to each senator was P200 million (4.27 million USD), while
congressmen received P65 million (1.39 million USD) each year of discretionary funds
for projects of their own choosing. 73 In 2009, each of the country’s 24 senators are
receiving P200 million each and the 238 House members are allocated P70 million each
(1.5 million USD). 74
In the Philippines, the debate around the PDAF leans towards issues of ‘pork-barrel
politics’ and political party funding reform. Some analysts have suggested that the use of
‘pork-barrel’ politics increases substantially in periods when the president is vulnerable
politically and in need of Congressional support. Similar to the African debate, there is
also serious concern with how the fund impacts on MP-constituent relations. The view of
some voters is that legislators should be evaluated on their ability to make laws and
contribute to the legislative debate; others evaluate candidates on their ability to bring in
benefits to their constituency. 75
As in Tanzania and Uganda, the timing of the release of the funds prior to election is
also a hot issue. Journalists have noted the tendency of politicians to hold back or save
their PDAF funds until just prior to an election. “A few months before the 2004 elections,
a publicist of several members of the House estimated that more than half of all
congressmen had not touched their pork for projects, saving it instead for reelection
purposes.” 76
73
1 USD = 46.8384 PHP (http://www.xe.com/
74
Philippine Centre for Investigative Journalism. “PCIJ Investigation – Arroyo Sons, Friends, Foes Get Big
Public Works Deals.” (30 April 2009). Available at http://www.pcif.org/stories/2009/arroyo-sons2.html
75
Chua, Y and Cruz, B. (2004). “Pork is Political, not a Developmental Tool”. Philippine Centre for
Investigative Journalism. (7 September 2004). Available at http://pcij.org/stories/2004/pork.html
76
Chua, Y and Cruz, B. (2004). “Pork is Political, not a Developmental Tool”. Philippine Centre for
Investigative Journalism. (7 September 2004). Available at http://pcij.org/stories/2004/pork.html
- 44 -
Appendix G. Pakistan
In Pakistan, Parliament first allocated funds for use in constituencies in 1985. The funds
were largely perceived as an attempt to bribe the legislators and to increase their
influence and clout. The elections which brought them to power were boycotted by the
major political parties and the new legislators’ credibility was weak. With project
determination largely left to the discretion of the politicians, the funds were “basically
used to reinforce the power base of politicians without any systematic procedure of
accountability or public involvement.” 77 Since then the government has retained or
increased the CDF funds, through a variety of changes in the last two decades to the
local government system. 78
In 2000 former President Pervez Musharraf introduced a new local government system
which includes the Union Council at the lowest level (representing approximately 20
villages, 15 000 to 25 000 people) and the District, one level up. The Union Councils are
the local government structures closest to the grassroots and with the most direct
relationship to citizens, thus there is substantial accountability at the level of Union
Councils. Typically each District has 2-3 National Assembly members, although this
differs according to the size and population of the district. 79
The CDF monies are allocated to each National Assembly and Provincial Assembly
legislator. In the recent budget, the amount for each National Assembly member was
doubled from 10 to 20 million rupees (240 356 USD). 80 In addition, each Provincial
Assembly member receives 5 million rupees in CDF funds. 81 The total amount of funds
flowing through the CDF system is approximately 5% of the 2009/10 development
budget, but not a sizable slice (in relative terms) of the overall national budget. 82
Over the last 25 years, the system of CDFs has become entrenched with legislators
exercising full control over project selection. Legislators choose projects which are
implemented by the relevant line department. Decisions on project approval and
selection of contractors are more often than not made on political grounds, rewarding
family members or influential supporters. There are no checks on the process and no
transparency regarding selection criteria. 83
Without separate legislation governing the CDF system, the funding flows through the
regular budget processes of the Public Sector Development Programme. The funds
77
Khan, Shadiullah. (2006). “Local Government and Participatory Rural Development: The Case study of
District Government in North Western Pakistan.” Thesis submitted for Phd. Department of Public
Administration/Gomal University, Dera Ismail Khan.
78
Phone interview with Ali Asghar Khan, Executive Director, Omar Asghar Khan Development
Foundation, Islamabad, Pakistan. 25 September 2009.
79
Phone interview with Ali Asghar Khan.
80
1 USD = 83.21 PKR (http://www.xe.com/ 5 October 2009)
81
As a result, a typical district with 2 National Assembly and 3 Provincial Assembly members could
receive up to 70 million rupees. In contrast, the development funds allocated to each Union Council Nazim
are only 0.25 million. With typically 50 Union Councils in each district, the development funds under the
Union Council’s control therefore amount to approximately 12.5 million in comparison.
82
83
- 45 -
remain within the government system and are not transferred to a separate bank
account or to the MPs personal account (as in Uganda). However, although the funds
remain within the government system, the bureaucracy takes instruction from the
legislator on how the funds will be allocated and projects implemented.
The degree of public participation and establishment of a committee system to control

the funds is entirely up to the National or Provincial Assembly member, as there are no
legislated requirements for organized public participation or representative governance
structures.
The system has created frustration on the part of the local government bureaucracy
upset by the establishment of a parallel source of funding to the local level. Given that
the local government system of Union Councils and Districts is non-party based,
conflicts and competition for resources can arise between the local government and the
legislators, who are perceived as less accessible to citizens compared to the Union
Council members who are directly answerable to their constituents. 84
Civil society efforts, including those of the Omar Asghar Khan Foundation, have focused
recently on advocating for more representative, responsive local government structures
and participatory development processes, given that the local government system itself
is under attack. The term of the local government structures established under
Musharraf will expire in October 2009 and moves have been made to reshape local
government with more central appointments and less democratic representation. From a
civil society perspective, the CDF monies would be better directed to the Union Councils
or Districts, who are less influenced by party political interests and have greater
accountability to the grassroots.
84
- 46 -

Constituency Development Funds Scoping Paper

Uploaded by

Copyright:

Available Formats

Constituency Development Funds Scoping Paper

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Constituency Development Funds Scoping Paper

Uploaded by

Copyright:

Available Formats

Constituency Development Funds:

Constituency Development Fund (CDF) schemes are decentralisation initiatives which

From a transparency and accountability perspective, the separation of powers is most

1. The existence of a legislative and regulatory framework which:

1. Defining and understanding CDFs ............................................................... 1

2. Advantages and disadvantages in practice ................................................. 5

3. Proposed reforms and alternatives to CDFs.............................................. 17

Appendix A. Summary of country experience of CDFs .......................................... 24

 Southern Sudan  Jamaica  Zambia

Second, a key argument in favour of CDFs—that they enable communities themselves

Even if a legislative framework is put in place to limit MP participation, ensure adequate

The purpose of this scoping paper is therefore to:

1.2. Methodology and sources

1.3. Typology of CDFs

Appendices B-F contain descriptions of the CDF schemes in a select number of

From a transparency and accountability perspective, the separation of powers is the

CDF Design Elements Impacting the Degree of MP Control

Authority to appoint CDC Signatories on

 Potential members of  Legislation

Table 1 Amounts allocated per MP in USD

Sudan 58.44 41.35 2.3094 733,333 Pound $ 317,543

2. Advantages and disadvantages in practice

2.1. Impact on MP-constituent relations

 The National Anti-Corruption Campaign Steering Committee (NACCSC) report

2.2. Public participation

The Kenyan CDF cites low/non-involvement of local communities in project identification

The degree and manner of public participation in project selection/prioritization appears

Source: National Anti-Corruption Campaign Steering Committee (June 2008). “The

In summary, while Tanzania and Kenya have legislated a degree of community

2.3. Selection of non-priority projects due to MP influence

Figure 2 Perceptions by Constituents of CDF Resource Distribution (Kenya)

Source: National Anti-Corruption Campaign Steering Committee (June 2008). “The

2.4. Development expenditure in the absence of coherent, long-term

In order to maximize the impact of resource utilization, each constituency should

2.5. Intergovernmental coordination to avoid project duplication

2.6. Targeting of allocations to poor areas

In some countries an equal amount of funds is allocated to each constituency. Other

2.7. Corruption and mismanagement of funds

 Collusion in the awarding of tenders and committee officials/MPs acting as

Corruption is a result of other weaknesses in CDF management—such as lack of

2.8. Need for greater professionalisation and improved project

2.9. Additionality and sustainability

In Kenya, where legislation mandates 2.5% of ordinary government revenue is allocated

2.10. Administrative costs and impact on local authorities

2.11. Transparency and public access to information

Efforts to access information are substantially impeded by the non-existence of Freedom

As described in Appendix B on India, utilization rates of CDF funds in India sharply

3. Proposed reforms and alternatives to CDFs

3.1. Views of donor organisations

3.2. Reforms proposed by civil society organisations

It is also suggested that representation on the committees should be opened up to other

3.3. Directions for future research and advocacy

communities, in concrete tangible ways: increasing health facilities, building classrooms

If arguments opposing the establishment of a CDF on the basis of violation of separation

1. The existence of a legislative and regulatory framework which:

According to Gikonyo, “A CDF project can be said to be successful if it enjoys public

Africa Leadership Institute. “Development or Politics of Patronage?: A study report of the

Bagaka, Obuya. “Fiscal Decentralization in Kenya: The Constituency Development Fund